The fourth Russian of Physics for Everyone by L.

Landau and
A. Kitaigorodsky was published in 1978 as two separate books:
Physical Bodies (Book 1) and Molecules (Book 2). They were
published in English in 1979. This is the first publication of
Book 3 of this series. It is called Electrons and was written by
A. Kitaigorodsky as a sequel to Physics for Everyone.
This book deals with physical phenomena in which our attention
is focussed on the next level in the structure of matter - the
electrical structure of atoms and molecules. Electrical and radio
engineering, without which the existence of today's civilization
is inconceivable, are based on the laws governing the motion
and interaction of electrical particles, primarily electrons, the
quanta of electricity.
The main subjects of this book are electric current, magnetism
and electromagnetic fields.
------
Physics for Everyone
Book 3
A, I, Kitoigorodsky
ELECTRONS
Translated from
the Russian
by Nicholas Weinstein
Mir Publishers Moscow
<%>H3HKa JlnSl BCeX
KHHra 3
A. H.
3neKTpoHbl
Hanarenscrao HayK3:t
First published 1981
H3Jla. TeJlbCTBO eHayxa, 1979
English translation,
Mir Publishers, 1981
PREFACE
The first book of the series Physics for Euenjone dealt
with the laws of motion of large bodies and with gravi-
tational forces. The second is about the molecular 'struc-
ture of matter and molecular motion.
The present, third, book of the series discusses the
electrical structure of matter, electric forces and electro-
magnetic fields.'
The next, fourth, book is concerned with photons,
structure of the atomic nucleus and nuclear forces.
Thus, the four books of the series contain information
on ali the basic concepts and laws of physics. Specific
facts presented in the series have been selected to il-
lustrate physical laws in the clearest possible way, to
demonstrate the techniques most commonly used by
physicists in investigating phenomena, t.o give the read-
er an idea of the evolution of physical theory and,
finally, to substantiate, in a most general way, the fact
that physics is the foundation of all natural science and
engineering.
Physics has drastically changed in the lifetime of a
single generation. Many of its chapters have grown
into independent branches of science whose applications
are of tremendous significance. In my opinion, one cannot
consider that his education is complete today if he has
mastered only the fundamentals of physics. Physics
for Everyone is intended as a series of books enabling one
to acquire some specific knowledge of the principles of
physics and to find out what new advances have been
made in the physical sciences during the last decades.
Preface
This series should, of course, prove to he of greatest
interest as a teaching aid and as a supplement to the
textbook for physics students.
I remind the reader again that this is not a formal text-
book. It was written for the layman, and its purpose is
to render physics intelligible to the nonspecialized read-
er. The amount of space devoted to any subject in a
textbook depends upon the difficulty with which the
student understands the presented material. A book
on science written for the general reader does not follow
this rule. Hence, various pages do not read with equal
ease. Another essential difference is that we can permit
ourselves to expound certain traditional chapters in
a less detailed manner, condensing older material to
make room for new developments.
A few words about the present book, Electrons. Some-
what unusual use has been made of the necessity for
reminding the reader of the definitions for the simplest
concepts employed to describe electrical phenomena.
I have tried to give an idea of the phenomenological
approach to physics.
Two out of the six chapters deal with applied physics..
Electrical engineering is presented as a summary. A
detailed description would require us to resort to. draw-
ings and diagrams. It was considered feasible, there-
fore, to limit the text to a presentation of only the basic
principles of electrical engineering and of important
facts that everyone should know.
The same is true of the chapter on the radio. The small
size of the book restricted the material to a brief history
of discoveries and developments, and an account of the
fundamentals of radio engineering.
A. Kitaigorodsku
CONTENTS
Preface
1. Electricity
Electric Current 9. Stationary Electricity i 7. Electric
Fields i9. What Is Basict 25. Evolution of Eledricity
Theory 30.
1. Electrical Strudure of Matter
Minimum Quantity of Electricity 33. Ion Flow 35. Elec-
tron Beam 37. Millikan's Experiment '0. Model 01 the
Atom 45. QuantiZing Energy '7. Mendeleev's Periodic
Law 49. Eledrical Structure of Molecules 52. Dietedrlcs
56. Conduction In Gases 66. Self-Maintained Discharge
72. Matter in the Plasma State 77. Metals Ii. Electron
Emission from Metals 86. Thermoelectric Phenomena 88.
Semiconductors 90. p-n Jundion 96.
3. Electromagnetism
Measure of Magnetic Field Intensity iOt. Eftects of
Uniform Magnetic Field 1Ot. Effeds of a Nonuniform
Magnetic Field 114. Amptrian Currents 1t6. Electron
Cloud of the Atom 121.Magnetic Moments 01Particles 123.
Electromagnetic Induction 130. Direction of Induced Cur-
rent 134. Discovery of the Law of Electromagnetic In-
dUdion 137. Induced Eddy Currents i39. Inductive Surge
i4i. Magnetic Susceptibility of Iron i.l. Domains i.7.
Diamagnetic and Paramagnetic Bodies 150. Earth's Mag-
netic Field t 52. Magnetic Fields of the Stars 156.
4. Summary of Electrical Engineering
Sinusoidal Emf 158. Transformers 167. Machines that Pro-
duce Elecfric Current i70. Electric Motors i76.
Contents
5. Electromagnetic Fields
Maxwell's Equations 183. Mechanical Models of RadIa-
tion 190. Two Aspects of an Electromagnetic Field 196.
Photoelectric Effed 200. Hertz's Experiments 204. Classi-
fication of Electromagnetic Radiation 213.
6. Radio
Some History 217. Vacuum-Tube Triode and Transistor
226. Radio Transmission 230. Radio Reception 234. Ra-
dio-Wave Propagation 237. Radar 2.0. Television 24.3.
Microelectronic Circuits ~ 7 . '
8
1. Electricity'
Eledric Current
Using the study' of electricity as an example, it is
feasible (and necessary) to acquaint readers showing an
interest in physics with the phenomenological approach
to the investigation of nature. The word "phenomenon"t
from a Greek word meaning "to appear", is defined by
the Webster Dictionary as "any fact, circumstance, or
experience that is apparent to the senses and that can
be scientifioally described or' appraised". The approach
mentioned above, consists in the following. The inves-
.ttgator is not interested in the "nature of things". He
uses words only to tell about facts. His aim IS not to
"explain", hut only to describe phenomena. Almost
all the terminology that he introduces is meaningful
to him only if he can indicate some way to numerically
evaluate the corresponding concepts.
He resorts to certain auxiliary names only to facili-
tate a" verbal account of the facts. But the role of these
names is absolutely secondary, they could just as well
be replaced .by other names or by sim-ply saying "some-
thing)) or "thingumajig".
The phenomenological method plays an immense role
in natural science. Electrical phenomena are exception-
ally suitable examples for explaining the essence of
this- method to the reader.
At the end of this chapter I shall review briefly the
actual sequence of events in the history of electricity
theory.At the present, I wish to present a certain ideal-
Electrons 10
ized outline for the evolution of the phenomenological
theory of electrical phenomena.
First, let us combine into a single mythical personage
such scientists as Charles Augustin de Coulomb (1736-
1806), Alessandro Volta (1745-1827), Georg Simon Ohm
(1787-1854), Andre Marie Ampere (1775-1836), Hans
Christian Oersted (1777-1851), Heinrich Friedrich Emil
Lenz* (1804-1865) and others of that ilk. Imagine that
our corporate investigator is endowed with a capacity
for modern scientific thought, and let us equip him with
a complete set of up-to-date terminology. This composite
scientist is to play the chief role in our further discus-
sion,
He begins to concei ve a phenomenological theory of
electricity by carefully examining a storage battery.
The first feature he takes notice of is that the battery has
two "poles". Touching them with his two hands, he in-
stantly learns that this is something to avoid (the shock
may be quite severe). But after this first experiment,
it probably occurs to him that something passed through
his body; we shall name this "something" electricity.
Being extremely careful not to suffer another shock,
be begins to connect the poles (actually terminals) by
various wires, rods and cords. He then discovers the
following: of the articles brought into contact with the two
poles, some are intensely heated, others become only
warm and, in some cases, no heating is observed.
Selecting appropriate words to describe his discovery,
our investigator decides to speak of it as follows: "When
I connect the poles with a wire, electricity flows through
the wire. I shall call this phenomenon an electric cur-
rent. Experiments indicated that items of different ma-
terials are differently heated by the current. Those which
.. Known as Emil Christianovich Lenz in his native Russia.
1. Electricity 11
are only slightly heated evidently "conduct" electric-
ity poorly or offer high resistance to the current. They
can be called insulators or dielectrics."
Next our investigator begins experimenting with liq-
uids. Again he finds that different substances behave
differently. Finally, he makes an interesting discovery:
taking a solution of copper sulphate and immersing car-
bon electrodes (as we call the items connected to the poles)
into the bath, our scientist finds a reddish copper deposit
on one of the carbons.
At this stage, our investigator becomes convinced
that the phenomenon he is studying has to do with the
flow of some kind of fluid. It obviously makes sense
to speak of the direction of flow. Let us agree upon mark-
ing the electrode on which the copper is deposited
with a minus sign and consider the other electrode to be
positive. Since it is longish and inconvenient to say
"negative electrode" and "positive electrode" each time,
the terms cathode and anode are suggested instead. Cur-
rent passage is from the plus to the minus, i.e. from the
anode to the cathode.
But the value of the discovery is far from being ex-
hausted at this point. It is further established that an
equal mass of copper is deposited on the cathode each
second. Evidently, the copper atoms carry the electric
fluid. Therefore, the investigator introduces two new
terms. First, he supposes that the mass M of the copper
is proportional to the quantity q of electricity flowing
through the circuit, i.e. he introduces the equation
q=kM
where k is the proportionality factor. Next, he proposes
that the quantity of electricity passing along the cir-
cuit in unit time be called the current strength or, sim-
ply, current:
Electrons 12
OUf investigator has .become substantially' enriched.
He can now characterize the current by two measurable
quantities: by the amount of heat evolved by a definite
portion of the circuit in unit time; and by the .current,
the quantity of electricity flowing in unit time.
This leads to a new opportunity: he can compare the
currents produced by various so.urces. He measures the
current I and then the energy Q generated in the form of
heat ,by -a single piece of wire. Repeating the experiment
with various conductors, the investigator finds that the
ratio of the amount of heat generated to the current in
the wire differs for various current sources. I t remains
to invent a term.for this ratio. It was called the voltage.
The higher the voltage, the .more the heat generated.
This last statement can, to some small extent, justify
our choice of terms. The more the tension required to pull
a loaded wagon (and tension is sometimes used to de-
note voltage, e.g. high-tension current), the hotter we
get. .Denoting the voltage by V, as is customary today,
we obtain
Q
V ::::c -., or Q= VI 't
q
Thus, the first steps have been taken. Two phenomena
have been discovered. An electric current deposits mat-
ter in passing through certain liquids, and an electric
current generates heat. Heat is something we are capable
of measuring. The method for measuring the quantity
of electricity has been devised, i.e. a definition of this
concept has been given. Definitions have also been given
for derived concepts: the current and the voltage.
A number of simple equations have been written, but
notice that they cannot be called laws of nature, For
1. Electricity -t 3
instance, our investigator called the ratio Q/q the-volt-
age, .instead of finding. that Q/q is equal to the voltage,
'Now, he begins to search for the law of nature. Two
quantities can be measured for the same conductor:
the current and heat evolved, or the: current andv.olt-
age (which, in principle, mean the same).
An investigation of the dependence of the current on
the voltage leads to the discovery of an important law.
The vast majority of conductors comply with the law:
V=IR
.The quantity R can be called the resistance; this fully
agrees with the initial qualitative observations. The
reader has, of course, recognized this equation; it is
Ohm's law. Substituting the value of the current from
Ohm '8 law into our preceding .Iormula, w ~ obtain
VI
Q.=[[ 't
You will not be confused, I hope, .by the possibility of
writing the expresslen for the .. energy .evolved by a con-
ductor in the form of heat in: a. different way:
Q = ]IR-c
It follows from the first of the last two formulas that
the amount of heat is inversely proportional to the
resistance: This is true if we add: at constant voltage,
This is the case we had in mind when we :fi.rst used the
term "resistance". The second formula, contending that
the heat is directly proportional to the resistance, re'::
quires the condition: at constant current.
These two equations are evidently recognized by the
reader as the law named after J ames Prescott Joule
( 1 8 1 8 - ~ 8 8 9 ) and Lenz (Le.. the J oule-Lenz tlaw).
Finding thus that the voltage and current are propor-
tional, thereby enabling .the resistance of a -conductor
Electrons 14
to be determined, our investigator naturally poses the
question: How is this important quantity related to
the shape and size of the conductor and to the material
of which it is made?
Experiments lead to the following discovery. It is
found that
1
R=PT
where l is the length of the conductor and A is its cross-
sectional area. This simple equation is valid when we deal
with a linear conductor of constant cross section along
its' whole length.. Resorting, if necessary, to more complex
mathematical operations, we can write the resistance
formula for a conductor of any shape. What, here, is the
factor p? It characterizes the material of which the conduc-
tor is made. The value of this quantity, named the re-
ststiuitu, varies in an extremely wide range. The resis-
tivity of various substances may vary by thousands
of millions. of times.
Let us carry out several formal transformations that
will prove useful further on. Ohm '8 law can be written
as
1
- VA
- pI
A frequently employed quantity is the ratio of the
current to the cross-sectional area of the conductor.
It is called the current density and is usually denoted by
the letter j. Then the same law can be written as
. 1 V
J-=-PT
At this point, it seems to our investigator that he has
found everything 'related to Ohm '8 law. If he has at his
disposal an unlimited number of conductors of known
t. Electricity
15
resistance, our investigator can reject the cumbersome
technique of determining voltage by means of a calorim-
eter: he now knows that the voltage equals the product
of the current by the resistance.
Our investigator soon finds, however, that this state-
ment is in need of refinement. Using the same current
source, he connects its poles through various resistances.
The current naturally differs in each experiment. But,
he finds, the product 1R of the current and resistance
does not remain constant. When he studies this, as yet
unexplained, phenomenon, the investigator finds that
with an increase in the resistance the product I R tends
to a certain constant value.
Denoting this limit by ~ , we derive a formula that
does not coincide with the one established when measur-
ing the voltage and current. The new formula is
e= I(R +r)
What a strange contradictionl
After some thinking, we come to the conclusion that
there is, of course, no real contradiction. When we di-
rectly measured the voltage with a calorimeter, we were
concerned only with the conductor that connected the
storage battery terminals. It must be clear, however,
that heat is also evolved in the battery itself (we can
make sure by touching the batter)? with our fingers).
The storage battery has its own resistance. The meaning
of quantity r, found in the new formula, is obvious,
it is the internal resistance of the current source. Quan-
tity ~ requires a special name. One cannot contend that the
nama selected is particularly appropriate. Quantity (S
IS called the electromotive force (emf), though it has
neither the meaning nor the dimenslonality of force.
Bot.h formulas continue to be called Ohm'8 law (ob-
serving, so to speak, historical fairness), only the first is
Electrons 18
called Ohm's law for a portion of a circuit, and the 'second,
Ohms law for the- whole: circuit.
Now, everything seems to hecleared up. The laws of
direct current have heen established.
But our investigator is still unsat.isfied. Even without
direct measurement of the voltage with a calorimeter,
its determination remains cumherso.me. Imagine weigh-
ing the cathode with its copper deposit each timel
You must admit that this is extremely inconvenient, to
say the least.
One truly fine day, our investigator quite accidentally
placed a compass near a conductor in which there was a
current. This turned out to be a great discovery. The
magnetic needle turned violently when current passed
through the, conductor, turning in the opposite direction
when the current was reversed.
The moment of force acting on the magnetic needle
can be readily determined. Hence, a measuring instru-
ment can he devised on the basis of this observed phenom-
enon. All that is required is to establish the dependence
of the- moment on the current. Our .investigator solves
this problem and designs excellent pointer-type instru-
ments for- measuring the current and 'voltage.
Our account of the research carried out bv our com-
posite investigator during the first half of the nineteenth
century in studying the laws of direct .currents would
Dot be complete if we- did not mention that .lre discovered
the interaction of currents. Parallel conductors with
currents' travelling in the same direction attract each
other. They repel each other when. thecurrents travel
in .opposite directions. Of course, this phenomenon can
also be .used to measure the' current.
I .shallcertainly Dot, limit myself to the last paragraphs
in presenting -the laws of electromagnetism; a whole
chapter- is devoted to. this, subject. But I had to remind
the reader of these important facts to achieve the aim
t. Electricity f 7
of the present chapter, which is to show how the basic
quantitative concepts and units of measurement describ-
ing electrical phenomena-current., charge and field-
are introduced.
Stationary Electricity
We shall assume that our idealized investigator has a
comprehensive knowledge of the wide variety of phe-
nomena that were said to be electrical in the remote past.
The special properties of amber, or a glass rod rubbed with
fur, the initiation of sparks jumping between two bodies
that have been electrified were studied (or, rather, were
used for striking demonstrations) for a sufficiently long
time. Naturally, our investigator studying electric cur-
rents asked himself: Is the fluid passing through a wire
and the one that can exist in a stationary state on some
body until the body is "discharged" one and the same
"something"?
Even if we digress from the information that has been
accumulated previously, is it not necessary to pose the
question: If electricity is "something" that flows
through a conductor like a liquid, can it be "poured into
a glass"?
If our investigator wants a direct answer to this ques-
tion, he should proceed as follows. First he obtains a
current source of sufficiently high voltage (so far we have
not mentioned units of measurement, and so the reader
should wait a while before asking what we mean by a
high voltage, a heavy current, etc.). One of the battery
poles is grounded and a tiny hollow bead, made of very thin
~ l u m i n i u m foil, is placed on the second pole. This bead
IS suspended by a silk thread. The same is done with
another such bead.
Next the two tiny beads are brought close together
(say, to a distance of 2 mm between their centres).
2-2259
Electrons 18
To his extreme delight or amazement (or any other epithet
you care to use), our investigator finds that the beads
repel each other. From the angle between the threads
and the known mass of the beads, he can compute the
force acting between the beads.
The investigator also establishes the following fact:
if the beads are charged by touching the same pole of
the battery, they repel each other. If one bead obtains
electricity from one pole and the other bead from the other
pole, they attract each other.
This experiment confirms the right to speak of electric-
ity as something behaving like a liquid and shows that
we can deal both with electricity in motion and at rest.
Since our investigator can determine the quantity of
electricity by weighing the mass of the copper deposited
on the cathode, he can find "how much liquid was poured
into the glass", i.e. the quantity of electricity taken by
the bead from the battery electrode.
First of all, the investigator observes the following.
If a charged bead is earthed, or grounded, i.e. connected
to the earth by a wire or other conductor, the bead loses
its charge. Next, he can show that the charge "drains"
through the wire, i.e, that a current passes through the
wire. Finally, he can measure the amount of copper depos-
ited on a cathode in an instrument with an electrolyte
(conducting solution), inserted into the path of the charge
to the earth. This, in other words, is a measurement of
the quantity of static (stationary) electricity that was
on the charged bead.
This quantity of electricity is called the charge on the
bead by our investigator who ascribes a sign-positive or
negative-depending upon which electrode the electric
fluid was taken from.
Now we can commence our next series of experiments.
Various quantities of electricity can be obtained by
beads or balls of various sizes from the poles of various
t. Electricity 19
batteries. Locating the balls at different distances from
one another, he can measure the force of interaction
between them. This leads him to the following impor-
tant law of nature:
F=KWJ.
r'J.
which states that the force of interaction is directly
proportional to the product of the charges on the balls
and inversely proportional to the square of the distance
between them. The reader recognizes it as the expression of
Coulomb's law which was established differently than
we have described. But our investigator is no historical
character.
Electric Fields
Our investigator knows of forces of two different kinds.
One kind is developed upon direct contact of one body
with another. Such is the case in pulling or pushing.
With respect to forces that act at a distance, so far he
knew only of the force of gravity or, in a wider aspect,
the force of universal gravitation.
Thus a new force was added to the remotely acting
ones: the force of Coulomb attraction or repulsion be-
tween two charged bodies. It closely resembles the grav-
itational force. Even the formulas resemble each other.
The force of gravity exerted by the earth on a body
does not lead to any particular difficulties in calcula-
tions. As for Coulomb, or, as they are also called, elec-
trostatic, forces, we may encounter cases in which electric
charges are distributed in space in some complex and,
even worse, unknown manner.
But we can manage without knowing the way in which
the charges are distributed. We know that these charges
"sense" one another at a distance. Why not say that the
Electrons 20
charges set up an electric fieldi It may seem at first that
difficulties arise from the fact that we cannot see the electric
field. "But I think," says our investigator, "that the
electric field should not be regarded as mathematical
fiction for facilitating calculations. If a charge located
at some point is subject to a force, this means that this
point (in space) is in a special state. An electric field is
a physical reality, i.e. it exists by itself even though
we cannot see it." Since he made this guess at the begin-
ning of the nineteenth century, our investigator could
not prove it. But the future showed that it was correct.
Coulomb's law establishes a formula by means of
which we can determine the effect of one small ball on
another. One of the balls can be fixed and the other
moved to various points in space. At all points the mov-
able (test) ball is subject to a force. Today, this fact
is differently formulated: a ball charged with electricity
sets up a field of electric forces or, more concisely, an
electric field about itself.
The source of an electric field may be charged bodies
of any shape. Coulomb '8 law is not valid in this case, but,
using a test ball, we can measure the" electric field sur-
rounding a charged body and specify this field in an en-
tirely comprehensive way by indicating the magnitude
and direction of the forces. To avoid the dependence of
the specification of an electric field on the charge of the
test ball, an electric field is characterized by its inten-
sity:
F
E=-
q
where q is the electric charge of the test ball.
There is a visual way of describing an electric field
by lines of force. Diagrams showing these lines of force
may differ greatly depending upon the shape of the charged
bodies and their mutual location. The simplest diagrams
1. Electricity
21
Figure f.f
of electric fields are illustrated in Figure 1.1. Their
meanings are the following. A tangent at any point to a
line of force indicates the direction of the electric force
at that point. The number of lines per unit of area per-
pendicular to the lines of force is entirely arbitrary
provided that it is proportional to quantity E. When,
however, we speak of the number of lines of force without
using a diagram, it is assumed that this number is simply
equal to E.
If a free electric charge is placed in an electric field,
it moves along the lines of force, unless other forces, for
instance gravity, interfere.
Of simplest appearance are force fields set up by bodies
of spherical shape. If two spheres or two charges that
can be regarded as points approach each other, their
fields are superimposed. Their field intensities are added
by the parallelogram method. We can determine the di-
rection of the line of force at any point A and find the
field intensity by constructing the parallelogram of
forces as shown in Figure 1.1.
If the charged bodies are in the form of plates, the
field resembles that shown below in the drawing. By
Electrons 22
bringing the plates closer together and increasing their
area, we can obtain almost ideal field uniformity, with
only a slight edge, or fringe, effect. It can be said of two
closely located plates that they condense the field. Ac-
cordingly, such a device is called a condenser. It is now
generally known as a capacitor.
As we know, work done in moving a body subject to a
force is equal to the product of the force by the length
of path. To transfer a charge from one plate of a capaci-
tor to the other along a line of force, the work required
is equal to qEl. The work required to carryover a unit
quantity of electricity equals E1.
Now let us connect the two plates of a capacitor with a
conductor. The amount of energy generated when quan-
tity q of electricity is passed through the conductor equals
qV. Since we already realize that there is no difference
in principle between the motion of a charged ball in an
electric field and the passage of the electric "fluid"
along the metal conductor, we can equate these two ex-
pressions for the energy expended by the field. Thus
qEl=qV
The validity of this equation can be readily checked
by moving the plates of a capacitor farther apart and mea-
suring the force exerted on a test charge.
This measurement can be done by an elegant procedure
without resorting to a charged ball suspended by '8 silk
thread.
Everybody knows that light bodies fall considerably
slower than heavy ones. It will be recalled that this is
what led wise men of antiquity and the Middle Ages to
suppose, previous to Galileo's experiments, ths t the ve-
locity of motion (and not the acceleration) of a body is
proportional to the force acting on it. This point of view
was proved false only after it had been demonstrated
that 8 piece of paper and a metal ball drop side by side
1. Electricity 23
in a vertical glass tube from which the air has been pumped
out. It was found that all bodies gain velocity in
falling at the same rate, i.e. fall to the earth with the
same acceleration. For our purposes, however, we can
utilize the influence of air resistance to make our light
hollow metal ball (or bead), used to demonstrate Cou-
lomb's law, fall very slowly.
If this ball is dropped between the plates of a capac-
itor, we can vary the voltage between the plates until
the electric field that is set up stops the ball from fall-
ing further. Equilibrium is achieved when the force
of gravity equals the field force: mg = qE. This equation
can be used to determine the field intensity and to
prove the validity of our theoretical reasoning.
The number of lines of force passing through any imag-
inary or real surface in an electric field is selected so
that in our chosen units of measurement it equals the
electric flux. Can we find the electric flux passing through
a closed surface that encompasses charged bodies?
Let us first consider the simplest case. The field is
set up by one small ball. We describe a sphere around
the ball. If the radius of the sphere is R, then the inten-
sity at any point of the surface on the sphere is KqIR2.
The area of the sphere is 41tR2. Thus the electric flux
through the sphere is 41tKq. It is clear, however, that
the flux remains the same regardless of the kind of sur-
face we consider.
Next we complicate the picture by assuming that the
field is set up by a large number of charged bodies of any
shape. These bodies, however, can be imagined to be
broken down into tiny portions, each being equivalent
to a point charge. We enclose the system of charges with
a surface of arbitrary shape. The flux from each charge
equals 41CKq. It is quite natural to assume that the
fluxes can be added together arithmetically. Hence, the
total flux through any closed surface, encompassing all
Electrons 24
the charges, is proportional to the sum of the charges
of all the bodies enclosed in this surface.
This statement is the basic law governing electric
fields (one of Maxwell '8 four equations, see Chapter 5).
Note that we neither derived nor proved this formula.
We simply guessed that this is the way things are and
not differently. This means we are dealing with a general
law of nature whose validity is established by an exper-
imental confirmation of any consequences that follow
from the general law.
It is vitally important to know the general rule that
holds for any system. On the basis of the written law,
an electronic computer requires only seconds to calcu-
late the required data on an electric field set up by a
most complex system of charged bodies. We shall re-
strict ourselves to the modest problem of deriving a for-
mula of practical importance for determining the ca-
pacitance of a capacitor (demonstrating, at the same
time, techniques of theoretical physics applied to an ele-
mentary case).
First we define this well-known quantity. The capac-
itance of a capacitor is the ratio of the charge that ac-
cumulates on its plates to the voltage applied across the
plates. Thus
In a capacitor, the lines of force do not pass out beyond
the edges of the plates to any appreciable extent. They
emerge from the positive plate and enter the negative
one. Neglecting distortion of the field at the edges of
the capacitor, we can write the product EA for the flux.
It follows from the general law that
EA=41tKq
f. Eledricity 25
from which the field intensity between the plates is
E = 41tK ~
On the other hand, the field intensity of a capacitor can
be written as
V
E=(f
","here d is the distance between the plates. Equating
the last two equations and recalling that C=qIV, we ob-
tain the formula for the capacitance of a capacitor:
A
C = 41tKd
In their actual design, capacitors are strips of metal
pressed against the faces of strips of mica or paraffin-
impregnated paper. The last two are insulating mate-
rials. What do we gain by inserting a dielectric between
the capacitor plates? Experiments show that the capac-
itance C of a capacitor is related to its capacitance
Co when it has no separating medium by the formula
C=Cos.
The constant E is called the permittivity, or dielec-
tric constant. For air, mica, water and Rochelle salt,
the permittivities are equal to 1, approximately 6, 81
and 9000.
Whit Is Basic!
Ohm's law and the Joule-Lenz law relate energy,
current, voltage and resistance. We can say that the volt-
age is equal to the product of the current by the resis-
tance. But we can also say: the current is the voltage
divided by the resistance. Both definitions are found
in textbooks and both have the shortcoming that they
Electrons 26
prove convenient only in cases when Ohm'8 law is valid.
And, as mentioned before, this law is not always true.
Consequently, it is better to proceed as we did, namely,
to assume t.hat the derived quantity is the resistance
of the conductor, and that it is defined as the ratio of
the voltage across the ends of the conductor to the cur-
rent passing through it.
Since the energy of an electric current can be measured
on the basis of the energy conservation law-by the
thermal or mechanical effects of the current-it is clear-
ly expedient to define the current or voltage as quanti-
ties derived from the energy. The most natural procedure
is to determine the current from electrolysis phenomena
and the voltage across the ends of a portion of a circuit
as the quotient of the generated energy divided by the
quantity of electricity.
The reader must clearly perceive, however, that this
is not the only possible system of determining these
electrical quantities. Instead of electrolysis, i.e. depo-
sition of metal on an electrode, the current can be deter-
mined on the basis of any other of its effects, such as
the effect of a current on a magnetic needle or on another
current.
There is nothing wrong in principle with another pro-
cedure: we select a certain standard current source and
then the voltage of any other current source is deter-
mined as the number of equal standard elements. Such
a proposal was made, and the standard element is called
the Weston cell.
Still another possibility is to devise the system of
definitions and units of measurement by selecting a cer-
tain standard resistance. Then, as above, we find how
many standard resistances can substitute for the given
conductor. At one time, a column of mercury of speci-
fied length and cross section was used as such a unit of
resistance.
t. Electricity 27
It is helpful to understand that the sequence in which
physical concepts are introduced is a random matter.
It does not in any way, of course, alter the laws of nature.
So far we have dealt with phenomena concerned with
direct electric current. Even within this group of phe-
nomena it is feasible to devise various systems of con-
cepts and, consequently, various systems of units of mea-
surement. Actually our choice is even more extensive
because electrical phenomena are not at all restricted
to those dealing with direct electric current.
Many physics textbooks, including ones recently pub-
lished, define the concept of the quantity of electric charge
(or the quantity of electricity, which is the same thing)
from Coulomb's law. Then the textbooks treat of voltage
and only later, after finishing the discussion of electro-
statics, do they introduce the concepts of the current
and electrical resistance. As you have seen, we went
about this matter in a different way.
There is even more freedom in selecting the units of
measurement. The investigator has the right to proceed
as he finds convenient. He should never forget, howev-
er, that the selection of the unit.s of measurement affects
the proportionality factors he must use in various for-
mulas.
There is nothing wrong (again, in principle) with se-
lecting the units of current, voltage and resistance inde-
pendently of one another. But this requires introducing
into Ohm'8 law a factor with specified dimensions. Until
quite recently, before the habitually familiar calorie had
been banished forever from physics by a stern interna-
tional committee, the formula of the Joule-Lenz law had
a numerical factor. This was due to the fact that the units
of measurement for current and voltage were defined
quite independently of the selected units for energy (heat
and work).
In the preceding paragraphs, I have written only
Electrons 28
two formulas in the form of proportionalities, all the
rest are equations without numerical coefficients. One
of these two formulas relates the mass of the substance
deposited on the electrode with the quantity of electric-
ity, and the other is Coulomb's law. This was not done
by accident, but because physicists seem to be unwill-
ing, so far and to some extent, to go over to the Inter-
national System of Units, SI, accepted as a law, and con-
tinue to employ the absolute physical system (cgs units)
in their work (less and less, it is true, due to the pressure
exerted on them by the editors of their articles and books).
In the absolute physical system, the quantity K in the
Coulomb formula for the interaction of charges in a va-
cuum is set equal to unity. When this is done, the value
of the so-called "absolute" unit of the quantity of elec-
tricity is predetermined (the charge equals unit.y if two
identical charges, located at unit distance from each
other, interact with unit force).
After measuring the mass in grams, it would be neces-
sary, to remain consistent, to calculate the value of fac-
tor k in the law of electrolysis, indicating how much
matter is deposited on the electrode in the passage of
one absolute unit of charge. But it is of no avail to look
through your textbooks in search of such a value of this
factor. Knowing the obstinate reluctance of engineers
to reject the ampere and coulomb, physicists made use in
the law of electrolysis of a number indicating the mass
of matter deposited when one coulomb of electricity
passes through the electrolyte. Thus, two units were given
in books for the same quantity. It was also clear that it
was convenient to employ one or the other in entirely
different cases, because the coulomb is equal to three
thousand million absolute units.
It is convenient, of course, to have K equal to unity,
but engineers drew attention to the fact that formulas
for the electric flux, capacitance of a capacitor and others
f. Electricity 29
include the unwanted coefficient 41t. They contended that
it would be a good thing to eliminate it.
As is usual in such controversies, victory was cele-
brated by those closer to practice than to theory. The sys-
tem accepted today proceeded along the course that en-
gineering has long followed. Supporters of the SI system
insisted on using a single unit of energy in all branches
of science and also required that the current, or current
intensity, be accepted as the only basic electrical concept.
Thus, we begin the study of electricity with the joule
as the unit of energy. We select the coulomb, equal to
the ampere-second, as the unit of the quantity of electric-
ity. We propose that the ampere be defined from the
force of interaction of currents. This definition (given
on p. 108 in the chapter devoted to electromagnetism)
is devised so that factor k in the electrolysis formula
remains the same one that everybody has long become
accustomed to. Still, we must understand that in the SI
system of units this factor does not determine the value
of the coulomb. With the increase in the accuracy of
measurement, we will be obliged to revise this value so
as to retain the definition of the ampere (I do not think
such a time will come, because I cannot conceive that
the measurement of electrodynamic forces can be more
precise than the measurement of mass).
The 81 system next follows the course I plotted for
our composite investigator. First came the unit of volt-
age, called the volt, equal to one joule divided by one
coulomb; then the unit of resistance, called the ohm,
equal to one volt divided by one ampere; the unit of re-
sistivity, one ohm multiplied by one metre, etc.
Now we have reached Coulomb's law and find that we
?o longer have the right to dispose of factor K. The force
IS measured in newtons, the distance in metres and the
charge in coulombs. Factor K acquires dimensionality
and has a certain value determined experimentally.
Electrons 30
Coulomb's law is rarely required and the equation
for the capacitance of a capacitor is the working formu..
la in many engineering calculations. To eliminate fac-
tor 41t in formulas for the electric flux, capacitance of
a capacitor and others, engineers replaced. factor
K by the expression 1/41C&o a long time ago. For readily
understandable reasons, the quantity Eo can be called
the electrical permittivity of vacuum. It equals
E = 8.85 X 10-12 coulomb
2
o newton-metre"
Hence, the flux of lines of force is expressed by
1
E; (ql +q2 + .)
and the capacitance of a capacitor is written as
C = e6
0
A
d
The unit of capacitance, one farad, equals one cou-
lomb divided by one volt.
Evolution of Electricity Theory
The actual evolution of electricity theory was not at
all along the lines followed by our mythical composite
investigator.
Electrostatic phenomena were well known in ancient
times. It is difficult to say today whether the ancient
Greeks knew what bodies in addition to amber (the Greek
word for amber, elektron, is the source of such words as
electricity, electron, etc.) acquire special properties after
being rubbed and attract small bits of straw. Not until
the beginning of the 17th century did Sir William Gil-
bert (1540-1603) show that this strange property was
possessed by diamond, sealing wax, sulphur, alum and
t. Electricity 31
many other substances. This outstanding English phys-
icist, who was also court physician to Queen Elizabeth I,
was evidently the first to devise instruments that could
be used to observe the interaction of electrified bodies.
By the 18th century it was already known that certain
bodies are capable of retaining charges, while the
charges "leak away" from other bodies. Few scientists of
this century doubted that electricity was something re-
sembling a fluid. The first electrostatic machines were
built. They produced sparks and could "shock" a line
of people holding hands when one of the end persons
touched a terminal of an operating electrostatic machine.
It was considered good taste for the courts of many coun-
tries to visit the laboratory of a scientist as they would
go to a circus. And the scientists, in their turn, made
every effort to dramatize the phenomena being demon-
strated.
In the 18th century we can already speak of electro-
statics as a science. A great many different kinds of elec-
troscopes were invented and made; Coulomb began his
quantitative measurements of the interaction forces be-
tween charges.
In 1773, the Italian physician and physicist Luigi
Galvani (1737-1798) began to investigate the contrac-
tion of the leg muscles of a dead frog when a voltage was
applied across them.
Continuing Galvani's experiments, Volta came to the
Conclusion at the end of the 18th century that an elec-
tric fluid passed through the frog's muscles. The next
remarkable step forward was the invention of the first
Current source, the galvanic cell, and later, the voltaic
pile.
At the very beginning of the 19th century, Volta's dis-
Coveries were known by the whole world of science. This
started off extensive investigations of electric current.
New discoveries followed, one after the other.
Electrons
32
A number of investigators studied the thermal effects
of current. This is what Oersted was engaged in when,
entirely by chance, he discovered the effect of a current
on a magnetic needle.
The brilliant investigations of Ohm and Ampere were
conducted at about the same time: in the twenties of
the nineteenth century_
Ampere's work quickly won him worldwide fame. Ohm
had no such luck. His scientific articles, combining neat
and elegant experiments with precise calculations, dis-
tinguished by their rigorous substantiation and syste-
matic introduction of phenomenological concepts, paying
absolutely no attention to the "nature" of things, were
not appreciated by his contemporaries. When anyone
mentioned Ohm's works, it was only to ridicule the "mor-
bid fantasy of the author, who is trying to belittle the
dignity of nature". (These words belong, evidently, to
the physicist De Ia Rive, who made no contribution what-
soever to science.)
It is extremely difficult to read the original works
of physicists written in those days. Experimental dis-
coveries were described in language alien to us today.
It is even impossible, in many cases, to understand what
the author meant by certain words. The names of famous
scientists are retained in the memory of posterity only
due to the efforts of historians of science.
2. Electrical Structure
of Matter
Minimum Quantity of Electricity
For many years, all the information on electrical phe-
nomena known to physicists consisted in the certainty
that electricity was something like a liquid. The fol-
lowing joke was still enjoying success at the turn of the
century. Wishing to make a laughing-stock of a poorly
prepared student, the examiner says, "Well, since you
can't answer all my other questions, try this simplest
one: What is electricity?"
The student answers, "On my word of honour, I knew,
Professor Jones, but I have forgotten."
At this the examiner exclaims, "What a loss for man-
kind! There was only one person in the whole world who
knew what electricity is, and he has forgottenI"
The first hints that electricity consists of special par-
ticles instead of being a continuous fluid, and the cer-
tainty that these electrical particles are related in some
way to atoms were obtained in studying electrolysis.
In conducting experiments on the dissociation of sub-
stances dissolved in water when a current is passed
through the solution, the English scientist Michael Fa-
raday (1791-1867) found that the same electric current
deposits various amounts of substances on the electrodes
depending on what chemical compound is dissolved
in the water. Faraday discovered that 96 500 coulombs
pass through the electrolyte to deposit one gram-atom
of a monovalent substance, and that twice as much is
required to deposit one gram-atom of a bivalent
substance.
S-2259
Electrons .34
Maybe you think that when Faraday obtained these
results he cried, "Eureka!" and announced that he had
discovered the essential character of electricity? Not at
all. The gifted experimenter permitted himself no such
illusions. Faraday, in any case, with respect to electric
current, behaved like our mythical investigator of the
preceding chapter. He considered -it feasible to make
use of only those concepts that can be characterized by
numerical values.
"How so?" the reader may ask. It has been shown that
6.023 X 10
23
atoms (you recall that this is Avogadro'8
number) carryover 96 500 coulombs of electricity. Con-
sequently, if we divide the second figure by the first,
we obtain the quantity of electricity carried by any mo-
novalent atom. The quotient is 1.6 X 10-
19
coulombs.
This then is the minimum quantity of electricity, or the
"atom of electricity", or the "elementary charge".
But Avogadro's number was not determined until 1870.
.It was only then (just think of it; a mere hundred
years ago) that physicists who like to devise hypothe-
ses (their frame of mind and mentality greatly differen-
tiate them from the investigator who tries to keep with-
in the bounds of the phenomenon being studied) decid-
ed that the following assumption is highly probable.
Along with electrically neutral atoms, particles exist
that carry one or several elementary charges of electric-
lity (positive or negative). Atoms carrying a positive
charge (cations) are deposited on the cathode during elec-
trolysis; those carrying a negative charge (anions) are
deposited on the anode.
Molecules of salts dissolved in water are dissociated
into anions and cations. A molecule of common salt,
sodium chloride, for instance, dissociates into a positive
ion of sodium and a negative ion of chlorine, rather than
into an atom of sodium and an atom of chlorine.
2. Electrical Structure of M....'
Ion Flow
35
Naturally, electrolysis only suggests to the investi-
gator the idea that electrical particles exist.
Many procedures were proposed at the turn of the last
century for converting molecules into charged fragments
(a phenomenon called ionization). Methods were found
for producing directed streams of charged particles and,
finally, procedures were worked out for measuring the
charge and' mass of ions. Physicists gained their first
knowledge of ion flow when they connected a glass tube
with rarified gas into a d-e circuit. At a low voltage across
the electrodes sealed into the glass of the tube, no cur-
rent is observed. But it turned out to be quite simple
to convert the rarified gas into a conductor. The gas can
he ionized by the effects of X-rays, ultraviolet light or
radioactive radiation. We can even manage without such
special measures if we apply a higher voltage to the tube
terminals.
Gas thus becomes a conductor of electric current. 'Ve
can assume that the molecules are broken up into anions
and cations. The anions travel toward the positive elec-
trode and the cations toward the negative one. Animpor-
tantbreakthrough in investigating this phenomenon was
the production of a stream of particles. This is done by
making a hole in one electrode and accelerating the ions
of 8 single sign passing through the hole with an electric
field. By means of a diaphragm, we can obtain a narrow
beam of anions or cations travelling at considerable ve-
locity. If such a beam is directed to a screen of the type
on a 'TV picture tube, we see a bright spot. By passing
the ion beam through two mutually perpendicular electric
fields and varying the voltage over the capacitors setting
up these fields, we can make the spot wander over the
screen.
II.drolt 36
Using a similar device, we can determine one of its
most important parameters: its charge-to-mass ratio.
In an accelerating field, ions gain energy equal to
the work done by the electric forces, i.e.
mrP =eV
2
We know the applied voltage, and the velocity of the
particles can be measured by various essentially differing
ways. \Ve call, for instance, measure the deviation of the
spot of light on the screen. It is clear that the longer
the path of the particle and the lower its initial velocity,
the greater the deviation. This problem can be quite
rigorously solved. It resembles the calculation of the path
of a stone thrown horizontally.
There are also methods for direct measurement of the
time it takes the ion to travel over its whole path.
Hence, we know the voltage and the ion velocity. \Vhat
can we calculate from the results of this experiment?
It follows from the equation that we can find the ratio
of the charge of the particle to its mass. It is a pity,
however, that we cannot separate the charge from the
mass no matter how we change tho conditions of the ex-
periment, or what deviations and accelerations of the
particles we use. Only on the basis of data obtained by
chemists, and the value of the elementary charge obtained
in electrolysis, can we come to the reliable conclusion:
the charges of all monovalent ions are the same, the
charges of bivalent ions are twice as much, those of tri-
valent ions are three times 8S much, etc. The differences
in the charge-to-mass ratios, which can be measured with
exceptional precision, indicate a method for measuring
the mass of the ion.
This is why the instrument, of exceptional significance
in chemistry and chemical technology and based on the
2. Electrical Structure of Mldter 37
Figure 2.t
principle of the simple experiment described above, Is
called the mass spectrograph (see Book 4), though, in
fact, it measures the charge-to-mass ratio of ions.
Electron le8m
We shall not trace the zigzag course of historical events
that led physicists to the unshaken conviction that a
minimum portion of electricity not only exists, but has
a material carrier, which they called the electron. Instead
we shall describe an experiment that is demonstrated
in school lessons in physics.
The apparatus used in this experiment was once called
a cathode, or cathode-ray, tube. Now it is called an
electron-beam tube or an electron gun or an oscillograph.
If your school days are far behind and you had no op-
portunity to become acquainted with this instrument,
no matter, you have seen and see plenty of them now:
an electron-beam tube is the main component of your
television set. On the somewhat flattened end, or screen,
of the tube, the electron beam draws the moving pictures
that we. sometimes view with pleasure and which always
enable us to kill time.
Electrons 38
Let us return, however, to the school experiment. A
diagram of such a tube is illustrated in Ftgure 2,1. The
tube is ideally evacuated; all molecules th.at could dis-
sociate have been pumped out. IIowever, after a current
(called the cathode current) heats the filament to incan-
descence, the filament heats the cathode. Then, when
the cathode and anode are connected to the corresponding
terminals of the voltage source, you can observe a bright
spot on the fluorescent screen. You can now establish by
means of a measuring instrument that an electric current
passes from the anode to the cathode. It is quite naturally
called the anode current.
Since the current passes through a vacuum, you must
come to the conclusion that the incandescent cathode
emits negatively charged particles. This is called ther-
moelectron, or thermionic, emission. Any Incandescent
body has this property.
These particles (we shall not hide from the reader
that they are electrons) travel toward the anodes, which
are designed as cylinders closed at one end and with a
tiny round hole in the centre of the bottom. The electrons
are emitted as a narrow beam, which can be investigated
in the same way as described above for' a beam of
ions.
Convinced by the spot on the luminous screen that the
cathode is emitting electrons, we proceed to find their
charge-to-mass ratio by means of deflector plates. The
results are the following. This ratio for the electron is
1840 times greater than for the Iightest ion, namely,
the hydrogen ion. This leads us to tho conclusion that
the hydrogen ion is 1840 times heavier than an electron.
Consequently, the mass of the electron is 9 X10-
28
grams,
Here our reader may rightly object, saying that we are
in too much of a hurry. On the face of it, you' cannot:
come to the conclusion that the mass of an. electron is
less than that of an ion only by measurmg their charge-
2. Electrica' Structure of Matter 39
to-mass ratio. Maybe the charge of the positively charged
ion and that of the electron are entirely different?
The charge-to-mass ratio of the electron was deter-
mined way back at the end of last century by the brilliant
English physicist Sir Joseph John Thomson (1856-1940).
His friends called him J. J. (This abbreviation, so often
found in memoirs and biographies, is due less to' the
fact that the English like abbr.eviations so much as to
the fact that another famous British physicist of the same
name lived and worked during the 19th century. He
was William Thomson (1824-1907). A peerage was be-
stowed upon him for his scientific merits and he became
Lord Kelvin.) The cathode tube used by J. J. Thomson
was, of course, not as highly perfected as an up-to-date
oscillograph. Thomson was perfectly aware that his mea-
surements only showed that the discreteness of electric
charge was quite feasible and that a minimum portion
of electricity could exist.
Strange as it may seem, even though many physicists
had observed the behaviour of cathode and anode beams,
there were still many supporters of the hypothesis that
these beams were of a wave nature. These investigators
found no necessity for admitting that the currents passing
through a metal conductor, through a Iiquid , and through
a gas or vacuum are near of kin. They insisted on more
direct proofs. We can, of course, understand their point
of view: circumstantial evidence is insufficient to 'con-
vert a hypothesis into a fact.
Hence, it was essential, first of all, to substantiate
the validity of the hypothesis by direct measurements
of the size of the electric charge of particles. These
measurements, far from unsuccessful, were begun by Thorn-
son and his colleagues in the early years of the 20th cen-
tnry. The most precise measurements were made in 1909
by the American .physicist Robert Andrews Millikan
(t$68-1953).
Electrons 40
Millikan's Experiment
The idea of the discreteness of electricity seems quite
bold, and the determination of the elementary charge,
with an account of which we began this chapter, can
be treated in a different way. What is wrong with con-
tending that anions actually exist and that negative elec-
tricity is a liquid entrained by the positive ion? One
ion picks up one amount of this liquid, another ion
picks up another amount, and the experiment indicates
some average value. This is a reasonable explanation,
is it not?
As mentioned above, Thomsons experiment.s were a
powerful, but not decisive, argument in favour of the
existence of the electron. No need then to prove the vital
importance to physics of an experiment that could dem-
onstrate the existence of the elementary charge of electric-
ity so obviously that all doubts were instantly dis-
carded. Such an experiment was devised in 1909 by Ro-
bert Millikan. I shall not mention here the other works
of this famed scientist, but this single investigation was
sufficient to put his name in all physics textbooks.
The pririciple of this ingenious experiment is based
on a simple fact. lust as a glass rod rubbed by a piece
of fur acquires electrical properties, other substances be-
have in a similar way. This is called electri fication by
friction. But, strictly speaking, what reason have we
to think that this property is inherent in only solid
bodies? If we spray tiny droplets of oil into a chamber,
will they become electrified by the friction as they pass
the orifice of an atomizer? We find that they will. This
can be demonstrated by what is, in principle, quite
simple apparatus. We spray a str.eam of fine oil droplets
into the space between two horizontal capacitor plates
enclosed in 8 chamber. Then we arrange a suitable mi-
croscope to enable us to observe the droplets. Until an
2. Electrical Structure of M8tter 4t
electric field is set up, the droplets naturally fall down-
ward by gravity. The droplets are very light and the
force of gravity is almost immediately counterbalanced
by the air resistanee so that they drift downward at uni-
form velocity. As soon as we apply a voltage across the
plates, we observe a definite change in the behaviour
of the droplets. Their motion is either accelerated or
decelerated, depending upon the direction of the electric
field. Millikan chose the direction that slowed down the
droplets. By gradually increasing the intensity of the
field, he could, so to speak, hold the droplet at rest be-
tween the plates. For hours, he observed 8 single oil drop-
let. Varying the field, he could control its motion or stop
it at will.
What can we calculate by means of this experiment?
First we shall discuss the data obtained in observation
before setting up the electric field. Th.e equality of the
force of gravity and the air resistance can be expressed
by the equation
mg=av
The density of the oil can be determined by an indepen-
dent experiment; the diameter of the droplet is measured
by the microscope. After this the mass of the droplet
is readily calculated. The droplet drifts slowly downward
and by engraving scale divisions on the microscope lens
we can measure the velocity v of the droplet with 8 stop-
watch with a sufficient degree of accuracy. Then, substi-
tuting these quantities into the preceding equation, we
obtain the resistance coefficient a.
Next we switch on the field. It proves most convenient
to select a field intensity for 'which the droplet' begins
to rise at uniform velocity. We have thus added a third
force to the two previous ones; it is the force exerted by
an electric field whose intensity E is known (the ratio
of the voltage across the plates to the distance between
Electrons 42
them). Upward motion 'at uniform velocity means that
the three forces counterbalance one another..The condition
for such equilibrium is given by
qE-mg=av'
The new 'Value v' of the velocity- can be measured with
the same microscope and stopwatch, All the quantlties
in the equation are known now except the 'chargeof the
droplet. We next calculate the charge' and' enter the
value in a journal for experimental data of the kind
kept' by 8-11 scrupulous experimenters.
Now we have come to the main idea 01 the oil-drop
experiment. The current in an electrolyte, reasoned Milli-
kan, is carried by ions of different signs. But Ions can
be formed in a gas as well. Air can be ionized by various
procedures. 'Va can, 'for instance, place out whole arrange-
ment near an X-ray tube. Then the X-rays ionize the
air. This was well known -In 1909-. But if' a droplet -is
charged it attracts ions of the opposite sign. As- soon as
all ion becomes attached to a droplet, the charge of the
latter is changed, and, with its charge, its velocity is
also changed. This new velocity is immediately deter-
mined by another measurement.
Observations proved the correctness of the idea. When
the X-ray tube was switched on, every now and then
various droplets would abruptly change their velocity,
Keeping his' eye on a single droplet, the observer' mea-
sured the difference in velocities before and after the X-rays
were switched on. From the formula given above, the'
value of q was readily found.
Have you not guessed yet 'what all this' is being done
for? Think again. If there is such a thing as an elementary
electric charge, then the measured value should equal
it if a single monovalent ion joins the droplet or be a'
multiple of the elementary charge if several ions become
attached to the droplet,
2. Electric.' of Mllfter 43
Conducting his experiment with droplets of oil, water,
mercury and glycerine, and reversing the. charge of the
droplets, Miflikan filled his data journal with hundreds
of values of q, and they all were equal to or a multiple
of a single value, the same that had been found by the
investigators of electrolysis.
When Millikan published his results, even his most
skepttcalopponents were convinced that the electric charge
is found in nature in discrete portions. Strictly
ing; however, Millikan'8 experiments do not directly
prove the existence of the electron as a particle.
But hypotheses foreshadow facts. Some scientists were
sure that electricity is of a granular nature as far back
as the beginning of the 19th century. The charge of the
ion was first calculated in 1881 by the Irish physicist
George Johnstone Stoney (1826-1911) and ten years later
he first proposed the term electron, not for the particle,
but for the charge of a monovalent negative ion,
J. J e. Thomson '8 experiments compelled the great ma-
jority of physicists to believe in the existence of the
electron as a particle. The German physicist Paul Karl
Ludwig Drude (186;3-1906)' was the first to unambiguously
define the electron as a particle carrying an elementary
charge of negative electricity.
Thus the electron had been generally recognized before
it was "seen"..
Direct proof of the existence of electrons was obtained
later by precise experiments. A weak beam of these
particles was directed onto a screen where they can be
counted one by one. Each electron causes a flash on a
luminescent screen. Luminescent screens have long "been
replaced for this purpose by special counters named after
their inventor, the German physicist Hans Wilhelm Gei-
ger (1882-1945). In a word, the idea of such a counte.r
is that a single electron, like the trigger of a revolver,
initiates a strong current pulse, which can be readily
Electrons 44
registered. This enables the physicist to establish the
number of electrons entering a trap per second. If this
trap is a metallic bulb into which the electrons fly,
the bulb is gradually charged with a quantity of electric-
ity large enough to be precisely measured. To find the
charge of the electron it is sufficient to divide the quantity
of electricity by the number of captured electrons.
Only after this can we contend that the existence of
the electron is no longer 8 hypothesis; it is a fact.
We have reviewed at racing-car speed the discoveries
that laid the foundations of modern physics. Such, how-
ever, is their fate. New matters crowd out the old, and
even cardinal events, occurring during the construction
of the cathedral of Science, become only material for
historians.
Now, perhaps, we can answer the question: What is
electricity? An electric fluid is a current of electrical
particles. A body is electrically charged when the number
of particles of one sign exceeds that of the other
sign.
"Well, what a feeble explanation," retorts our indig-
nant reader. "And what then is an electrical particle?"
"Isn't it perfectly clear? A particle is said to be elec-
trical if it interacts according to Coulomb '8 law."
"And is that all?" asks the puzzled reader.
"All that concerns the answer to your question," an-
swers the physicist. "But ahead of us are answers to many
other interesting questions. We have not yet mentioned
the cases in which we shall find elementary particles
of positive electricity. We shall also find out that elec-
trical particles are characterized by other properties be-
sides their charge and mass."
First, however, we shall discuss the structure of the
atom.
2. Electrica. Structure 01 Mder 45
Model of the Atom
How is an atom built up of electrical particles? The
answer was obtained by means of rays emitted by radium.
We shall discuss this wonderful substance and the ex-
tensive family of natural and artificial radioactive' ele-
ments in Book 4 of this series. For the time being it is
sufficient for us to know that radium constantly emits
penetrating electromagnetic radiation (gamma rays), 8
stream of electrons (formerly called beta rays) and alpha
rays, which consist of doubly charged ions of helium
atoms.
In 1911, the eminent New Zealand-born physicist Sir
Ernest Rutherford (1871-1937) proposed the so-called plan-
etary model of the atom, based on his careful investi-
gations of the scattering of alpha particles by various
substances. Rutherford conducted his experiments with
thin gold leaf, only a tenth of a micron in thickness.
He found that out of 10 000 alpha particles only one was
deflected through an angle exceeding 10.
In these strikingly simple experiments, the passage
of each separate particle was recorded. Up-to-date tech-
niques, of course, enable such measurements to be made
entirely automatically.
It becomes clear then that the atom consists m.ainly
of empty spacel The rare head-on collisions should be
interpreted as follows: inside the atom there must be
a positively charged nucleus, This nucleus is surrounded
by electrons. They are extremely light and therefore
present no appreciable obstacles to the alpha particles.
Th.e electrons slow down the alpha particles, but a colli-
sion with each separate electron cannot deflect the par-
ticle from its path.
Rutherford assumed that tho forces of interaction be-
tween the atomic nucleus and alpha particle, which are
of like charge, are Coulomb forces. .Assuming further
Electrons 46
that the mass of the atom is concentrated in its nucleus,
he calculated the probability of the particle being de-
flected through the given angle and obtained splendid
coincidence of theory with experimental data.
That is how physicists cheek the models they ha-ve
devised.
"So the model predicts the experimental results?"
"Yes. "
"Which means that it represents reality?"
"Why so categorical? A model explains a series of
phenomena and so is a good one. Its refinement. iS8
matter for the future."
The results of Rutherford's experiments swept aside
all doubts as to the valtdity of the following 'statement:
electrons are subject to Coulomb forces and travel about
the nucleus.
The theory also led to certain quantitative estimates
which were later confirmed. The size of the very smallest
atomic nuclei turned out to be about 10-
13
em while the
size of the whole atom is of the order of 10-
8
em.
Comparing the experimental results with the calcula-
tions, physicists were able to estimate the charges :of
colliding nuclei" These estimates played a vital, ,if .not
the main, role in interpreting the periodic law of .the
struct ure of the elements.
Thus, we have built the model of the atom. But the
following question is immediately posed: Why dont the
electrons (negatively charged particles) fall Into the
nucleus (positively charged)? Why is -the atom stable?
"There is nothing strange about this," reasons the
reader. "The planets do not fallon the SUD. A force of
electrical origin, like the gravitational force, is centri.pe-
tal and provides for circular motion of the electrons
about the nucleus."
But the point is that the analogy between the plane-
tary system and an atom is only superficial. As we shall
2. Electrical Structureof Matter 47
find out further on, an atom must emit, from the point
of view of the general laws of electromagnetic fields,
electromagnetic radiation. But this is something we
can ascertain without a knowledge of the theory of elec-
tromagnetism. Matter, .i.e.atoms, is capable of radiating
light and heat. Since this is so, the atom loses energy
and the electron must finally fall into the nucleus.
What is the way out of this dilemma? It is quite simple.
We must reconcile ourselves to the facts and elevate
them to the rank of a law of nature. This step was taken
in 1913 by one of the greatest physicists of our century,
Niels Henrik David Bohr (1885-1962).
Qu.ntlzing Energy
Like all first steps, this step was relatively timid.
We shall now present the new law of nature that not
only saved Rutherford's atom, but forced us to the con-
clusion that the mechanics of large bodies is inapplicable
to particles of small mass.
Nature is arranged so that a number of rnechanical
quantities, such, for instance, as angular momentum or
energy, cannot have a. continuous series of values for
any system of interacting' particles. On the contrary, the
atom, now being discussed, or the atomic nucleus, to be
discussed later, has its own sequence of energy levels
inherent. in only the given system. There is the lowest
level (ground state). The energy of the system cannot be
less than.. this value. In the case of the atom this means
that there is a state in which the electron is at a certain
minimum distance from the nucleus.
Changes in the energy of an .atom can occur only in
jumps. If the jump is "upward"; it means that the atom
absorbed energy; if it is "downward", the atom radiated
energy.
We shall see further on how nicely, on this basis, the
Electrons 48
radiation spectra of various systems can be deciphered.
):'The formulated law is called the law of the quantiza-
tion oj energy. We can also say that energy is of a quan-
tum nature.
It should be noted that the law of quantizing is of an
entirely general nature. It is applicable not only to the
atom but to any item consisting of thousands of millions
of atoms. When we deal with large bodies, however, we
may not "notice" the quantizing of energy. The point
is that for an item consisting of a million million million
.atoms the number of energy levels are increased, roughly
speaking, by a million million million times. These ener-
gy levels are so close to one another that they practically
merge into a single band. Consequently, we cannot ob-
serve the discreteness of the energy values. Thus, the
mechanics we discussed in Book 1 remains practically
unchanged when we deal with large bodies.
In Book 2 we found that allergy can be transferred from
one body to another in the forms of work and heat. Now
we are in a position to explain the difference between
these two forms of energy transfer. III mechanical action
(for example, in compression), the energy levels of the
system are displaced. This displacement is very slight
and can be detected only by precise experiments if the
pressure is high enough. As to heat action, it consists
in converting a system with a lower energy level to a
higher one (in heating) or one with a higher energy level
to a lowor one (in cooling).
The quantizing of energy, like that of other mechanical
properties, is 8 general law of nature. A great variety
of consequences that follow strictly from this law can
be confirmed experimentally.
Maybe you wonder why energy is quantized? There is
no answer. Nature just happens to be made that wayl
Any explanation is a reduction of a particular fact to
a more general one. We know today of no statement
2. Electrical Structure of Matter 49
sufficiently general for the quantizing of energy to follow
from it as a consequence. This does not imply, of course,
that at some future date, in principle at least, sufficiently
broad and comprehensive laws will be discovered for the
principles of quantum mechanics to be their corollaries.
However that may be, the law of quantizing is today
one of the few great laws of nature that require no logical
substantiation. Energy is quantized because-it is quan-
tized.
This law was established in such a general form during
the years 1925-1927 in the works of the French physicist
Louis Victor Pierre Raymond, Prince de Broglie (b. 1892)
and the Austrian and German physicists Erwin Schro-
dinger (1887-1961) and Werner Karl Heisenberg (1901-
1976). The branch of physics based on the quantizing
principle (I forgot to mention that the word quantum
means a specified quantity or portion) was named quan-
tum, or wave, mechanics. But why wave? This will be
explained later on.
Mendeleev's Periodic Law
In 1869, the famed Russian chemist Dmitri Ivanovich
Mendeleev (1834-1907) published the periodic law he
discovered in the order of the chemical elements. We
shall not give Mendeleev's periodic table here; it can
be found in any school chemistry textbook. We need
recall here only that when he arranged the elements in
the order of their atomic weights Mendeleev noted that
their chemical properties and certain physical features
vary periodically in accordance with their atomic weight.
In the table devised by Mendeleev, each element be-
longs to one of nine groups and to one of seven periods.
He arranged the elements of the groups in columns so
that the elements whose symbols are arranged one below
the other possess the same chemical properties. He found
4-2259
Electrons 50
that this could be done only if he assumed that certain
as yet undiscovered elements existed. He left empty spaces
(gaps) for them in his table. The exceptional foresight
of this great scientist also led him to put the nickel
atom in its "proper" place following cobalt, notwith-
standing the fact that the atomic weight of cobalt is
somewhat higher.
Some of the gaps were filled during Mendeleev's life.
This brought him world renown because it became clear
that the compilation of this table was the discovery of
a great law of nature rather than a mere formality.
The essence of the atomic number assigned by the
table to each chemical element became evident after the
physicists had no more doubts about the validity of
Rutherford's planetary model of the atom and of the
quantizing of energy. What does this number mean? The
answer turned out to be extremely simple. The number
of an element in the periodic table, its atomic number,
is equal to the number. of electrons rotating about the
nucleus. We can say that the atomic number of an ele-
ment is the positive charge of its nucleus expressed in
units of charge of its electrons.
Mendeleev's periodic law, the principle of energy quan-
tization and a study of the characteristic optical and
X-ra.y spectra of atoms (to be discussed further on)
cleared up the reason for the same chemical behaviour of
atoms in anyone column of Mendeleev's table.
The energy of an atom is the energy of interaction
between the electrons and the nucleus. Since energy is
quantized, it is logical to assume that the electrons of
each atom can be arranged in a series according to their
energies. The first electron is most strongly bonded to
the nucleus, the second, more weakly, and the third,
still more weakly, etc. Thus the electrons of an atom are
arranged in energy steps. We have not been misled by
our logical reasoning, but investigations have more ac-
2. Electrica' Structure of Matter 5t
curately defined this concept. In the first place, each
energy step can be occupied by two rather than a single
electron. True, these electrons are identical; they differ
from each other in a property known as spin; This is
a vector quantity. Hence, people who prefer a more
visual concept can imagine that a filled step has two
"tiny dots" with arrows, one arrow pointing "downward"
and the other "upward".
The word "spin" is of the following origin. One of its
definitions is to rotate swiftly, like a top. To explain
the difference between two electrons occupying the same
step, one was to imagine one electron as rotating clock-
wise and the other counterclockwise about their axes.
This modal was suggested by the superficial resemblance
between the atom and the planetur y system. If an electron
is something like a planet, then why not allow it to rotate
about its axis? Again I must distress my readers: it is
quite impossible to visualize the spin of an electron.
But it can be measured, as we shall see in the next chapter.
This, however, is not the only significant conclusion
that we can reach (by a careful examination of atomic
spectra). A second conclusion is that the energy steps
can be at unequal distances from one another and can
be divided into groups.
Following the first step, called the K-Ievel, is an energy
gap. This is followed by a group of eight electrons, de-
noted by the letter L, and then a group of 18 electrons,
denoted by the letter M, etc. We shall not describe here
the locations of the levels and the order in which they are
filled for all atoms. It is not a simple picture and its
description would require much space. Details play no
appreciable role in our small book. I mentioned the ener-
gy steps only to explain the similarity of atoms located
in a vertical column of Mendeleev's table. It was found
that they have an equal number of electrons in the upper
group of steps.
Electrons 52
This clears up the chemical concept of the valency of
an atom. Thus, lithium, sodium, potassium, rubidium,
cesium and francium have a Ringle electron in the upper
group of steps. Beryllium, magnesiurn, calcium, etc. have
two electrons in the upper group of steps. The valence
electrons are the most weakly bonded in the atom. There-
fore, in ionizing atoms of the first column, singly charged
particles are most easily formed. Ions of beryllium,
magnesium and others usually carry two charges, etc.
Electrical Structure of Molecules
Chemists call the molecule the tiniest representative
of a substance. Most physicists use this word only when
this tiniest representative really exists as a separate
small body.
Do molecules of common salt exist? "Of course!" an-
swers a chemist and writes the formula NaCI. Common
salt is sodium chloride. A molecule consists of one atom
of sodium and one of chlorine. This answer, however,
is only formally valid. Actually, we cannot find pairs
of these atoms behaving as a single whole neither in crys-
tals of common salt, nor in a solution of salt in water,
nor in the vapour of sodium chloride. As we mentioned
in Book 2, each atom of sodium ill a crystal is surrounded
by six chlorine neighbours. All of these neighbours pos-
sess equal rights and there is no way to find which one
"belongs" to the sodium atom.
Let us dissolve common salt in water. We find that the
solution is an excellent conductor of electric current.
Rigorous experiments, discussed above, can be conducted
to demonstrate that the electric current is a stream of
negatively charged atoms of chlorine travelling in one
direction and a stream of positively charged atoms of
sodium travelling ill the opposite direction. Hence, in
2. Electrical Structure of Matter 53
a solution, atoms of chlorine and sodium do not form
strongly bonded pairs of atoms.
After we have established the model of the atom, it
becomes clear that an anion of chlorine is an atom of
chlorine with an "extra" electron. A cation of sodium,
on the contrary, has a "shortage" of one electron.
Can this lead us to the conclusion that a solid body is
also built up of ions rather than atoms? Yes, it can.
This can be proved by many experiments on whose
description we shall not dwel! here.
What about the vapour of sodium chloride? Here again
we find no molecules. The vapour of sodium chloride
consists of ions or of various extremely unstable groups
of ions. We can speak of the molecules of ionic compounds
only in the chemical meaning of the word.
Ionic compounds all dissolve in water. Such solutions,
classical examples being simple metal salts, such as sodi-
um chloride, have good conductivity and are therefore
called strong electrolytes.
We can give several examples of substances built up
of genuine molecules, molecules in the physical meaning
of the word. They are oxygen, nitrogen, carbon dioxide,
hydrocarbons, carbohydrates, steroids, vitamins, etc. This
list could be continued to great length.
All classifications are always somewhat arbitrary. There-
fore, I must warn the reader that sometimes we find
cases in which the substance consists of physical mole-
cules when it is in one state of aggregation, but does not
when it is in other states. These substances include such
vital ones as water. Water vapour molecules are undoubt-
edly separate bodies. But it is a difficult matter to
"outline" a single molecule in ice crystals and to contend
that these atoms of hydrogen are bonded only to this
particular atom of oxygen.
Be that as it may, the class of molecular crystals is
vastly extensive.
Electrons
In Book 2 we have already discussed the structure of
molecular crystals. Recall that in a crystal of carbon
dioxide' (C0
2
) the carbon atom has two very close oxygen
neighbours. In all other cases, in investigating the struc-
ture of a molecular crystal, we can readily see that it
is possible to break down the crystal into intimately
arranged groups uf atoms.
If they are so closely located, they must be bonded by
high forces. This is true. The forces bonding atoms belong-
ing to a single molecule are, roughly speaking, a hundred
or even a thousand times greater than the forces Acting
between atoms of neighbouring molecules.
lust what is this intramolecular bond? It is sufficiently
cleat' that we shall not be able to manage by using only
concepts of the mutual attraction of electrically charged
negative and positive ions. We know that the molecules
of oxygen, nitrogen and hydrogen consist of identical
atoms. It is impossible to assume that one atom loses
and the other atom gains an electron. Why on earth
should an electron prefer to stay near either one of two
identical atoms?
The principle of the intramolecular bond was grasped
only with the evolution of quantum mechanics. We have
just mentioned that the energy of any system is quan-
tized and that two electrons of opposite spin can occupy
a single energy level. In addition, an interesting conse-
quence follows from one of the basic hypotheses of quan-
tum mechanics. It was found (and this is no hypothesis,
but a strict mathematical derivation which we do not
give because of its complexity) that the lowest energy
value assumed by an electron depends upon the size of
the region within which it travels. The greater the size,
the lower the energy of this "zero level".
Now imagine that two atoms of hydrogen approach
each other. If they merge into a single system, the "apart-
ment" for each electron is approximately doubled. Two
2. Electrica. Structure of M....r 55
electrons with opposite spin can peacefully coexist in
a single "apartment". Consequently, such life together
is expedient. The region in which the electrons exist
has appreciably increased. This means that the total
energy of the system has decreased after the two atoms
merge together. By now we have completely. mastered
the fact that any system, if there are any such possibili-
ties, tends to go over to the state with minimum energy.
For the same reason, a balJ, left to itself, rolls down-
hill.
Thus, the formation of a chemical bond implies the
collectivization of the electrons. There is a certain num-
ber of electrons (said to be inner-shell electrons) that
rotate about the nuclei of the atoms, but some (outer-
shell) electrons include at least a pair of nearest atoms
in their motion or even may travel among all the atoms
of the molecule.
We can recognize a substance built of molecules by
its electrical properties. A solution of such a substance
conducts no current. The molecules do not break down
into parts and a whole molecule is electrically neutral.
In liquids and vapours, the molecules retain their struc-
ture; the whole group of atoms travels as a single whole,
with translatory or rotary motion. Atoms belonging to
such molecules can only vibrate about their equilibrium
positions.
A neutral molecule carries no electric charge. But do
not hasten to the conclusion that such a molecule does
not set up an electric field. If the molecule is asymmetric,
the centres of gravity of its positive and negative charges
almost certainly do not coincide. It is intuitively clear
that the centres of gravity of the two charges coincide
in such molecules as oxygen and nitrogen, consisting of
two identical atoms. Nor is it difficult to understan.d
that in a molecule of a gas such as carbon monoxide
(CO) these centres may be displaced with respect to each
Electrons 56
other. If there is such a displacement, the molecule is
said to have a dipole moment.
This term has the following origin: a dipole molecule
behaves like a system of two point charges (one point is
the centre of gravity of the negative charges and the
other is the same for the positive charges). A dipole is
specified by the size of the charge and the dipole arm,
i.e. the distance between the centres of gravity.
Do not require proof that an asymmetric molecule has
an electric dipole moment. We can dispense with a the-
oretical discussion because the existence of a permanent
(or, as they also say, rigid) dipole moment can be readily
demonstrated by experiments.
Dieledrics
We can place equality signs between a dielectric, a
nonconductor of current and an insulator.
Dielectrics include molecular gases, molecular liquids
and solutions of solids made up of molecules. Solid dielec-
trics include glass, both organic and nonorganic (sil-
icate, borate, etc.): polymer substances, built up of
giant molecules; plastics, molecular crystals, as well as
ionic crystals.
We reminded the reader in Chapter 1 that the capac-
itance of a capacitor increases when we insert any di-
electric into the space between the plates. Imagine that
the capacitor was connected to a source of constant volt-
age. The capacitance increased even though the voltage
remained the same. This means that an additional charge
approached the capacitor plates. It would seem that the
field intensity must also increase. It has not changed,
however, because it is the quotient of the voltage divided
by the distance between the plates. What is the way
out of our quandary? The only way is to assume that
2. Electrical Structure of MaHer 57
an electric field of the opposite direction is set up in the
insulator. This is called the polarization of the dielectric.
What are the peculiar charges that are produced inside
a dielectric? How can we interpret the failure of attempts
to draw the charge of a dielectric off to the earth? Even
with no knowledge of the electrical structure of matter we
can say that these charges are "bound" instead of being'
free as in a metal. But when we have at our disposal
information on the structure of the molecule, we are
capable of comprehensively explaining polarization phe-
nomena and the mechanism of formation of the counter
field. All other things being equal, the higher the per-
mittivity, or dielectric constant, E, the stronger this
field.
Before going any farther, we must answer the question:
What can an electric field do to an atom or a molecule?
An electric field can shift the electrons of a neu traI
atom or ion in the direction opposite to that of the field.
The atom or ion is thereby converted into a dipole and
sets up a field in the opposite direction. Thus, the polar-
ization of a substance results from the polarization of
the atoms, ions or molecules of which it is made.
The polarization mechanism just described is called
the process of producing induced dipoles. If there is no
field, there are no dipoles. The stronger the field, the
greater the displacement of the centre of gravity of the
electrons, the higher the induced dipole moment, and
the more the polarization.
The formation of induced dipoles cannot depend upon
the temperature. Experiments indicate that there are
dielectrics which are not influenced by temperature.
Consequently, the above-described mechanism is valid for
them.
Well, and what can we do about the cases in which
the permittivity distinctly depends upon the tempera-
ture? A careful investigation of the relationship between
Electrons 58
the structure of molecules and the behaviour of the given
substance in an electric field, as well as the nature of
the temperature dependence of e (polarization always
drops with a rise in temperature) lead us to the following
conclusion. If molecules have a dipole moment even in
the absence of a field (permanent dipole) and can change
their orientation, then this explains the temperature de-
pendence of permittivity.
As a matter of fact, the molecules are oriented haphaz-
ardly when there is no field. The dipole moments are
added vectorially. Hence, the resultant moment for a
volume containing many molecules equals zero. An elec-
tric field seems to "comb" the molecules, making them
face mainly in one direction. They are subject to two
antagonistic forces: thermal motion, introducing disorder
in the arrangement of the molecules, and the ordering
effect of the field. It can be readily understood that the
higher the temperature, the more difficult it is for the
field to "handle" the molecules. From this it follows
that the permittivity of such' substances drops with an
increase in temperature.
These ideas can be more easily visualized and remem-
bered by studying the drawings in Figure 2.2. The upper
drawing shows that the polarization of an atom consists
in the displacement and deformation of its electron shells.
The farther an electron from the atomic nucleus, the
greater the effect of the field on this electron. The layers
represented in these schematic drawings by dots show
where the electrons are located. It must be borne in mind
that the drawings are extremely rough approximations
because various electrons have differently shaped regions
in which they exist in molecules (cf. p. t22).
The middle drawing illustrates the behaviour of 8
symmetric diatomic molecule. In the absence of a field
it has no moment. The field induces an electric moment.
This moment may vary in magnitude in accordance with
2. Electrical Structure of Matter
Figure 2.2
59
the angle of the molecule with respect to the field. The
moment is formed due to the deformation of the electron
shells.
Finally; the lower drawing shows the behaviour of a
molecule which has a dipole moment even when there is
no field. In our case, the molecule was simply turned
when a field was applied. In the general case, however,
both mechanisms of polarization are found in substances
whose molecules have a dipole moment in the absence
of a field: in addition to rotation of the molecule its
electrons may also be displaced. These two effects can
Electrons 60
be readily distinguished by making measurements at
very low temperatures where the thermal motion has
practically no influence on the molecules.
If our model is valid, we should observe no tempera-
ture dependence in the permittivity of substances having
symmetric molecules, for example, molecules of oxygen
or chlorine. On the other hand, if the diatomic molecule
consists of two differeut atoms, for instance, a molecule
of carbon monoxide (CO), the permittivity E should be
temperature-dependent. This is actually the case. Nitro-
benzene is one of the substances whose molecules have a
very high dipole moment.
What will happen to ordinary dielectrics when the
strength E of the applied electric field is increased?
Evidently, the polarization of the substance is increased
as well. This takes place due to stretching of the dipoles:
in an atom this is a displacement of the electron cloud
with respect to the nucleus; in 8. molecule it may be due
to the pulling apart of two ions. This naturally poses
the question: IIp until what point does an electron,
pulled far away from the nucleus by the applied field,
remain an electron of the given atom, or 'do two ions,
pulled sufficiently far apart, still form a molecule? Such
a limit doubtlessly exists, and, at a sufficiently high field
strength, or intensity, E, the so-called breakdown of
the dielectric occurs. Th.e order of magnitude of this
field intensity is several thousand kilovolts per metre.
In any case, such breakdown implies the release of elec-
trons or ions, i.e. the production of free current carriers.
The dielectric ceases to be a dielectric; it conducts elec-
tric current.
We most frequently find such a breakdown when a
capacitor is disabled in a television or radio set. We can,
however, observe another kind of breakdown: an electric
discharge in a gas. Electric discharges in gases are to be
specially discussed later. For the present, we shall become
2. Electrical Structure of Matter 61
Figure 2.3
acquainted with two distinguished members of the fam-
ily of dielectrics: piezoelectric and ferroelectric crystals.
Quartz is the chief representative of the class of
piezoelectric crystals. Members of this class (which in-
cludes, besides quartz, such substances as sugar and
tourmaline) must have a definite symmetry. Shown in
Figure 2.3 is a crystal of quartz. The principal axis of
this crystal is a threefold symmetry axis. Three twofold
symmetry axes lie in a perpendicular plane.
A plate about 2 em thick is cut out of this crystal as
shown in the drawing. We see that it is perpendicular
to the principal axis and that the twofold symmetry axes
lie in a plane of this plate. Then a thin wafer, about 0.5
mm thick, is cut out of the thick plate in a direction
perpendicular to one of the twofold symmetry axes.
Interesting experiments can be conducted with the thin
piezoelectric wafer obtained in this manner (it is shown
displaced downward in the drawing in the middle of
Figure 2.3).
Let us compress the wafer in the direction of arrow A,
perpendicular to the symmetry axes, and connect an
electrometer (an instrument used to detect electric charges)
Eledrons 62
to the side surfaces of the wafer. These surfaces, ac-
tually faces of the wafer, must be silver-plated to en-
sure proper electric contact. We shall find that compression
has induced unlike charges on the faces of wafer. If
tension is applied instead of compression, the signs of
the charges are reversed: where a positive charge ap-
peared in compression, a negative charge appears in tension
and vice versa. This phenomenon-the induction of elec-
tric charges by applying pressure or tension-is called
piezoelectricity.
Piezoelectric quartz crystal devices are exceptionally sen-
sitive: electric devices enable us to measure charges in-
duced on the quartz by extremely low forces that could not
be measured in any other way. Apiezoelectric quartz crys-
tal is also capable of detecting extraordinarily rapid
variations in pressure, a feature unattainable with other
instruments. Consequently, the effect. described above
is of immensely practical significance as a method of
electrically detecting all kinds of mechanical action,
including sounds. It is sufficient to blow softly at a pi-
ezoelectric quartz wafer, and the electric indicating In-
strument instantly responds.
Piezoelectric quartz wafers are used in medicine to
listen to murmurs of the heart. They are also employed
in a similar manner in engineering to test the operation of
machinery in which they can detect all "suspicious" noises.
Quartz is used as a source of the piezoelectric effect in
the tone arms (pickups) of record players. The motion
of the needle in the groove of the record leads to com-
pression of the piezoelectric crystal, which, in turn, pro-
duces the electric signal. The electric current is amplified
and is fed to a dynamic loud-speaker to be converted
into sound.
So far we have discussed substances that are electri-
cally polarized by an electric field and (in certain cases)
by mechanical deformation. When the external effect is
2. Electrlca' Structure of Matter
Figure 2.4
63
removed, the substance becomes electrically neutral
again. But, along with this widespread behaviour, we
find certain bodies possessing a total electric moment in
the absence of external forces. We obviously cannot find
such substances among liquids and gases because thermal
motion, which is not withstood by the ordering effect
of a field, must inevitably lead to disorder in the arrange...
ment of the dipole molecules. Crystals, however, can
be conceived of in which the atoms are arranged so that
the centres of gravity of the anions and cations within
each unit cell are identically displaced. Then all the di-
pole moments face the same direction. Such substances
could be expected to possess maximum possible polar-
ization and, consequently, a permittivity of huge value.
Such crystals do exist. The phenomenon was first
discovered using Rochelle salt and this class of sub-
stances are called [erroelectrics (or ferroelectric crystals).
Of greatest practical value among the ferroelectrics
is barium titanate. Using it as an example, we shall dis-
cuss the exceptionally singular behaviour of this class
of substances.
The unit cell of a barium titanate crystal is illustrated
in Figure 2.4. We have chosen the barium atoms for the
corners of the cell. The small light-coloured spheres are
Electrons
~
+ + --
y y y'"
I I I
t:J::f:f
1 f
l
I
, I I
~ ' . ~ ~ ~
-- ++
~
Figure 2.5
anions of oxygen, and the large sphere in the centre is
a cation of titanium.
The drawing looks as if the cell were a cubic one. A
strictly cubic cell of the substance actually does exist,
but only at temperatures above 120C. Obviously, a
strictly cubic cell is symmetrical and can have no dipole
moment. Consequently, the special properties of barium
titanate disappear above this temperature, which is
called the Curie point. Above this temperature the sub-
stance behaves as an ordinary dielectric.
When the temperature is lowered below 120C, the
ions of oxygen and titanium are displaced in opposite
directions by an amount of the order of 0.1 angstrom.
At this, the cell acquires a dipole moment.
Note the following especially important fact. This
displacement can with equal success take place in any
of three directions: along the three axes of the cube.
Displacement leads to distortion of the cell. Hence,
not all ways of dividing up the crystal regions within
which all the dipole moments are arranged in a single
direction turn out to be equally expedient.
Feasible ways of dividing a crystal into ideally polar-
ized regions (called domains) are shown in Figure 2.5.
2. Electrica' Structure of M....r 65
Along with the case in which the whole crystal is a single
domain-when the maximum electric field is obtained-
there may be less expedient versions and even one (at the
extreme right) in which the external field equals zero.
How does a ferroelectric behave when an external field
is applied to it? It was found that the polarization mech-
anism consists in the growth of a domain that faces
the "required" diraction. by spreading its boundaries.
Domains oriented with their dipole moment at an acute
angle to the field "devour" domains oriented at an o.b-
tuse angle. When' located in very strong fields, inversion
of the: domains may also be observed.
Barium titanate is the main industrial ferroelectric.
It is obtained by the firing of two powders: titanium
dioxide and barium carbonate, The result is a kind of
ceramic material.
Ceramic ferroelectrics are extensively applied in elec'"
trieal and radio engineering. Their main feature is that
they drastically increase the permittivity of capacitors.
Moreover, as is clear from the described polarization
mechanism, the value of E increases with the intensity
of the electric field. A capacitor is thus converted into a
varactor, or variable capacitor, by means of which fre-
quency modulation can be most easily accomplished.
This is a process that takes place in any radio or TV
set.
For many purposes, ferroelectric ceramics have re-
placed quartz. They can be used to produce more power-
ful sounds. In the same way, the gain is higher for ul-
trasounds as well. The only field in which quartz has DO
competitors is the' stabilization of radio frequencies.
The great majority of textbooks begin their chapters on
electricity by describing electric charges produced on
glass or hard rubber rods by rubbing them. The explana-
tion of this phenomenon is usually evaded. Why?
First we should emphasize the fact that the electri-
5-2259
Electrons 66
fieation of dielectrics by friction is not related (at least,
directly) to the polarization of insulators that we have
just discussed. As a matter of fact, polarization consists
in the formation of bound electric charges whose special
feature is that they cannot be "drawn off" the dielectric.
The charges produced on the glass or hard rubber by
rubbing them with cat fur are obviously free charges
.and, of course, that means electrons.
The general picture of what takes place is more or
less clear, but not entirely. Apparently, the scanty
amount of free electrons possessed by the insulator are
bound to its molecules by forces of various magnitudes
for various dielectrics. Hence, if two bodies are held
in tight contact, electrons may pass over from one to
the other. At this, electrification occurs. But tight con-
tact here means to bring the surfaces within distances
equal to interatomic ones. Since no atomically smooth
surfaces exist in nature, rubbing helps to eliminate all
kinds of projections and increases the area, so to say,
of trne contact.
Transfer of electrons from one body to another may
occur for any pair of bodies, whether they are metals,
semiconductors or insulators. But only insulators (or
nonconductors) can be electrified because only in such
bodies do the charges remain at the places to which they
were rubbed off the other body.
I cannot express especially full satisfaction with this
theory. It does not explain the advantages of using hard
rubber, glass and cat fur for this purpose. We can pose
heaps of more questions that have no convincing answers.
Conduction in Gases
If we fill a glass tube with a gas, seal electrodes into
the tube and apply a voltage across them, we obtain
a simple device for studying the conduction of electric-
2. Electrical Structure of Matter
67
Figure 2.6
ity in gases.. We can change the substance through
which the current passes, we can vary the pressure of
the gas and the applied voltage.
Investigations of the conduction in gases played an
immense role in developing OUf concepts of the electrical
structure of matter. The main part of this research
was conducted in the nineteenth century.
Tubes of various shapes, used to study the phenomena
we are discussing, are illustrated in Figure 2.6. Since
all ancient statues and paintings of the old masters have
long since been 'purchased and repurchased, idealers in
antiques are now offering connoisseurs obsolete laboratory
equipment. In curiosity shops in Western Europe and
America, you can buy (usually at a .good price) one of
the rare items shown in Figure 2.6.
An electric current is initiated in a gas because the
neutral molecules are broken down into anions and ca-
tions. Moreover, an electron may break away from mole-
cules and atoms. The current is produced by a beam of
positive ions and beams of negative ions and electrons
travelling in the opposite direction..
Electrons 68
To make a gas conduct a current it is necessary to con-
vert the neutral molecules or atoms into charged particles.
This may be accomplished either by an external ionizer,
or by collision of the particles. As mentioned above,
external means of. ionization may be ultraviolet, X-,
cosmic or radioactive rays. High temperature also leads
to the ionization of a gas.
The passage of a current through a gas is often ac-
companied by some light effect. The character of the glow
varies with the gas used, its pressure and the voltage
applied. Investigations of this glow played a vital
role in the history of physics, particularly as a source of
information on the energy levels of the atoms and the
laws of electromagnetic radiation.
Conduction in a gas does not obey Ohm's law. It is
characterized by a curve showing the dependence of
the current on the voltage. This curve is called the
current vs voltage characteristic (not only for gases,
but for any conducting systems that do not conform to
Ohm's law).
Let us consider the phenomena, typical of any gas,
that occur when voltage applied to a gas-discharge tube
is increased. This behaviour of the gas, now to be de-
scribed, is found in a wide pressure range. We shall except
only such low pressures for which the free path of the
molecules becomes commensurate with the dimensions
of the gas-discharge tube. Our discussion will also not
refer to such high pressures at which the density of gases
approaches that of liquids.
We begin by applying a low voltage across the tube.
If we do not use some ionizer, there is no current through
the tube. If an ionizer is available, it produces charged
particles-ions and electrons-in the gas. As soon as a
field is set up in the tube, the particles are directed by
the field to the electrodes of the tube. The velocity at
which the charged particles travel toward the elec-
2. Electrical Structure of M8fter
69
trodes depends upon many circumstances and primarily
on the field intensity and gas pressure.
Chaotic motion is superimposed on the ordered motion
of the ions and electrons resulting from the constant elec-
tric force. A particle accelerated by the field travels only
a short distance. Its short run inevitably ends in a col-
lision. At low velocities, these collisions conform to the
law of elastic collisions.
The mean free path is determined first of all by the
pressure of the gas in the tube. The higher the pressure,
the shorter the mean free path and the lower the average
velocity of ordered motion of the particles. Voltage
applied across a gas-discharge tube has the opposite
effect: higher voltage raises the average velocity of or-
dered motion of the particles.
If no voltage is applied across the tube, the following
events occur in the gas. The ionizer produces ions and
then ions of opposite signs join each other when they meet.
This is called" ion recombination. Since it takes a pair of
particles to recombine, the rate of this process is pro-
portional to the square of the number of particles.
When the ionizer operates continuously, equilibrium is
set up between the two processes: ionization and recom-
bination. This is what happens in the ionosphere surround-
ing our earth. Depending upon the time of day and the
season, the number of ionized particles in a cubic cen-
timetre varies from a million to hundreds of millions of
ions and electrons. Thus, the degree of ionization is a
quantity of the order of one per cent (recall the number
of air molecules in unit volume at high altitudes).
But let us return to the ionized gas in the tube subject
to a voltage. Equilibrium is upset, of course, between
ionization and recombination because a part of the ions
reach the electrodes before they are recombined. As the
voltage is raised, a larger and larger part of the ions
'produced in unit time reach the electrodes, increasing
Electrons
Figure 2.7
70
the current through the gas. This continues until there
is no time left for recombination and all the ions produced
by the ionizers reach the electrodes. It is obvious that
no further increase in voltage can increase the current
(this is called the saturation current and is represented
by the horizontal portion of the current-voltage charac-
teristic in Figure 2.7).
The less the density of the gas, the lower the fie-ld
voltage at which the saturation current is reached.
The saturation current is equal to the charge of the
ions formed by the ionizer in one second throughout the
volume of the tube. The saturation current is usually
not very large: of the order of a microampere or less. Its
magnitude depends, of course, on the number of destruc-
tive missiles with which the ionizer bombards the gas.
When we operate the tube at a point on the current-
voltage characteristic within the limits of the saturation
current and protect the gas from the effects of an exter-
nal ionizer, the current ceases to flow. This is said to be
a non-self-maintained gas discharge.
If the voltage is raised again, new phenomena occur.
:At a certain instant, the velocity of the electrons becomes
2. Electrical Structure of Matter 71
high enough for them to knock electrons out of the neu-
tral atoms and molecules. For this the voltage across the
tube should reach a value at which an electron acquires
energy sufficient in its free path to ionize a molecule.
The initiation of collision ionization has a marked effect
on the current-voltage characteristic: the current begins
to increase again because an increase in voltage leads to
an increase in the velocity of the electrons. An increase in
velocity, in its turn, raises the ionizing capacity of the
electrons so that a greater number of ion pairs is pro-
duced and the current increases. The current-voltage
characteristic CU1've turns sharply upward. In compari-
son with the saturation value, the current increases by
hundreds and thousands of times and the gas begins
to glow.
Now, if we eliminate the action of the external ion-
izer, the current continues to flow. We have gone over
to the region of a self-maintained gas discharge. The
voltage at which this qualitative change occurs is called
the breakdown voltage or the ignition voltage of a gas
discharge.
The sharp increase in current in passing over this
critical limit is due to the avalanche-type increase in
the number of charges. One released electron destroys
a neutral molecule and produces two charges of such
high energy that they are capable of breaking down
another pair of molecules that they encounter. Thus, two
charges form four, four form eight, etc. You must agree
that the use of the word "avalanche" is fully justified.
A quantitative theory has been proposed that predicts
the shape of the current-voltage characteristic curves
for gases with a fair degree of accuracy.
Electrons 72
Self-Maintained Discharge
There are many types of discharges in gases. We shall
discuss only a few.
Spark discharge. A spark passing through the air
between two electrodes can be readily observed in the
most elementary experiments. Sparks can be produced
simply by bringing two wires, across which a voltage
has been applied, sufficiently close together. What do
we mean by "sufficiently"? If air is the medium through
which the spark is to pass, the required field intensity
is 30 thousand volts per centimetre. This means that 8
potential difference of 3000 volts is sufficient for a small
gap of one millimetre. Any reader has observed such
small sparks in everyday life when repairing faulty
electric wiring or when bringing two leads from the
terminals of a storage battery close together (here the
gap between the ends of the leads must equal the thick-
ness of a safety razor blade).
The breakdown voltage depends upon the density of
the gas. It also depends upon the shape of the electrodes.
A spark can pass through dielectric liquids and sol-
ids as well as through a gas. It is important for an elec-
trician to know the breakdown voltage of all the materials
he uses in his work.
Today it seems quite obvious to us that lightning is a
spark jumping between clouds charged with electricity
of opposite sign. This was not always the case, however,
and Mikhail Vasilievich Lomonosov (1711-1765) and
Benjamin Franklin (1706-1790) made great efforts to
prove this statement. Georg Wilhelm Richmann ( 1 7 1 1 ~
1753), a Russian physicist that collaborated with Lo-
monosov, lost his life in an attempt to ground lightning
through a twine that conducted electricity from a kite
that he flew during a thunderstorm.
Interesting data are available on the spark discharge
2. Electrical Structure of Metter 73
in a bolt of lightning. The voltage between the cloud and
the earth is about 10S to foe volts and the current ranges
from tens to hundreds of thousands of amperes. The
diameter of the spark channel ranges from 10 to 20 em.
The duration of 8 flash of lightning is extremely short:
of the order of 8 microsecond. We CBD readily estimate
that the quantity of electricity passing through the light-
ning channel is comparatively small.
These sparks from the heavens have been studied in
detail by means of high-speed motion picture photog-
raphy. A bolt of lightning quite frequently consists
of a series of spark discharges following a single path.
Lightning has a sort of "leader" which pierces the most
convenient, always freakishly branched way for the elec...
trio discharges.
Also observed to some extent is ball lightning. This
type is not fully understood and cannot, unfortunately,
be reproduced under laboratory conditions. It is an in-
candescent sphere of gaseous plasma, 10 to 20 em in
diameter. Ball lightning moves slowly and is sometimes
even stationary. It exists several seconds, or even min-
utes, and then disappears with a powerful explosion.
It must be admitted that so far no comprehensive theory
of this interesting phenomenon has been proposed.
Arc discharge. An arc discharge was first obtained
by the' famed aussian physicist and electrical engineer
Vasily Vladimirovich Petrov (1761-1834) as far back
as 1802. He struck the arc by bringing into contact two
pieces of carbon across which a powerful voltage had been
applied from a source of electric energy. Then he moved
the carbons (electrodes) apart. This procedure is still
being employed today. True, special carbon rods of
pressed graphite powder are used as the electrodes. The
positive rod burns away more rapidly than the negative
one. By just looking at the electrodes we can immedi-
ately find- 'to which one the positive pole is connected: a
Electrons 74
hollow, called the crater, is formed at its end. The tem-
perature in the crater at standard pressure may reach
4000 C. If the pressure is increased, the temperature
can be raised almost 6000C, i.e. the temperature at
the surface of the sun. An arc struck between: metal elec-
trodes has a flame with a considerably lower temperature.
A low voltage, of the order of 40 to 50 volts, is suf-
ficient to maintain the arc. The current may reach hun-
dreds of amperes because the resistance of the incandescent
gaseous column is low.
How do we explain the high electrical conductivity
of the gas at such low voltages? The molecules are not
accelerated to 'high velocities,and their collisions cannot
play much of a role in producing a heavy current, The
explanation is as follows: at the first instant the arc'
is struck, a large quantity of heat is evolved at the point
of contact, drastically raising the temperature at the
ends of the electrodes. This initiates the process of ther-
mal electron (thermionic) emission in which the cathode
ejects an immense number of electrons'. From this it
follows, by the way, that only the high 'temperature of the
cathode is of importance; the anode can remain cold.
The mechanism of an arc discharge of this type is
entirely different from that of a spark discharge.
No need, evidently, to remind the reader how important
this phenomenon is in engineering practice. Arc dis--
charges are widely employed in welding .and cutting me-
tals, as well as in electrometallurgy.
Glow discharge. This kind of self-maintained dis-
charge is also of great practical value, since it takes place
in gas-discharge tubes or, as they are also called, day-
light lamps. The tube is designed and is filled with gas
(at a pressure substantially below atmospheric pres-
sure) 'so that it operates at a voltage exceeding the igni-
tion voltage. The electric current in gas-discharge lamps
is produced by the ionization of the. gas m.oleculesby
2. Electrical Structure of Matter 75
electrons, and also because electrons are knocked out
of the cathode. A gas-discharge lamp does not ignite
instantaneously. This is evidently due to the fact that
the first impulse must be obtained from the small amount
of charged particles that are always present in any gas.
Corona discharge. A corona is observed at atmo-
spheric pressure in a highly nonuniform field, for exam-
ple, in the vicinity of a wire or sharp point. The field
intensity should be high: of the order of a million volts
per metre. It makes no difference to which pole the sharp
point is connected. Thus, there are both positive and
negative coronas. Since the field intensity is reduced
as we depart from the sharp point, the corona disappears
at a short distance away. We can contend that a corona
discharge is an incomplete breakdown of the gas gap.
A corona is produced by electron avalanches travelling
either toward the point or from the point to external space.
Besides electrons, the corona region contains negative and
positive ions-products of disintegration of the neutral
air molecules. The corona glows only in the small space
near the point in which there is an electron avalanche.
The initiation of a corona depends on the atmospheric
conditions and, primarily, the moisture.
The atmospheric electric field may cause the tops of
trees and tips of ship's masts to glow. Formerly, this
phenomenon was called St. Elmo's fire. It was supposed
to be an evil omen. This has a reasonable explanation.
It may well be that such a glow is observed just before
a storm breaks.
We can learn a moral lesson from an event that took
place quite recently. Two amateur investigators, Mr.
and Mrs. Kirlian, had spent some years studying the
following phenomenon. A person lays a hand, connected
to one terminal of a high-voltage power supply, on a
photographic film that is separated by a layer of insu-
lation from the other electrode of the same power cir-
Electrons
cuit. When the voltage is switched on, a blurred image
of the palm and fingers is obtained on the film. This
image is due to a corona discharge. Naturally, the volt-
age must be less than that required for a spark break-
down of the insulation.
These experiments drew the attention of experts in the
field of so-called parapsychology, the general name for
a group of "theories" that the great majority of phys-
icists and psychologists consider to be pseudo-scientific.
This attention was due to the fact that the Kirlians
and their followers related the kind of photograph ob-
tained to the "psychic" state of the subject.
As a result of the extensive publicity received by such
an extravagant interpretation of these experiments, a
group of physicists and psychologists from various uni-
versities in the USA decided to check the experiments
more carefully. They searched for a simpler explanation
of the undoubted fact that the appearance of the pho-
tograph thus obtained really does differ for different per-
sons or even for the same person if the photograph is
taken under different conditions.
The investigators reached the following conclusion:
"Photographic images obtained by the Kirlian tech-
nique are principally a record of corona activity during
an exposure interval. Most of the variations in the im-
ages of the corona of a living subject who is in contact
with the photographic film can he accounted for by the
presence of moisture on or within the subject's surface.
During exposure, moisture is transferred from the sub-
ject to the emulsion surface of the photographic film and
causes an alteration of the electric charge pattern on
the film, hence the electric field at the surface of the
subject. "
The investigators proposed further application of this
technique, which they preferred to call "corona-discharge
photograph", .' ~ f o r the detection and quantification of
2. Electrical Structure of MaHer 77
moisture in animate and inanimate specimens through
the orderly modulation of the image due to the various
levels of moisture".
This interesting piece of information, published in
the December 1976 issue of the journal Scientific Amer-
ican with the title "Sweaty Palms", leads to two conclu-
sions. In the first place, any real phenomenon deserves
attention and it is quite possible that it may prove useful
in practice. Secondly, when an investigator discovers a
new fact, he should begin by overcoming the tempta-
tion to interpret it in a way that does not fit in with up-
to-date scientific concepts. The discovery can be made
public and presented for judgement by specialists only
after it has been comprehensively shown that existing
theories are incapable of explaining the new fact.
Real facts, which are falsely interpreted and explained,
can be called (after an old joke) cockroach exper-
iments. According to the joke, the legs are pulled off
one cockroach which is then placed on a table side by
side with a cockroach having all his legs. The "inves-
tigator" then knocks on the table. The uninjured cock-
roach runs away while the "cripple" remains. This proves,
contends the "investigator", that a cockroach hears
with its legs.
Every year publications appear that describe such
"cockroach experiments". I consider it good policy to
warn my readers against them.
Metter inthe Plasma Stille
The term Plasmenzustand was proposed as far back
as 1939 by two German scientists whose article was trans-
lated into the Russian by the author for the Soviet jour-
nal Advances in Physical Sciences. This seems to be a
suitable name. As a matter of fact, plasma is neither a
Electrons 78
solid, nor a liquid, nor a gas. It is a special state of
matter.
Thermal ionization of gases, i.e. stripping the atoms
of their electrons and the disintegration of neutral mol-
ecules into ions, hegins at temperatures exceeding 5
or 6 thousand degrees. Is it worthwhile, then, to discuss
this problem? No materials exist that can withstand a
higher temperature.
It certainly is worthwhile! The great majority of ce-
lestial bodies, like our sun, are in a plasma state. An
example of plasma is the ionosphere. Plasma can be con-
fined in a limited volume by properly shaped magnetic
fields, so-called magnetic bottles, under laboratory con-
ditions. We can, in addition, speak of the plasma of a
gas discharge.
The degree of ionization of a gas depends on the pres-
sure as well as the temperature. Hydrogen at a pressure
of the order of 1 mm of mercury column is practically
completely ionized at a temperature of 30 thousand de-
grees. Under such conditions, there is only one neutral
atom per 20 thousand charged particles.
Hydrogen in the plasma state consists of a mixture
of chaotically travelling and colliding particles of two
gases: proton "gas" and electron "gas". Plasma formed
of other substances is a mixture of many "gases". In such
a plasma we find electrons, bare nuclei, various ions,
as well as a negligible quantity of neutral particles.
A plasma with a temperature of tens or hundreds of
thousands of degrees is said to be cold. A hot plasma has
a temperature of millions of degrees.
But one must be careful with the concept of a plasma
temperature. As the reader knows, temperature is unique-
ly determined by the kinetic energy of the particles.
In a gas consisting of heavy and light particles, a state
of equilibrium is set up only after the heavy and light
particles have acquired the same average kinetic energy.
2. Electrical Structure of Matter 79
This means that in a gas existing for some time under
stable conditions, the heavy particles travel slowly and
the light particles, rapidly. The time required to set
up equilibrium depends on what we had "at the begin-
ning". Other conditions being equal, the greater the
difference between the masses of the particles, the long-
er the time required to attain equilibrium.
These are exactly the conditions we find in a plasma.
The mass of the lightest nucleus is almost two thousand
times greater than that of an electron. In each collision,
an electron transmits only a small part of its energy
to a nucleus or ion. Thus the average kinetic energies
of all the particles of plasma are equalized only after
an immense number of collisions. Such a plasma is said
to be isothermal. It is the kind existing in the interior
of the sun and other stars. The time required to reach
equilibrium in a hot plasma ranges from fractions of
a second to some seconds.
The plasma in a gas discharge (spark, arc, etc.) is a
different matter. Here the particles not only travel chaot-
ically, but also produce an electric current. In their
path between the electrodes, the rapid electrons do not
have the opportunity to transmit any appreciable part
of their energy to the leisurely paced ions. This is why
the average velocity of the electrons in a gas discharge
is much higher than that of the ions. Such a plasma is
said to be nonisothermal and must be specified by two
temperatures (or even three if we take the neutral par-
ticles into account). Naturally, the electron temperature
is substantially higher than the ion temperature. Thus,
in an arc discharge, the electron temperature is from 10
to 100 thousand degrees, while the ion temperature is
close to 1000 degrees.
The behaviour of the particles in a plasma can be de-
scribed by means of the same quantities that are used
in the kinetic theory of gases. Many techniques have
electrons 80
been devised for directly or indirectly determining the
mean free path of the particles, the mean free time, and
the concentrations of particles of various kinds.
To povide the reader with an idea of the orders of mag-
nitude found in plasma physics, we present certain data
describing a high-concentration hydrogen plasma (10
20
ions per cubic metre). It was found that in a cold plasma
(at a temperature of 10 000 C), the mean free path equals
0.03 cm and the mean free time is 4 X 10-
10
s. If the same
plasma is heated to a hundred million degrees, the re-
spective figures are 3 X 10
6
em and 4 X 10-
4
s.
In citing such data, it is absolutely necessary to spec-
ify additionally the kind of collisions we have in mind.
The data given are for encounters between electrons and
ions.
It is evident that a volume containing many particles
is electrically neutral. But we may be interested in the
behaviour of the electric field at some definite point in
space. The field varies rapidly and drastically because
ions and electrons alternate in rushing past the point.
The rapidity of these variations can be calculated, as
can the average intensity of the field. Plasma complies
with the condition of neutrality with exceptionally high
precision. Strictly speaking, we should employ the term
"quasi-neutrality", i.e. near-neutrality. But what do
we mean by this "near"?
Rather simple calculations indicate the following. Con-
sider a length of one centimetre in the plasma. Calculate
the electron and ion concentrations for each point in
this length. Quasi-neutrality means that these concen-
trations should be "nearly" equal. Next, let us imagine
that in one cubic centimetre we have an "extra" amount
of electrons that are not neutralized by positive ions.
If this is so and the particle density is equal to the air
density at the earth's surface, a field with an intensity
of about 1000 V/cm is set up along the length being con-
2. Electrica' Structure of Matter 81
sidered, even if the difference in ion and electron con-
centrations equals only one thousand millionth of one
per cent! Here is what the word "near" signifies.
But even this negligible lack of equality of positive
and negative charges lasts only for an extremely short
instant. The field that forms excludes all the super-
fluous particles. This automatism is already operable
for regions measured by thousandths of a centimetre.
We shall return to plasma in magnetic bottles again
in Book 4. The reader has undoubtedly seen accounts,
and perhaps descriptions, of installations of the Toko-
mak type (in the USSR). A whole army of scientists are
working on their improvements. The point is that if
a high-temperature plasma could be produced and prop-
erly handled, it would lead to controlled fusion of light
atomic nuclei, which would be accompanied by the gen-
eration of titanic amounts of energy. Physicists have
learned to realize this process (uncontrolled fusion) in
bombs. Will it be possible to produce a plasma with a
sufficiently high temperature and sufficiently long du-
ration to initiate a chain reaction of the kind accom-
plished in a nuclear reactor? There is no answer yet to
this question.
Metals
The subdivision of solids into various classes accord-
ing to their electrical resistance is based on the mobility
of their electrons.
An electric current is a stream of moving charged par-
ticles. When we deal with streams of ions or electrons,
we literally visualize an electric current. An electric
current also reveals itself distinctly in passing through
liquids because matter is deposited on the electrodes.
But as for solids, there is only circumstantial evidence
of the passage of an electric current.
6--2259
Electrons 8 2 ~
A series of facts are available that enable us to make)
the following statements. No displacement of the atomic
nuclei occurs in solids. An electric current is produced
by electrons. The electrons travel due to the energy sup-
plied by the current source. This source sets up an elec-
tric field inside the solid.
The equation relating the voltage and the electric
field intensity is valid for any conductors. Therefore,
combining the equations on pp. 14 and 22, we can write
Ohm's law for a solid conductor in the form
j=oE
where a=1/p is called the electrical conductivity.
The electrons of a solid can be divided into bound and
free electrons. The bound electrons belong to definite
atoms; the free electrons form a kind of electron gas.
These electrons move around in the solid. When no volt-
age is applied to the solid, the electrons have random mo-
tion. The more the motion of the free electrons is impeded,
the oftener they collide with fixed atoms and with one
another, the higher the electrical resistance of the body.
The vast majority of electrons in a dielectric have an
owner that is either an atom or a molecule. The number
of free electrons is negligible.
In metals each atom donates one or two electrons for
common use. This electron gas is the current carrier.
On the basis of a roughly approximate model we can
estimate the electrical conductance and thereby check
the model.
In exactly the same way as when we discussed a mo-
lecular gas, we shall assume that each electron manages
to travel a path of length l without collisions. The dis-
tance between the atoms of a metal equals several ang-
stroms. It is logical, then, to assume that in order of mag-
nitude the mean free path of the electrons equals 10 A,
i.e, 10-
7
em,
2. Electrica' Structure of MaHer 83
In its motion the electron is subject to the accelerating
force eE during the time ll u, where v is its velocity. The
chaotic velocity of electrons can be estimated on the
basis of data obtained in investigations of thermionic
(thermal electron) emission. This velocity is of the order
of 10
8
em/s.
To determine the velocity of ordered motion of elec-
trons, i.e. the velocity of the motion that produces 8
current, the acceleration elilm. is to be multiplied by the
mean free time. This assumes that each collision discon-
tinues motion of the electron, after which it begins to
pick up speed again. Multiplying, we obtain the velocity
of the electrons that produce the electric current:
eEL
u=-
mv
Next, we attack the problem of determining the resis-
tivity of a metal. If we obtain the correct order of mag-
nitude, then we can presume that our model "works".
We shall leave to our readers the task of showing that
the current density j can be written as the product of
the number of electrons in unit volume by the charge
of the electron and by the velocity of ordered motion
(drift) of the electrons. Thus j=neu. Substituting into this
equation the drift velocity, we obtain j = ne
2
l E. Then
mv
the electrical conductivity is
ne
2
1
a=-
mv
If we assume that each atom contributes one electron
for common use, then we find that a conductor has a re-
sistivity of the order of 10-
5
ohm-me A very reasonable
valuel It confirms both the validity of our roughly ap-
proximate model and the proper choice of the values
of the parameters in our "theory". I place the word
Electrons -
theory in quotation marks only because it is a crude
approximation and elementary. This example, however, il-
lustrates the typical physical approach in interpreting
phenomena.
According to the theory of a free electron gas, the elec-
trical resistance should decrease with a drop in temper-
ature. But do not hasten to relate this circumstance
with the change in the velocity of chaotic motion of the
electrons. This velocity has nothing to do with the mat-
ter because it depends only slightly on the temperature.
The reduction in resistance is due to the' reduced ampli-
tude of vibration of the atoms. As a result, the mean
free path of the electrons increases.
This fact can also be expressed in other words: upon
an increase in the amplitude of vibration of the atoms,
the electrons are scattered to a greater degree in various
directions. Consequently, the component velocity in the
direction of the current is reduced, i.e. the resistance
should increase.
The increase in electron scattering also explains the
increased resistance of a metal (and not only a metal)
when impurities are added to it. The impurity atoms
act like defects in crystal structure and therefore facil-
itate electron scattering.
Electric energy, as we know, is transmitted by wires.
Owing to their resistance, the wires draw energy from
the current source. Such energy losses are enormous and
their prevention is one of the most vital of engineering
problems.
There is hope that this problem can be solved on the
basis of the remarkable phenomenon known as super-
conducti vity.
In 1911, the Dutch physicist Heike Kamerlingh Onnes
(1853-1926) found that at temperatures close to abso-
lute zero certain bodies suddenly lose practically all of
their electrical resistance. If a current is induced in a
2. Electrlce' Structure of M8fter 85
ring-shaped superconductor, it continues to flow for days
without diminishing. Of pure metals, the highest tem-
perature at which superconductive properties are found
is about 9 K, the metals being niobium (8.9 K) and tech-
netium (9.3 K). No need to mention what a vast army
of scientists are engaged in the search of superconductors
that would acquire this wonderful property at a higher
temperature. So far they have not been any too suc-
cessful, though an alloy has been discovered that is
claimed to become superconductive at about 20 K.
There is reason to believe, however, that this limit
can be raised (perhaps even to room temperature). The
search is being made among special polymer substances
and among complex lamellar materials in which a di-
electric alternates with a metal. The significance of this
problem -can hardly be overvalued. I take the liberty
of regarding it to be one of the cardinal problems in
modern physics.
The search for superconductors that acquire this prop-
erty at sufficiently high temperatures especially gained
in scope after a theory had been proposed for this phe-
nomenon. This theory suggested new courses along which
the required materials might he found.
It is characteristic that much time passed between
the discovery of the phenomenon and its explanation. The
theory was advanced in 1957. It might be well to point
out that- the laws of quantum physics, on which the theo-
ry of superconductivity is based, were established as
far back as 1926. It follows that the explanation of the
phenomenon is far from simple. In this book I can only
start, as you might say, from "the middle of the story.
It seems that with the slowing down of vibrations in
the atomic lattice, certain electrons become "paired"
Such a "pair" behaves in coordination with each other.
When the pairs are scattered by the atoms (and this
scattering, as mentioned above, is the cause of res is-
Electrons 86
tance), the rebound of one of the members of the pair to
one side is compensated for by the behaviour of its
"friend". This compensation is in the sense that the total
momentum of the electron pair remains constant. Thus,
electron scattering does not disappear but no longer in-
fluences the passage of the current.
In addition to paired electrons, a superconductor also
carries ordinary electron gas. Hence, two fluids seem
to exist simultaneously: one is ordinary and the other
is superconductive. As the temperature of the super-
conductor is raised from absolute zero, thermal motion
breaks apart more and more electron pairs, and the per-
centage of the ordinary electron gas increases. Finally,
at the critical temperature, all the paired electrons dis-
appear.
In Book 2 we made use of the two-fluid model, with
one- ordinary fluid and one special fluid, to explain su-
perfluidity observed in liquid helium. These two phe-
nomena are closely related: superconductivity is super-
fluidity of the electron fluid.
Each electron pairt" of the kind mentioned above, has
a total spin equal to zero. Particles with a spin equal
to zero or a whole number (i.e. with integral spin) are
called bosons. Under certain known conditions, large
amounts of hosons can occupy the same energy level.
In such cases their motions become ideally coordinated
and nothing can impede their displacements. We shall
return to this phenomenon again in Book 4.
Electron Emission from Metals
Since a part of the electrons behaves like a gas of rapid
particles, it is natural to expect that electrons are capa-
ble of emerging to the surface of a metal. For the electron
to leave the metal entirely, it must overcome the attrac-
2. Electrica. Structure of Matter 87
tive forces of the positive ions. The work done by the
electron for this purpose is called the work function.
The higher the temperature of the metal, the greater
the kinetic velocity of the electrons. If the metal is heat-
ed to incandescence, an appreciable number of electrons
will be able to escape from it.
Thermionic emission, as the ejection of electrons from
a metal is called, can be investigated b.y a simple exper-
iment. An additional electrode is sealed into an elee-
trio light bulb. A sensitive instrument can be used to
measure the current set up by the part of "evaporating"
electrons that reach the new electrode (a part, and not
all, because the electrons fly out of the incandescent
filament in all directions).
To evaluate the work function we resort to a "barrier"
voltage, i.e. we connect the sealed-in electrode to the
negative pole of a battery. Gradually raising the volt-
age, we reach a value at which the emitted electrons
can no longer arrive at the electrode.
The electron work function for tungsten is about 5
electron volts. Special coatings can, if required, reduce
the work function to one electron volt.
What is this unit of work called the electron volt?
It is, as the name implies, the energy acquired by an
electron in travelling over a portion of its path with a
voltage of one volt applied across this portion. One
electron volt equals 1.6 X 10-
19
joules. The thermal ve-
locity of electrons is quite considerable, but their mass
is very small. Therefore, the given harrier height is
extremely large. Theory and experiments show that elec-
tron emission depends drastically on the temperature. A
temperature rise from 500 to 2000 K increases the emission
current a thousandfold.
Owing to thermal motion the emission of electrons
from a metal is, so to speak, a natural process. But elec-
trons can also be knocked .out .of a metal..
Electrons 88
In the first place, this can be done by bombarding the
metal with other electrons. This is called secondary (elec-
tron) emission. It is made use of to multiply the number
of electrons in certain engineering instruments.
Vastly more essential is the extraction of electrons
from solids by light. This phenomenon is called the
photoelectric effect.
Thermoelectric Phenomena
Very long ago (in the evolution of mankind this time
is a mere instant, but in the development of science it
is almost as long as eternity), over 150 years in the past,
a simple fact was discovered. If an electric circuit is
made up of a piece of copper wire and a piece of bismuth
wire by soldering them together at two junctions, current
flows through this circuit. This happens only when the
temperature of one junction is higher than that of the
other. This is the thermoelectric effect.
What makes the electron travel along our combined
circuit? The explanation of this phenomenon is not at
all simple. The electromotive force is' due to two factors.
In the first place, we have a contact electric field,
secondly, we have a temperature electric field.
We have just mentioned that work is required to remove
an electron beyond the limits of a metal. It is natural
to assume that work function A differs for various metals.
Hence, there is a voltage across the junction of the two
metals equal to
_I (AI -A
2
)
e
The contact voltage can be detected experimentally.
By itself, however, this voltage cannot set up a current
in a closed circuit. Such a circuit obviously consists of
two junctions and the contact voltages oppose and cancel
2. Electrical Structure of Metter 89
each other. But why does the difference in" the tempera-
tures of the junctions produce an electromotive force? The
answer is the only logical one. Evidently the contact
voltage depends upon the temperature. Heating one of
the junctions makes the voltages unequal and sets up a
current. But here we must take another phenomenon into
account. It can naturally be assumed that there is an
electric field between the ends of a conductor if these
ends have different temperatures because the electrons
travel faster at higher temperatures. This being the case,
diffusion of the electric charges begins and continues
until a field is set up that counteracts the tendency to
uniform distribution.
Experiments leave no doubt of the fact that both phe-
nomena are simultaneously present and both are to be
taken into account in proposing a theory.
Thermoelectromotive forces are small: of the order of
a millivolt for a temperature difference of 100 degrees.
Such voltages are readily measured. Consequently, the
thermoelectromotive effect isemployed for measuring tem-
peratures. You cannot, of course, insert a liquid-in-glass
thermometer into molten metal. For such purposes a
thermocouple (as the thermoelement used for measuring
temperatures is called) is an excellent instrument. A
thermocouple has, in addition, many other advantageous
features. How vitally important it may be to measure
temperatures at great distances. And the exceptional sen-
sitivity. Electrical measurements are always precise, and
it was found that differences in temperature as small
as a millionth of a degree can be sensed by a thermo-
couple.
This high sensitivity enables thermoelements to be
applied for measuring heat flow from extremely remote
objects. The reader can himself estimate the possibilities-
of a thermoelement. Suffice it to say that a tenth of an
erg per second is no limit.
Electrons 90
Like storage cells, thermoelements are frequently as-
sembled into banks to form a thermal battery. If the
power requirements are not very high, such a battery can
serve as a generator of energy and can find application
in radio communications.
semiconductors
A great many substances, both elements and chemical
compounds, have conductivities filling the wide range
between conductors and insulators. Such substances were
first discovered a long time ago. But a mere twenty years
ago it was hardly probable that anyone could foresee that
semiconductor physics would give rise to a powerful
branch of industry whose importance in world econom-
ics cannot be overrated.. Without semiconductors, up-
to-date electronic computers, TV sets and tape recorders
would be unfeasible. Radio engineering is inconceivable
today without semiconductors.
Insulators (nonconductors) have a conductivity ranging
between 10-
8
and 10-
18
ohnrv-rrr", the range of conductiv-
ity of metals in the same units is from 10
2
to 10'. The
conductivity of semiconductors lies between these two'
ranges. As we shall see, not only their resistance is of
interest in dealing with semiconductors.
Like in metals, no chemical changes occur in semi-
conductors when we pass an electric current through them.
This indicates that the ions of these substances, forming
the frame of their crystal lattice, are not moved around
by the action of the electric field. Therefore, as in metals,
the motions of the electrons are responsible for electric
conduction.
Though this seems obvious, physicists, in the early
stages of semiconductor research, decided, in any case,
to find which charges were the current carriers. For 501-
2. Electrical Structure of Matter 9f
ids, this can be done by means of the Hall effect, dis-
covered in 1879 by Edwin Herbert Hall (1855-1938).
In the next chapter I shall remind you that a magnetic
field deflects positive and negative particles in different
directions. If a solid in which charges are travelling is
made in the form of a plate and is placed in a magnetic
field of the proper direction, a voltage appears across
the edges of the plate. This arrangement is shown sche-
matically in Figure 2.8.
The physicists were surprised to find that certain
bodies, when investigated in this manner, behaved some-
times as if positive particles were travelling along the
conductor, and at other times, as if the carriers of electric
charge were negative. We can readily find names for
this behaviour. We call the first case positive (p-type)
conduction and the second, negative (n-type) conduction.
The point, of course, is not in the name, but in an expla-
nation of the matter. There is no doubt that electrons are
travelling inside the semiconductor. How can we recon-
cile this contradiction? How can we explain the positive
conduction?
Imagine a formation of athletes. For some reason, one
man drops out of line. This leaves an empty space. Though
it doesn't sound any too aesthetic, we can say that a
Electrons
"hole" is formed. To dress the ranks, the commanding
officer tells the neighbour of the "hole" to move over
one place. It is absolutely clear that this forms a new
empty space. It can be filled by commanding the next
man to move Into the "hole". If the athletes shift, one
by one, to the. left, the "hole" moves to the right. This
is the ~ ..mechanism' that explains positive conduction of
semlcoriduotors.
The concentration of ,free electrons is very low in semi-
conductors. Hence, the definition itself of conduction
(recall the formula we derived a few pages hack for the
current density) should suggest that the majority of atoms
in a semiconductor are not in the form of ions, but are
neutral atoms. Still, a semiconductor is not an insulator.
Consequently, a small number of electrons are freed.
These electrons travel as in a metal and are 'responsible
for the negative, i.e. electron, conduction. But a positive
ion, surrounded by neutral atoms, is in an unstable state.
As soon as an electric field is applied to the solid, a pos-
itive ion tries to "lure" an electron from its neighbour;
the next atom proceeds in exactly the same way. A
positive ion is quite similar to a "hole", This interception
of electrons may overcome the motion of free electrons.
This is how positive, or hole, conduction occurs.
Maybe you do not care for this model? I can suggest
another. As we have mentioned, the energy of a particle
is quantized. This is a fundamental law of nature. All
phenomena occurring in semiconductors can be explained
if we assume that the electrons are distributed among
energy levels in a solid as they are in the atom. But since
there are so many electrons in a solid, the levels merge
into energy bands.
If there is only weak interaction between t he electrons,
the bands are quite narrow. Therefore, the fact that atoms
are a part of a solid has practically no effect on the inner
electrons.
2. Electrical Structure of Metter
93
Al4C !
Iii ...
I
. -
___
Figure 2.9
Outer electrons are a different matter. Their levels
form the bands. The width of these bands and the "dis-
tance" between them differ in different bodies (actually
we should call them energy gaps; the word "distance"
in this context is simply physical shoptalk or jargon).
This picture clearly explains the division of solids
according to their electrical conductivity into metals,
semiconductors- and insulators (Figure 2.9). When a band
is completely filled with electrons and the width of the
gap to the next higher empty band is large, the bod.y is
an insulator, or nonconductor. If the upper band is only
Electrons 94
partly filled with electrons, we have a metal because
even the weakest electric field can kick an electron to a
slightly higher energy level. Typical of a semiconductor
is the fact that its upper band is separated from the next
lower one by a narrow gap. In contrast to nonconductors
and metals, thermal motion is capable in semiconductors
of transferring electrons from one band to another. In
the absence of a field, the number of such transitions
upward and downward are the same. A rise in temperature
only increases the electron concentration in the upper
hand.
But what happens when a field is applied to semi-
conductor?
Now a free electron in the upper band begins to move
and contributes to the negative conduction. But the equi-
librium of upward and downward electron transitions is
violated. Therefore a hole is formed in the lower band
and begins to move, due to the field, in the opposite
direction. Such semiconductors are called conductors with
mixed (positive-negative) conduction.
The band theory of semiconductors is an orderly and
consistent one. The reader should not consider the de-
scribed model to be artificial or far-fetched. It simply
and clearly explains the principal difference between a
metal and a semiconductor, namely, their particular
behaviour with a change in temperature. As mentioned
in a preceding section, the electrical conductivity of
metals drops with a rise in temperature because the elec-
trons more frequently collide with obstacles. An increase
in the temperature of a semiconductor leads to an increase
in the number of electrons and holes, and a consequent
increase in conductivity. Calculations indicate that this
effect substantially exceeds the loss in conductivity due
to collisions with obstacles.
Of prime importance in engineering are semiconductors
with impurities. By adding impurities, called doping, we
2. Electrica' Structure of Matter 95
can produce a body that has only negative or only posi-
tive conduction. The idea is extremely simple.
The most extensively employed semiconductor mate-
rials are germanium and silicon. These elements are tet-
ravalent. Each atom is bound to four neighbours. Ideally
pure germanium is a semiconductor of the mixed type.
The number of holes and free electrons per 1 em" is very
small, namely, about 2.5 X10
13
This comes to about
one free electron and one hole per thousand million atoms.
Now let us replace an atom of germanium with an atom
of arsenic. Arsenic is pentavalent (having five valence
electrons). Four of its electrons bind it to atoms of the
host-germanium-and the fifth is free. This material
possesses electron (negative) conduction because the ad-
dition of an arsenic atom does not lead to the formation
of holes.
Even if only a trace of arsenic is added, one atom per
million atoms of germanium, the conductivity of the
germanium increases by a thousandfold.
It should be clear now what is needed to convert the
germanium into a p-type conductor. This can be done by
replacing a germanium atom with a trivalent atom, for
example, indium.
This leads to the following situation. An atom of
germanium adjacent to the guest is converted into a
positive ion because it must form a bond with the indium,
which has one electron too few. 'But we already know
that a positive ion plays the role of a hole. Owing to
the field, the holes move and there is motion of free elec-
trons.
No wonder that the semiconductor industry has greatly
influenced the techniques of growing pure crystals. It,
could not be otherwise when even a trace of impurities,
as small as a millionth part, makes all the difference.
It would be incorrect to suppose that there is no hole
conduction in n-type semiconductors. There are holes,
Electrons 96
of course, but their number is substantially less than the
number of free electrons. In n-typ{' semiconductors, the
electrons are the majority carriers of current, while the
holes, constituting a minority, are called minority carriers.
On the contrary, the majority carriers in p-type semi-
conductors are holes, and the minority carriers are elec-
trons.
p-n Junction
Now that we know what p- and n-type semiconductors
are, we can look into an interesting effect that is vitally
important in up-to-date electronics. This effect occurs
in the transition region between p- and n-type semicon-
ductors tightly joined together. This region is commonly
called a p-n junction, though the word transition has
served as the basis for naming a whole class of devices
(called transition-region devices) operating on the p-n
junction principle. What happens when we take two
blocks of the same cross section, one made of Ge doped
with In (p-type semiconductor) and the other of Ge
doped with As, grind one face on each block to an excep-
tionally high degree of smoothness and flatness! and join
the ground faces very tightly together? We shall have,
actually, a single crystal of germanium, but one half has
an excess of free electrons and the other, an excess of
holes.
To keep the explanation as simple as possible" we
shall, for the moment, forget about the minority current
carriers. At the initial instant of time (see upper drawing
in Figure 2.10), both halves of the crystal are electrically
neutral. But the n-type half has (notwithstanding its
electrical neutrality) "surplus" electrons (black dots) and
the p-type half has "surplus" holes (circles).
Both the electrons and holes can freely pass through
the boundary. The reason for such t.ransitions is the same
2. Electrical Structure of MaHer
Figure 2.tO
Electrons 98
as in mixing two gases whentheir containers are connect-
ed together. But, in contrast to gas molecules, electrons
and holes are capable of recombining.
At first we had six black dots at the left and six circles
at the right. As soon as the transition begins, a circle
and a dot annihilate each other. The next drawing shows
that there are less electrons left in the left half than
are required for it to remain electrically neutral. The
right half has one circle less. Depleting the left half
of one electron, we give this half a positive charge.
For the same reason, the right half acquires a negative
charge.
Transition through the boundary of the next holes
and electrons becomes difficult. They have to move
against the electric fields set up by the charges. Transition
continues for some time, as' long as the thermal motion
is able to surmount the constantly increasing barrier.
Finally, dynamic equilibrium is reached. i!.
Now, what happens if we apply a voltage across....our
composite p-n crystal as shown in the third drawing
in Figure 2.101 Evidently, we submit additional energy
to the current carriers of an amount sufficient to enable
them to clear the barrier.
On the contrary, if we connect the positive pole to
the n-type half, it remains impossible for electrons and
holes to surmount the barrier.
In short, a p-n junction allows current to flow in only
one direction, thereby possessing rectifying properties.
Today, rectifiers (for which valves and diodes are
synonyms) find most extensive application in many dif-
ferent fields of engineering. Their principle has just been
described.
Our explanation, however, was extremely crude. In
no detail did it take into account the behaviour of the
holes and electrons capable of surmounting the barrier
without recombination. The main drawback was our disre-
2. Electrical Structure of Matter 99
gard for the minority carriers that do not allow a p-n
crystal to completely rectify a current. Actually, a weak
current does flow when we apply the voltage as shown in
the lower drawing of Figure 2.10.
Let us consider in more detail the events that occur at
the boundary, or junction, when dynamic equilibrium is
reached.
We first discard the simple assumption we made above
and recall the existence of minority carriers.
In establishing dynamic equilibrium, the situation is
as follows. Approaching the boundary from the depth of
the p-type crystal, the hole current gradually increases.
Contributing to this current are holes that manage to
reach the p-n junction and jump across it without recom-
bining with electrons.
Of course, these holes must also possess sufficient ener-
gy to .overcome the potential barrier.
In passing through the transition region, this current
gradually fades due to recombination with electrons. At
the same time, a hole current gradually increases from
the depth of the n-type crystal, flowing in the opposite
direction. There are much fewer holes in this part, but
they do not have to surmount the barrier to get into the
p-type part. We can say that the barrier is arranged so
that the forward and reverse currents compensate each
other.
The above also concerns electron current. True, the
hole and electron currents may differ greatly in magni-
tude because the p- and n-type parts are differently doped
with impurities and possess, consequently, different
amounts of free carriers. If, for instance, there are a great
many more holes in the p-type part than electrons in
the n-type part, the hole current is much higher than
the electron current. Such a p-type part is called the emit-
ter of free current carriers, and the n-type part, the base.
After this more detailed account of the events hap-
7
Electrons 100
pening at the p-n boundary, we can understand why the
current cannot be completelyIrectlfied.
As a matter of fact, if the positive pole is connected
to the p--type crystal (or half), the barrier is lowered.
The voltage drives the electrons ahead. If the positive
pole is connected to the n-type part, the electric- field
set up by the power source coincides in direction with
that of the barrier. The field in the junction is increased.
Now the number of electrons capable of surmounting the
barri.er, as well as the number of holes capable of moving
in the opposite direction, are reduced. This raises the
resistance in the junction region, leading to the so-called
asymmetric volt-ampere characteristic.
As we can now see, a more thorough consideration
clearly explains why the rectification in the junction
layer cannot be complete.
3. Electromagnetism
Measure of MagneticFieldIntensity
The interaction between bars and needles made of
certain kinds of iron ores has been observed since ancient
times. These items have a certain peculiar property: one
end of the needles (or bars) points to the north. Two
poles, north and south, can be ascribed to such needles.
It can he readily shown that like poles repel and unlike
poles attract each other.
The behaviour of such special bodies was comprehen-
sively investigated by William Gilbert (1540-1603). He
explained the laws of their behaviour at various regions
of the earth and the rules of interaction between one
another.
On July 21, 1820, a Danish physicist Hans Christian
Oersted (1777-1851) published and widely acclaimed his
discovery in an article with the extremely strange name:
"Experiments Concerning the Effect of an Electrical
Conflict on a Magnetic Needle". The small article, only
four pages long, announced to the reader that Oersted
(to be more exact, one of the members of the audience
at his lecture) noticed that a magnetic needle is deflected
when placed near a wire through which an electric current
is passed.
This was soon followed by another discovery. The
eminent French physicist Andre Marie Ampere (1775-1836)
found that electric fields interact.
Thus, magnets affect other magnets and currents, and
currents affect other currents and magnets.
These interactions as well as electrical phenomena can
be conveniently described by introducing the concept
Electrons 1 0 ~ ;
of a field. We shall contend that electric currents, and
natural and artificial magnets set up magnetic fields.
It is worth emphasizing here that the reality of elec-
tric and magnetic fields or, in other words, the fact that
a field is a form of matter, can be proved only by investi-
gating variable fields. For the time being, a field is
simply a convenient concept and nothing more. As a
matter of fact, the sources of a magnetic field can be hid..
den behind a screen, and we can make sure that it exists
in space by observing the actions it performs.
In the presence of a magnetic field the same systems
react that set up the field, i.e. a magnetic field affects
magnetic needles and electric currents. The first question
that arises before an investigator studying magnetism
is the "probing" of the'"space in which the magnetic
field exists. When we specified an electric field, we de-
termined at each point of the field the force acting on
unit charge. How do we go about describing a magnetic
field?
In the general case, a tiny magnetic needle behaves
in a quite complex way. It turns in a definite manner,
but sometimes moves in a straight line. To characterize
a magnetic field, we should not permit motion of the
needle. First of all, we should find out the direction its
north pole points to (i.e. the end that points north when
there are no currents or magnetic objects in the vicin-
ity).
We explained above that a convenient graphical way
of describing an electric field is by means of lines of force.
The direction of the electric lines of force indicates which
way the positive charge is deflected. The density of the
lines corresponds to the magnitude of the force. We
can proceed in a similar way in describing a magnetic
field. The end of a freely rotating magnetic needle indi-
cates the direction of the lines of force.
What are we to take as the measure of the "intensity"
3. Electrom.gnetlsm 103
of a magnetic field? By means of a simple device, we can,
of course, measure the moment of force acting on the
magnetic needle. It is worthwhile, however, to search
for a better way. A magnetic needle is a kind of "thing in
itself". In conducting experiments with a magnetic needle,
we must simultaneously search for the measure of "in-
tensity" of the magnetic field and a measure character-
izing the needle. Physicists prefer to avoid such situa-
tions. As the saying goes: If you run after two hares, you
will catch neither. It is better to chase one hare first and
then the other.
Hence, for the time being, we shall restrict the func-
tion of the magnetic needle to an indication of the
pattern of the lines of force. To find a qualitative mea-
sure of the "intensity" of a magnetic field, we shall
consider one of Ampere '8 experiments. As far back as
1820 Ampere discovered that a current loop behaves much
as a magnetic needle. Namely, a current loop turns in a
magnetic field so that a normal to the plane of the loop
points in the same direction as a magnetic needle would,
i.e. along the lines of force. The role of the north pole
is played by the face of the loop you are looking at when
the current is flowing counterclockwise in the loop.
In contrast to a magnetic needle, a current loop is not
an item we find difficulty in characterizfng. The prop-
erties of a current loop are uniquely specified by the
current, loop area and the direction of the normal to
this area. We can expect that such a loop should be a
suitable instrument for probing a magnetic field.
We decide, then, to use the torque acting on a current
loop as a measure of the "intensity" of a magnetic field.
There is no reason to suppose that this instrument is
less convenient than a magnetic needle. A resourceful
investigator can make a loop of tiny area and devise a
simple way to counterbalance the rotation of the field by
compressing- '8 calibrated spring.
Electrons
Figure 3.1
t04
First we examine the behaviour of various test loops
at some definite point in a constant magnetic field.
These investigations yield the following result: the
moment of force is proportional to the product of the cur-
rent by the area of the loop. This means that' the test
loop is characterized by the product of the current by the
area and not by each separately.
In addition to this product, we must also know how the
normal to the loop is located with respect to the direc-
tion of the field. The loop behaves similar to a magnetic
needle, as mentioned above. Hence, if it is located so
that its positive normal, i.e, the vector emerging from
3. Electromagnetism 105
its north face, is directed along the lines of force, the loop
remains in this position (the moment of force being equal
to zero) as shown below in Figure 3.1. If the loop is
located with its normal perpendicular to the lines of
force (above in Figure 3.1), the moment of
its maximum value. -
It follows from the foregoing that it would be ex-
pedient to introduce a new concept at this point. This
concept, as we shall see, proves to be of prime signifi-
cance. We shall characterize the current loop by the vector
M, which we shall call the magnetic moment (see Figure
3.1). The magnitude of the magnetic moment is taken
equal to the product of the current I by the loop area
A=ld. Thus
M=IA
Vector A has the direction of the positive normal to
the plane of the loop.
Now we have an instrument that can be used to mea-
sure a field. It is most convenient to measure the maximum
moment of force acting on the test loop.
Going from one point of the field to another or chang-
ing the field by moving its sources or varying the cur-
rents setting up the field, we obtain different values
of the moment of the couple of forces F acting on the
test loop. The maximum moment of force can be written
as
N=BM
where B is a quantity that we take as the measure of the
field. It is called the magnetic induction. Thus, the
magnetic induction is equal to the maximum moment of
force acting ona test loop with unit magnetic moment.
We take the density of lines of force, i.e. their number
perunit area, proportional to value B. Vector B is directed
along the lines of force.
Electrons tOO
The magnetic moment, magnetic induction and our
old friend the moment of force are all vectors. After think-
ing it over, we must admit that these vectors differ
from the vectors of displacement, velocity, acceleration,
force, etc. As a matter of fact, the velocity vector, for
example, of a body indicates the direction of motion of
the body, the vectors of acceleration and force indicate
the direction of attraction or repulsion. In these cases,
the arrowhead at the end of the line symbolizlng the vec-
tor has"'an entirely objective and real meaning. As to
our new'acquaintances and the moment of force, these are
horses of another colour. The vectors are directed along the
axis of rotation. It is clear that an arrowhead put at one
or the other end of a line symbolizing an axis of rotation
is of a wholly arbitrary nature. It is necessary, however,
to agree upon the direction of the vectors. An arrowhead
at the "end" of an axis of rotation is meaningless. But
the direction of rotation has an objective meaning, and
this is what we should try to specify by the arrowhead.
It has been agreed that the arrowhead is put on the axis
in such a manner that when we face the arrowhead, rotation
is either clockwise or counterclockwise. Physicists are
accustomed to the latter.
These two types of vectors have expressive names that
speak for themselves: polar and axial vectors.
After measuring the fields of various systems, we can for-
mulate the following rules. Magnets always have two
poles: a north pole near which each line of force starts
and a south pole near which it ends. Naturally, we cannot
determine by such experiments what happens to the lines
of force inside the magnet.
~ With respect to the magnetic fields of currents (Fig-
ure 3.2), we discover the following law: the magnetic
lines of force are circular and surround the current con-
ductor. If we look along the conductor in the direction
of the current, the lines of force have the clockwise
3. Electromagnetism
Figure 3.2
107
direction. A point or an X in diagrams or drawings
indicates (and this is universally accepted) that the cur-
rent flows toward or away from us.
As is clear from the formula, the magnetic moment is
measured in amperes multiplied by the loop area in
square metres.
The unit of magnetic induction is the tesla. One tesla
(T) is equal to 1 kg/(A-S2).
Magnetic fields are set up by currents and permanent
magnets. Magnetic fields affect currents and permanent
Electrons 108
magnets. If for some reason the investigator does not
wish to resort to the concept of a magnetic field, he can
classify all kinds of interaction in which magnetic fields
participate into four groups. These are: magnetic, i.e.
the action of a magnet on another magnet; electromag-
netic, i.e. the action of a current on a magnet; magne-
toelectric, i.e. the action of a magnet on a current; and,
finally, electrodynamic, i.e. the action of a current on a
current.
This terminology is mainly used in engineering. An
instrument is said to be magnetoelectric, for instance,
if the magnet is fixed and the current-carrying loop is
movable.
The electrodynamic interaction is the basis for the
modern definition of a unit of current. This definition
is as follows: an ampere is the equivalent of a constant
current which, in passing along two parallel and straight
conductors of infinite length and negligible cross sec-
tion, located in a vacuum at a distance of one metre
from each other, would develop a force between these
conductors equal to 2 X 10-
7
newton per metre of length.
In the International System of Units (81), accepted
allover the world, the ampere is one of the fundamental
units. Correspondingly, the coulomb is defined as an
ampere-second. I must admit to the reader that I prefer
the system in which the quantity of electricity is the fun-
damental unit and is expressed in terms of the mass of
deposited silver in electrolysis. But metrologists must
know best. Evidently, the above definition must have
some merits, though, it seems to me that in practice
a measurement of the electrodynamic force with high
accuracy is a far from simple matter.
Since he now knows how to determine the direction
of a magnetic field, as well as the rules for finding the
direction of the force exerted on the current by a magnetic
field (to be discussed a little further on), the reader
3. Electromagnetism 109
can deduct for himself that currents flowing parallel
in the same direction attract one another, and those
flowing in opposite directions repel one another.
Effects of Uniform MagneticField
A magnetic field is said to be uniform if its effect on
any device indicating its presence is the same at various
places in the field. Such a field can be set up between
the poles of a magnet. Naturally, the closer to each
other the poles are located and the larger the flat end
faces of the magnet, the more uniform the field.
We have already discussed the effect of a uniform
magnetic field on a magnetic needle and on a current
loop: if there is no counterbalancing spring, then they
locate themselves in the field so that their magnetic
moment coincides with the direction of the field. Their
"north pole" faces the "south pole" of the magnet. This
fact can be expressed in the words: the magnetic moment
becomes aligned with the lines of force of the magnetic
field.
Let us consider the effect of a magnetic field on moving
charges.
It is utterly simple to show that such an effect exists
and that it is no small one. We just take a plain horse-
shoe magnet of the kind used in school physics classes
and bring it near an electron beam produced by an elec-
tron gun. The bright spot on the fluorescent screen is
displaced and moves from place to place as we move the
magnet.
From a qualitative demonstration of this phenomenon
we can pass on to a quantitative investigation. We find
that the force exerted by a magnetic field of magnetic
induction B on an electron travelling at the velocity v
at right angles to the lines of force is
F=evB
Electrons ue
N
Figure 3.3
where e is the charge of the particle (this law is valid,
not only for electrons, but for any charged particles).
When, however, a charged particle travels along a
line of force in a magnetic field, the field hae no effect
on the particle. Readers familiar with trigonometry
should have no difficulty in writing an equation for
the force exerted on a particle travelling at some angle
to the direction of the field. We shall not clutter up the
text with equations that will not be required further on.
But we have not yet said anything about the direction
of the exerted force, though this is of prime impor-
tance. Experiments show that the force is perpendicular
both to the direction of motion of the particle and to the
magnetic induction. In other words, the force is per-
pendicular to a plane passing through vectors v and B.
But this is not all. Every medal has two sides. How do
3. Electromagnetls. 11t
they differ in our case? In the direction we must rotate
one' vector so that it coincides with the other. If we see
the rotation of vector v toward vector B through the angle
less than 180 as being counterclockwise, we are on the
side of the positive normal.
The simple vector diagrams at the left in Figure 3.3
indicate that a positively charged particle is deflected
by the field in the direction of the positive normal.
An electron is deflected in the reverse direction.
Consider now the interesting result that follows from
this law for an electron flying at right angles into a
constant magnetic. field (at the right in Figure 3.3). Try
to figure out the path described..by the electron in the
field. It will, of, course, travel in a circle. The force
exerted by the field is a centripetal force and we can
readily calculate the radius of the circle by equating
mv2lr and evB. Thus, the radius of the path is
mv
T= eB
Note that we could calculate the properties of a par-
ticle from its behaviour. But again we find ourselves
in the same predicament as when we investigated the
motion of particles in an electric field. We cannot deter-
mine the electric charge and the mass of the particle
separatelyI Here as well we determine the ratio elm.
Thus, a particle travels along a circle if its velocity
is directed at right angles to the magnetic field; 8, par-
ticle travels by inertia if its velocity is directed along
the magnetic field. What does it do in the general case?
Your answer is ready, of course. The particle travels
along a helix whose axis is a line of force. The helix
consists of tightly or loosely wound coils, depending upon
the initial angle of entry of the electron into the magnetic
field.
Since a magnetic field acts on a moving particle, it
Electrons ff2
should also exert a force on each piece of wire carrying
a current. Consider a portion of length l of an electron
beam. Assume that there are n particles in this portion.
The force exerted on a wire of the same length, along
which the same number of particles travel at the same
velocity, is equal to nevB. The current is equal to the
total charge passing through the wire in unit time. The
time 't during which the electrons being discussed travel
a path I equals
l
't=-
v
This means that we can write the equation for the current
as
1
- ~ _ n e v
-'t-l
Substituting the velocity from this equation
Il
V=-
ne
into the formula for the force acting on a portion of an
electron beam, we obtain the force exerted on a conductor
of length l, Thus
F= IlB
This is valid only when the wire is perpendicular to the
field.
The direction that a wire carrying a current. is deflected
can be determined by the diagram shown in Figure 3.3.
In deference to the investigators working in the 19th
century, I have included Figure 3.4. As a matter of fact,
this drawing is not of only pure academic interest. It
can serve as an aid in remembering the rule for the deflec-
tion of currents in magnetic fields. The drawing shows how
the field set up by a current (flowing away from us)
3. Electromagnetism 113
Figure 3.'
combines with the external field. The result of this
combination is illustrated at the right. If we conceive
of lines of force as being tensioned ether and having
material proper.ties (a widespread point of view last
century), the direction the conductor isdisplaced can he
visually interpreted: the .conductor is simply pushed out
by the field.
Next we shall demonstrate that the effect of a magnetic
field on a moving charge and on a piece of current-ear-
rying conductor is the same phenomenon with which
we began our discussion of the- effects of a magnetic field.
Let us return to Figure 3.1. It shows the forces acting
on a current-earrying loop. No forces are exerted on the
sections of the wire that are aligned with the field. The
force couple acts on the other two sections. We can see
from the drawing that the moment of this couple is pre-
cisely equal to the product of the force by the arm:
N=IlBd=IAB= MB
Thus the expression for the moment of force as the
product of the magnetic moment of the loop by the
8-2259
Eledrons
Figure 3.5
1 1 ~
magnetic induction follows directly from the equation
for the force exerted on a charge.
The formula F=evB with which we hegan this section
is called the Lorentz equation after Hendrik Antoon
Lorentz (1853-1928), the Dutch physicist who proposed
it in 1895.
Effeds of Nonuniform Magnefic Field
It is quite easy to set up a nonuniform magnetic field.
We can, for instance, impart curved shapes to the faces of
the poles as shown in Figure 3.5. Then the paths of the
lines of force are as illustrated.
Assume that the poles are sufficiently far away from
each other and place a magnetic needle near one of the
poles. As mentioned before, such a needle can move in a
straight line, in the general case, as well as rotate. We
observe rotary motion alone of a needle or current loop
J. Electromagnetism
115
only if the magnetic field is uniform. Both kinds of
motion occur in a nonuniform field. The needle turns so
that it is aligned with the lines of force and is attract-
ed toward a pole (see Figure 3.5). It is pulled into the
region where the field is stronger. (The artist has, of
course, overdone it; it is improbable that even an extremely
strong field can split the compass in two.)
To what is this behaviour due? Evidently, to the Iact
that not only a force couple acts on. the needle in a non-
uniform field. The "forces" exerted on the north and
south poles of a needle placed in a nonuniform field are
not the same. The end located in a stronger region of the
field is subject to a larger force. Therefore, after the nee-
dle turns, the arrangement of the forces is as shown in
the drawing. There is an excess force exerted toward
the side where the field is stronger.
True, a current loop of exceptionally thin wire b e ~
haves in exactly the same way. Thus, in beginning with
a model of a needle with two "poles", my aim was only
to provide a more pictorial representation.
What, then, is the law of nature pertaining to our
phenomena? What is the force equal to? Experiments
and calculations indicate that, for any system having
the magnetic moment M, this force is the product of the
moment of the system by the slope of the curve repre-
senting the increase in the field strength (magnetic in-
duction).
Assume that the magnetic needle aligns itself with a
line of force. The magnetic induction values differ at
the places where the north and south poles of the magnetic
needle are located. \\Te plot a curve of the field along the
line passing through the poles. For the sake of simplicity,
we replace the section of the true field curve between
the poles with a straight line. The shorter the needle, i.e.
the closer its poles to each other, the more accurate our
approximation. The slope of the curve, i.e. the tangent
Electrons
f16
of the angle our straight line on the diagram makes with
the. horizontal axis, is equal to the quotient of the dif-
ference in the values of the field strength divided by
the length of the needle. Thus
F=M ( B ~ - B s )
l
where l = length of the needle; and B
N
and B
s
= field
strengths at the north and south ends of the needle.
(Do not be surprised that the tangent of an angle turns
out to be a dimensional quantity.)
If we replace the fraction by the value of the tangent
of the angle of inclination of the straight line tangent
to the curve representing the field at the point where the
particle interesting us is located, the "poles disappear"
and the formula is valid for any particle or system of
particles.
In conclusion, we can say that in a nonuniform field a
system or a particle with a magnetic moment is attracted
to the poles of the magnet or repelled by them in ac-
cordance with the direction of the magnetic moment:
either along with or opposite to the lines of force.
But can the magnetic moment align itself against the
direction of the field? It certainly canl We shall discuss
below the cases in which this occurs.
Amperlan Currents
Right up to the nineteenth century it was no difficult
matter to advance a physical theory. If a body becomes
heated, that means it contains more caloric. If a medicine
makes you fall asleep sooner, it contains a soporific
power. Certain rods, made of iron ores, point to the
north. A strange behaviour, but immediately under-
stood if we say that such rods and needles possess a mag-
netic spirit. We know that magnetic needles have ren-
3. Electromagnetism tt7
dered good service to seafarers since ancient times.
But sometimes they played tricks. So what; no problem:
evil spirits are to blamel Likewise, it was no surprise
that iron, steel and certain alloys could be magnetized.
These are simply bodies that a magnetic spirit (or soul)
can easily be imparted to.
After the discoveries made by Oersted and Ampere,
it became clear that a bridge could be constructed be-
tween electrical and magnetic phenomena. At one time,
two theories enjoyed equally wide support. From one
point of view, everything became clear if we assume that
the wire along which an electric fluid flows is converted
into a magnet. The other point of view was expounded by
Ampere. He contended that the magnetic spirit of iron
ores consists of microscopic electric currents.
Ampere's point of view seemed to be more logical.
No particular importance was attached to this theory
because, in the first half of the nineteenth century, no
one even dreamt of the feasibility of really discovering
such currents. They even doubted that the world is built
up of atoms and molecules.
But in the twentieth century, when physicists con-
ducted a whole series of ingenious experiments that proved
that the world around us consists of atoms and that atoms
consist of electrons and atomic nuclei, the existence of
Amperian currents was finally taken to be a real fact
that could be employed in an effort to understand the
magnetic properties of substances. Most scientists agreed
that the "molecular currents" proposed by Ampere are
formed by the motion of electrons about atomic nuclei.
It seemed possible to explain magnetic phenomena by
means of these conceptions. As a matter of fact, an elec-
tron travelling around a nucleus can be likened to an
electric current; we have the right to ascribe a magnetic
moment to this system and relate it to the angular mo-
mentum of the moving charged particle.
Electrons 118
It is extremely simple to prove this last statement.
Assume that an electron revolves in a circle of radius
r, Since the current equals the charge carried in unit
time, the revolving electron can be likened to a current
I=Ne, where N is the number of revolutions per second.
The velocity of the particle can be related to the number
of revolutions per second by the equation v=NX 21tr.
Then the current equals
1= .!!!-
21tr
It is natural to call the magnetic moment of an elec-
tron revolving about a nucleus its orbital moment. It
equals
ve 1
M = I A = - 1tr
2
= - evr
21tr 2
We remind the reader (see Book 1) that the angular
momentum of a particle is L=mvr and find that between
the angular momentum and the magnetic moment we
have the following relationship that is of great significance
in atomic physics:
M= ~ L
It follows that atoms must have magnetic moments.
By various procedures, on which we shall not dwell
here, we can obtain the atomic gas of various substances.
Using two sli ts in a gas chamber we can produce beams
of neutral atoms of hydrogen, lithium, beryllium, etc.
They can be passed through a nonuniform magnetic
field and traces of the beam can be observed on a screen.
The question we put to nature is the following: Will
the stream of atoms be deflected from a straight line,
and if so, how?
The atom has an orbital moment and, consequently,
behaves similar to a magnetic needle. If the magnetic
3.,. 1t 9
is directed along the field, the atom is de-
flected toward the region of a strong field. In antiparallel
arrangement of the moment and field, the atom is de-
flected toward the region of a weak field. The amount of
deflection can be calculated by a formula similar to the one
given on p. 116 for the force acting on a magnetic needle.
The first thing that occurs to us is that the magnetic mo-
ments of atoms are arranged at random. If this is so, we
should expect that the trace of the beam is blurred. But
experiments yielded entirely different results. The beam
of atoms is never blurred; it splits into two, three, four or
more components, depending upon the kind of atoms.
This splitting is always symmetrical. Sometimes the
components of the beam include an undeflected beam,
sometimes there is no undeflected beam, and sometimes
the beam does not split up at all.
It follows from this experiment, one of the most im-
portant ever conducted in physics, firstly, that the
motion of electrons about an atom can really be likened
to a closed-circuit electric current. It can be likened
in a narrow and quite definite sense: like closed-circuit
currents, atoms can be ascribed a magnetic moment.
Further, the magnetic moments of the atoms can make
only certain discrete angles with the direction of the
magnetic induction vector. In other words, the projec-
tions of the magnetic moments on this direction are
quantized.
A great triumph for theoretical physics was the fact
that these results had been predicted in elaborate detail.
It follows from theory that the angular momentum and
the magnetic moment of the electroD, due to the motion
of atomic electrons in the. field of the nucleus (these mo-
ments are said to be orbital*), are antiparallel and their
* This name is of historical origin: the development of atomic
theory began with the supposition that an atom resembles the
solar system.
Mz = mJ1;a
and
Electrons
I
projections on the direction of the field can be written fn
the form
I
where m=a whole 'number that can take the values 0,
1, 2, 3, ... ; h/21t=the smallest value of the projection
of the angular momentum; and J.1B=the smallest value
of the projection of the magnetic moment. The values
of k and J.1B are determined from experiments:
h = 6.62 X 10-
27
erg- and 10-
20
erg/Gs
We may add that these important constants of physics
have been named after the great scientists that laid the
foundations of quantum physics: h is called Planck's
constant after the German physicist Max Karl Ernst
Ludwig Planck (1858-1947) and J.1B is the Bohr magneton
after the Danish physicist Niels Henrik David Bohr
(1885-1962).
The postulates of quantum mechanics were not suf-
ficient, however, to provide a comprehensive understand-
ing of the different ways that the beams of various atoms
are split. Even the simplest of atoms, those of hydrogen,
behaved unexpectedly. It became necessary to add another
exceptionally vital hypothesis to the laws of quantum
mechanics. We have already mentioned it once. It con-
sists in ascribing its own (intrinsic) angular momentum,
called spin, and the corresponding intrinsic magnetic
moment to the electron (or to any elementary particle,
as was found later). To understand why it is inevitable
to liken the electron to a magnetic needle, we must
first investigate in more detail the motion of atomic
electrons.
Electromagnetism 12t
Cloud of the Atom
\It is impossible to observe the motion of an electron.
And what is more, we cannot even hope that the advance
of science can ever provide the opportunity of seeing
an electron. The reason is sufficiently clear. To "see"
something, we must first "illuminate" it. But "illumi-
nation" means to subject the electron to the energy of
some kind of ray. The electron is of such tiny mass that
any interference, by means of an instrument for observ-
ing it, inevitably causes the electron to leave the place
it was previously located at.
Not only the meager information about the structure of
the atom which we are about to convey to the reader,
but the whole consistent doctrine of the electronic struc-
ture of matter, is the result of theoretical rather than
experimental investigations. We are, however, sure of
its validity because of the innumerable amount of effects
observed in experiments and rigorously derived from
theory by logical reasoning. We establish the picture
of electronic structure, invisible to us, with the same
degree of assurance that Sherlock Holmes estahlished the
picture of a crime from the clues left by the criminal.
Primarily, a vast source of confidence in this theory
is the fact that the picture of electronic structure is
predicted by means of laws of quantum mechanics that
were established by other experiments.
We have already mentioned that the atomic number of
a chemical element in Mendeleev's periodic table is none
other than the charge of its nucleus Of, what is the same
thing, the number of electrons belonging to a neutral
atom. An atom of hydrogen has one electron, an atom of
helium has two, lithium has three, beryllium has four,
etc.
How do all these electrons travel? This question is far
from simple and its answer is of an approximate nature.
Electrons 122
Difficulties arise from the fact that the electrons interact
with one another and not only with the nucleus. For-
tunately, the mutual repulsion (avoidance) of the elec-
trons plays a smaller role than the motion that is due
to the interaction of the electron with the nucleus. Only
this circumstance enables us to draw conclusions on the
nature of motion of electrons in various atoms.
Nature has allocated to each electron a spatial region
in which it travels. According to their shape these regions
of the electrons are divided into categories denoted by
the letters s, p, d, and [,
The simplest is the "apartment" of the s-electrons.
It is a spherical layer. We know from theory that most
of the time the electron is at the middle of the spherical
layer. Hence, any talk about a circular orbit of such
an electron is a crude simplification.
The region of space through which the p-electron trav-
els is entirely different. It resembles a dumbbell used
for physical exercise. Other categories of electrons have
even more complex regions of existence.
Electronic theory can indicate (though not without re-
sorting to experimental data) how many electrons of each
kind each element of Mendeleev's table contains.
Is this distribution of electrons according to their
types of motion related to their distribution among the
K, L, M, ... energy levels discussed in the preceding
chapter? It is, and most directly. Theory and exper-
iments show that electrons belonging to the K-level
can only be of the s-type; to the L-IeveI, of only the s-
and p-types; to the M-Ievel, of the s-, p-, and d-types; etc.
We shall not discuss the electronic structure of atoms
in any particular detail, restricting ourself to an account
of the structure of the first five elements of the table.
Atoms of hydrogen, helium, lithium and beryllium have
only s-type electrons. The boron atom has four s-electrons
and one p-electron.
3. Electromagnetism 123
The spherical symmetry of the region of space in which
the s-electron travels casts a doubt on our discussion of
the magnetic moment of an atom containing a single
electron. As a matter of fact, since the angular momentum
can take on identical values directed to all sides with
equal probability, the average rotational moment and,
consequently, the magnetic moment of such a system
should equal zero. Quantum physics also reaches this
same natural conclusion: atoms containing only s-elec-
trons cannot have a magnetic moment.
But if this is so, the beams of atoms of the first four
elements of Mendeleev's table should not be deflected in
a nonuniform magnetic field. Is this what we observe?
It was found that these predictions are upset for the
atoms of hydrogen and lithium. Beams of these atoms
behave in an exceptionally strange manner. In both,
the beam of atoms is split into two components deflected
in opposite directions the same distance from the initial
direction. Incomprehensible!
Magnetic Moments of Particles
Electron spin made its first appearance on the scene
in 1925. The necessity of including it in the participants
of the events taking place in the microcosm was first
revealed by Samuel Abraham Goudsmit (b. 1902) and
George Eugene Uhlenbeck (1900-1974). Proposing that
the electron has its own (intrinsic) angular momentum,
these two investigators showed that all the confusion that
had accumulated by that time in the interpretation of
atomic spectra could be readily eliminated by this new
concept.
The experiments mentioned above for splitting atomic
beams were performed somewhat later. When it became
clear that here also only the concept of spin could pro-
Electrons 124
vide a complete explanation of the observed facts, all"
physicists finally accepted it.
A short time elapsed and it was found that intrinsic
angular momentum, or spin, is a property possessed by all
elementary particles and not only electrons.
We have already mentioned that the name spin is
witness to a natural tendency toward visualization.
Since the angular momentum was first introduced in
physics as a property of a rotating solid, certain phys-
icists immediately resorted to the graphic picture of a
particle rotating about its axis when it was found neces-
sary to ascribe a certain value of the angular momentum
to elementary particles in order to save the conservation
laws. This naive concept holds no water: we have no
more right to speak of an elementary particle rotating
about its axis than we would of a mathematical point.
The supporters of visualizability were able to assess
the size of the electron on the basis of certain circum-
stantial evidence. To be more exact, they established that
if this concept is applicable to the electron, its size
must be less than a certain definite value. The value of
spin is known; we shall give it on the following page.
Assuming a shape for the electron, we can calculate with
what velocity "points on its surface" rotate. This veloc-
ity was found to be greater than that of light. Thus,
if we are persistent in advocating such particle rotation,
we must throw overboard the theory of relativity.
Perhaps, the most irrefutable argument against vi-
sualizability is the fact that the neutron, which carries
no electric charge, possesses spin. Why is this argument
decisive? Judge for yourself.
If a particle can be conceived of in the form of a charged
sphere, its rotation about its axis should produce some-
thing resembling an Amperian current. But if a neutral
particle also has angular momentum as well as a magnetic
moment (we shall mention these properties of the neu-
3. Electromagnetism 125
tron briefly in Book 4), any analogy with an Amperian
current is out of the question.
It does not pay, of course, to pose as a prophet and
state that spin and magnetic moment of elementary
particles will never be made clear, even on the basis
of some more general, as yet undiscovered, law. This
problem has been partly solved by the theory of the bril-
liant English physicist Paul Adrien Maurice Dirac (b.
1902). But we cannot give our reader even a general idea
of this theory; it is so abstract. But, as for today, we must
consider the "arrows" representing the angular momentum
and magnetic. moment of a particle to be primary concepts
(not reducible to something simpler).
About fifty years ago, the majority of physicists up-
held the point of view of Einstein, who wrote: "Every
physical theory should be such that it can be .illustrated,
apart from any calculations, by means of simple images."
Alas, this opinion of the great thinker turned out to be
mistaken. For many years, physicists have been calmly
applying theories containing measurable quantities that
we cannot associate with any visual image.
The electron and other elementary particles have no
"poles". In many cases, we confidently speak of them as
point particles, agreeing that the idea of shape is not
applicable to elementary particles. Nevertheless, we are
obliged to ascribe to them two vector properties, an an-
gular momentum (spin) and a magnetic moment. These
two vectors always lie along a single line and are parallel
in some cases and antiparallel in others.
Experiments show that the general formulas for the
projections of the angular momentum and the magnetic
moment, given on p. 120, are also valid for the intrinsic
moments as well. All experiments, both in spectral
analysis and in splitting atomic beams in a nonuniform
magnetic field, can be irreproachably interpreted if quan-
tity m in the formula for the projection of the angular
Electrons 126 ;
momentum of an electron is allowed to take two values:
1/2. As for the formula for the projection of the mag-
netic moment, quantity m can also have two values:
1.
Electron spin has the numerical value (1/2) (h/21t) and
may be disposed only in two directions: along the field
or against the field. As to the magnetic moment of the
electron, it imitates the spin, also having only two
orientations in the field. Its numerical value equals one
Bohr magneton.
Now let us return to the experiments with atomic
beams. We shall show how easily all the specific features of
atomic beam splitting can be explained by the concept
of spin.
Indeed, how can we explain the fact that beams of
helium and beryllium atoms do not split? As follows.
The electrons of these atoms have no orbital moment
because they are of the s "kind". As to the spins of the
electrons, they are in opposite directions. As a matter
of fact, this statement does not follow from anything,
though intuitively seems quite natural. The principle,
according to which a pair of electrons in an atom estab-
lish themselves so that their spins are opposed, is called
Pauli's exclusion principle, after the Austrian-Swiss
physicist Wolfgang Pauli (1900-1958).
So many hypotheses! Yes, quite a few. But taken all
together they form the shapely structure of quantum
physics from which so many consequences follow that
there is not the least uncertainty but that spin must be
ascribed to the electron, that the value of 1/2 must he
given to the spin quantum number and that the spins
of a pair of electrons must obey the Pauli principle. Not
even a single physicist has a shadow of a doubt on these
matters. The sum of these hypotheses represents the
structure of the microcosm.
Let us return to our atomic beams. We have just
3. Electromagnetism t27
explained why no splitting is observed in beam-s of he-
lium and beryllium atoms. .
But why do hydrogen and lithium behave differently?
The hydrogen atom has a single electron. Its orbital
moment equals zero because it is an s-electron. The projec-
tion of its spin can have only two values: plus 1/2 and
minus 1/2, Le, the spin can be aligned either opposite
to or along the direction of the magnetic field. This is
why an atomic beam splits into two components. The
same occurs with lithium atoms because two of its elec-
trons "compensate for" their spin and the third behaves
like the single electron of the hydrogen atom.
The atoms of other elements that have a single un-
paired electron in their outer shell behave in exactly
the same way.
To explain why the atomic beams of other elements
split into a large number of components it would be
necessary to cite several other theorems without giving
their proof. They are proved in quantum physics. By
taking into account the facts that only s-electrons have
no orbital moment and that the spin of an electron is
manifested only when the electron is alone at its energy
level, physicists managed to comprehensively explain
the behaviour of atomic beams of all kinds. After study-
ing this fascinating chapter of physics even the staunch-
est skeptic becomes convinced that all the unproved
assumptions accepted in quantum physics are general
laws of nature.
I fear that many readers may remain unsatisfied by
these statements. Experiments on the deviation of atomic
beams in a nonuniform magnetic field are insufficient in
themselves, of course, to introduce such a strange con-
cept as spin. This book is too small, however, for me to
cite the vast number of facts that demand full rights
for spin among the authentic phenomena of the physical
sciences.
Electrons
128'
For example, how worthy as such proof is the phenom-
enon of magnetic resonances, having nothing in common
with the aforesaid. Radio waves of the centimetre range
are absorbed by a substance if they have to flip (invert)
the spin. No difficulty is encountered in calculating the
energy of the interaction between the magnetic moment
of an electron and the constant magnetic field into which
the substance is placed in magnetic resonance experi-
ments. This energy constitutes the difference between
two energies .(parallel and antiparallel arrangement),
which is equal to a quantum of the absorbed electromag-
netic wave. We can determine the value of the wave
frequency in this experiment with exceptionally high
precision, verifying the fact that it absolutely coincides
with the frequency we calculated on the basis of the
known induction of the field and the value of the elec-
tron'8 magnetic moment.
This phenomenon forms the foundation for a large
branch of science: the study of electron resonance. It
is remarkable that the same events" hut, naturally, in
another wavelength range, are observed for atomic nu-
clei. Nuclear magnetic resonance is the most essential
method used in studying the chemical structure of sub-
stances.
Before going on, it will, perhaps, be helpful to sum up
all the facts concerning systems that set up magnetic
fields and respond to the presence of a magnetic field.
First of all, we must emphasize again that Ampere's
hypothesis was only partly substantiated: magnetic
fields are set up not only by moving electric charges.
Other sources of magnetic fields are elementary particles,
-primarily electrons, which have an intrinsic magnetfe
moment. The technical classification of the kinds of
interaction, given on p. 108, turns out to be inexact.
Magnetic fields are set up, by natural and artificial mag-
nets, by electric currents (including streams of electrical
3. Electromagnetism
Figure 3.6
t29
particles in a vacuum), as well as elementary particles.
The same systems, as well as the particles, respond to
the action of magnetic fields.
The main quantity characterizing a magnetic field
and its effects is the magnetic moment vector. For cur-
rents, this vector depends upon the shape of the current
loop. The moment of a needle is related in a complex
manner to the atomic structure of the substance, but it
9-2259
Electrons
Figure 3.7
t30
can be readily measured. Electrons travelling in the
field of the nucleus have an "orbital" magnetic moment
as if (note, please, this "as if") their motion about the
nucleus generated an electric current. Finally, the in-
trinsic magnetic moment is a primary property that char-
acterizes elementary particles.
Figure 3.6 should help you to remember this funda-
mental information. This drawing represents the sum
of our knowledge today about the "magnetic soul" or,
if you please, the magnetic heart. The French for a mag-
net is "aimant" (from the verb "aimer"-to love). The
drawing underlines the fact that macroscopic current,
a bar magnet, the orbital motion of the electron and the
electron itself are all characterized by a single physical
concept.
Electromagnetic Induction
Experiments show that a beam of electrons travelling
in a magnetic field is deflected from a straight line. As
~ . Electromagnetism i3t
mentioned on p. 109, the force responsible for this deflec-
tion, called the Lorentz force, is directed perpendicular
to the magnetic lines of force and to the velocity vector
of the electrons. It is determined by the formula !i'=evB.
This is the simplest expression for the Lorentz force
and is valid when the direction of the electron velocity
makes a right angle with the direction of the magnetic
field.
If to this fact we add our certainty that a metal con-
ductor contains free electrons, we can, by simple reason-
ing, come to the conclusion that upon certain motions
of a conductor in a magnetic field an electric current
should be produced in the conductor.
This phenomenon, which, we might say, is fundamen-
tal for all modern engineering, is called electromagnetic
induction. We shall proceed to derive its formula.
Illustrated in Figure 3.7 is a current-conducting loop
consisting of rod AC of length l, which rolls along met-
al wires between thelpoles of a magnet without break-
ing the circuit of the loop. If the rod is rolled in the
direction perpendicular to the lines of force, its electrons
are subject to a force and an electric CUITent flows along
the loop circuit.
We reach a conclusion whose importance cannot be
overestimated: an electric current can be produced in a
conductor of a closed circuit even if the circuit contains
no storage battery or other current source.
We can calculate the emf, i.e. the work required to
carry unit charge along the closed loop. Work is the
product of a force by the path. It is performed only on
the part of the loop moving in the field. The length of
the path is l and the force per unit charge is vB.
This developed electromotive force is called the in-
duced emf. Its value is determined by the formula
~ l n d = vBl
Electrons 132
It is desirable to generalize this formula 'So that it
is suitable for any motion of any conducting loops. We
derive this generalization as follows. During time t the
rod-type conductor travels the distance x, its veloci-
ty v being equal to xl, The area of the conducting loop
is reduced by the amount A = xl. Then the formula
for the induced emf takes the form
What does the numerator in this formula signify?
This is sufficiently clear: BA is the amount of change
in the magnetic flux (the number of lines ~ f force) through
the loop.
OUf proof has, of course, been carried out for a very
simple case. The reader will have to take my word that
an entirely rigorous proof can be carried out for any
case. The formula obtained is of most general application
and the law of electromagnetic induction can be formu-
lated as follows: an induced emf is always developed
when the number of lines of force through the loop is
changed. Here the value of induced emf is numerically
equal to the change of magnetic flux per unit time.
There may be movements of a loop in a magnetic
field that induce no current. There is no current when
the loop moves in a uniform field parallel to the lines
of force. If the loop is rotated in a uniform magnetic
field, a current is induced. This also occurs when the
loop is moved away from or toward a pole of a bar mag-
net.
Experiments indicate that our generalization is even
more significant than we have found so far. We consid-
ered cases in which the current loop and the source
of the magnetic field changed their relative positions.
The last formula we derived says nothing about any
3. Electromagnetism
FI,9ure 3.8
133
CURRENT
MAGNETIC
FIELD IS
~ ~ - ~ INTENSIFIED
motion. The only factor is the change of magnetic flux.
But a change in the magnetic flux through a conducting
loop does not necessarily involve a mechanical displace-
ment.
As a matter of fact, we can use, as the source of the
magnetic field, not a permanent magnet, but a loop or,
even better, a coil through which a current from any
outside source is passed. By means of a rheostat or in
some other way we can vary the current in this primary
coil which is the source of the magnetic field. Then the
magnetic flux through the loop changes even though the
source of the magnetic field and the conducting loop are
stationary (Figure 3.8).
Will our generalization be valid in this case as well?
This question can be answered by an experiment, and
the answer is "yes". Regardless of how the number of
lines of force is changed, the formula for the induced
emf, given on the preceding page, is. valid.
Electrons 134
Direction of Induced Current
Next we shall demonstrate that a simple universal rule
exists for the direction of induced currents. Let us consider
several examples from which we shall draw a general
conclusion.
Returning to Figure 3.7 we can note the following.
When we reduce the area of the loop, the magnetic flux
through the loop is reduced. The direction of the current
shown in the drawing is such that the magnetic moment
of this induced current is directed along the lines of force.
This means that the intrinsic field of the induced current
is directed so 8S to "hinder" the reduction of the mag-
netic field . We reach the same conclusiorain the reverse case.
When the area of the loop is increased, the flux through
the ~ < ? ~ P is also increased. But now the magnetic moment
or the loop is directed against the lines of force. Again
we find that the field of the induced current hinders the
action that causes it.
Another example. Assume that the loop is located
between the poles of the magnet in such a manner that
the flux through it is equal to zero. Then we begin to
turn the loop clockwise and counterclockwise. Both
cases are illustrated in Figure 3.9. Solid lines show the
projection of the loop in the initial position; dash lines
show the projections in the turned positions when the
current is induced. Using the left-hand rule, we find
the direction of the induced current in each case. In
our drawing, the north pole is shown at the left. Con-
sequently, when the loop is turned clockwise, the mag-
netic moment of the induced current faces downward;
when the loop is turned counterclockwise, it faces up-
ward. As the angle of rotation is increased, the intrinsic
magnetic field of the loop (in either case) reduces the
field causing the induction more and more. Again the
same rule is valid.
3. Electromagnetism
Figure 3.9
135
Next we shall see how our loop behaves in nonuni-
form fields. Return to Figure 3.8. Assume that the
current of the electromagnet is constant. What happens
when we displace the loop? When we move the loop
toward the north pole, the magnetic moment is direct-
ed against the lines of force. If the loop is moved away
from the pole, the intrinsic field of the induced current
strengthens the field. We can predict such behaviour
by again resorting to the left-hand rule.
How about magnetic fields set up by alternating cur-
rents? Any increase or decrease of the current in the
primary coil leads to a change in flux. An emf is induced
in the loop (see Figure 3.8 again).
How can we determine the direction of the current?
We can no longer use a hand rule because there is no
motion. This is where our generalization comes in handy.
It was found that in this case as well the direction of
the current, induced by reducing or increasing the Dum-
ber of lines of force through the loop, obeys the same
rule: the induced current sets up a field in the direction
that compensates for the change in the magnetic field
causing the induction. ~
Electrons 136
Michael Faraday 11791-1867J-famous English physicist; he dis-
covered electromagnetic induction (in 1831). This was no chance
discovery; Faraday searched for it. Faraday's laws of electromag-
netic induction are the foundation of electrical engineering.
It is difficult to overestimate the importance of the laws of elec-
3. Electromagnetism 137
Discovery of the Law
of Eledrom8gnetlc Induction
The discovery of electromagnetic induction belongs
to those rare events that have a decisive influence on
the progress of mankind. It would therefore be unpar-
donable not to dwell to some extent on the history of
this discovery. It was made a long time before the in-
vestigation of the behaviour of an electron beam in a
magnetic field. The historical course of events does not
coincide in any way with the sequence we chose in the
preceding sections for expounding this subject. Logic
and the continuity of thought from item to item are not
at all obliged to proceed in parallel with the historical
succession of events.
Up to the time when Michael Faraday began his ex-
periments that led to the discovery of electromagnetic
induction, the situation in the science of electric and
magnetic fields was the following.
By this time, the production of a direct current and
the laws of its behaviour in electric circuits posed no
serious problems to physicists. The effects of current
trolysis established by Faraday. This great scientist introduced
and explained such terms, so widely used today, as anode, cathode
anion, cation, ion and electrolyte.
Faraday proved that the intermediate medium influences electri-
cal interaction.
Mention must be made of the discovery of magnetic rotation of
the plane of polarization. The fact that all bodies belong either
to paramagnetic or to diamagnetic materials was also established
by Faraday.
The world had never known a more gifted experimental physicist
than Michael Faraday,
Electrons iSS
on a permanent magnet and the interaction of eurrents
had been established. It became clear that a direct cur-
rent sets up a magnetic field surrounding the con-
ductor. This field can be measured either with a magnet
or by means of another current. This raised the question
of the inverse phenomenon: Can a magnetic field pro-
duce a current in a conductor?
In an entry made in 1821 in his diary, Faraday set
himself the task of corwertmg magnetism into electric-
ity. It took this great scientist ten years to achieve this
aim. He had failed for so many years because he tried to
obtain a current by placing a conductor in a constant
field. But in 1831, his persistent efforts finally succeed-
ed. The quotation given below, from an article writ-
ten by Faraday in 1831, is the first description of this
phenomenon.
"10. Two hundred and three feet of copper wire in one
length were coiled around a large block of wood; other
two hundred and three feet of similar wire were inter-
posed as a spiral between the turns of the first coil, and
metallic contact everywhere prevented by twine. One
of these helices was connected with a galvanometer, and
the other with a battery of one hundred pairs of plates
four inches square, with double coppers and well charg-
ed. When the contact was made, there was a sudden
and very slight effect at the galvanometer, and there
was also a similar effect when the contact with the battery
was broken. But whilst the voltaic current was contin-
uing to pass through one helix, no galvanometric ap-
pearances nor any effect like induction upon the other
helix could be perceived, although the active power of
the battery was proved to be great, by its heating the
whole of its helix and by the brilliancy of the discharge
when made through charcoal."
The discovery of electromagnetic induction was the
first stage in the next twenty years of Faraday'S inves-
3. Electromlgnetlsm tOO
tigations whose aim was to find a unified relationship
between all electrical and magnetic phenomena.
In discussing electromagnetic induction, we should also
mention the names of other prominent physicists. The
American physicist Joseph Henry (1797-1878) discov-
ered self-induction in 1832. If the current in a coil
changes, the magnetic field set up by this current also
changes, thereby changing the flux of the field passing
through this same coil and inducing an emf in its "own"
circuit.
And who discovered the law governing the direction
of the induced emf? The most comprehensive answer
to this question can be found in the works of Lenz. Lenz's
law determines the direction of the induced current.
"If a metal conductor is moved near a current or mag-
net, a galvanic current is induced in the conductor.
The direction of this current is such that if the wire were
at rest, it would begin to move in the direction directly
opposite to the actual movement. It is assumed that
the wire can move both in the direction of its actual
motion and in the opposite direction."
After 1840, a unified concept of electromagnetism was
developed, step by step. The discovery of electromagnet-
ic waves was the last and, evidently, most brilliant
step.
Induced Eddy Currents
If currents can be induced in wire conductors, it would
seem quite natural that they can also be induced in heavy
solid pieces of metal. Each piece of metal contains free
electrons. If the metal moves in a constant magnetic
field, the free electrons are subject to the Lorentz force.
The electrons describe circular paths, i.e, form eddy
currents. This phenomenon was first observed in 1855
by the French physicist Jean Bernard Leon Foucault
(1819-1868).
Electrons 140
The laws of electromagnetic induction are equally
valid whether the magnetic flux changes due to relative
displacement of the metal and the source of the field
or whether the magnetic field changes due to motion
of the electric current setting up the field. Consequently,
eddy currents are induced when the magnetic field
changes with time and not only when there is a relative
motion. The most convincing experiment illustrating the
latter is dropping a coin between the poles of a strong
magnet. It drops as if in viscous oil rather than with
the usual acceleration. The idea of this experiment is
obvious: eddy currents are induced in the coin. Their
direction, according to Lenz's law, is such that their
interaction with the primary magnetic field brakes the
motion that causes the induction.
Among the useful applications of eddy currents are
the following. In the first place, they are used in the
so-called induction furnaces for heating to high temper-
atures and even melting metals. Secondly, they pro-
vide for "magnetic damping" in many indicating instru-
ments.
An ingenious invention (and this is thirdly) is the
electric meter. You have, of course, noticed that its
main component is a rotating disk. The more lights
you turn on or electric appliances you plug in, the
faster the disk rotates.
The principle of this device consists in having two
currents. One is in a circuit parallel to the load and the
other is the load current itself. These currents flow in
coils wound on iron cores. The alternating current mag-
netizes the iron cores. Since we have an alternating
current, the poles of the electromagnets are continually
being reversed. A sort of running magnetic field is set
up between their poles. The coils are arranged so that
the funning field formed by both coils induces eddy
currents in the body of the disk. The direction of these
3. Electromagnetism 141
eddy currents is such that the running magnetic field
pulls at the disk, rotating it.
The speed of rotation depends upon the currents in
both coils. As can be shown by exact calculations, this
speed is proportional to the product of the current by
the voltage and by the power factor (cosine of the phase
angle) Of, in other words, to the consumed power. We
shall not dwell on the simple mechanical transmission
connecting the rotating disk to the counter for indicat-
ing the watt-hours of energy used.
In the majority of cases, however, efforts are made
to eliminate eddy currents. This is one of the concerns
of designers of electric machinery. As all other currents,
eddy currents utilize some energy of the system. These
energy losses may reach such high values that it becomes
necessary to resort to all kinds of contrivances. The
simplest method of combatting eddy current losses is
to replace large solid pieces of metal in electric ma-
chines with laminated sheet stock. In sheet metal, the
eddy currents lack sufficient "elbowroom", their mag-
nitudes are substantially lower and we obtain a corre-
sponding drop in heat losses.
The reader has undoubtedly noticed that transformers
become heated. At least one-half of the heating is due
to the effect of eddy currents.
Inductive Surge
Highly perfected methods of measuring magnetic fields
can be devised by making use of electromagnetic induc-
tion. Thus far we proposed using a magnetic needle for
this purpose, or 8 test loop with a known direct current.
The magnetic induction was determined by the mag-
nitude of the moment of force acting on the test loop
or needle whose magnetic moment is equal to unity.
Now we proceed in. a different manner. After connect-
tnectrons 142
ing a tiny current loop to a measuring instrument, we
locate it perpendicular to the lines of force and then,
with a rapid motion, turn the loop 90 As the loop is
turned, a current is induced in it and a quite definite
quantity Q of electricity, which can be measured, flows
through the loop circuit. In what way is this quantity
of electricity related to the strength of the field at the
point where we located the test loop?
The necessary calculations are sufficiently simple. Ac-
cording to Ohm's law, the current I is the quotient of
the induced emf by the resistance, i.e.
t
1= R ~ i n d
If we make use of the expression for the law of electro-
magnetic induction ~ i n d = BA/'t and recall that Q = !'C,
then the magnetic induction is
iS1nd't ItR QR
B=-A-=-A-=7
We repeat again that this formula is valid, of course,
when in the final position the lines of force do not pass
through the loop, and in the initial position they inter-
sect the area of the loop at right angles. As a matter
of fact, it does not matter whatsoever which we call
the initial and which the final position. Only the direc-
tion of the current is changed, but not the quantity of
electricity flowing in the loop.
The sensitivity of this method of measurement is in-
creased n-fold if we use a coil instead of a single turn
(loop). The quantity of electricity is proportional to
the number of turns n. Good experimental physicists
manage to wind coils only one millimetre in size, enabl-
ing them to examine a field in great detail by the induc-
tive surge method.
Probably the most expedient application of this meth-
3. Electromagnetism t43
od is for measuring the magnetic permeability of iron
bodies. We shall now discuss this important property
of iron.
Magnetic Susceptibility of Iron
We found in the preceding chapter that atoms have
magnetic properties. Single electrons have a magnetic
moment, and orbital magnetic moments are developed
by the movement of electrons about the nucleus. The
nuclei of atoms have magnetic moments. Therefore, when
we put a body in a magnetic field, this must affect the
field in some way. The opposite is also true: the pres-
ence of a magnetic field affects the behaviour of solid,
liquid and gaseous bodies to a more or less extent.
Iron has absolutely outstanding magnetic properties,
as have certain of its alloys and some substances akin
to iron. This small class of substances is said to be fer-
romagnetic. We can, for instance, conduct the fol-
lowing experiments. We suspend a small rod, about the
size of a wooden match, free to turn, by a thread and
bring a magnet near to it. Whatever other substances,
besides iron, we make the rods from: wood, glass, plas-
tics, copper, aluminium, etc., we cannot reveal the mag-
netic properties of these substances by bringing a mag-
net up to them. To prove that any substance has mag-
netic properties, it is necessary to perform precise, care-
ful experiments, which are to be discussed below.
But -iron bodies behave themselves in an entirely dif-
ferent manner. They move obediently to follow' even
the weakest bar magnet found in school physics labo-
ratories.
To convince the reader of the sensitivity of iron
bodies to the presence of a magnetic field, I wish to relate
the following true story, instructive in every sense, of
which I was the main character.
"Electrons 144:
Several years ago I was asked to become acquainted
with the experiments of a Czech "magician" He had
won world fame and was called the "Czech Merlin"
by American reporters who have a weakness for sensa-
tional news. The act of this wizard included several
dozens of experiments that supposedly could not be ex-
plained in any rational way. The Czech Merlin attrib-
uted the results of these experiments to his psychic
powers.
'One of his leading items was to magnetize a wooden
match. First he showed that the wooden match, suspend.
ed by a thread, is not deflected by a magnet. After this, he
began to "hypnotize" the match, making certain mys-
terious passes with his hands. As an indispensable ele-
ment of this performance, he brought the wooden match
into contact with a metal idol which, as Merlin ex-
plained, was the receptor of his psychic energy.
After several weeks of work, I was able to show that
all the experiments, without exception, of this Czech
magician could be rationally explained by known facts
of science. But how did he manage to magnetize the
match? After making the passes and touching the idol
he suspended the match again from the same thread.
Now the match began to obediently follow a magnet
he held and moved around. How could this be?
Merlin's "psychic power" had the following expla-
nation. When he touphed the metal idol, a negligible
amount of fine iron dust was transferred to the end of
the match. I demonstrated that one thirty-millionth of
a gram of iron is sufficient to impart appreciable mag-
netic properties to the match. Here we have another
case of "cockroach experiments".
This striking example demonstrates with sufficient clar-
ity that, in the first place, we should not believe in
"miracles" that contradict the laws of nature, and, sec-
ondly, and this is what interests us at the moment,
3. Electromagnetism f45
that the magnetic properties of iron are particularly
extraordinary.
The classical experiment characterizing the magnetic
properties of iron is performed in the following way.
An electric circuit is wired that consists of two coils,
one inside the other. The primary coil is connected to
a storage battery circuit and the secondary coil is con-
nected to an instrument that measures the quantity of
electricity. If the primary circuit is closed, the magnet-
ic flux passing through the secondary coil varies from
zero to a certain limiting value w
O
This magnetic flux
can be measured to great accuracy by the inductive
surge method.
The magnetic properties of substances are investigat-
ed by means of the device just described. A rod is made
of the substance and is inserted into the coils. The re-
sults of the two measurements, with and without the
rod, are compared. If the rod is made of iron or some
other ferromagnetic material, the quantity of electrici-
ty measured by the instrument is increased by several
thousand times.
The ratio of the magnetic fluxes measured with and
without a rod can be taken as an indication of the mag-
netic properties of the rod material. This ratio J.L = tb/tlJ
o
is called the magnetic susceptibility of the sub-
stance.
Thus, an iron body drastically increases the flux of
lines of force. This can only have a single explanation:
the iron body adds its intrinsic magnetic field to that
set up by the electric current in the primary coil.
The difference c1) - ~ o is usually denoted by the
letter J. Thus, J = (Jl - 1) <1>0 is the additional mag-
netic flux produced by the substance itself.
After we have completed the experiment for measur-
ing the magnetic susceptibility and the rod is pulled
out of the coils, we find that an iron rod retains its mag-
IG-2259
Electrons f46
netization. It" 'is less than J, but is still quite consid-
erable.
The remaining, or remanent, magnetism of the iron rod
can be eliminated. This is called demagnetization and is
done by inserting the rod again into our experimental rig,
but so that the intrinsic field of the metal and the mag-
netic field set up by the primary coil electric current
are opposed. We can always select a primary current
such that an inductive surge in the opposite direction
eliminates the magnetic properties of the iron and re-
turns it to the initial state. For historical reasons that
we shall not go into, the strength of the demagnetizing
field is called the coerci ve force, or coerci vi ty.
This peculiar property of ferromagnetic materials, by
means of which they retain magnetism in the absence
of a current, and the possibility of eliminating this
remanent magnetism by an electric current of the proper
direction, is called hysteresis. What is the origin of this
word? It comes from the Greek word hysteros, meaning
behind or later. But what has this to do with the phe-
nomenon we are discussing? One cannot know beforehand
what the susceptibility J! of a definite piece of iron is
equal to. It depends on previous events-whether the
specimen was magnetized previously, and if" so how strong-
IY. In short, the magnetic permeability depends on the
history of the specimen. If we plot a curve of the mag-
netization versus the magnetizing field strength and re-
verse the magnetizing field, we find that the two S-
curves for the two directions of magnetizing field do not
coincide, with the magnetization lagging behind, and
thereby the curves form a loop, known 8S the hystere-
sis loop. This lagging behind gave the name to the hys-
teresis phenomenon.
Engineering requirements may specify ferromagnetic
materials with various properties. In the magnetic alloy
Permalloy, the magnetic susceptibility J.1 approaches
3. Electromagnetism t47
100 000; the maximum value for soft iron is only one-
fourth as much.
The feasibility of increasing the nux of magnetic
lines of force an immense number of times, by inserting
an iron body into a wire coil, enables electromagnets to
be produced. The capacity of ail electromagnet, i.e. its
capability of attracting and holding iron items of great
weight, increases, of course, with the current passed
through its winding. This process, however, is not unlim-
ited; there is such a phenomenon as magnetic saturation,
though it is not easy to reach saturation when we are
concerned with- heavy powerful magnets.
In recent years exceptionally strong magnetic fields
are being set up by employing a superconductive wind-
ing. Engineers face formidable technical difficulties
in working at extremely low temperatures. But in this
range, we can be sure that we obtain from ferromagnetic
materials all that they are capable of because J.L drops with
an increase in temperature.
In heating, the ferromagnetic properties disappear
when we reach 8 certain temperature limit, for instance,
767 Cfor iron and 360 C for nickel. At this the magnetic
permeability approaches unity as for all other bodies.
This limit was found .in 1895 by the French physicist
Pierre Curie (1859-1906) and is called the Curie point of
the particular substance.
DOm8lns
The main .feature of ferromagnetic materials is their
domain structure. A domain is a region that has been
magnetized to the limit. Inside the domain, all the atoms
are aligned so that their magnetic moments are parallel
to one. another.
The behaviour of magnetic (ferromagnetic) domains is
exactly the same as that of ferroelectric domains in fer-
I.
Electrons
t48
roelectric materials. The linear dimensions of ferromag-
netic domains are not especially small, namely, of the
order of 0.01 mm. Therefore, using a simple contrivance,
domains can be observed in an ordinary micro-
scope.
To render visibility to domains, a drop of colloidal
suspension, consisting of a carefully crushed ferromagnetic
substance of the magnetite type, is applied to the pol-
ished surface of a ferromagnetic monocrystal. The colloidal
particles concentrate along the boundaries of the domains
because the magnetic fields are especially strong at these
places (in the same way that ordinary magnets accumulate
magnetic particles in the regions adjacent to their
poles).
As in ferroelectric substances, the domains in ferro-
magnetic materials exist even when the material is
unmagnetized and not only when an external magnetic
field is applied.
The arrangement of the domains in an unmagnetized
monocrystal is such that the total magnetic moment of
the crystal equals zero. But this does not imply that the
domains are located haphazardly. Again, as an exact
analogy to what was mentioned on p. 64, the nature of
the- crystal structure dictates certain directions in which
the magnetic moments are most easily aligned. Crystals
of iron have a cubic unit cell and the axes of the cube are
the directions of easiest magnetization. In other fer-
romagnetic metals, the moments are aligned along the
diagonals of the cube. In any case, there is complete order
in the arrangement of the domains in an unmagnetized
crystal. There are just as many domains with magnetic
moments pointing in one direction as there are with mag-
netic moments pointing in the opposite direction. Exam-
ples of domain structure have already been illustrated
in Figure 2.5.
Magnetization, . like polarization, consists in the
3. Electromagnetism t49
"devouring" of domains whose magnetic moments are
oriented at an obtuse angle with the field.
The struggle between the tendencies toward order and
toward disorder in atomic arrangement is a necessary
feature of any state of matter. This has been discussed
in detail in another book by the author, published by the
same publishers. The book is called Order and Disorder
in the World of Atoms.
As we found in Book 2 of this series, the tendency
toward order is the tendency to a state of minimum energy.
If there is not much thermal motion, the particles, left
to themselves, form that marvel of atomic architecture.
the crystal. The crystal is the symbol of ideal order in the
world of atoms. The tendency toward disorder is dictated
by the law of degradation of energy, or increase of
entropy.
When the temperature is raised, entropy tendencies
gain the upper hand and disorder becomes the prevalent
form of existence of matter.
In dealing with ferromagnetic materials we have the
following. As the temperature is raised, the magnetic
moments begin to swing from side to side. First these
oscillations keep in time without disturbing the estab-
lished order. Then first one and another atom swivel into
an "incorrect" position. The number of such atoms that
drop out of the ordered ranks continually increases.
Finally, at 8 strictly definite temperature (the Curie
point) magnetic order completely falls apart.
It is difficult in a book of this size to explain why
so few substances have ferromagnetic properties. What
particular details in the structure of their atoms have
put these substances in an exclusive class? I feel that the
reader would be too demanding if he wanted to find an-
swers to all questions in this small book for the layman.
Let us discuss the behaviour of other substances.
Electrons 150
Diamagnetic andParamagnetic Bodies
It has already been mentioned that, with the exception
of ferromagnetic materials, all other substances have a
magnetic permeability very close to unity. Substances
for which J.1 is slightly higher than unity are said to be
paramagnetic; substances for which it is less than unity
are said to be diamagnetic. Examples of both classes of
substances and their magnetic susceptibilities are. listed
in the following table:
I
J.L
! I
IJ
Aluminium 1.000023
I
Silver 0.999981
Tungsten 1.000175 Copper 0.999912
Platinum 1.000253 Bismuth 0.999824
Even though the values differ only slightly from unity,
extremely precise measurements can be made. Generally
speaking, we could use the inductive surge method with
which we began our discussion on the magnetic mea-
surements of the properties of substances. But the most
exact values are obtained by means of a magnetic balance.
A hole is made in one of the pans of a microbalance
(which, as is known, is capable of measuring forces with
an accuracy within one ten-millionth of a gram). Sus-
pended by a thread passing through this hole is the spec-
imen, which hangs between the poles of a magnet.
The pole shoes of the magnet are designed to set up a
nonuniform field. Then the specimen is eith-er attracted
into Of repulsed from t ~ e region with a strong field. It
is pulled in when the magnetic moment of the specimen
tends to align along the field; it is pushed 'out in the
reverse case. The formula for the force is given on
p. 116.
3. Electromagnetism 1St
The specimen is counterbalanced by weights when there
is no magnetic field. When the field is switched OD,
equilibrium is upset. If we are measuring paramagnetic
substances, weights will have to be added to restore
equilibrium. With diamagnetic substances, some of the
weights must be removed. We can readily calculate that
we can cope with our difficult task if we use a good balance
because (in the easily achieved case with a field non-
uniformity of the order of some hundredths of a tesla
per centimetre) the force acting on 1 em" of the sub-
stance will equal about one milligram.
Both kinds of properties, paramagnetic and diamagnetic,
are sufficiently simply explained.
Diamagnetism is the direct result of the fact that in
a magnetic field each electron of the substance describes
a circle. These circular currents develop their own mag-
netic moments directed against the field that causes the
rotation.
Diamagnetism is a property common to all substances'.
Paramagnetism and, to an even greater extent, ferro-
magnetism "suppress" the diamagnetic properties of
substances.
Paramagnetic materials are ones whose atoms or ions
have a magnetic moment. This moment may be due to
orbital motion of the electrons, to spin of a single electron
or to both causes acting together.
Atoms of diamagnetic substances have no magnetic
moment in the absence of a magnetic field. Atoms of para-
magnetic substances have magnetic moments, but, due
to thermal motion, they are arranged in complete disor-
der, exactly like those of ferromagnetic bodies above the
Curie point. When a field is applied, a struggle begins
between the ordering forces of the field and the disorder
introduced by thermal motion. As the temperature drops,
more and more atoms locate themselves so that their
magnetic moment makes an acute angle with the diree-
Electrons f52
tion of the field. This makes it quite clear why the mag-
netic susceptibility of paramagnetic bodies increases with
a drop in temperature.
Earth's MagneticField
People today are accustomed to the fact that any
instrument is devised on the basis of some physical theo-
ry. When the instrument has been developed, engineers
concern themselves with it; physicists have no more to
do. The nature of the phenomenon, on which the princi-
ple of the instrument is based, was well understood even
before it had been developed.
Matters were entirely different in the case of the com-
pass. It was probably developed in China in the 11th
century and was used as the chief nautical instrument
for hundreds of years before somebody really understood
the principle of its operation. Why does one end of the
needle "always point to the north? Most wise men of that
time thought that the behaviour of the needle was due to
extraterrestrial forces, for example, attraction of the
needle tip by the North Star (Polaris).
The brilliant work of William Gilbert (1540-1603)
called About the Magnet, Magnetic Bodies and a Great
Magnet-the Earth was published in 1600. A strictly
scientific approach enabled this English physicist to come
up close to an understanding of magnetic phenomena.
Gilbert shaped a piece of magnetic iron ore, called a
lodestone, into a sphere and thoroughly investigated the
orientation of a magnetic needle suspended over various
parts of the sphere. He found complete analogy with
the orientation of a compass needle at various points
on the earth. He drew the conclusion that the action of
a compass can be excellently explained if we assume that
the earth is a spherical permanent magnet whose axis
is directed along the earth's axis.
3. Electromagnetism i53
\
,
\
TOWARD EARTHS
CENTRE \'"
" E A R T H ~ AX'S
Figure 3.10
From this moment, the study of geomagnetism was
raised to a new level. More precise investigations showed
that a magnetic needle does not point exactly north.
The deviation of a compass needle from the meridian
passing through the given point is called the magnetic
declination. The magnetic pole's are displaced with re-
spect to the earth's axis by 11.5 (Figure 3.10). The needle
is not exactly in a horizontal plane but points downward
(in the northern hemisphere) through an angle, varying
at various latitudes, called the magnetic dip. After mea-
suring the magnetic dip at various points on the earth, we
come to the conclusion that the magnetic "dipole" is
deep within the earth. I t sets up a nonuniform field,
the nonuniformity reaching 0.6 X 10-
4
T at the magnetic
poles and 0.3X 10-
4
T at the equator.
What is this "magnet" inside the earth? The magnetic
"dipole" is in the earth'5 core, which consists mainly of
Electrons i54
molten iron. Even in the molten state iron remains a good
conductor of electricity. On this basis, a model con-
stituting a sort of "magnetic dynamo" has been proposed
to explain the earth's magnetic field . We shall not de-
scribe this model. It is sufficient to point out that the "ter-
restrial magnet" is the result of currents passing through
the molten iron core.
The earth's magnetic field varies. The' magnetic poles
are continually being displaced at. a rate of "5 or 6 km
per year. This is a negligible displacement on the scale
of the whole earth and is a phenomenon that can be ob-
served only in the course of "hundreds of years. This is
why it has been called the secular variation of the earth's
magnetic field.
It is needless to demonstrate how vitally important it
is to have an exact knowledge of all the elements of ter-
restrial magnetism at any point on our planet. The .tnag-
netic compass still serves navigators. This being the case,
they must be furnished with maps showing the magnetic
declinations and dips. Near the poles, as is shown in
Figure 3.10, the north end of a magnetic needle no longer
points toward the north. Near the equator, it is also
difficult to manage without a magnetic map. The mag-
netic equator does not coincide at all with the line of
zero latitude (true equator).
A precise knowledge of the earth's magnetic field is
also of immense interest on land, as well, because it can
be of service in geological surveys. But we cannot dwell
on these problems. Geological physics, or geophysics,
is a significant and extensive field of science and deserves
a special discussion.
We shall devote a few words to so-called paleomagnetic
investigations that give us an idea of the earth '8 magnetic
field in ancient and prehistoric times. These investiga-
tions are based mainly .on a' study of remanent (residual)
magnetization of rock, etc.
3. Electromagnetism t55
The following is the essence of methods employed
for the prehistoric period. Bricks and clay pottery have
a small remanent magnetism that is developed in the
hot clay when it is being fired. The direction of the mag-
netic moment coincides with that of the magnetic field
at the time the item was fired and cooled. Sometimes it is
possible to determine the position of the item during its
manufacture.
Another example of similar investigations is finding the
geographic direction of the magnetic moment of an ore,
its age being determined by the amounts of radioactive
isotopes.
Paleomagnetic investigations are the most rigorous
proof of continental drift. It was found that the magnetism
of iron ore deposits, formed several hundreds of millions
of years ago on the various continents, can be directed
along the lines of force of the earth's magnetic field if the
continents are brought together into a single vast con-
tinent Panganea which split into two supercontinents
Laurasia and Gondwanaland. Later these continents
were further split up, dividing Gondwanaland, for in-
stance, into Africa, Australia, Antarctica and South
America, which then gradually drifted apart.
So far we have mentioned only the intraterrestrial
origin of magnetism, and this is actually its main source.
Certain changes occur, however, in the magnetic field
of the earth due to charged particles arriving from space.
These are chiefly streams of protons and electrons emit-
ted by the sun. The charged particles are carried by the
field to the magnetic poles and are rotated there in a
circle by the Lorentz forces. This leads to two phenomena.
In the first place, the moving charged particles set up
a supplementary magnetic field consisting of magnetic
storms. Secondly, they ionize the molecules of atmospher-
ic gases, producing the aurora borealis, commonly called
the northern lights (this is in the Northern Hemisphere:
Electrons 156
the lights seen in the Southern Hemisphere are called
the aurora australis}. Severe magnetic storms occur peri-
odically (after 11.5 years). This period coincides with
the periods of intensive solar activity.
Direct measurements by means of spacecraft indicate
that the bodies nearest to the earth-the mOOD, and the
planets of Venus and Mars-do not have their own mag-
netic field similar to that of the earth. Of the other plan-
ets of the solar system, only Jupiter and, evidently,
Saturn have their own magnetic fields. A field of a strength
up to 10 Gs and a number of typical phenomena (magnetic
storms, synchrotron radio-frequency radiation, etc.) were
discovered on Jupiter.
MagneticFields of the Stars
Magnetism is found, not only on planets and extinct
stars, but on incandescent heavenly bodies as well. Since
the sun is our nearest star, we know more about its mag-
netic field than that of other stars. The magnetic field
of the sun can be visually observed during solar eclipses.
Particles of solar material that have a magnetic moment
align themselves along the lines of force, thereby sketch-
ing a picture of these lines. The magnetic poles are
clearly seen and the strength of the magnetic field can
be estimated. In regions of a size of the order of tens
of thousands of kilometres, this field is of a strength a
thousand times greater than that of the earth's field.
These regions are called sunspots. Since the spots are
darker than the rest of the sun's surface, the tempera-
ture here must be lower, namely, 2000 degrees below
the "normal" temperature of the sun.
Without doubt, the lower temperature and the stronger
magnetic field are related in some way. But no proper
theory relatin.g these two facts has yet been proposed.
What about the other stars? The advances in astro-
3. Electromagnetism 157
physics have been so amazing in recent years that it is
now feasible to establish the existence of magnetic fields
on the stars. It was found that these "stellar magnetic
spots" have a temperature of about 10 000 C and can
change their position or even disappear entirely in the
course of several months. It is simpler to explain these
changes if we assume that the whole star rotates rather
than have spots change their location.
The presence of magnetic fields is indicated by the
anomalous intensity of certain spectral lines. It would
seem that magnetic stars have an increased iron content
on their magnetic equator.
Magnetic fields in space are very weak (a millionth of-a
gauss). This needs no explanation because an extremely
high vacuum reigns in space. When stars are formed of
atoms scattered throughout the universe, the condensing
of the stellar material is accompanied by a ".condensing"
of the magnetic field. Why, then, do not all stars have
a magnetic field?
The earth has existed for thousands of millions of
years. It follows that the magnetic field of the earth
is continually maintained by the electric currents flowing
in its depths. Certain stars, having no magnetic field,
have evidently cooled to an extent that they no longer
have electric currents inside them. It is improbable,
however, that this explanation is a universal one.
4. Summary
of Electrical Engineering
Sinusoidal Emf
Storage batteries and dry batteries are sources of direct
current. Electricjpower mains, on the other hand, provide
us with alternating current. The words "direct" and
"alternating" refer to the values of the voltage, "emf
and current. If these quantities remain unchanged as the
current passes through a circuit, the current is said to
he constant, or direc t (d-c); if they vary, the current
is said to be alternating (a-c).
An electric current may vary in different ways with
time in accordance with the device producing the cur-
rent. The curve described by the change in electric cur...
rent can be obtained by means of a cathode-, or electron-
beam, tube. The electron ray is deflected by the fields
of two mutually perpendicular parallel-plate capacitors.
Applying various voltages across the capacitor plates, we
can make the luminous spot produced by the ray wander
allover the screen.
To obtain a picture of an alternating current, we pro-
ceed as follows. A so-called sawtooth voltage is applied
across one pair of plates. The curve of this voltage is
illustrated in Figure 4.1. If the electron beam is subject
only to its action, the spot travels uniformly across the
fluorescent screen and then returns with a jump to its
starting point. The position of the spot provides infor-
mation on the instant of time. When the alternating
voltage being investigated is applied across the second
pair of plates, the spot is given an up-and-down motion
as it sweeps across the screen, producing a waving line
4. Summary of Electric.. Englne.rlng
f59
Figu,. ... -
in the same manner that mechanical vibrations could
be made visible by a simple device as shown in Book 1.
It was no slip of the tongue when I mentioned "'vi-
brations" (or "oscillations"). Most of the quantities char-
acterizing an alternating current oscillate according to
the same harmonic or sine-curve law that a pendulum
complies with when deviated from its equilibrium posi-
tion. This can be readily shown by connecting a-c city
mains to an oscillograph.
Either the current or the voltage may be plotted along
the vertical" direction. The characteristics of the current
are the same as parameters of mechanical vibrations.
The interval of time after which the picture of current
change is repeated is called the period T. Current frequen-
cy v, the reciprocal of the period, is usually equal to
50 or 60 cycles per second.
When we examine a single sine curve (as the curve
obtained for alternating voltage or current is called),
the selection of the initial instant of time is of no con-
sequence. But if two sine curves are superimposed, as
illustrated ,in Figure 4.2, we must specify the fraction
of a cycle that they are shifted in phase. The phase is
expressed by the angle cp=21ttYT.. Thus, if the curves are
Electrons 160
Figure 4.2
displaced with respect to each other by one-fourth of a
period, we say they have a phase shift of 90
0
; if by one-
eighth of a period, the phase shift is 45
0
, etc.
When we are discussing several sine curves, shifted in
phase, engineers speak of current or voltage vectors.
The length of the vectors represents the amplitude of
the sine curve, and the angle between them represents
the phase shift. Many engineering devices provide a cur-
rent whose curve is the sum of several sine curves dis-
placed with respect to one another, rather than a simple
sinusoidal current.
We shall show that a simple sinusoidal current is pro-
duced when the conducting loop rotates at constant speed
in a uniform magnetic field.
Upon an arbitrary direction of the rectangular loop
with respect to the lines of force, the magnetic flux pass-
ing through the loop equals
<I> = <l>max sin cp
where cp is the angle between the plane of the loop and
the direction of the field. This angle varies with the time
according to the law <p=2nt/T.
4. Summary of Electrlca' Engineering 161
The law of electromagnetic induction enables us to
calculate the induced emf. Let us write the equations
of the magnetic fluxes for two instants, differing by the
extremely small time interval
= sin t and = sin (t +t)-
The difference between these two equations is
cos (t + ;)sin T)
Since is very small, the following approximate equations
are valid:
(2ft t) 2ft t 21t ( t ) 271:
SID .T2" :::: 12 and C,OS T t +2 t
The induced emf is equal to this difference referred to the
time. Then
[Bind = cos t = (})lDax sin t -
We have shown that the induced emf is expressed by a
sine curve shifted in phase with respect to magnetic
flux sine curve by 90. As for the maximum value of
induced emf, its amplitude, it is proportional to the prod-
uct of the magnetic flux amplitude by the frequency
of rectangular loop rotation.
The law for the current is obtained by dividing the
induced emf by the resistance of the circuit. But we shall
make a crude error if we equate the resistance to alter-
nating current, in the denominator of the equation

J
a
.
c
= - -
n.;
to the ohmic resistance, i.e. the quantity we have dealt
with previously. It turns out that R
a

c
is determined not
only by the ohmic resistance hut also by two more param-
11-2259
Electrons 162
Figure '.3
eters of the circuit: its inductance and the capaci-
tances connected into the circuit.
The fact that Ohm'8 law is more complicated for al-
ternating current than for direct current is demonstrated by
the following simple experiment. Illustrated in Figure 4.3
is an electric circuit with the current passing through an
electric lamp and a coil into which an iron core can be
inserted. First we connect the lamp circuit to a d-e
source. Next we move the core into the coil and pull it
out. No effect whatsoeverl The resistance of the circuit
remains constant and so does the current. We repeat the
experiment for the case when the circuit is connected
to an a-c source. The result is spectacular, is it not?
Now the Iamp burns brightly when the core is not insert-
ed into the coil, and gradually dims as it is inserted.
Thus, at a constant external voltage, and a constant
ohmic resistance (depending only on the mat.erial, length
and cross section of the wire), the current varies with
the position of the iron core in the coil.
What does this mean?
Recall that an iron core drastically increas.es the mag-
netic flux passing through the coil (by thousands of
times). For an alternating. current, the flux varies con-
Summary of Electrical Engineering 163
tinually. But if it varied without the iron core from zero
to some arbitrary unit, with the core it varies from zero
to several thousand units.
As the magnetic flux varies, the lines of force thread
through the turns of their "own" coil. This produces a
current of self-induction in the coil. According to Lenz's
law, the induced current is in the direction that weak-
ens the effect that causes it: the external emf meets with
an obstacle that did not exist when the current was
direct. In other words, an alternating current is subject
to an additional resistance because the magnetic field,
threading through the wires of its own circuit, produces
a back electromotive force, called the emf of self-in-
duction, which weakens the average current. This ad-
ditional resistance is called inductive reactance, to dis-
tinguish it from a true (ohmic) resistance.
Experiments indicate (and this will seem quite natural
to the reader) that the magnetic flux threading through
the coil (or, more generally speaking, the whole
current-earrying circuit) is proportional to the current:
<1>=1. As for the proportionality factor L, called the
inductance, it depends upon the geometry of the circuit
or coil and what kind of cores it encompasses. As is ob-
vious from the formula, the numerical value of the induc-
tance is equal to the magnetic flux when the current
equals one ampere. The unit of measurement of L is
the henry (1 H=t ohm-s).
We can theoretically derive and prove by experiments
that the inductive reactance R
L
is expressed by the for-
mula
R
L
= 21tvL
If the ohmic resistance (with which we are already
acquainted) and the capacitive reactance (which we shall
discuss below) are small, the current in an a-c circuit
equals
Electrons tM
I = ~
RL
To be able to judge what is meant by "small" or
"large", we can calculate the inductive reactance for the
frequency of ordinary city power supply current and an
inductance of 0.1 H. We obtain about 30 ohms.
Now, what would a coil with an inductance of one
henry be like? The following formula, which we give
without derivation, is used to assess the inductance
of coils and chokes (coils with an iron core):
L= ~ A and =4 10-1 joule
flail l 110 1t X amperes-metre
where n = number of turns; l = length of the coil; and
A = cross-sectional area. Thus, an inductance of 0.002 H
can be obtained, for instance, with a coil having the
following parameters: l = 15 em, n = 1500 and A = 1 em",
If an iron core with J1 = 1000 is inserted into this coil,
its inductance equals 2 henries.
An emf of any origin and, consequently, an emf of
self-induction, performs work. This work, as we know,
is equal to ~ l . If the current is alternating, the values
of i and I vary each instant. Let their values equal II
and II at the instant t and It and 1
2
at the instant (t+t).
The magnetic flux threading the turns of a coil of
inductance L is equal to LI. At the instant t, its value
is Ll
1
and at the instant (t+t), it is LI
2
What amount
of work is required to increase the current from lIto
/21 The emf is equal to the change in magnetic flux
divided by the time during which the change took place:
~ = L (12-11)
't
This equation is to be multiplied by the time and the
current, i.e. ~ I t , to obtain the amount of work. But by
4. Sum...." of Electric.' Engineering 165
what current? The average value: (II +1
2)/2.
Hence,
the work of the emf of self-induction equals
L ~ L 2 L 2
""2 (1
2+11)(12-11)=2
[2-21 1
This arithmetical result can be expressed in words: the
work performed by the emf equals the difference in the
values of L[2/2 at two instants of time. This means that
energy is not dissipated by an inductive reactance; it
is not converted into heat as in circuits with ohmic re-
sistance, but is transferred into the "reserve". For this
reason, we can rightly call the quantity 1
2/2,
the mag-
netic energy of the current.
Now let us discuss the effect of introducing a capac-
itor into an alternating current circuit.
If we include a capacitor in a d-e circuit, there is no
current. Introducing a capacitor is the same thing as
breaking the circuit. But the current is not interrupted
by a capacitor in an a-c circuit.
The difference is what interests us; what is it? The ex-
planation is simple enough. When the circuit is connected
to an a-c power source, an electric charge begins to accu-
mulate on the plates of the capacitor. The positive charge
goes to one plate and negative to the other. Assume
that the inductive reactance and ohmic resistance are
low. Charging continues until the voltage across the ca-
pacitor plates reaches a maximum and is equal to the
emf of the power source. At this instant, the current
equals zero. Then the voltage of the source begins to
drop and the capacitor is "discharged".
If we measure the current with some instrument in a
circuit containing a capacitor, we shall find that the
current varies in accordance with two quantities. In
the first place, it can be proved (both by experiments and
by theoretical reasoning) that the current decreases with
the frequency. This means that the capacitive reactance
Electrons 1M.
is inversely proportional to the frequency. This result
is quite natural because the lower the frequency, the mora
alternating current, so to speak, approximates direct
current.
If we change the geometrical parameters of the capac-
itor, i.e. the distance between the plates and their
area, we shall find that the capacitive reactance is also
inversely proportional to the capacitance of the capac-
itor.
The formula for capacitive reactance is of the form:
1
R
c
= 2nvC
At the frequency of city power supply current, I.e,
50 cps (cycles per second), a capacitor with a capacitance
of 30 microfarads constitutes a capacitive reactance
of about 100 ohms.
I do not intend to relate how the total resistance
(called the impedance) is calculated in complex circuits
consisting of ohmic resistances, and inductive and capac-
itive reactances. I only wish to warn the reader that
the impedance is not the sum of the separate resistances
and reactances.
The electric current in a part of a circuit that includes
an ohmic resistance, a capacitor and an inductive coil,
and the voltage across this part can be measured in the
ordinary way by means of an oscilloscope (electron-
beam tube). We see both the current and the voltage on
the screen in the form of sine curves. It should not sur-
prise us to find that these sine curves are shifted with
respect to each other by 8 certain phase angle <p. (The
reader readily comprehends that this must be so if he
recalls that, for instance, the current equals zero in a
circuit with a capacitor when the voltage across the ca-
pacitor reaches its maximum value.)
The value of the phase shift cp is of prime importance.
4. Summary of Electricel Engineering 167
The power transmitted equals the product of the current
by the voltage. If the current and voltage sine curves
coincide, the power has its maximum value. But if they
are shifted as they would be in a circuit including only
ca-pacitive or only inductive reactance, the power equals
zero. This can easily be demonstrated by plotting two
sine curves with a phase shift of 90, multiplying their
ordinates together at each point and adding together
these products over a single period. It can be strictly
demonstrated that the average power per period of alter-
nating current equals
P = IV cos cp
One of the main concerns of the electrical engineer
is to increase cos cp (called the power lactor).
Tl1IllSformers
You have just purchased an electric refrigerator. The
salesperson warned you that the refrigerator is designed
for a mains voltage of 220 V. But in your home the
mains voltage is 127 V. Is this a hopeless situation?
Not at all. You must simply spend a little more and buy
a suitable transformer.
A transformer is a very simple device enabling you to
either raise or lower the voltage. It consists of a laminated
iron core on which two coils have been wound. Each coil,
or winding, has a different number of turns. .
When we plug one of the windings into a socket with the
mains voltage, we find with a voltmeter that the voltage
across the ends of the second winding differs from that
of the power mains. If the primary winding has WI turns
and the secondary, W
2
turns, the ratio of the voltages
is
Vi WI
V
2
= WI
Electrons 168
Thus, a transformer steps up the voltage when the
primary voltage is connected to the winding with the
smaller number of turns, and steps down the voltage in
the opposite case.
Why is this so? The fact is that practically the whole
magnetic flux is within the iron core. This means that
both windings are linked by an equal number of lines of
force. A transformer can operate only with an alternating
voltage across the primary winding. A sinusoidal cur-
rent variation in the primary winding induces a sinu-
soidal emf in the secondary winding. Each turn of the
primary and secondary windings are in identical con-
ditions. The emf per turn of the primary winding is equal
to the emf of the power mains divided by the number
of turns of the primary winding, i.e. Vt/Wt, and the emf
of the secondary winding is equal to the product of
Vt/w
l
by the number W
2
of turns of the secondary winding.
In principle, each transformer can be employed either
as a step-up or a step-down device depending upon the
winding to which the primary voltage is applied.
One frequently deals with transformers in everyday life
(Figure 4.4). Besides the transformers we employ willy-
nilly because some electric appliance is designed for one
voltage and the city mains for another, we may also have
to deal with the spark coil of an automobile. The spark
coil is a step-up transformer. To produce the sparks that
ignite the fuel mixture in the cylinders, we require a
high voltage. It is obtained from the storage battery
of the automobile, after converting its direct current
into alternating current by means of an interrupter (con-
tact-breaker) and then stepping up the voltage with a
spark coil.
It can readily be seen that when the voltage is stepped
up, the current is reduced and vice versa, with an ac-
curacy within the energy losses due to the heating of
the transformer.
4. Sum....ry of Electrical Englne.rlng 169
Figure 4.4
Welding apparatus requires step-down transformers.
An exceptionally heavy current is required for welding
Electrons f70
operations, and the' transformer has only a single out-
put turn.
You have probably noted that the core of a trans-
former is made of thinsheets of steel. This is done to pre-
vent energy losses in converting voltages. As mentioned
previously, eddy currents have a considerably weaker
effect on sheet stock than on solid material.
In the home you deal with small low-power trans-
formers. Powerful tr.ansformers are huge structures. In
them the core with Its windings is inserted into a tank
filled wi th cooling oil.
Machines that Produce Electric Current
Machines that' convert mechanical motion into electric
current were first. developed only some hundred and fifty
years ago.
The first generator of electric current was the machine
built by Michael Faraday and consisted of a rectangular
loop of wire rotating in a field set up by permanent mag-
nets. Soon somebody (but not Faraday) got the idea of
replacing the single loop with a whole coil, thereby add-
ing all the emf induced in all the turns. Only in 1851
were the permanent magnets replaced by electromagnets,
i.e. coils wound on an iron core. Originated with these
devices was the expression "exciting the machine" because
it was necessary first to "vitalize" the electromagnet
before an electric current could be produced. The early
generators were separately excited from some outside
power source.
The next stage in generator development was the dis-
covery of the principle of self-excitation, which elimi-
nated the necessity of a separate power supply to excite
the electromagnets. It proved sufficient to connect the
excitation, or field, winding of the generator in some
manner to the main winding. By the end of the eighties
Summery of Electrical Engineering t7t
Figure '.5
of last century, electric machines had already acquired
the principal features they have today. The simplest
model of a d-e generator is illustrated in Figure 4.5. If
the loop is rotated in a field of permanent magnets, a
sinusoidal emf is induced in it.
If you wish to obtain a direct current from the alter-
nating current in the loop, it will be necessary to equip
the generator with a special device called a split-ring
commutator. This commutator consists of two half-
rings, A and B, insulated from each other and mounted
on a common cylinder (see Figure 4.5). This cylinder
rotates together with the rectangular loop. Applied to the
half-rings are contacts P and Q(called brushes), by means
of which the current is conducted to the external cir-
cuit. Upon each half-revolution of the loop, its commu-
tator half-rings pass from one brush to the other. Hence,
notwithstanding the reversal of the current in the loop
itself, the current in the external circuit has only a single
Electrons
I
172
Mikhail Osipovich Dolivo-Dobrovolsky If862-1919)-distinguished
Russian scientist and engineer; he developed the three-phase
current systems, which are the foundation of all modem
electrical engineering. He worked out all the elements of three-
phase circuits with alternating currents. In 1888 he built the first
three-phase a-c generator with a rotating magnetic field.
Summary of Electrical Engineering t73
Figure '.6
direction. Since the rotating component (rotor) of a real
machine consists of a great number of loops, or sections,
displaced through a definite angle with respect to one an-
other, and the commutator consists of the corresponding
number of segments, we obtain a practically constant emf.
D-e generators are being built today with a power
rating from fractions of a "kilowatt to several thousand
kilowatts. Powerful generators are employed for
the power supply in electrolysis in the chemical in-
dustry and in nonferrous metallurgy (aluminium and
zinc production). They are designed for a large current
and a relatively low voltage (120 to 200 V and 1000 to
20 000 A). Dvc generators are likewise used for electric
welding.
But d-e generators are not the main producers of
electric power. Alternating current with a frequency of
50 Hz (or cps) is applied in the USSR for the production
and distribution of electric energy. An a-c generator,
called an alternator, is designed so that it simultaneous-
ly produces three emf's of the same frequency, differing
in phase by the angle 2n/3.
Such a three-phase generator is shown schematically
in Figure 4.6. In the drawing each coil is represented by
Electrons t 74
a single turn. The wire of" one turn is denoted by C
t
-
c; the second, C
2-C6
and the third, c.c; If the current
enters C
I
, it returns through C., etc. (Of course, at in-
stants corresponding to various relative positions of
the rotor and stator, any of these leads, or wires, may be
points of entry or exit of the current.) The emf in the
stationary turns of the stator is induced because they cut
across the magnetic field of the rotating electromagnet,
the rotor. As the rotor rotates at constant speed, a peri-
odically varying emf is induced in the windings of each
phase of the stator. The three induced emf's are of the
same frequency and differ from one another in phase by
the angle 120
0
as a result of their spatial displacement.
The three turns of the winding can be either star or
delta connected. These circuits were worked out and put
into operation in the nineties of last century by Mikhail
Osipovich Dolivo-Dobrovolsky (1862-1919). In a star
connection, the ends of all three windings of the generator
stator, C4" C. and C
e
, are connected together at a single
point, which is called the neutral, or zero, point. Four
leads connect the generator to the consumer of the ener-
gy: three line wires, from the starts of the windings,
C
I
, C
2
and CSt and a neutral wire from the neutral point
of the generator. This is called a four-wire system.
The voltage between the neutral point and the start
of a phase is called the phase voltage. The voltage be-
tween the starts of the windings is called the line voltage.
These voltages are related by the equation
VI = V3 Vph
If the loads (I, I I and I I I) in the three phases are the
same, the current in the neutral wire equals zero. In this
case, the neutral wire can be done away with, using what
is called a three-wire system. Diagrams of star connections
are illustrated in Figure 4.7.
A delta connection also allows a three-wire systemto be
Summary of Electrical Engineering
LINE WIRE
LINE WIRE
Figure 4.7
t7S
(a)
(6)
employed. Here the end of each winding is connected
to the start of the next one so that the windings form a
closed triangle. The line wires are connected to the ver-
tices of this triangle. Here the line voltage equals the
phase voltage and the currents are related by the equation
11 = V3/
r
h
Three-phase circuits have the following advantages:
more economical energy transmission than in single-phase
circuits, and the feasibility of obtaining two voltages-
phase and line-from a single installation.
Eledrons 176
. The a-c generator, or alternator, described above be-
longs to the class of synchronous machines. These are
machines in which the speed of rotor rotation coincides
with the speed of rotation of the magnetic field set up
by the stator.
Synchronous generators are the main producers of elec-
tric energy. They exist in several design versions, depend-
ing on the method of driving the rotor.
The reader may ask: If certain machines are said to
be synchronous, are there asynchronous machines as well?
Yes, there are! But they are employed as motors and
we shall discuss them in the next section. There we shall
also consider why the magnetic field rotates in a three-
phase a-c machine.
Eledric Motors
More than half of all the electric energy produced is
converted by electric motors into mechanical energy to
satisfy the needs of industry, agriculture, transportation
and the household. The most extensively used is the
simple, reliable, inexpensive, easily maintained asyn-
chronous, or induction, electric motor, invented in 1889
by the same talented Russian engineer Dolivo-Dobro-
volsky. Its main features have been retained up to the
present time. Motors of this type are used for the drives
of various machine tools, pump and compressor installa-
tions, forging and pressworking machinery, materials
handling and transportation equipment and other mecha-
nisms.
The prototype of the induction motor was the model
devised by the French physicist Dominique Francois
Jean Arago (1786-1853). In 1824, Arago demonstrated
before the Paris Academy of Sciences a phenomenon
which he called "magnetism of rotation". He showed
that a magnetic needle suspended over a rotating copper
Summary of Electrical Engineering
if


117
Figure ....
disk rotates with the:disk. The inverse is also true: a
disk mounted on a pivot rotates if it is in the field of a
rotating permanent magnet. This idea was brilliantly
utilized by Dolivo-Dobrovolsky, who combined it with
the features of a three-phase system of currents. This
enabled rotation of the magnetic field to be obtained
without using any additional devices.
Let us examine the schematic diagrams in Figure 4.8.
For the sake of extreme simplification only three turns are
shown (actually, of course, electric machines have wind-
ings with a great number of turns). The cross and black
dot indicate whether the current flows away from us
12-2259
Electrons 178
or toward us in each turn at a definite instant of time.
The. three turns make angles of 120
0
with one another.
Figure 4.8a shows the phase relations of the three cur-
rents it, i
2
and i
3
flowing in the turns. Of interest to
us is the resultant magnetic field set up by the three
"coils". Figure 4.8b shows the lines of force of the resul-
tant field for instant t
1
(current flowing into C
2
, C
a
and
C
4
) . Similar diagrams are shown in Figure 4.8c and d for
instants t
2
and t
s
Hence, as we can see, the field we are
interested in rotates (note the positions of the crosses),
and rotates in the full sense of the word! The axis of the
field in the middle of the system is aligned along the axis
of the turn (phase) in which the current is maximum at
the given instant of time.
The diagrams we have just discussed give an idea of
how a three-phase a-e winding is arranged in the stator
of a three-phase induction (asynchronous) electric motor.
The rotor (Figure 4.9), set into motion by the rotating
magnetic field, is short-circuited, i.e. we see neither the
start nor the end of the winding. Such a rotor, also called
an armature, resembles a squirrel cage. It consists of
copper bars joining two copper rings. Compare it with
a d-e electric machine. How much simpler this isl We
need only to connect a three-phase a-c power supply to
the stator. This sets up a rotating magnetio field in the
machine. The magnetic lines of force of this field link
the rods of the rotor, inducing currents in the rods and
giving the motors their common name. As a result of
the interaction between the rods, along which currents
flow, and the magnetic field, the rotor begins to rotate
at a speed near to that of the field, but does not reach
this speed. This is just as needs be as otherwise the rods
of the rotor would not cut the magnetic lines of force of
the rotating field of the stator, and the rotor would stop.
This is why such machines are said to be asynchronous.
The lag of the rotor is called slip.
4. Summary of Electrical Engineering
179
Figure A.t
Induction electric motors are available in a large pow-
er range: from fracti.ons of a watt to hundreds of kilo-
watts. There are even more powerful induction motors:
up to 6000 kW, operating on a voltage of 6000 V.
Asynchronous micromachines are used in automatic
control systems as actuating mechanisms, or power units,
to convert the input electric signal into mechanical mo-
tion of a shaft. They can also be used as tachogenerators
for converting rotation into an electric signal.
The synchronous machines previously considered and
d-e machines can also be electric motors. This follows
from the obvious principle of convertibility of electric
machines: any electric machine can operate either as
a4 generator or as a motor.
For example, the Kiev hydraulic power system on the
Dnieper River includes a hydraulic accumulator station
equipped with convertible units that can operate either
as pumps or as turbines. When there is excess electric
energy in the power system, the units pump water into
the storage basin. At this time, the synchronous machines
coupled to the units operate as drive motors. Upon peak
consumers' load, the units are operated by the stored
water as turbines and generators.
IU.drons faO
Synchronous motors are used in metallurgical plants,
mines and refrigerators to power pumps, compressors,
fans and other mechanisms that operate at constant
speed. Miniature synchronous motors are extensively ap-
plied in automatic control systems. They have a rating
from fractions of 8 watt to several hundred watts. Since
the speed of these motors is rigidly coupled to the fre-
quency of the power supply mains, they are most expedi-
ently employed where a constant speed of rotation is
required. Such applications include: electric clock
mechanisms, tape-moving mechanisms of recording instru-
ments, film-advancing devices of motion-picture projec-
tors and cameras, radio apparatus, programming devices,
as well as systems of synchronous communication, where
the speed of rotation of the mechanisms is controlled
by varying the frequency of the supply voltage.
In the principle of its design, a d-e motor in no way
differs from a d-e generator. The machine has a stationary
system of poles whose excitation (field) winding is con-
nected in one or the other way with the armature winding
(either in series or in parallel). The machine can also
be separately excited from a power supply. The armature
has a winding properly distributed in its slots and con-
nected to a source of direct current. D-c motors like
d-c generators have a commutator whose purpose is to
"straighten out" the torque, I.e, to make the machine,
if it is a motor, rotate continuously in one direction.
Series-wound d-e electric motors (with the field winding
in series) are especially suitable for electric traction,
cranes and hoisting machinery. Such installations re-
quire the speed of the drive motor to drop drastically
at heavy loads and its torque to increase substantially.
Series-wound d-e motors are distinguished particularly
for these features.
The first experiments with electric traction in Russia
were conducted by Fyodor Apollonovich Pirotsky (1845-
4. Summary of Electrical Engineering 181
1898). As far back as 1876, he adapted ordinary railway
tracks for transmitting electric energy, and in 1880 he
started to run an electric streetcar on an experimental
line of the horsecar railway near Rozhdestvensky Park
in St. Petersburg. As cars for the first streetcar line, he
used the double-decker horsecars, mounting an electric
motor under the body.
The first electric trolley-car line in Russia was opened
in Kiev for the general public in 1892. Its electric motor
was supplied from an overhead contact wire. The munic-
ipal construction committee agreed to build the street-
car line only after they had been convinced by calcula-
tions of the technical advantages of electric traction
over horse traction on the steep hilly streets of Kiev.
These streets proved to be beyond the capacity of either
horse or steam traction.
The first experiments in "electric navigation" were
carried out in 1838 by Boris Semyonovich Yakobi (1801-
1874). He demonstrated an electric boat, holding fourteen
passengers, on the Neva River. It was powered by a
550-watt electric motor. Yakobi supplied the motor from
320 galvanic batteries. This was the first time in history
that an electric motor was used for traction.
The term "turbo-electric ship" appears more and more
often in the press in recent years. It simply means that
steam is used to drive powerful d-e turbogenerators (tur-
hodynamos) and that the screws are mounted on the
shafts of electric motors. But isn't this too complicated?
Why not mount the screws directly on the turbine shafts?
The point is that a steam turbine generates maximum
power only at strictly definite speeds. Powerful turbines
run at 3000 rpm. If the turbine is slowed down, the gen-
erated power drops. If screws were mounted directly
on the turbine shafts; a ship powered in this way would
have poor seafaring properties. A d-e electric motor, on
the other hand, has an excellent traction performance
Electrons 182
curve: the higher the forces of resistance, the higher the
tractive effort it develops. Moreover, such a motor can
develop higher power at low speed at the moment the
ship is just getting under way.
Thus a d-e generator and a d-e motor, arranged between
the turbine and screw of a turbo-electric ship, operate
like an automatic stepless (infinitely variable) gearbox,
developed to a high state of perfection. It may seem that
such a system is somewhat bulky, but at the high power
ratings of modern turbo-electric ships, any other would
take up just as much space and would be less reliable.
The power installation of a turbo-electric ship can
be substantially improved in an entirely different way.
It proves highly efficient to replace cumbersome steam
boilers by a nuclear-power reactor. There is a huge saving
in the volume of the fuel required for each voyage. The
well-known Lenin nuclear-power icebreaker of the Soviet
arctic fleet is the first of its kind. The nuclear-power
installation of this turbo-electric ship provides for voy-
ages as long as a year without refueling.
D-e electric motors are installed on main line electric
locomotives, suburban electric trains, streetcars and trol-
ley buses. They are supplied by energy from stationary
electric power stations. In the USSR, electric traction is
powered by either direct current or by single-phase al-
ternating current with a commercial frequency of 50 Hz.
Silicon rectifiers are being widely used in the traction
substations of streetcar, trolley bus and subway lines.
As to electric railways, the current may he rectified
either in substations or in the electric trains themselves.
5. Electromagnetic Fields
M8xwell's Equations
By the fifties of last century, much information had
accumulated on electricity and magnetism. It was dis-
connected in the main, however, sometimes contradictory
and, in any case, did not fit into a single orderly system.
But a great deal was known. In the first place, physi-
cists knew that electric charges at rest set up an electric
field; secondly, that electric currents set up magnetic
fields; and, thirdly, the results of Faraday's experiments
were published and generally recognized. In them he
proved that a variable magnetic field generates an elec-
tric current.
In those times, a number of scientists, and primarily
Faraday, doubtlessly became convinced that some kind
of events occur in the space surrounding electric currents
and charges. This group of investigators supposed that
magnetic and electric forces are transmitted from point
to point. Attempts were widely made to draw diagrams,
similar to a system of meshing gears, which could visually
demonstrate the mechanism of the transmission of electric
energy. But certain men of science advocated the theory
of long-range interaction. They denied the physical pro-
cess of transmitting electric and magnetic forces. They
said that the concepts of a field and lines of force are
to be treated as only geometric images that do not re-
present any possible kind of reality.
As has frequently occurred in the history of science,
the truth turned out to be somewhere in the middle.
Attempts to reduce electromagnetic phenomena to the
motion of a special kind Qf matter-the ether-were
Electrons 184
James Clerk Mlxwen 1ill1-1179J-famed Scottish physicist;
he was the founder of theoretical electrodynamics. Maxwell '8
equations describe the behaviour of electromagnetic waves and an
electromagnetic field regardless of its origin. Maxwell is the author
of the electromagnetic theory of Iight. The theoretical value of
the velocity of light follows automatically from his equations,
5. EledromBgnetic Fields t85
found to be unsound. But, on the other hand, investiga-
tors that proposed that electromagnetic interaction is
transmitted instantaneously from one charge or current
to another were also wrong.
The famed Scottish physicist James Clerk Maxwell
(1831-1879) published a work called On Faraday's Lines
of Force when he was only 26. As a matter of fact, this
work already contained the laws he discovered. Several
more years were required, however, for him to discard
his mechanistic conceptions and to formulate the laws
of an electromagnetic field in a form that requires no
naive graphical illustration.
In this connection Maxwell once said that for the
benefit of people with different kinds of minds scientific
truth should be presented in various forms and be regard-
ed as equally scientific whether "it is presented in the
clear form and lively colours of physical illustration or
in the simplicity and pallor of symbolic expression.
Maxwell's laws belong to the fundamental general laws
of nature. They are not derived by logical reasoning and
mathematical calculations. Fundamental laws of nature
are generalizations of" our knowledge. Laws of nature
are discovered, found or ascertained. It is a matter of
great interest to historians of science and to psycholo-
gists to follow out the' chain of guesses and creative
perception by means of which a genius discovers a law
of nature. This, however, is a subject for a special book.
Maxwell '8 theory is the basis for the relationship between the per-
mittivity and the refractive index, the orthogonality of the elec-
tric and magnetic vectors in a wave, and the existence of light
pressure.
Of no less importance was Maxwell's contribution to the kinetic
theory of gases. Maxwell worked out the velocity distribution law
for gas moleculest
Electrons
As for the reader and myself, nothing is left but to
a certain outline of consecutive assumptions that lead-
to Maxwell's laws.
What was known to Maxwell when he set himself the
task of expressing in concise symbolic form the laws
governing the behaviour of electric' and magnetic fields?
In the first place, he knew that any point in space in
the vicinity of an electric charge can be specified by a'
vector of electric force (intensity), and any point near
an electric current, by a vector of magnetic force.
But are stationary electric charges the only source of
an electric field? And are electric currents the only source
of a magnetic field?
Maxwell answered these questions in the negative and
proceeded in his search for the laws of an electromagnetic
field along the following series of guesses.
Faraday showed that an electric current is induced
in a wire loop which is threaded by an alternating flux
of magnetic lines of force. But a current is produced
when an electric force acts on electric charges. This being
so, we can express Faraday's law by the following phrase:
an electric field is set up in a wire loop through which
an alternating magnetic flux passes.
But is it essential that the magnetic flux is encom-
passed by a wire loop? Is it not all the same to the electric
field where it is set up: in a metal conductor or in empty
space? Let us assume that it is all the same. Then the fol-
lowing statement is valid: a closed electric line of force
appears near a variable flux of magnetic lines of force.
Thus the first two of Maxwell's laws concerning an
electric field have been formulated. We contend that
an electric field is set up in two ways: by electric charges
(in which case the lines of force start at positive and
end at negative charges) and by a variable magnetic
field (in which case the electric line of force is closed
and encompasses the variable magnetic flux).
5. Electromagnetic Fields 187
Next we attempt to find the laws concerning a magnetic
field. A magnetic field is set up by currents; this Max-
well knew. Direct (constant) current is the source of a
constant magnetic field, and alternating (variable) cur-
rent sets up a variable magnetic field. But we know that
an alternating current is produced in wire by a variable
electric field. And what if there is no wire, but only a
variable electric field existing in empty space? Is it not
logical to assume that a closed magnetic line of force
is produced near a variable flux of electric lines of force?
Its symmetry makes this picture very attractive: a vari-
able magnetic flux sets up an electric field, and a variable
electric flux, a magnetic field.
Thus, the two laws concerning electric fields are supple-
mented by two more that govern the behaviour of mag-
netic fields. That a magnetic field has no sources (there
being no magnetic charges) is the third law, and that
a magnetic field is produced by electric currents and
a variable electric field is the fourth law.
Maxwell's four laws can be written with exceptional
elegance in the form of mathematical equations. With
respect to them and quoting a line from Goethe, the
great Austrian physicist Ludwig Eduard Boltzmann
(1844-1906) wrote: "Was it a god who wrote these lines...."
Unfortunately, I cannot acquaint the reader with the
meaning expressed by these equations. This would re-
quire from him a knowledge of mathematics at a level
above that used in this series (Figure 5.1).
Maxwell '8 laws indicate that a variable magnetic field
cannot exist without an electric field, and a variable
electric field, without a magnetic field. This is why the
two adjectives, electric and magnetic, are not separated
by a comma. An electromagnetic field is a single entity.
As we move away from the charges that are the sources
of an electromagnetic field, we deal with electromagnetic
matter, so to speak, in the pure form. It is not necessary
Electrons
Figure S.t
188
divD=4.1CE
curl E=-lQ1 }
div B=oc in
curLH= j+J
---:/
~ ~ ~
to consider pencils of lines of force. Maxwells laws can
be written so that they are applicable to a point in space.
Then they sound especially simple: at a point where the
electric vector varies with time, there exists a magnetic
field vector that also varies with time,
"But is not all the aforesaid pure imagination?" asks
the reader. In practice, as a matter of fact, it is abso-
lutely impossible to measure the magnitudes of rapidly
j. Iledi'omaQn6tlc Fields t89
varying electric and magnetic field vectors at a point
in space.
Truel But the grandeur of laws of nature is assessed
by the consequences that follow from them. The conse-
quences, in this case, are legion. It is no exaggeration
whatsoever to contend that all of electrical and radio
engineering is contained in Maxwell's laws.
It is imperative, however, to tell about one vital con-
clusion based on Maxwell's equations. It can be shown
by faultlessly rigorous calculations that electromagnetic
radiation must exist.
Assume that there are charges and currents in some
bounded region of space. Various energy transformations
may occur in this system. Mechanical or chemical sources
produce electric currents, and the currents, in their turn,
drive mechanisms and evolve heat that is liberated by
the wires. Now let us calculate all the debits and credits.
But they do not tally! The calculations show that some
fraction of the energy of our system has escaped into
space.
Has theory anything to say about this "radiated"
energy? It seems that it has. The solution of the necessary
equation is of quite complex form near the source, but
at distances substantially exceeding the size of the "ra-
diating" system, the picture becomes highly clear-cut
and, what is most important, lends itself to experimental
proof.
At great distances electromagnetic radiation, as we
call the energy deficiency in our system of moving charges,
can be characterized at each point in space by its
direction of propagation. In this direction the electro-
magnetic energy travels at a velocity of about 300 000
km/s. This figure follows from the theoryI The second
conclusion of the theory is that the electric and magnetic
vectors are perpendicular to the direction of wave propa-
gation and perpendicular to each other. Thirdly, the
Electrons 190
Intensity of electromagnetic radiation (energy per unit
area) decreases inversely proportional to the square of
the distance.
Since it was known that light travels with about the
same velocity that had been calculated for electromag-
netic radiation, and sufficiently complete information on
the polarization of light had been acquired to indicate
that luminous energy has some "transverse" properties,
Maxwell came to the conclusion that light is a form of
alectromagnetic radiation.
Ten years after Maxwell's death, at the end of the
aighties, the famous German physicist Heinrich Rudolf
Hertz (1857-1894) experimentally substantiated all the
conclusions from Maxwell's theory. Following these ex-
periments, Maxwell's laws and their equations were firm-
ly established forever and ever as one of the few cor-
nerstones supporting the structure of modern physics.
Mechanical Modelsof Radiation
Mechanical models are contrasted with mathematical
models. Mechanical models can be contrived by using
balls, springs, strings, rubber cords, etc. A mechanical
model helps to make some phenomenon visual. By build-
ing a mechanical model and demonstrating its operation
we help a person to understand the phenomenon by say-
ing: "Such and such a quantity behaves similar to such
and such a displacement." Far from all mathematical
models can be contrasted with a mechanical one.
Before we take up electromagnetic radiation, a fact
established by innumerable experiments and following
by strict logic from Maxwell 's equations, we should
discuss the feasible mechanical models that can be con-
trived for radiation.
There are two such models: corpuscular and wave.
We can make a toy which "radiates" streams of tiny
5. Electromlgnetlc Fields t91
particles-peas or poppyseed-in all directions. This is
the corpuscular model because the word "corpuscle" means
a particle.
A particle flying at a certain velocity and having a
certain mass should conform to the laws of mechanics.
Particles are capable of colliding, changing their directions
of motion, but only if the collision obeys the laws of con-
servation of energy and momentum. Some bodies may
prove impenetrable to the particles. In such cases the
particles must bounce off the bodies acoording to the
law: the angle of incidence equals the angle of reflection.
Particles can be absorbed by the medium. If particles can
travel easier in one medium than in another, we can
readily explain refraction. After passing through a hole
in an opaque screen, a stream of particles emerging from
a point source should travel within a cone. True, a small
amount of scatter is possible because a small fraction
of the particles are reflected by the edges of the hole.
Of course, these reflected particles can only be chaotic
and cannot produce any regular pattern that extends
outside the geometric shadow.
The wave model is usually demonstrated by a water
bath. The water can easily be made to oscillate at some
point. From this point, as from a stone dropped into
water, concentric waves spread outward in ever-widening
circles. We can see the wave-like surface of the water.
Energy is propagated in all directions and a wooden
chip, floating at some distance, begins to oscillate with
the frequency of the point to which we supply energy.
Sound vibrations are somewhat more difficult to see.
But we can conduct absolutely convincing experiments
to show that sound is transmitted by the mechanical
displacement of the medium from point to point.
A great many phenomena are equally well explained
by both the wave and corpuscular models. Both modeJs
are equally suitable, however, only when an additional
Electrons f92
condition is complied with: a wave acts just like a stream
of particles if the obstacles and holes it encounters are
much smaller than the wavelength.
As we can readily calculate from the main formula,
used to describe the wave model, c = VA, the average fre-
quency of a human voice of 1000 Hz corresponds to a
wavelength of 30 em. Such a wave can go around a corner
if it has to pass through holes one metre in size. But if
the hole has a size of the order of one centimetre, we can
contend that a sound beam passes through it only when
a straight line connecting the source and receiver of the
sound meets no obstacles.
Suppose that a radio set is speaking in a room with
open windows high up on the wall. A man sitting outside
on a bench under the window can hear what is going
on. If the windows are tightly closed and the walls
are thick, the sound can pass only through the keyhole
of the door. Now, even the most sensitive detector re-
ceives a signal only if the source of the sound, the keyhole
and the detector are on a straight line. Here sound
energy is transmitted like a stream of particles.
It can readily be shown by both reasoning and experi-
ments in a water bath that the law of reflection from a
wall of a roughness less than the wavelength is complied
with by the wave model as well.
The reader knows quite well how sound or any other
wave is reflected by a smooth plane surface. Interesting
problems may be posed when the reflecting surface is
of curved shape.
One such problem follows. What kind of surface is
required to collect a wave, emit.ted by a point source,
back into a single point? The shape of this surface must
be such that rays from one point, incident to the surface
at various angles, are all reflected to another single point.
What is this surface?
The reader may recall one remarkable property of the
5. Electrom.gnettc Fields t93
curve called an ellipse. The distance from one focus of
the ellipse to any point of the curve plus the distance
from this point to the other focus is the same for all
points on the ellipse. Now imagine that the ellipse is
rotated about its major axis. The rotating curve describes
a surface which is said to be an ellipsoid. (Its shape
resembles a symmetrical egg.) Another property of an
ellipse is the following. If an angle is drawn with its
vertex at a point of the ellipse ~ a n d its sides passing
through the foci, the bisector oflthis angle is normal
to the ellipse at the point. This means that if a wave
or a stream of corpuscles is emitted at one focus of the
ellipsoid. it is reflected by the internal surface and arrives
complete only at the other focus.
As far as sound waves are concerned, the walls and
ceiling are sufficiently smooth. If the ceiling is vaulted,
a special case of sound reflection can be observed. Since
the shape of the arched ceiling is close to that of an ellip-
soidal surface, sound emitted at one focus is reflected
by the ceiling to the other focus. This property <?f vaulted
ceilings was known in ancient times. In the Middle Ages,
during the inquisition, it was made use of to eavesdrop.
Two persons, disclosing their thoughts to each other in
a quiet voice had no idea that they were being overheard
by the dozy monk sitting at the other end of the tavern
(Figure 5.2).
The corpuscular and wave models are equally suitable
to explain this phenomenon. But a phenomenon like
the collision of billiard balls cannot be explained by
the wave model.
There are, on the other hand, several vital facts that
the corpuscular model cannot cope with.
Primarily, this concerns interference, i.e. a summation
in which the sum may turn out to be less than the addends
or even equal to zero. If two waves arrive at a point and
are added together, of prime importance is their differ-
lS-2259
Electrons
t ~
Figure 5.2
ence in phase at this point. If the crest of one wave
coincides with the crest of the other, the waves add up.
But if the crest of one wave coincides with the trough of
the other and their amplitudes are the same, addition
bas a zero result. The waves arriving at a single point
cancel out each other. "Then two wave fields are superim-
posed, at certain places the waves are added together and
at others they are subtracted one from the other. This
is interference. This is the first phenomenon that is ab-
solutely impossible to deal with using the language of
streams of particles. If radiation behaves like a stream
of peas, then the superimposed fields must intensify each
other at all places and at all times.
The second important phenomenon is diffraction, I.e.
bending around corners, or passing around obstacles. A
stream of particles cannot act in this way, but a wave
s. Electromagnetic FI.1eIs i95
must behave in exactly this way_ In school, diffraetion
is "demonstrated by generating waves in a small bath
of water. Then a partition with a hole is put into the
path of the waves and the bending around corners becomes
visible to the naked eye. The cause of this phenomenon
is entirely natural. The particles of water begin to vi-
brate in the plane of the hole. Each point lying in the plane
of the hole produces a wave, all the points having the
same rights as the initial source of radiation. Nothing
hinders this secondary wave from "bending around a
corner".
Interference and diffraction can be demonstrated with-
out difficulty if the condition mentioned above is com-
plied with: the wavelength must be greater than or at
least commensurable with the size of the obstacles or
holes. We shall refine this condition and deal with dif-
fraction and interference in more detail in Book 4.
Next we shall discuss the change in the frequency of
waves observed when the source of radiation is in motion.
That this phenomenon is a necessary consequence of the
wave model was shown by the Austrian physicist Christian
Johann Doppler (1803-1853) in the early days of theo-
retical physics.
We shall derive the equation of the Doppler effect,
as it is called; it shall prove useful later. Assume, for
the sake of clearer representation, that an automobile is
approaching a marching brass band. The number of peri-
odic compressions of the air reaching the driver's ear
per unit time is greater, than if the automobile were
at rest, by the ratio (c +u)/u, where c is the velocity of
propagation of the waves and u is the relative velocity
of the source of the waves and their detector (the driver
in our case). Hence
IS
Electrons 196
This means that the heard frequency v' is higher when
the automobile and band approach each other (the pitch
is higher when u > 0) and drops when they are moving
away from each other (the pitch is lower and u< 0).
Getting a little ahead of our story, we can say that for
light waves this conclusion is the following: a "red shift"
occurs when the source is receding. The reader will ap-
preciate the importance of this conclusion when we dis-
cuss the observation of the spectra of distant stars.
From a long time in the past and down to the twenties
of our century, philosophers and scientists have often
argued whether some case of energy transmission is of
a wave or a corpuscular nature. Experiments indicated
that any kind of radiation has two aspects. Only a com-
bination of these two aspects properly represents reality.
Theory has elevated this fact to the rank of a principal
law of nature. Wave mechanics, quantum mechanics and
quantum physics are equivalent names for the modern
theory describing the behaviour of fields and particles.
Two Asped's of an Electromagnetic Field
In certain phenomena electromagnetic radiation be-
haves like waves and in others like a stream of particles.
In this sense, Maxwell '8 laws are prone to "error";
they only outline the wave aspect of electromagnetic
radiation.
In complete agreement with experimental investiga-
tions, a solution of Maxwell's equations leads to the
conclusion that we can always conceive of electromag-
netic radiation as being the sum of waves of various
lengths and intensities. If the radiating system is an
electric current of fixed frequency, the radiation consists
of a "monochromatic" (single-colour) wave.
An electromagnetic wave is illustrated in Figure 5.3.
To understand the changes that take place in space in
5. Electromagnetic Fields
197
Figure 5.3
the transmission of electromagnetic energy, we must "pull"
our figure as a rigid whole along the axis of abscissas.
This picture is obtained as a solution to Maxwell '8
equations. It is what enables us to speak of electromag-
netic waves. In employing this term and in resorting
to the analogy between an electromagnetic wave and
one spreading outward in water when we drop a stone,
we must proceed with extreme caution. Pictorial repre-
sentations can easily mislead us. A wave on water is
only a model of an electromagnetic wave. This means
that electromagnetic waves and waves on water behave
the same only in certain respects.
But are not Figure 5.3, illustrating an electromagnetic
wave, and a water wave, which alternately raises and
lowers a wooden chip thrown into the sea, alike as two
peas? Nothing of the kind! Consider carefully the essence
of the drawing. Plotted along the vertical axis are the
electric field vectors and not some kind of spatial dis-
placements!
The magnitude of the ordinate at each point indicates
the force that an electric charge would be subject to if
Electrons 198:
it were placed at that point. Strictly speaking, nothing
is moved from its place during the travel of an electromag-
netic wave. It is absolutely impossible in practice to
conduct an experiment showing the variation in value of
an electromagnetic wave at some point, even for very
slow vibrations.
Hence, the concept of an electromagnetic wave is of
a theoretical nature. We confidently speak of the existence
of electromagnetic waves because we listen to the
radio. We have no doubt that an electromagnetic wave
has a certain frequency because we must tune our radio set
to a definite frequency for the reception of a definite
broadcasting station. We are sure that the concept of
length is applicable to an electromagnetic wave. Not
only can we measure the wave velocity and calculate
the wavelength by formula c = VA, which relates the fre-
quency, wavelength and velocity of propagation, but we
can obtain an idea of the length of electromagnetic waves
from diffraction, i.e, bending around a corner or an
obstacle. The principle involved in this last kind of mea-
surement is the same as for waves travelling on wa-
ter.
It is absolutely necessary to warn the reader that he
should not attempt to work out a visualizable conception
of an electromagnetic wave because, as mentioned at
the beginning of this section, electromagnetic radiation
not only "resembles" waves, but in many cases "reminds"
us of the behaviour of a stream of particles. To try to
imagine something that simultaneously resembles a stream
of particles and a wave is entirely contrary to reason.
We are discussing physical processes that cannot be re-
presented by chalk on a blackboard. This does not imply,
of course, that we cannot obtain exhaustive knowledge
of an electromagnetic field. We must keep in mind that
graphic pictures are only a teaching aid, 8 method for
more easily memorizing symbols. But we should not
5. Electrom8gnetlc Flelels 199
forget that the wave picture is only a model of electro-
magnetic radiation, and nothing morel Where feasible,
we employ this model, but we should not be surprised
that this model can only lead us into error in some cases.
In exactly the same way, the corpuscular aspect of
an electromagnetic field is not always observed. It would,
of course, be simpler if the conditions under which these
two aspects are manifested were mutually exclusive. But
no, this is not so. Even in describing one and the same
experiment, it is often necessary to speak in two lan-
guages.
It is nevertheless simpler (or, it is better to say, it
previously was simpler)' to observe the corpuscular aspect
of electromagnetic radiation for short waves. In an ion-
ization chamber and other similar instruments, we can
observe the collision of particles of electromagnetic ra-
diation with an electron or other "honest" particles.
The collision may resemble that of billiard balls. It is
absolutely impossible to understand such behaviour if
we only resort to the wave aspect of electromagnetic
radiation.
Using the language of Maxwell's theory, let us consider
the production of electromagnetic radiation. A system
of charges oscillates with a certain frequency. The electro-
magnetic field varies in time with these oscillations.
The frequency v of oscillations of the field divided by
the velocity of propagation (300 000 km/s) gives the
wavelength of the radiation.
When we deal with the same phenomenon in the lan-
guage of quantum physics, it is described as follows.
We have a system of charges characterized by a system
of discrete energy levels. For some reason the system
got into an excited energy state, but did not exist long
in this state and went over to a lower energy level. The
energy E,-E
1
=hv evolved in this transition is radiated
in the form of particles called photons. We araalready
Electrons 200'
acquainted with constant h (see p. 120). It is Planck '8'
constant .
..If the energy levels of the system are very close togeth-
er, the photon has low energy, low frequency and, con-
sequently, a long wavelength. Here the quantum corpus-
cular aspect of the electromagnetic field is hardly noticea-
ble and manifests itself only in absorption phenomena,
which are associated with extremely small changes in
the energy of electrons or atomic nuclei (magnetic res-
onance). In the case of waves of long length we do not
succeed in observing collisions of a photon with particles,
like the collisions of billiard balls.
We shall briefly mention facts that, so to speak, drove
physicists into a corner, forcing them to agree that wave
theory (which they had believed in for many years as the
complete and comprehensive truth) is incapable of ex-
plaining all the facts concerning electromagnetic fields.
There are many such facts but, for the time being, we
shall restrict ourselves to a phenomenon called the pho-
toelectric effect. After the reader has consented that a
comprehensive picture of the electromagnetic field cannot
be created without its corpuscular aspect, we shall dis-
cuss Hertz '8 wonderful experiments on which all of radio
engineering is based. We shall show how the wave as-
pect of the electromagnetic field was depicted, not only
in its general form, but in all its details as well.
PhotoeJectrlc Effed
The fine sonorous name "photon" appeared on the
scene somewhat later than the product of Planck '8 con-
stant h by the frequency v of an electromagnetic wave.
As mentioned above, the transition of a system from one
energy state to another is accompanied by the absorption
or emission of a portion of energy equal to hv. The first
to come to this conclusion, a ~ the turn of the century,
5. Electromagnetic Fields 20t
was the famous German physicist Max Karl ErnstLud-
wig Planck (1858-1947). He showed that this was the
only way that the radiation of incandescent bodies could
be explained. The argument concerned electromagnetic
waves produced by a method having nothing in common
with radio engineering. It had not yet been proved at
that time and was not recognized by all physicists that
what is valid for light is true for radio waves as well.
This was the situation even though Maxwell's laws def-
initely proved that there is no difference in principle
between radio waves and any other electromagnetic
waves, including light. An understanding and the experi-
mental proof of the universal validity of Planck's state-
ment came later.
Planck's work concerned the radiation of light in dis-
crete portions, i.e. in quanta. The work did not note,
however, that the quantum nature of radiation made
it inevitable to introduce t.he corpuscular aspect of the
electromagnetic field into the discussion. A field was
said, at that time, to radiate in portions, but a portion
was thought to be a certain wave train.
The most important step, i.e. the recognition of the
fact that the radiated portion of energy hv is the energy
of a particle which was immediately named the photon,
was made by Einstein. He showed that only corpuscular
concepts could explain the photoelectric effect (photo-
effect), i.e, the ejection of electrons from solids under
the influence of light.
Shown schematically in Figure 5.4 is the apparatus
employed at the end of last century to begin a detailed
study of the phenomenon called the extrinsic photoelectric
effect.
In 1888, Heinrich Hertz reported that light in some
way affects the electrodes of a vacuum tube. He was
evidently the first to notice this phenomenon. Working
at about the same time, the Swedish chemist Svante
Electrons 202
Figure 5.'
August Arrhenius (1859-1927), the German physicist
Wilhelm Ludwig Franz Hallwachs (1859-1922), the Ital-
ian physicist Augusto Righi (1850-1921) and the out-
standing Russian physicist Alexander Grigoryevich Sto-
letov (1839-1896) showed that illumination of the cathode
produces an electric current. If no voltage is applied
across the tube shown in Figure 5.4 (it is called a photo-
electric cell), only a negligible part of the electrons, ejected
from the cathode through the impact of light, reach
the opposite electrode. A weak accelerating voltage (with
a negative photocathode) increases the current. Finally,
as the voltage is increased, the current reaches its satura-
tion value: all the electrons (whose number is quite
definite at the given temperature) reach the anode.
The magnitude of the photocurrent is strictly propor-
tional to the light intensity. The light intensity is
5. EI.ctro....gRellc Fields 203
uniquely determined by the number of photons. It immedi-
ately occurs to us (and this is confirmed by rigorous cal-
culations and experiments) that each electron is ejected
by one photon.
The energy of the photon is expended in liberating an
electron from the metal and imparting a velocity to it.
This is expressed by an equation first derived by Albert
Einstein in 1905. Here it is:
mv
2
hv= y+A
where A is the work function (see p. 88).
The energy of the photon must at least exceed the work
function, i.e. the minimum energy it is necessary to
expend to eject an electron from the metal. This means
that for the photon each energy (and the energy is unique-
ly related to the "chromaticity") has its threshold value
of the photoelectric effect ..
Photoelectric cells, based on the extrinsic photoeffeet,
are extensively used in photorelays, television and sound
motion pictures.
The sensitivity of a photoelectric cell can he increased
by filling it with gas. Here the current is increased
because the liberated electrons break up the neutral mol-
ecules of gas and unite them with the photocurrent.
The photoelectric effect, true, not the one we have
described, but the so-called internal photoelectric effect,
occurring at the boundary of the p-n layer in semicon-
ductors, is of exceptional importance in modern engi-
neering. To avoid interrupting our discussion, however,
we shall postpone the treatment of the application of
the photoelectric effect until Book 4. Here we mentioned
this phenomenon only to demonstrate the inevitability
of recognizing that an electromagnetic field has corpus-
cular properties.
For many years photons were the forlorn stepsons of
Electrons 204
physics. Proof of the existence of photons and the investi-
gation of the laws of the photoelectric effect anticipated
the founding of quantum physics by about 20 to 30 years.
On.ly by the end of the twenties, when these laws had
been established, it became clear why the same numerical
constant, Planck's constant h, appears both in the for-
mula for the energy of 8 photon and in the formula,
discussed on p. 120, that determines the possible values
of the angular momentum of particles.
The value of this constant is determined from many
different kinds of experiments. The photoelectric effect,
the Compton effect (change in the wavelength of X-rays
in scattering), the radiation produced upon the annihi-
lation of particles and many other experiments yield
the same numerical value.
Hertz's Experiments
Now let us consider the way in which the hypotheses
concerning the wave aspect of the electromagnetic field
were proved.
Logic and mathematics extract many consequences from
Maxwells laws. These consequences may turn out to
be valid or they may not be confirmed by experiments.
A physical theory becomes a part of science only after
being experimentally checked. The establishment of elec-
tromagnetic field theory-from separate facts to a gen-
eral hypothesis, from the hypothesis to the consequences
and to the final stage, an experiment that provides
decisive proof-is the only proper procedure for a natu-
ralist. This highway to scientific truth is especially
defined in the case of the laws of the electromagnetic
field .
. For this r e ~ s o n We shall discuss Hertz's experiments
In more detafl , They are still being used by teachers
and lecturers to' .delllonstrate to students of schools and
5. Electromagnetic Fields
Figure 5.5
205
institutes how a scientist gains confidence in the valid-
ity of laws of nature.
Electrons
206
w ~ must begin our story in 1853, when the famous
British physicist Lord -Kelvin proved mathematically
that electric oscillations occur when a capacitor is dis-
charged through an inductance: the charge on the capac-
itor plates, the voltage at any point in the circuit and
the current all vary according to the law of harmonic
oscillations. If we assume that the resistance of the cir-
cuit is negligible, these oscillations will continue for-
ever.
The pictures in Figure 5.5 demonstrate the phenomena
that occur in a so-called oscillatory circuit. At the initial
instant of time the capacitor is charged. Current flows
as soon as the circuit is closed. After one-fourth of a
period, the capacitor is completely discharged. Its energy,
q2/2C, is converted into the energy of the coil's mag-
netic field. The current is at a maximum at this instant.
Continuing in the same direction, the current gradually
decreases. After a half-period, the current drops to zero,
the magnetic energy, L1
2/2,
disappears, the capacitor
is completely charged again and its energy is restored.
But the sign of the voltage is reversed. Then the pro-
cess is repeated in the reverse direction. After a certain
time T (the period of oscillation), everything returns
to the initial state and the process begins again.
Electric oscillations would continue to infinity except
for the inevitable resistance to current flow. Energy is
lost in each period due to resistance. As a result,
the oscillations decrease in amplitude and are soon
damped.
The obvious analogy with a load suspended by a spring
enables us to dispense with an algebraic analysis of the
process and to reason out the period of oscillations of
such a circuit. (The reader should refresh his memory
of the corresponding pages in Book 1.) It is sufficiently
evident, as a matter of fact, that the electric energy
of the capacitor is equivalent to the potential energy
. ~ \
5. Electromagnetic Field. 207
of the compressed spring, and the magnetic energy of
the coil, to the kinetic energy of the weight.
Associating analogical quantities, we "derive" the for-
mula for the period of oscillations occurring in the cir-
cuit: q2/2C is the analogue of k:rJ'/2, LP/2 of mvl/2,
k of tIC and L of m, Hence the frequency of oscillations
is v = 1/21tVLC because the corresponding formula
for mechanical oscillations is of the form
v= -.!-.. /. k
21t JI m
Now let us try to guess the train of thought of Hertz,
who made it his aim to prove the existence, without
leaving his laboratory, of electromagnetic waves prop-
agating at a velocity of 300 000 km/s. For this purpose,
an electromagnetic wave is required with a wavelength
of the order of 10 m. If Maxwell was right, the electric
and magnetic vectors should oscillate with a frequency
of 3 X 10
8
hertzes; pardon me, cycles per second. At
that time Hertz did not know that his name would be
perpetuated as the unit of frequency.
"What shall I begin with?" he probably asked him-
self. In the first place, since the oscillations are damped,
it is necessary to devise an arrangement that re-
commences the process when the current stops. This is
not a difficult task. The required circuit is illustrated
in Figure 5.6. An alternating voltage is applied across
the primary winding of transformer T. As soon 8S it
reaches the breakdown voltage between the metal spheres
connected to the secondary winding, a spark is ini-
tiated. This spark is what closes the oscillatory circuit
OC, acting like a switch key, and dozens of oscillations,
with a decreasing amplitude and a more or less high fre-
quency, pass through the circuit.
But a high frequency is required. What can be done
to obtain one? We can reduce the self-induction and the
Electrons
208
Heinrich Rudolph Hertz (t857-18941-brilliant German phys-
icist; using his U oscillator" and "resonator" he proved in exper-
iments that an oscillatory discharge produces electromagnetic
waves in space. Hertz showed that electromagnetic waves are
refleeted, refracted and subject to interference, thereby confirming
s. Eledromagnetic Fields
FigUre 5.6
209
capacitance. But how? We replace the coil by a straight
wire, and we spread the capacitor plates apart and rer:
duce their area. Into what does the oscillatory circuit
degenerate? It is converted into two simple rods ending
in metal spheres ~ e t w e e n which sparks jump.
Thus Hertz got the idea for his oscillator which can
serve both as a source and as a detector of electromag-
netic waves.
It was difficult for Hertz to predict beforehand the
inductance and capacitance of this unusual "circuit"
of which almost nothing remained. The inductance and
capacitance of the oscillator are not concentrated in
Maxwell's theory. Hertz's experiments laid the foundation of
radio engineering. In his first radio broadcast in 1895, Alexander
Stepanovich Popov, inventor of the radio, transmitted two words:
"Heinrich Hertz".
14-2259
Electrons 2tO
some definite place in the circuit but are distributed along
the rods. Some other theory is required.
A discussion of this new approach to electric circuits
with currents of very high frequency would take us too
far from our main subject. The reader can believe our
word that oscillations of high-frequency current really
do occur in .Hertz 's oscillator.
The "transmitter" and "receiver" of waves, used by
Hertz, were practically the same. In the "transmitter"
the waves were produced by sparks periodically jumping
the gap between the spheres in accordance with the op-
eration of the transformer that applied voltage to the
oscillator. The width of the spark gap could be regulated
by a micrometric screw. Serving as the receiver, or de-
tector, was either a rectangular turn of wire interrupted
by a spark gap or two small rods that could be adjusted
as required to obtain a gap within a fraction of a milli-
metre.
In his first work, published in 1885, Hertz showed that
oscillations of extremely high frequency can be obtained
by the above-described method and that these oscilla-
tions actually do set up an alternating field in adjacent
space. The existence of this field, he ,*rote, can be dem-
onstrated by sparks jumping across a gap in the "re-
ceiver". Hertz called this receiving oscillator. a resona-
tor. He immediately grasped the principle of detecting
an electromagnetic field. This principle is the basis of
modern radio engineering.
Incidentally, we may note that neither in Herta's
works nor in others published in the next few decades
do we find the terms "electromagnetic waves" or "radio
waves". They speak only of electric waves or of waves
of electrodynamic force.
In his next work Hertz showed that, in accordance
wi th the requirements of Maxwell's theory, the dielec-
tric medium (a bar of sulphur or paraffin) affects the
5. EI.ctro....gnetic Flelels 2t t
frequency of the electromagnetic field. When he re-
ceived this article, Helmholtz, the editor of the scien-
tific journal, sent Hertz a postcard with the message:
"Your manuscript has been received. Bravol I shall send
it to the press on Thursday."
Hertz's work in which he showed that electromagnet-
ic waves can be reflected made a tremendous impres-
sion on the physicists of his day. The waves were reflect-
ed by a zinc screen, 4 X 2 m in size. The oscillator was
13 m from the screen and at a height of 2.5 m above the
floor. The tuned resonator was arranged at the same
height and could he moved between the oscillator and
screen. Setting the resonator at various distances from
the screen and observing the intensity of the sparks,
Hertz established that there were maximums and mini-
mums typical of the interference phenomena that set
up standing waves. The wavelength was found to be ap-
proxiDnately 9.6 m.
I should like to emphasize the fact that at that time
no one could say what material could serve as a mirror
for electromagnetic waves. Today we know that waves
of such length do not penetrate metals, but are reflected
from them.
In trying to obtain additional proof of Maxwell's theo-
ry, Hertz reduced the geometrical size of his apparatus
until he reached a wavelength of 60 em. In 1888 he pub-
lished the work On the Beams 0/ Electric Force. He was
able to concentrate electromagnetic energy by means
of parabolic mirrors. The rods of the oscillator and
resonator were arranged at the foci of the mirrors. Em-
ploying the mirror-type transmitter and receiver, Hertz
showed that electromagnetic waves do not pass through
metals, but wooden screens do not check them.
Figure 5.7 shows how Hertz demonstrated the polar-
ization of electromagnetic waves. He' arranged grating
C, made of parallel copper wires, in the path of the elec-
I"
Electrons 212
Figure 5.7
tromagnetic beam produced by oscillator AA. Hertz
showed that when he turned the grating, the intensity of
the spark in resonator BB 'was varied. When the grating
wires were parallel to the electric vector and perpendic-
ular to the axes of the oscillators, the beam did not
pass through the grating. Thus, the transverse charac-
ter of an electromagnetic wave was proved.
Finally, to study the refraction of the waves, Hertz
made a prism weighing over one tonne from an asphalt
compound. The refractive index of the asphalt for waves
60 em in length could he measured with high accuracy.
It was found to equal 1.69.
The proof of the existence of electromagnetic waves,
measurement of their length, and the establishment of
the laws of their reflection, refraction and polarization
are all the results of only three years of research! Truly
a capacity for work to be admired.
213 5, Electromagnetic Fields
Classification
of Electromagnetic Radiation
Physicists deal with electromagnetic radiation over a
huge range. The electromagnetic radiation of current
of city mains frequency is absolutely negligible. Prac-
tically, the feasibility of detecting electromagnetic ra-
diation begins with frequencies of the order of dozens
of kilohertzes, i.e. at wavelengths equal to hundreds of
kilometres. The shortest waves have a length of the or-
der of ten thousandths of a micron, i.e. a frequency of
the order of a thousand million gigahertzes.
Radio umues are in the range of electromagnetic ra-
diation produced by electrical engineering means, i.e.
as a result of the oscillation of electric currents. The
shortest radio waves have a wavelength of hundredths of
a millimetre. r
Beginning with several hundred microns and smaller
is. the wavelength range of radiation due to energy tran-
sitions inside molecules, atoms and atomic nuclei. As
we can see, this range substantially overlaps the radio
range.
Visible light occupies a narrow range. Its limits are
0.38 to 0.74 micron. Radiation of longer wavelength,
not obtained by radio engineering techniques, is said to
be infrared; that of shorter wavelength is called ultra-
violet and extends down to a wavelength of about 0.1
micron.
Electromagnetic radiation produced in X-ray tubes
overlaps ,the range of ultraviolet waves and extends down
to a wavelength of 0.01 micron, where it overlaps the
gamma ray range. Gamma rays are produced in nuclear
decay, in nuclear reactions and in collisions. between
elementary particles.
The principal characteristic of any electromagnetic ra-
diation is its spectrum. The spectrum is a graph on which
Electrons
Figure 5.8
I
2i4
the intensity (i.e. the energy per unit time per unit area)
is plotted in the vertical direction and the wavelength
or frequency in the horizontal direction. The simplest
spectrum is that of monochromatic ("single-colour") ra-
diation. Its graph consists of a single line of extremely
narrow width (Figure 5.8, above). The degree of mono-
chromaticity of the line is specified by the ratio A / ~ A . A
broadcasting station produces almost purely monochro-
matic radiation. For a short-wave station, for instance,
operating in the 30-m range, A/&l equals about 1000.
Excited atoms, for example, atoms of gases in daylight
lamps (the excitation being due to the collisions of
positively and negatively charged particles travelling to-
ward the anode and cathode), produce a spectrum con-
sisting of a great number of monochromatic lines with
a relative width of 100000-
1
Lines are observed in mag-
netic resonance with a relative width as small as 10-
7

Strictly speaking, no continuous spectra exist. How-

5. -Electromagnetic Fields 215
ever, -if- the lines overlap, the experiment yields an in-
tensity curve of the kind shown below in the same il-
lustration.
"Information on an electromagnetic spectrum can be
obtained either by investigating radiation or by study-
ing absorption. In general, both experiments provide
the same information. This is evident from the princi-
pal law of quantum physics. In radiation, the transi-
tion of the system is from a higher to a lower -energy
level; in absorption, it is from 8. lower to a higher level.
But the difference in energy, which determines the fre-
quency of radiation or absorption, is exactly the same.
Which spectrum we investigate, radiation or absorption,
is, simply a matter of convenience.
In characterizing a radiation spectrum, we can, of
course, use either wave or corpuscular terminology. Em-
ploying the wave aspect of radiation, we state that the in-
tensity" is proportional to the square of the wave ampli-
tude. Regarding radiation as a stream of particles, we
count up the intensity as the number of photons.
We repeat again that we should in no way be dis-
turbed by this alternating application of the two aspects-
of radiation. Radiation resembles neither waves nor a
stream of particles. Both concepts are only models that
can be conveniently applied in explaining various phe-
nomena.
We have not presented a scale of electromagnetic
waves, but have mentioned with sufficient clarity that the
names of its various ranges are, to some degree, arbit-
rary. In any case, we find occasions in which waves of
the same length are differently called, depending on the
way in which they were produced.
Today, the scale of electromagnetic waves is contin-
uous. There are no ranges that have not been obtained
by one or another method.
But the overlapping of the infrared and radio waves,
. E 1 ~ r o n s 216
the gamma and X-rays, etc. was discovered a relatively
short time ago. For a long time there was a gap between
the short radio waves and the infrared range. Waves with
a length of 6 mm were first obtained by the brilliant Rus-
sian physicist Pyotr Nikolayevich Lebedev (1866-1912)
and heat (infrared) waves of a length to 0.34 mm, by
the German physicist Heinrich Rubens (1865-1922).
,In 1922 this gap was closed by the Russian physicist
Alexandra Andreyevna Glagoleva-Arkadyeva (1884-1945)
when she obtained electromagnetic waves in a range
fr.om 0.35 mm to 1 em by nonoptical techniques.
At the present time, wa ves of this length are obtained
by radio technicians without any trouble whatsoever.
Glagoleva-Arkadyeva had to employ much ingenuity and
resourcefulness to develop the required apparatus, which
she named the mass radiator. The source of the electro-
magnetic radiation was metal filings in suspension in
transformer oil. A spark discharge was passed through
this mixture.
6. Radio
Some History
1ust as Faraday had no idea that the discovery of
electromagnetic induction would lead to the founding
of electrical engineering, and Ernest Rutherford con-
sidered idle talk and rank ignorance the feasibility of
extracting energy from the atomic nucleus, so Heinrich
Hertz, after discovering electromagnetic waves and show-
ing how they can he detected at distances of several
metres, had no notion of radio communication and even
denied its possibility. Three amusing facts, are they not?
But we shall leave their discussion to the psychologists.
Therefore, restricting ourselves to a mere statement of
this remarkable circumstance, we shall see how events
developed after Hertz's early death in 1894.
Hertz's classical experiments, which we have described
in. 'such detail, attracted the attention of scientists
all over the. world. Professor N. G. Yegorov of the St.
Petersburg University accurately repeated these exper-
iments. The spark in the resonator was barely visible.
It could be seen only in complete darkness and then only
with a magnifying glass.
Alexander Stepanovich Popov (1859-1906), an unas-
suming lecturer in electrical engineering at the Kronstadt
Military Academy, set to work in 1889, at the age of
30, to improve Hertz's experiments. The sparks that he
obtained in his resonators were much fatter than those
other investigators succeeded in producing.
In 1894, a fall issue. of the English journal Electrician
published an article by the well-known English physi-
Figure 6.t
cist Sir Oliver Joseph 'Lodge 'in which he
claimed that Hertz's resonator. could be. improved by
using the Branley tube. The French scientist Edouard
Branley (1844-1940) was engaged in research on the con-
ductivity of metal filings. He found' that such filings
do not always offer the same resistance to electric cur-
rent. Loosely packed metal filings in a tube have prac-
tically infinite- resistance, but if the tube is placed in
the vicinity of ail operating Hertz resonator, the resis-
tance drops drastically. The explanation is that the small
filings cohere owing to the welding action of the tiny
sparks produced between them by the electromagnetic
wave. lightly tapped or shaken, the resistance
of the filings is restored.
This property of .metal filings was made use of by
Lodge. He wired a circuit consisting of the Branley tube
(which came to be known as a cohererv, a battery and a
sensitive galvanometer, The hand of the instrument was
deflected at the instant electromagnetic waves passed
by. Lodge succeeded in detecting radio waves at dis-
tances up to 40 ID.
This system was Inconvenient, however, because the
coherer immediately stopped operating. It was neces-
sary to find a way to restore the cohering (welded) filings
6. IMlo
2t9
Figure 6.2
to their initial state, with the required device designed
so that the shaking occurs automatically.
This problem was solved by Popov. He tried many
kinds of coherers and finally decided to use one designed
as follows. "Two strips of thin sheet platinum, Ail
and CD, are glued on the inside wall of a glass tube and
extend over almost the whole length of the tube. One
strip emerges at one end of the tube and the other strip
at. the other end. The strips are 8 mm wide and are ar-
ranged with a space of about 2 mm between them. The
inner ends, Band C, of the strips do not reach the corks
that plug the tube so that filings jammed between the
cork and the tube cannot form conducting chains that
cannot be broken up when the tube is shaken. This hap-
pened in some of the earlier models. The length of the
whole tube is from 6 to 8 em and its diameter is about
Electrons 220
Alexander Sfepanovich Popov 118S9-t906)-great Russian phys-
j cist and electrical engineer; invented the radio. Popov's scien-
tific work was highly appraised by his contemporaries. In 1900
he won a" Gold Medal at the World '5 Fair in Paris for his invention.
6. Radio 22f
1 cm. In operation the tube is arranged horizontally so
that the strips are at the bottom and are covered by
metal filings. Best operation is achieved when the metal
filings fill not ove-r one-half of the space in the tube."
Popov's coherer, just described in his own words, is
shown in Figure 6.1. He used iron or steel powder in it.
The main problem, however, was not to improve the
coherer, hut to invent a method of restoring its initial
state after detecting an electromagnetic wave. In Popov's
first receiver, whose wiring diagram is shown in Figure
6.2, this job was performed by an ordinary electric bell.
The striker of the bell replaced the galvanometer 'hand
and its hammer struck the glass tube when the striker
returned to its initial position.
What a simple solution of a puzzling problem! And
really simple. Do you realize, dear reader, that this
elementary arrangement, which neither Hertz nor Lodge
hit upon, was the first application of what engineors
now call a relay circuit? The negligible energy of radio
waves is not directly detected, but is used to control
a current circuit.
In the spring of 1895, Popov set up his .apparatus out-
side, 'in an orchard. He began to move the receiver far-
ther and farther away from the oscillator. At 50 m the
bell responded to the spark of the oscillator; at 60 m the
apparatus still worked, but at 80 m the bell refused to
ring. Then Popov took a coil of copper wire, threw one
end up into a tree and connected the other end to the
coherer. The bell rang. This is how the first antenna was
devised.
In the USSR, the 7th of May is celebrated each year as
Radio Day. On this day in 1895 Popov read a paper
with the unassuming name "On the Relationship of Met-
al Powders to Electric Oscillations" at the regular ses-
sion of the Russian Physico-Chemical Society. Many of
those present had watched a demonstration, several years
Electrons 222
previously, of Hertz's experiments in which the tiny
sparks could be seen only through a magnifying glass.
But, when they heard the brisk clangour of Popov's re-
ceiver, they realized that they were witnessing the birth
of the wireless telegraph, that a new and more efficient
method of transmitting signals over long distances had
been invented.
On March, 12, .1896, Popov transmitted the world '8
first radiogram. By pressing a key for short and long
intervals, the words "Heinrich Hertz" were transmitted
over a distance of 250 m from one building to another
and were recorded on telegraph tape.
By 1899, the range of radio communication between
the training ships of a mine-layer detachment had al-
ready reached 11 km. The practical importance of the
wireless telegraph was no longer doubted even by the
most skeptical officials.
The Italian electrical engineer and inventor Guglielmo
Marconi (1874-1937) began his experiments somewhat
later than Popov. Marconi carefully followed all the ad-
vances in the fields of electrical engineering and the study
of electromagnetic waves. He skilfully employed them
to improve the quality of radio reception and trans-
mission. His especial contribution was more in the line
of organization and management rather than the strictly
technical aspect. This is no small matter, however, and
Marconi's fame is well deserved. We should not forget
that the priority in the discovery of the radio, based on
the paper read on May 7, 1895, belongs to the unpreten-
tious Russian scientist Popov who always refused to
put his knowledge and research at the disposal of any
country except his native land.
Marconi did not mention Popov in his articles and
lectures.. But not everyone knows that in 1901 he offered
Prof. A. S. Popov a position in the commercial company
that Marconi had founded and was the president of.
6. -RMIlo
Fllure 6.3
'I'he range of radio reception grew at 8 rapid rate. In
1899, Marconi established radio communications be-
tween France and England, and in 1901 a radio signal
was sent from Europe to America.
What technical innovations facilitated these successes
and made radio broadcasting feasible?
Beginning with .1899, radio engineering no longer
based reception on the coherer. Instead of detecting radio
waves by the drop in the resistance of a circuit due to the
effect of an electromagnetic wave, an entirely different
technique can be made use of. A rectified pulsating eleo-
tromagnetic wave can he detected by the clicks heard in
an ordinary telephone receiver.
This began a search for rectifiers. The extensively
used contact detector, applied right up to the twenties
of our century, consisted of a crystal with one-way con-
duction..Such crystals had been known since 1874. They
include metal sulphides, copper pyrites and hundreds
Electrons
Figure 6.4
224
of different minerals. People of my age remember such
radio sets and the irritating procedure of searching for
"good contact" with a whisker (contact spring). Such
contact was achieved when the point of the whisker
had found a "suitable" spot on the crystal (Figure 6.3).
By that time many broadcasting stations were in op-
eration so that the set had to be tuned to the required
wave. This was done with a multicontact switch for a
small number of stations, or by steplessly varying the
capacitance of the capacitor, which is also employed
in up..to-date apparatus.
It was extremely difficult, or even impossible in some
cases, to operate at high power with the spark-gap broad-
casting stations because the spark-gap device became
overheated. Such stations were soon replaced by one
operating on the principle of an oscillating electric arc
or a high-frequency alternator. After this the power
ratings reached hundreds of kilowatts.
6. Radio 225
The real revolution in radio communications, enabling
speech and music to be transmitted instead of only teleg-
raph code, came with the development of the vacuum tube.
In October 1904, the English electrical engineer Sir
John Ambrose Fleming (1849-1945) showed that high-
frequency current could be rectified by a vacuum tube
consisting of a filament heated by the current and sur-
rounded by a metal cylinder. Its diagram is illustrated
in Figure 6.4. Fleming realized the value of the vacuum-
tube diode for converting electric oscillations into sound
(he called this device a "valve" because it opened and
shut the gate to the flow of electricity), hut he could
not achieve wide application of his detector.
The fame of inventing the electron lamp was won by
the American scientist Lee de Forest (1873-1961). In 1906,
he converted the two-electrode tube (diode) into a triode by
adding a third element. The tube was called an audion
(from the Latin word audire, meaning I hears, De Forest's
vacuum tube radio set received signals on the grid (third
element) in the lamp, rectified them and enabled them
to be heard as telegraph code in headphones.
The feasibility of using the electron tube as an amplifier
was evident to the American scientist. But only after
seven years had passed, in 1913, a triode WAS first ap-
plied in a generator circuit by the German radio engineer
Alexander Meissner (1883-1958).
Attempts to transmit speech, i.e. the modulation of an
electromagnetic wave, had been made before the electron
tube was used as a generator. But the difficulties were very
great: t.he band of frequency modulation could not be
made wide enough. Speech could be transmitted with
some little success, but not music. Only in the twenties
did radio transmitters and receivers, operating with
electron tubes, demonstrate the truly inexhaustible ca-
pabilities of radio communications for transmitting the
whole range of audio frequencies.
J5-2259
Electrons
Figure 6.5
226
The next revolutionary breakthrough occurred not
long ago, when semiconductor elements superseded elec-
tron tubes in radio circuits. This established a new branch
of applied physics dealing with the huge complex of
problems concerning the input, transmission and storage
of information.
Vacuum-Tube Triode andTransistor
Vacuum-tube triodes incited an upheaval in radio
engineering. But technology ages faster than people do.
Today, the electronic tube has become a pensioner Of,
as they say in the USA, a senior citizen. Not so many
years ago you could hear impatient prospective custom-
ers in shops selling TV sets demanding transistorized
models, i.e, sets in which semiconductors have replaced
electronic tubes.
But age demands respect. Moreover, the principles
underlying the two fundamental applications of tubes
and transistors, namely, the amplification and generation
of waves of a definite frequency, can be more simply
described by using tho electronic tube as an example.
6. Radio
-8 -6 -"f - ~ 0 l. *I 6 8
227
Figure 6.6
We shall therefore cover its operation in more detail than
that of the transistor.
Besides the plate (anode) and heated filament (cath-
ode), the bulb of a three-electrode tube has a sealed-in
third electrode, called the grid. The electrons freely
pass through the grid. Its openings are as much larger
than the electron as the earth is larger than a dust speck.
Figure 6.5 illustrates how the grid enables the plate
current to be controlled. It is evident that a negative
voltage impressed on the grid reduces the plate current
and a positive voltage increases this current.
Let us conduct a simple experiment. First we apply 100
volts across the cathode and anode (filament and plate).
Then we begin to vary the grid voltage as shown, for
instance, in Figure 6.6 in a range from minus eight volts
to plus five. Using an ammeter, we measure the current
in the plate circuit. From this data we can plot the upper
curve shown in Figure 6.6. This is called the tube char-
acteristic. Next we repeat the experiment with a plate
voltage of 90 V. We obtain a similar curve (lower curve
in Figure 6.6).
Take note of the following outstanding feature. As is
Electro...
Figure 6.7
228
obvious from the hatched triangle, we can increase the
plate current by 5 milliamperes in two different ways:
either by increasing the plate voltage by 10 volts or by
increasing the grid voltage by 2 volts. The introduction
of the grid makes an amplifier out of the vacuum-tube
triode. The amplification factor in our example equals 5
(ten divided by two). In other words, the effect of the
grid voltage on the plate current is five times that of
the plate voltage.
Now, we shall discuss how a triode enables us to gen-
erate waves of a definite length.
This is illustrated by the extremely simplified cir-
cuit diagram given in Figure 6.7. When the plate voltage
is applied (switched on), capacitor Cckt of the oscillatory
circuit is charged through the tube. The lower plate is
positively charged. The capacitor is immediately dis-
charged through the inductor Lett. This produces free
(natural) oscillations that would be damped if there were no
continuous energy input from the tube. How can we ensure
that this energy is supplied at the proper times so that
the oscillatory circuit builds up in the same way as the
amplitude of a swing when you push it at the right times?
This requires what is called feedback. The current of the
6. Redlo
229
E
SGHEMATIC DIAGRAMS
c E c
F'..,. 6.8
oscillatory circuit induces an emf in coil Ln, of the same
frequency as that of free oscillations. Thus the grid
produces a pulsating current in the plate circuit; this
current builds up the circuit with its natural frequency.
The two ingenious principles described above are the
basis on which radio engineering and allied fields have
been established. The electronic tube has become ob-
solete, making way for the transistor, but the idea of am-
plifying and generating electromagnetic oscillations has
remained the same.
As in a vacuum-tube triode, low power in the input
circuit of a transistor can control high power in the out-
put circuit. There is a difference in the way control is
accomplished. The plate current of a tube, as we have
seen above, depends upon the grid voltage. The current
Electrons
230
of a collector in a transistor depends on the emitter
current.
But we have not described a transistor yet. It has
three electrodes. The emitter corresponds to the cathode,
the collector to the plate (anode) and the base to the grid.
The lead from the emitter is the input and that from the
collector is the output.
As shown in Figure 6.8, the transistor consists of two
p-n junctions. At the left is p-n-p transistor with the
it-layer in the middle between two p-Iayers. We can also
have a p-Iayer in the middle, in which case we have an
n-p-n transistor (at the right).
We always bias the emitter positively so that it can pro-
duce a large number of majority carriers of charges.
When the low-resistance emitter circuit changes the cur..
rent in the high-resistance collector circuit, we obtain
amplification.
The ways in which transistors are connected into cir-
cuits and employed as amplifiers and generators are
similar, in the main, to the principles of the vacuum-
tube triode. But we shall not discuss this extremely
vital branch of up-to-date physics here.
Radio Transmission
The kinds of radio transmission can be classified on
the basis of the, power rating of the broadcasting sta-
tion. Large stations transmit signals at powers ranging
up to a megawatt. A miniature transmitter of the
walkie-talkie type, employed by a traffic cop to notify
his colleague that a Lada car with the license number
31416P has just jumped a stop light and is to be held
up, radiates about one milliwatt. Even less power is
sufficient for some purposes.
There are essential differences in the layo.ut and desi-gn
of stations operating with waves over several metres
6. 1"'.0 231
long and transmitters radiating ultrashort waves with a
length of dozens of centimetres or even only fractions of a
centimetre. Within each of the wavelength and power
ranges, the engineer designing the station can apply any of
a huge number of circuits and layouts that may be dic-
tated by the locality, specific aims, economic consider-
ations Of, simply, engineering intuition.
The basic unit of a radio transmitter is the radio wave
generator, or oscillator. What kind will you employ?
You have at least five choices. You can use a vacuum-
tube oscillator. Its range is exceptionally wide. Power
ratings may vary from fractions of a watt to hundreds
of kilowatts, and frequencies from dozens of kilohertzes
to several gigahertzes. But if your power requirement is
small, of the order of tenths of a watt, only a transistor
oscillator will suit you. On the contrary, you will have
to reject transistors for the time being (probably not for
long) if your power requirement exceeds several hundred
watts. If, however, the power rating is such that both
types of oscillators can be efficiently applied, the de-
signer will evidently prefer the transistor version. Such
an engineering solution is doubtlessly more elegant.
A transistorized transmitter occupies considerably less
space and, if necessary, can much more easily be designed
as a portable model than a transmitter with a vacuum-
tube oscillator.
Magnetron and klystron oscillators have more special-
ized application. The former can be extremely useful
in sending pulses of several megawatts into space. The
frequency band for which magnetron oscillators can be
used is much narrower: it lies approximately between
300 megahertzes and 300 gigahertzes.
Klystron oscillators are used for the same range of
ultrashort waves. But they find application only in low
power installations: not exceeding several watts in the
centimetre and several milliwatts in the millimetre bands.
Electrons
Figure 6.9
MODULATOR
------.-..----e-:--(AUDIO- fREQUENCY

232
The last two types of oscillator, as well as the fifth
type, the quantum oscillator, are highly specific and
require a special discussion. As to transistor and vac-
uum-tube transmitters, they resemble each other. There
is a clear-cut engineering rule enabling one to replace
a vacuum tube with an equivalent transistor.
The choice of the generator of electromagnetic oscil-
lations is by far not all that is required for designing a
transmitter. You must decide to how amplify the power
produced by the primary (or, as they say, master) oscil-
lator. You must also select a method for modulating the
carrier wave by the audio frequencies. There are also
many ways for transmitting power to the antenna field.
The arrangement of the antenna field itself provides abun-
dant opportunity for exercising engineering ingenuity.
Radio engineers frequently resort to so-called block
diagrams. Such a diagram consists of several rectangles
with captions. The contents of each rectangle is cleared
up as required. A block diagram of a radio broadcasting
station is illustrated in Figure 6.9. The master oscillat6r
generates continuous, almost harmonic, oscillations of
the same frequency and wavelength to which you tune
6. Radio 233
your radio receiver if you wish to listen to the given sta-
tion. The second unit is the power amplifier. Its name
speaks for itself and we shall not describe its design.
The task of the unit called the modulator is to convert
sound vibrations into electric oscillations and superimpose
them on the carrier wave of the broadcasting station.
Modulation can be accomplished in various ways.
The simplest kind to explain is frequency modulation.
In many designs the microphone is a capacitor whose
capacitance is varied by the sound pressure because- the
capacitance depends upon the distance between the plates.
Imagine now that such a capacitor is connected into the
oscillatory circuit which generates the wave. Then the
frequency of the wave varies with the sound pressure.
Since we have "invaded" the oscillatory circuit with
our microphone, a band of frequencies is transmitted into
space rather than a strictly definite frequency. It is suf-
ficiently evident that ideally this spreading should
include the whole audio interval of frequencies which,
as we know, equals about 20 kHz.
If the station is broadcasting on long waves, correspond-
ing to a frequency of the order of 100 kHz, the passband
is about one-fifth of the carrier frequency. It is clear that
long waves cannot provide for a large number of non-
overlapping broadcasting stations. Short waves are an
entirely different matter. For a frequency of 20 MHz,
the band width is only a fraction of one per cent of the
carrier frequency.
There is probably not a single home in the USSR that
has no plug socket for listening to the radio, You receive
these broadcasts from the so-called rebroadcasting net-
work. It is also called wire broadcasting.
The first single-program rebroadcasting network was
established in Moscow in 1925. The broadcast was trans-
mitted simultaneously through 50 loudspeakers.
Single-program rebroadcasting is carried out on audio
16-2259
Electrons
frequencies. From the broadcasting station, the program
is transmitted by wire to the central amplifying station.
From the central station it is transmitted, again by
wire, to the control points, where it is again amplified
and transmitted along the trunk feeder mains to the
transformer substations. From each substation the wires
branch off to substations of the next lower rank. Depend-
ing upon the size of the city or region, the number of
steps in the network and, consequently, the number of
times the voltage is stepped down, may vary. In the sub-
scriber's lines the voltage is equal to 30 V.
Since 1962, three-wire rebroadcasting is being installed
in the cities of the USSR. The transmission of two
additional programs is accomplished along independent
networks by amplitude modulation on carrier frequencies
of 78 and 120 kHz. These two broadcasts are demodulated
(i.e. the sound is separated out and the high frequency
is filtered out) in the home by turning the knob of your
Mayak wire-broadcasting plug-in receiver or some other
model.
Thus, in three-program rebroadcasting, a single wire car-
ries three programs simultaneously: the main program is
on audio frequencies and two are nondemodulated. There-
fore, the broadcasts do not interfere with one another. A
simple idea, but what excellent results! Economy, relia-
bility and high fidelity of the broadcasts are factors that
indicate that wire broadcasting has a great future, includ-
ing the installation of wire networks for television.
Radio Reception
Radio receiving sets are available in innumerable de-
signs. The field of radio electronics is being- developed at
an exceptionally rapid rate, so that radio sets soon become
obsolete and new items, better than the previous models,
are available each year in. the shops.
6. Reel 235
Figure 6.tO
What do we mean by "better" with respect to receiv-
ers? Each and every reader knows the answer even if
he does not understand the physics involved. A good
receiving set must separate out of the chaos of radio
waves that reach the antenna only those signals that are
required. This property is called selectivity. A radio
set must also be as sensitive as possible, i.e. it should
he able to receive even the weakest signals. Finally, it
must have high fidelity, i.e. it must reproduce the music
and speech broadcasted from the tuned-in station without
any distortion whatsoever.
Thus, sensitivity, selectivity and fidelity. We could,
perhaps, add one more requirement: the set should op-
erate well on all wave bands.
The block diagram of a receiver with straight ampli-
fication is sufficiently clear (Figure 6.10). It is necessary,
first of all, to separate out the required wavelength and
to amplify the radio-frequency oscillations produced
in the antenna by the wave of the broadcasting station.
Next, it is necessary to accomplish rectification, or demod-
ulation, which is the name of the process of "discard-
ing" the.. carrier frequency and sifting out the infor-
1 6 ~
Electrons
Figure 6.11
236
mation carried by the sound from the electric current.
Finally, it is necessary to provide one more amplifier,
but for audio frequencies. The concluding stage is to
convert these electric oscillations into sound. This is
done by means of a dynamic loudspeaker 01' by head-
phones. The latter are used by considerate people who do
not wish to be a nuisance to their neighbours.
The antenna of the receiving set is usually inductively
linked to the oscillatory circuits of several frequency
bands. When we turn the band switch knob, we perform
the operation shown schematically in Figure 6.11. Within
the limits of each band we usually tune the set by varying
the capacitance of the oscillatory circuit capacitor.
The capacity of the radio set to select the frequency in
the optimal way is determined by the resonance curve
of the oscillatory circuit.
I am looking at the specifications in the instruction
book of an automobile radio set. Its selectivity is within
9 kHz of resonance for the long- and medium-wave fre-
quency bands. This, of course, is not the limit that can
be reached.
The sensitivity of a radio receiver is characterized by
the minimum emf in its antenna that enables us to hear
~ R ~ ~ 2M
the broadcast with sufficient clarity (I cannot say that
this definition is very exact). In the automobile radio
set the sensitivity for long waves is better than 175 muV
and for ultrashort waves, bet.ter than 5 muV.
The sensitivity depends upon the amplification factor
and on set noise. The amplification factors of radio
sets vary from 10
6
to 10
8
This means that the broadcast-
ing station I want to listen to must produce an induced
emf of at least 1.0-
8
muV in the antenna of the receiver.
Radio-Wave Propagation
The simplest case is the propagation of radio waves
in free space. At a relatively short distance away the radio
transmitter can be regarded as a point. If so, then the
radio wave front can be assumed spherical. If we imagine
several spheres concentrically surrounding the trans-
mitter, it will be clear that, in the absence of absorption,
the energy passing through the spheres remains constant.
As we know, the surface of a sphere is proportional to
the square of the radius. Hence, the wave intensity,
i.e, , the amount of energy transmitted by the wave in
unit time through unit area perpendicular to the direc-
tion of wave propagation, decreases as we move away
from the source in inverse proportion to the square
of the distance.
This important rule is applicable, of course, if no spe-
cial measures have been taken to obtain a narrow directed
beam of radio waves.
Various techniques exist for producing directed radio
beams. One way is to use the proper beam, or antenna,
array. The antennas should be arranged so that the waves
they transmit are sent in the required direction, "hump
to hump". Reflectors of various shapes are also employed
for this purpose.
Radio waves travelling in space may deviate from the
Electrons 238
straight-line direction of propagation by being reflected,
dispersed or refracted if they meet with an obstacle
commensurate with the wavelength.
Of greatest interest is the behaviour of waves travel-
ling near the surface of the earth. Each case may prove
to be entirely unique, depending upon the wavelength.
Of cardinal importance are the electrical properties
of the earth's surface and of the atmosphere. If the sur-
face can conduct a current, it does not "release" the ra-
dio waves. Electric lines of force of an electromagnetic
field enter a metal (or, for that matter, any conductor)
at a right angle.
Now just imagine that the radio transmission is near
the surface of the sea. Sea water contains dissolved salts
and is therefore an electrolyte. Sea water is an excel-
lent current conductor. Consequently, it "holds" the ra-
dio waves, making them travel along the surface of the
sea.
Plains and timbered areas are also good conductors
for currents of not especially high frequency. In other
words, plains and forests behave like metals with re-
spect to long waves.
For these reasons, long waves are contained by the
whole surface of the earth and are capable of circumvent-
ing the globe. Incidentally, the velocity of radio waves
can be measured in this way. Radio engineers know that
it. takes a radio wave 0.13 second to go around the earth.
How about mountains? As far as long waves are con-
cerned, mountains are not so high, and a wave a kilometre
long can go around a mountain with more or less ease.
The feasibility of long-range reception of short waves
is based on the presence of the ionosphere. The sun '8
rays are capable of breaking up the molecules of air in
the higher regions of the atmosphere. The molecules
are converted into ions and form several charged layers
at altitudes from 100 to 300 km. Thus, the space through
~ R ~ ~ ~
which waves of short length travel is a layer of dielec-
tric sandwiched between two conducting surfaces.
Since plains and forest lands are poor conductors for
short waves, they are incapable of holding them. Short
waves depart on their journey into free space but encoun-
ter the ionosphere which reflects them like the surface
of a metal.
The ionosphere is nonuniformly ionized and this
varies from day to night. Therefore, short radio waves
may travel along various paths. They may reach your
radio set after being reflected time after time by the
ionosphere and the earth. The actual fate of any short
wave depends upon the angle it makes with the iono-
sphere layer. If this angle is close to a right angle, no re-
flection occurs and the wave goes through and off into
outer space. More frequently, however, total internal
reflection takes place and the wave returns to the earth.
The ionosphere is transparent to ultrashort waves.
Radio reception of such waves is therefore possible only
along a line of sight (with the receiver antenna within
sight of the transmitting antenna) or with the aid of
communications satellites. If we direct our wave at the
satellite, we can pick up the signals reflected from it at
enormous distances.
Satellites ushered in a new era in radio communica-
tions techniques by making radio and TV reception fea-
sible on ultrashort waves.
Transmission by centimetre, millimetre and submil-
limetre waves presents extremely interesting opportu-
nities. Waves of this length can be absorbed by the
atmosphere. It was found, however, that there are "win-
dows" and, if the wavelength is properly selected, we
can make use of waves that are already within the op-
tical band. The advantages of such waves are well known:
in a narrow wave band we can "fit in" a huge number
of nonoverlapping broadcasts.
Electrons 240
Radar
The principles of radar (radio detecting and ranging)
are sufficiently simple. We transmit a signal, it is re-
flected by the object that interests us and returns to our
receiver. If the object is 150 m away, the signal returns
in 1 microsecond, if the distance is 150 kID, it returns
in 1 millisecond. The direction in which the signal re-
turns is along a line from the point where the airplane,
rocket or automobile was at the instant it met the ra-
dio beam.
Naturally, the radio wave must be a pencil beam;
the angle of the beam, within which the greater part
of its power is concentrated, should be of the order of
one degree of arc.
The principle is really not at all complicated, hut
the apparatus required is far from simple. To begin
with, exceptionally high requirements are made to the
oscillator. Vacuum-tube oscillators are used for the me-
tre and decimetre bands (longer waves are obviously inap-
plicable), and klystrons and magnetrons, for the centi-
metre band.
The pulsed system of operation is evidently the most
natural one. Very short pulses are periodically trans-
mitted into space. The duration, or length (of time),
of the pulse ranges from 0.1 to 10 microseconds in up..
to-date radar transmitters. The frequency of the pulses
must be selected so that the echo signal has time to
return and be received in the pause before the next
pulse is sent.
The maximum range at which an airplane or rocket
can be detected is limited only by the condition that the
object must be on a line of sight from the radar set. The
reader knows, without doubt, that up-to-date radar
installations are capable of picking up signals reflect-
ed from any planet of our solar system. The waves they
6. Radio
24t
use must, of course, pass unimpeded through the iono-
sphere. Fortunately, shortening the wavelength also has
a direct effect on the range of radar" operation because
this range is proportional to the frequency of radiation
and not only to the energy of the transmitted pulse.
Traces of the transmitted and received pulses are
visible on the screen of an oscilloscope (cathode-ray
tube). If the airplane is flying toward the observer, the
trace of the echo signal moves toward the trace of the
transmitted pulse.
Radar installations do not necessarily have to op-
erate with a pulsed system. Assume that the airplane
is flying toward the transmitter antenna with the veloc-
ity v. The radio beam is being continuously reflected
from the airplane. Due to the Doppler effect, the fre-
quency of the received waves is related to that of the
transmitted waves by the equation
Vr = Vtr ( 1+2: )
Frequency values are determined with great precision
by radio engineering methods. By tuning into resonance,
we determine the value of v ~ and from it we calculate
the airplane '8 velocity u. If, for instance, the frequency
of the transmitted signal equals 10
9
Hz and the airplane
or rocket is approaching the radar antenna at a ve-
locity of 1000 km/h, the received frequency exceeds the
transmitted frequency by 1850 Hz.
The reflection of a radar beam from an airplane, rock-
et, steamship or automobile is not the same as from a
reflector. The wavelength is commensurate with or sub-
stantially smaller than the size of the reflecting object
which, moreover, is of a complex shape. As the rays are
reflected by various points of the object, they (the rays)
Electrons 242
"interfere with one another and are scattered to the
sides. Owing to these two phenomena, the effective re-
fleeting surface of the object differs considerably from
its true surface. Calculations here are extremely compli-
cated and only the skill and experience of the radar
operator enable him to determine what kind of object
is encountered by the radar beam.
You have, of course, seen radar antennas: large sphe-
rical reflectors of wire lattice structure, always in motion,
surveying space. A great many different kinds of motion
can be imparted to the radar reflector. For example, the
beam can be moved so that it scans space in lines or cir-
cles. With this mode of operation, the path of flight of
the airplane can be followed besides determining its
range.
This technique is used to home aircraft into an airport
under conditions of a total lack of visibility. This job
may be done either by a person or even by an automatic
device.
A radar set can be "deceived". In the first place, the
object can be covered by a material that absorbs radio
waves. Coal dust or rubber can be used for this purpose.
Moreover, to reduce the reflection factor, the coating may
be corrugated so that the major part of the radiation
is disorderly scattered in all directions. If packages of
aluminium foil strips or metallized fibres are thrown
overboard from the airplane, the radar set is completely
disoriented. This trick was first used by the English
flying forces during World War II. A third way is to
fill space with false radio signals.
Radar is an interesting field of engineering and has
found extensive applications for peaceful purposes. It
is impossible to conceive of defense today without the
use of radar.
A competitor of radar is the laser (light amplification
by stimulated emission of radiatio.n). The principles or
6. 1_10 243
location and ranging with a laser in no way differ from
those described above.
Communication between spacecraft and the earth is
based on the principles of radar. Radio telescopes are
located so that they keep the spacecraft in view. The
antennas of these telescopes are of huge size, some hun-
dreds of metres in diameter. Such large antennas are
required so that they can transmit powerful signals and
pick up extremely weak signals from a radio transmit-
ter. Of vital importance, naturally, is a narrow radio
beam. If the antenna operates with a frequency of 2.2
thousand million oscillations per second (the wavelength
being about 1 em), the beam diverges only to a diameter
of 1000 km over the distance to the moon. True, when
the beam reaches Mars (300 million km away), its diam-
eter is already equal to 700 000 km.
Television
Since 99 out of 100 readers daily spend an hour or two
watching some TV program, it is only fair to say a few
words about this wonderful invention. We shall discuss
only the principles involved in television broadcasting
and reception. .
The idea of sending pictures over some distance con-
sists in the following. The picture to be transmitted is
divided into small squares. A physiologist can tell us
how small the square must be for the eye not to be able to
distinguish variations of brightness within it (the square).
The luminous energy of each portion of the picture
can be converted by the photoelectric effect into an elec-
tric signal. Next a way was devised for reading these sig-
nals. This is, of course, done in a strictly definite sequence
as in reading a hook. The electric signals modulate
the electromagnetic carrier wave in exactly the same
manner as in radio transmission. What happens further
Electrons 244
on is also identical to radio communications. The mod-
ulated oscillations are amplified and detected. The TV
receiver reconverts the electric pulses into a visible image.
The televisor tube in the transmitter is called an im-
age iconoscope, superorthicon or a vidicon. By means of
a lens in the TV camera, the image is focussed on a photo-
cathode. The most extensively used are cesium-oxide
or cesium-antimony cathodes. The photocathode is mount-
ed in an evacuated tube together with the target screen.
It is possible, in principle, to transmit an image by
successively projecting the luminous flux from each ele-
ment of the image. In this case, the photocurrent should
flow only during the short time each element is being
transmitted. Such operations would be inconvenient,
however, and the camera tube contains a large number
of photocells, rather than a single one, this number equal-
ling the number of elements into which the image being
transmitted is divided. This receiving screen is called
the target and is of mosaic design.
The mosaic target is a thin plate of mica with a large
number of grains of silver, insulated from one another
and coated by cesium oxide, applied on one side. Each
grain is a photocell. The other side of the mica plate
is coated by a metallic film. A small capacitor is formed
between each grain of the mosaic and the metal, and
is charged by electrons emitted from the photocath-
ode. It is evident that the charge of each small capaci-
tor is proportional to the brightness of the correspond-
ing spot on the image being transmitted.
Thus, a latent electric image of the object is produced
in the metallic plate. How do we take it from the plate?
By means of an electron beam which is made to scan
the plate in the same way as the eye passes along the
lines of print on the page of a book. The electron beam
plays the role of a key switch which, for an instant,
closes the electric circuit through a microcapacitor. At the
6. Radio
245
instant the circuit is closed, the current is uniquely re-
lated to the brightness of the image.
Each signal can and should be amplified many times
by the ordinary means applied in radio engineering. In
transmitting the image, the eye should not be able to
distinguish the fact that the electron beam is successive-
ly scanning the various spots of the luminous screen.
A complete image, obtained on the screen of the kine-
scope in the TV receiver, during one cycle of electron
beam motion, is called a frame. The frames must change
at a frequency sufficiently high so that persistence of
vision eliminates flickering of the picture being watched.
What no-flickering frame frequency should be chosen?
It must be a number related to the current frequency
in the mains. The "fact is that the pulsating voltage ap-
plied to the grid of the cathode-ray tube produces dark
and bright lines on the screen. These lines are station-
ary and imperceptible only if the frequency with which
the frames change is equal to or a multiple of the power
mains frequency. Continuous motion is seen when the
frames change with a frequency of at least 20 Hz; in tel-
e,ision a frequency of 25 Hz is taken but a small flick-
ering is still noticeable. I t is undesirable to change
the frames with a frequency of 50 Hz and therefore TV
designers resorted to what is called interlacing in the
scanning process. The frequency is left at 25 Hz, but
the electron beam (also called the exploring spot) first
scans the odd numbered lines and then the even num-
bered lines. Thus, the frequency with which the semi-
frames change equals 50 Hz and any flickering in the
brightness of the image becomes unnoticeable.
The frequency with which the frames change and the
line-by-line scanning should be strictly synchronized.
The technical details of this synchronization are beyond
the scope of our book. Consequently, we shall not ex-
plain why the number of lines must be odd and consist of
Electrons 246
several whole factors. In the USSR the frame is divided
into 625 lines, i.e, 54. Since 25 frames are changed per
second, lines are scanned with a frequency of 15 625 Hz.
From this condition follows the width of the band of
TV signal frequencies.
The lowest frequency of 50 Hz is the semiframe fre-
quency. The highest frequency is determined by the time
required to transmit a single element.
Simple calculations, which we shall not carry out here,
indicate that it. is necessary to take 6.5 MHz as the high-
er frequency. It follows that the carrier frequency must
be at least 40 or 50 MHz, because the frequency of the
carrier wave must be at least 6 to 7 times the bandwidth
of the transmitted frequencies. Now we understand why
only ultrashort waves can be used in television trans-
mission and why the range is limited to the line of sight
from the transmitting antenna.
This, however, is a slip; I should have written: was
limited. A breakthrough, enabling TV reception to be
carried out at any distance, is the use of communica-
tions satellites. The USSR first employed satellites fgr
this purpose. "Today the whole -of the Soviet Union :fs
covered by communications accomplished by making use
of a number of satellites.
Without going into the design of powerful TV sta-
tions, we shall only cite some interesting figures that
characterize the huge possibilities possessed by up-to-
date radio engineering in the amplification of signals.
Before amplification, an ordinary video signal has a
power up to 10-
3
W; the power amplifier increases this
power by a million times. This pulse of 1Q3 W is fed to
a parabolic antenna about 30 m in diameter. This an-
tenna produces a narrow directed beam which is reflect-
ed by the satellite. After the electromagnetic wave has
travelled about 35 000 km to the satellite, its power
equals only 10-1.
1
w.
6.Radio
The amplifier mounted on the repeater satellite in-
creases the power of this exceptionally weak signal
to about 10 W. When this signal, 'reflected from the
satellite, returns to the earth, its power is 10-
17
w.
Amplification returns the power of this video signal to
its initial 10-
3
w.
I think that ten years ago not even the most optimis-
tic engineer would have believed these figures.
Circuits
It is impossible to end this chapter devoted to radio
engineering without at least a few words about the new
revolution that we are witnessing in this field.
We have in mind the fantastic miniaturization of
all electronic devices and apparatus. This became fea-
sible in going over from apparatus built up of separate
elements, such as resistors, transistors, etc., connected
together by wiring, to electric circuits "drawn" by spe-
cial techniques on a piece of silicon several millimetres
ilsize.
. his new technology (at least one of its versions) con-
si . in using various kinds of masks and a series of chem-
ical compounds enabling p-type and n-type impuri-
ties to be introduced at the required places of a crystal
of silicon or germanium. Ion-beam treatment is used
for this purpose.
An electric circuit consisting of tens of thousands of
elements (I) is arranged on an area with linear dimen-
sions of about two millimetres. When we mentioned the
"drawing" of the circuit, the reader may have gotten
the impression that the problem involved only some sur-
face treatment of a piece of semiconductor material.
This is not so, however. The techniques are much more
complicated. Each electronic element is of three-dimen-
sional structure. Several layers containing different
Eledrons 248
amounts of impurities must be built up on a tiny area
of silicon.
How is this done? First a layer of oxide is applied
to the surface of the silicon. This is coated with a light-
sensitive material. The obtained layer-cake item is ex-
posed to ultraviolet light through a mask of the de-
signed shape. After developing, recesses are formed in the
surface of the piece of silicon only at the places where
the light passed through the mask.
The next stage consists in treating the electronic cir-
cuit being manufactured with hydrofluoric acid. The acid
removes the silicon dioxide but has no effect on either
the primary surface (i.e. silicon) or the light-sensitive
layer. In the final stage another solvent is used to re-
move the light-sensitive layer. As a result, a piece of
semiconductor material is coated with an insulating layer
-silicon dioxide-wherever required by the design.
The recesses of the required shape are bare silicon. These
recesses are treated with an ion beam to dope the
silicon with the required amount of impurities.
The manufacture of microelectronic circuits is one".f
the most rapidly advancing fields of engineering tod .
New ideas and new discoveries in semiconductor ph s-
ics indicate that the fantastic results that have already
been achieved are only the beginning.