200 Puzzling Problems

200 Puzzling Physics Problems is aimed at strengthening a student’s grasp of
the laws of physics by applying them to situations that are practical and to
problems that yield more easily to intuitive insight than to brute-force methods
and complex mathematics. The problems are chosen almost exclusively from
classical (i.e. non-quantum) physics, but are no easier for that. For the most
part, these problems are intriguingly posed in accessible non-technical language.
This requires the student to select the right framework in which to analyse the
situation and to make decisions about which branches of physics are involved.
The general level of sophistication needed to tackle most of the 200 problems is
that of the exceptional school student, the good undergraduate or the competent
graduate student. The book should be valuable to undergraduates preparing for
‘general physics’ papers, either on their own or in classes or seminars designed
for this purpose. It is even hoped that some physics professors will find the more
difficult questions challenging. By contrast, the mathematical demands made are
minimal, and do not go beyond elementary calculus. This intriguing book of
physics problems should prove not only instructive and challenging, but also fun.
peter gnädig graduated as a physicist from Roland Eötvös University (ELTE) in Budapest
in 1971 and received his PhD in theoretical particle physics there in 1980. Currently, he is
a researcher (in high energy physics) and a lecturer in the Department of Atomic Physics
at ELTE. Since 1985 he has been one of the leaders of the Hungarian Olympic team taking
part in the International Physics Olympiad. He is also the Physics Editor of KÖMAL, the
100-year-old Hungarian Mathematical and Physical Journal of Secondary Schools, which
publishes several challenging physics problems each month, as well as one of the organisers
of the formidable Hungarian Physics competition (the Eötvös Competition). Professor
Gnädig has written textbooks on the theory of distributions and the use of vector-calculus
in physics.
gyula honyek graduated as a physicist from Eötvös University (ELTE) in Budapest in

1975 and finished his PhD studies there in 1977, after which he stayed on as a researcher and
lecturer in the Department of General Physics. In 1984, following a two-year postgraduate
course, he was awarded a teacher’s degree in physics, and in 1985 transferred to the
teacher training school at ELTE. His current post is as mentor and teacher at Radnóti
Grammar School, Budapest. Since 1986 he has been one of the leaders and selectors
of the Hungarian team taking part in the International Physics Olympiad. He is also a
member of the editorial board of KÖMAL. As a co-author of a physics textbook series
for Hungarian secondary schools, he has wide experience of teaching physics at all levels.
ken riley read mathematics at the University of Cambridge and proceeded to a PhD
there in theoretical and experimental nuclear physics. He became a research associate in
elementary particle physics at Brookhaven, and then, having taken up a lectureship at the
Cavendish Laboratory, Cambridge, continued this research at the Rutherford Laboratory
and Stanford; in particular, he was involved in the discovery of a number of the early
baryonic resonances. As well as being Senior Tutor at Clare College, where he has
taught physics and mathematics for over 30 years, he has served on many committees
concerned with teaching and examining of these subjects at all levels of tertiary and
undergraduate education. He is also one of the authors of Mathematical Methods for
Physics and Engineering (Cambridge University Press).
200 Puzzling Physics Problems
P. Gnädig
Eötvös University, Budapest
G. Honyek
Radnóti Grammar School, Budapest
K. F. Riley
Cavendish Laboratory, Fellow of Clare College, Cambridge
cambridge university press
Cambridge, New York, Melbourne, Madrid, Cape Town, Singapore, Sa~o Paulo,
Delhi, Dubai, Tokyo, Mexico City
Cambridge University Press

The Edinburgh Building, Cambridge CB2 8RU, UK
Published in the United States of America by Cambridge University Press, New York
www.cambridge.org
Information on this title: www.cambridge.org/9780521773065

c Cambridge University Press 2001
This publication is in copyright. Subject to statutory exception

and to the provisions of relevant collective licensing agreements,
no reproduction of any part may take place without
the written permission of Cambridge University Press
First published 2001

6th printing 2011
Printed in the United Kingdom at the University Press, Cambridge
A catalogue record for this publication is available from the British Library
Library of Congress Cataloguing in Publication data

Gnädig, Péter, 1947–
200 Puzzling Physics Problems / P. Gnädig, G. Honyek, K. F. Riley.
p. cm.
Includes bibliographical references and index.
ISBN 0 521 77306 7 – ISBN 0 521 77480 2 (pb.)
1. Physics–Problems, exercises, etc. I. Title: Two hundred puzzling physics problems.
II. Honyek, G. (Gyula), 1951– III. Riley, K. F. (Kenneth Franklin), 1936– IV. Title.
QC32.G52 2001
530 .076–dc21 00-053005 CIP
ISBN 978-0-521-77480-2 Paperback
Cambridge University Press has no responsibility for the persistence or accuracy of URLs
for external or third-party internet websites referred to in this publication, and does not
guarantee that any content on such websites is, or will remain, accurae or appropriate.
Information regarding prices, travel timetables and other factual information given in this
work are correct at the time of first printing but Cambridge University Press does not
guarantee the accuracy of such information thereafter.
Contents
Preface page vii

How to use this book x
Thematic order of the problems xi
Physical constants xiii
Problems 1
Hints 50
Solutions 69
v
Preface
In our experience, an understanding of the laws of physics is best acquired

by applying them to practical problems. Frequently, however, the problems
appearing in textbooks can be solved only through long, complex calcu-
lations, which tend to be mechanical and boring, and often drudgery for
students. Sometimes, even the best of these students, the ones who possess
all the necessary skills, may feel that such problems are not attractive enough
to them, and the tedious calculations involved do not allow their ‘creativity’
(genius?) to shine through.
This little book aims to demonstrate that not all physics problems are like
that, and we hope that you will be intrigued by questions such as:
• How is the length of the day related to the side of the road on which
traffic travels?
• Why are Fosbury floppers more successful than Western rollers?
• How far below ground must the water cavity that feeds Old Faithful
be?
• How high could the tallest mountain on Mars be?
• What is the shape of the water bell in an ornamental fountain?
• How does the way a pencil falls when stood on its point depend
upon friction?
• Would a motionless string reaching into the sky be evidence for
UFOs?
• How does a positron move when dropped in a Faraday cage?
• What would be the high-jump record on the Moon?
• Why are nocturnal insects fatally attracted to light sources?
• How much brighter is sunlight than moonlight?
• How quickly does a fire hose unroll?
vii
viii 200 Puzzling Physics Problems
• How do you arrange two magnets so that the mutual couples they
experience are not equal and opposite?
• How long would it take to defrost an 8-tonne Siberian mammoth?
• What perils face titanium-eating little green men who devour their
own planet?
• What is the direction of the electric field due to an uniformly
charged rod?
• What is the catch in an energy-generating capacitor?
• What is the equivalent resistance of an n-dimensional cube of resis-
tors?
• What factors determine the period of a sand-glass egg timer?
• How does a unipolar dynamo work?
• How ‘deep’ is an electron lying in a box?
These, and some 180 others, are problems that can be solved elegantly by an
appropriate choice of variables or coordinates, an unusual way of thinking,
or some ‘clever’ idea or analogy. When such an inspiration or eureka moment
occurs, the solution often follows after only a few lines of calculation or
brief mental reasoning, and the student feels justifiably pleased with him-
or herself.
Logic in itself is not sufficient. Nobody can guess these creative approaches
without knowing and understanding the basic laws of physics. Accordingly,
we would not encourage anybody to tackle these problems without first
having studied the subject in some depth. Although successful solutions to
the problems posed are clearly the principal goal, we should add that success
is not to be measured by this alone. Whatever help you, the reader, may seek,
and whatever stage you reach in the solution to a problem, it will hopefully
bring you both enlightenment and delight. We are sure that some solutions
will lead you to say ‘how clever’, others to say ‘how nice’, and yet others to
say ‘how obvious or heavy-handed’! Our aim is to show you as many useful
‘tricks’ as possible in order to enlarge your problem-solving arsenal. We wish
you to use this book with delight and profit, and if you come across further
similar ‘puzzling’ physics problems, we would ask you to share them with
others (and send them to the authors).
The book contains 200 interesting problems collected by the authors over
the course of many years. Some were invented by us, others are from the
Hungarian ‘Secondary School Mathematics and Physics Papers’ which span
more than 100 years. Problems and ideas from various Hungarian and
international physics contests, as well as the Cambridge Colleges’ entrance
examination, have also been used, often after rewording. We have also been
Preface ix
guided by the suggestions and remarks of our colleagues. In particular, we

would like to thank Masaaki Kato for several helpful observations and
suggested clarifications and Alfonso Diaz-Jimenez for an interesting note
on the launching of space probes (see solution 17). It is impossible to
determine the original authors of most of the physics problems appearing in
the international ‘ideas-market’. Nevertheless, some of the inventors of the
most puzzling problems deserve our special thanks. They include Tibor Bı́ró,
László Holics, Frederick Károlyházy, George Marx, Ervin Szegedi and István
Varga. We thank them and the other people, known and unknown, who have
authored, elaborated and improved upon ‘puzzling’ physics problems.
P.G. G.H.
Budapest 2000
K.F.R.
Cambridge 2000
How to use this book
The following chapter contains the problems. They do not follow each other
in any particular thematic order, but more or less in order of difficulty,
or in groups requiring similar methods of solution. In any case, some of
the problems could not be unambiguously labelled as belonging to, say,
mechanics or thermodynamics or electromagnetics. Nature’s secrets are not
revealed according to the titles of the sections in a text book, but rather
draw on ideas from various areas and usually in a complex manner. It is
part of our task to find out what type of problem we are facing. However,
for information, the reader can find a list of topics, and the problems that
more or less belong to these topics, on the following page. Some problems
are listed under more than one heading. The symbols and numerical values
of the principal physical constants are then given, together with astronomical
data and some properties of material.
The majority of the problems are not easy; some of them are definitely dif-
ficult. You, the reader, are naturally encouraged to try to solve the problems
on your own and, obviously, if you do, you will get the greatest pleasure.
If you are unable to achieve this, you should not give up, but turn to the
relevant page of the short hints chapter. In most cases this will help, though
it will not give the complete solution, and the details still have to be worked
out. Once you have done this and want to check your result (or if you have
completely given up and only want to see the solution), the last chapter
should be consulted.
Problems whose solutions require similar reasoning usually follow each
other. But if a particular problem relates to another elsewhere in this book,
you will find a cross-reference in the relevant hint or solution. Those requiring
especially difficult reasoning or mathematically complicated calculations are
marked by one or two asterisks.
Some problems are included whose solutions raise further questions that
are beyond the scope of this book. Points or issues worth further considera-
tion are indicated at the end of the respective solutions, but the answers are
not given.
x
Thematic order of the problems
Kinematics: 1, 3, 5, 36, 37, 38*, 40, 41, 64, 65*, 66, 84*, 86*.
Dynamics: 2, 7, 8, 12, 13, 24, 32*, 33, 34, 35, 37, 38*, 39*, 70*, 73*, 77, 78*,
79*, 80**, 82, 83, 85**, 90, 154*, 183*, 184*, 186*, 193*, 194.
Gravitation: 15, 16, 17, 18, 32*, 81**, 87, 88, 109, 110*, 111*, 112*, 116,
134*.
Mechanical energy: 6, 7, 17, 18, 32*, 51, 107.
Collisions: 20, 45, 46, 47, 48, 71, 72*, 93, 94, 144*, 194, 195.
Mechanics of rigid bodies: 39*, 42**, 58, 60*, 61**, 94, 95*, 96, 97*, 98,
99**.
Statics: 9, 10*, 11, 14*, 25, 26, 43, 44, 67, 68, 69*.
Ropes, chains: 4, 67, 81**, 100, 101*, 102**, 103*, 104*, 105**, 106*, 108**.
Liquids, gases: 19, 27, 28, 49*, 50, 70*, 73*, 74, 75*, 91*, 115**, 143, 200.
Surface tension: 29, 62, 63, 129, 130*, 131*, 132**, 143, 199*.
Thermodynamics: 20, 21*, 133, 135**, 136, 145, 146*, 147, 148.
Phase transitions: 134*, 137*, 138, 140*, 141*.
Optics: 52, 53, 54, 55, 56, 125*, 126, 127, 128*.
Electrostatics: 41, 90, 91*, 92, 113*, 114, 117*, 118, 121, 122, 123*, 124*,
149, 150, 151*, 152, 155, 156, 157, 183*, 192*, 193*.
Magnetostatics: 89**, 119, 120**, 153*, 154*, 172, 186*.
Electric currents: 22, 23, 158, 159, 160*, 161, 162*, 163*, 164*, 165, 169,
170*, 172.
Electromagnetism: 30, 31, 166, 167, 168*, 171*, 173*, 174*, 175*, 176, 177,
178*, 179, 180, 181*, 182*, 184*, 185*, 186*, 187*.
Atoms and particles: 93, 188, 189*, 190*, 191, 194, 195, 196, 197*, 198*.
Dimensional analysis, scaling, estimations: 15, 57, 58, 59*, 76*, 77, 126, 139,
142, 185*, 199*.
*, ** A single or double asterisk indicates those problems that require especially difficult reasoning or
mathematically complicated calculations.
xi
Physical constants
Gravitational constant, G 6.673 × 10−11 N m2 kg−2

Speed of light (in vacuum), c 2.998 × 108 m s−1
Elementary charge, e 1.602 × 10−19 C
Electron mass, me 9.109 × 10−31 kg
Proton mass, mp 1.673 × 10−27 kg
Boltzmann constant, k 1.381 × 10−23 J K−1
Planck constant, h 6.626 × 10−34 J s
Avogadro constant, NA 6.022 × 1023 mol−1
Gas constant, R 8.315 J mol−1 K−1
Permittivity of free space, ε0 8.854 × 10−12 C V −1 m−1
Coulomb constant, k = 1/4πε0 8.987 × 109 V m C−1
Permeability of free space, µ0 4π × 10−7 V s2 C−1 m−1
Some astronomical data

Mean radius of the Earth, R 6371 km
Sun–Earth distance (Astronomical Unit, AU) 1.49 × 108 km
Mean density of the Earth, ρ 5520 kg m−3
Free-fall acceleration at the Earth’s surface, g 9.81 m s−2
Some physical properties

Surface tension of water, γ 0.073 N m−1
Heat of vaporisation of water, L 2256 kJ kg−1 = 40.6 kJ mol−1
Tensile strength of steel, σ 500–2000 MPa
xiii
xiv Physical constants
Densitiesa , ρ (kg m−3 )

Hydrogen 0.0899 Titanium 4510
Helium 0.1786 Iron 7860
Air 1.293 Mercury 13 550
Water (at 4 ◦ C) 1000 Platinum 21 450
a Densities quoted in normal state.
Optical Refractive Indicesb , n

Water 1.33 Glass 1.5–1.8
Ice 1.31 Diamond 2.42
b At λ = 590 nm.
Problems
P1 Three small snails are each at a vertex of an equilateral triangle of

side 60 cm. The first sets out towards the second, the second towards the
third and the third towards the first, with a uniform speed of 5 cm min−1 .
During their motion each of them always heads towards its respective target
snail. How much time has elapsed, and what distance do the snails cover,
before they meet? What is the equation of their paths? If the snails are
considered as point-masses, how many times does each circle their ultimate
meeting point?
P2 A small object is at rest on the edge of a horizontal table. It is pushed
in such a way that it falls off the other side of the table, which is 1 m wide,
after 2 s. Does the object have wheels?
P3 A boat can travel at a speed of 3 m s−1 on still water. A boatman
wants to cross a river whilst covering the shortest possible distance. In what
direction should he row with respect to the bank if the speed of the water
is (i) 2 m s−1 , (ii) 4 m s−1 ? Assume that the speed of the water is the same
everywhere.
P4 A long, thin, pliable carpet is laid on the floor. One end of the carpet
is bent back and then pulled backwards with constant unit velocity, just
above the part of the carpet which is still at rest on the floor.
v=1
0 1
Find the speed of the centre of mass of the moving part. What is the
minimum force needed to pull the moving part, if the carpet has unit length
and unit mass?
1
2 200 Puzzling Physics Problems
P5 Four snails travel in uniform, rectilinear motion on a very large plane

surface. The directions of their paths are random, (but not parallel, i.e. any
two snails could meet), but no more than two snail paths can cross at any one
point. Five of the (4 × 3)/2 = 6 possible encounters have already occurred.
Can we state with certainty that the sixth encounter will also occur?
P6 Two 20-g flatworms climb over a very thin wall, 10 cm high. One of
the worms is 20 cm long, the other is wider and only 10 cm long. Which of
them has done more work against gravity when half of it is over the top of
the wall? What is the ratio of the amounts of work done by the two worms?
P7 A man of height h0 = 2 m is bungee jumping from a platform situated
a height h = 25 m above a lake. One end of an elastic rope is attached to his
foot and the other end is fixed to the platform. He starts falling from rest in
a vertical position.
The length and elastic properties of the rope are chosen so that his speed
will have been reduced to zero at the instant when his head reaches the
surface of the water. Ultimately the jumper is hanging from the rope, with
his head 8 m above the water.
(i) Find the unstretched length of the rope.
(ii) Find the maximum speed and acceleration achieved during the jump.
P8 An iceberg is in the form of an upright regular pyramid of which
10 m shows above the water surface. Ignoring any induced motion of the
water, find the period of small vertical oscillations of the berg. The density
of ice is 900 kg m−3 .
P9 The suspension springs of all four wheels of a car are identical. By
how much does the body of the car (considered rigid) rise above each of
the wheels when its right front wheel is parked on an 8-cm-high pavement?
Does the result change when the car is parked with both right wheels on
Problems 3
the pavement? Does the result depend on the number and positions of the
people sitting in the car?
P10∗ In Victor Hugo’s novel les Misérables, the main character Jean
Valjean, an escaped prisoner, was noted for his ability to climb up the corner
formed by the intersection of two vertical perpendicular walls. Find the
minimum force with which he had to push on the walls whilst climbing.
What is the minimum coefficient of static friction required for him to be
able to perform such a feat?
P11 A sphere, made of two non-identical homogeneous hemispheres
stuck together, is placed on a plane inclined at an angle of 30◦ to the
horizontal. Can the sphere remain in equilibrium on the inclined plane?
P12 A small, elastic ball is dropped vertically onto a long plane inclined
at an angle α to the horizontal. Is it true that the distances between con-
secutive bouncing points grow as in an arithmetic progression? Assume that
collisions are perfectly elastic and that air resistance can be neglected.
P13 A small hamster is put into a circular wheel-cage, which has a
frictionless central pivot. A horizontal platform is fixed to the wheel below
the pivot. Initially, the hamster is at rest at one end of the platform.
When the platform is released the hamster starts running, but, because of
the hamster’s motion, the platform and wheel remain stationary. Determine
how the hamster moves.
P14∗ A bicycle is supported so that it is prevented from falling sideways
but can move forwards or backwards; its pedals are in their highest and low-
est positions. A student crouches beside the bicycle and applies a horizontal
force, directed towards the back wheel, to the lower pedal.
(i) Which way does the bicycle move?
(ii) Does the chain-wheel rotate in the same or opposite sense as the rear
wheel?
(iii) Which way does the lower pedal move relative to the ground?
P15 If the solar system were proportionally reduced so that the average
distance between the Sun and the Earth were 1 m, how long would a year
be? Take the density of matter to be unchanged.
1m
Sun Earth
P16 If the mass of each of the members of a binary star were the same
as that of the Sun, and their distance apart were equal to the Sun–Earth
distance, what would be their period of revolution?
P17 (i) What is the minimum launch speed required to put a satellite
into a circular orbit?
(ii) How many times higher is the energy required to launch a satellite into
a polar orbit than that necessary to put it into an Equatorial one?
(iii) What initial speed must a space probe have if it is to leave the
gravitational field of the Earth?
(iv) Which requires a higher initial energy for the space probe – leaving the
solar system or hitting the Sun?
P18 A rocket is intended to leave the Earth’s gravitational field. The fuel
in its main engine is a little less than the amount that is necessary, and an
auxiliary engine, only capable of operating for a short time, has to be used
as well. When is it best to switch on the auxiliary engine: at take-off, or
when the rocket has nearly stopped with respect to the Earth, or does it not
matter?
P19 A steel ball with a volume of 1 cm3 is sinking at a speed of 1 cm s−1
in a closed jar filled with honey. What is the momentum of the honey if its
density is 2 g cm−3 ?
1 cm s−1
Problems 5
P20 A gas of temperature T is enclosed in a container whose walls are

(initially) at temperature T1 . Does the gas exert a higher pressure on the
walls of the container when T1 < T or when T1 > T ?
P21∗ Consider two identical iron spheres, one of which lies on a
thermally insulating plate, whilst the other hangs from an insulating thread.
Equal amounts of heat are given to the two spheres. Which will have the
higher temperature?
P22 Two (non-physics) students, A and B, living in neighbouring college
rooms, decided to economise by connecting their ceiling lights in series. They
agreed that each would install a 100-W bulb in their own rooms and that
they would pay equal shares of the electricity bill. However, both decided to
try to get better lighting at the other’s expense; A installed a 200-W bulb and
B installed a 50-W bulb. Which student subsequently failed the end-of-term
examinations?
P23 If a battery of voltage V is connected across terminals I of the
black box shown in the figure, a voltmeter connected to terminals II gives a
reading of V /2; while if the battery is connected to terminals II, a voltmeter
across terminals I reads V .
I II
The black box contains only passive circuit elements. What are they?
P24 A bucket of water is suspended from a fixed point by a rope. The
bucket is set in motion and the system swings as a pendulum. However, the
bucket leaks and the water slowly flows out of the bottom of it. How does
the period of the swinging motion change as the water is lost?
P25 An empty cylindrical beaker of mass 100 g, radius 30 mm and neg-
ligible wall thickness, has its centre of gravity 100 mm above its base. To
what depth should it be filled with water so as to make it as stable as
possible?
P26 Fish soup is prepared in a hemispherical copper bowl of diameter

40 cm. The bowl is placed into the water of a lake to cool down and floats
with 10 cm of its depth immersed.
A point on the rim of the bowl is pulled upwards through 10 cm, by a

chain fastened to it. Does water flow into the bowl?
P27 A circular hole of radius r at the bottom of an initially full water
container is sealed by a ball of mass m and radius R (> r). The depth of the
water is now slowly reduced, and when it reaches a certain value, h0 , the ball
rises out of the hole. Find h0 .
m R h0
P28 Soap bubbles filled with helium float in air. Which has the greater
mass – the wall of a bubble or the gas enclosed within it?
P29 Water which wets the walls of a vertical capillary tube rises to a
height H within it. Three ‘gallows’, (a), (b) and (c), are made from the same
tubing, and one end of each is placed into a large dish filled with water, as
shown in the figure.
H (c)
(a) (b)
H´
Does the water flow out at the other ends of the capillary tubes?
Problems 7
P30 A charged spherical capacitor slowly discharges as a result of the

slight conductivity of the dielectric between its concentric plates. What are
the magnitude and direction of the magnetic field caused by the resulting
electric current?
P31 An electrically charged conducting sphere ‘pulses’ radially, i.e. its
radius changes periodically with a fixed amplitude (see figure). The charges on
its surface – acting as many dipole antennae – emit electromagnetic radiation.
What is the net pattern of radiation from the sphere?
+ + +
+ +
+ +
+ +
+ + +
P32∗ How high would the male world-record holder jump (at an indoor
competition!) on the Moon?
P33 A small steel ball B is at rest on the edge of a table of height 1 m.
Another steel ball A, used as the bob of a metre-long simple pendulum,
is released from rest with the pendulum suspension horizontal, and swings
against B as shown in the figure. The masses of the balls are identical and
the collision is elastic.
A 1m
1m
Considering the motion of B only up until the moment it first hits the
ground:
(i) Which ball is in motion for the longer time?
(ii) Which ball covers the greater distance?
P34 A small bob is fixed to one end of a string of length 50 cm. As a

consequence of the appropriate forced motion of the other end of the string,
the bob moves in a vertical circle of radius 50 cm with a uniform speed of
3.0 m s−1 . Plot, at 15◦ intervals on the circular path, the trajectories of both
ends of the string, indicating on each the points belonging together.
P35 A point P is located above an inclined plane. It is possible to reach
the plane by sliding under gravity down a straight frictionless wire, joining P
to some point P on the plane. How should P be chosen so as to minimise
the time taken?
P36 The minute hand of a church clock is twice as long as the hour
hand. At what time after midnight does the end of the minute hand move
away from the end of the hour hand at the fastest rate?
P37 What is the maximum angle to the horizontal at which a stone can
be thrown and always be moving away from the thrower?
P38∗ A tree-trunk of diameter 20 cm lies in a horizontal field. A lazy
grasshopper wants to jump over the trunk. Find the minimum take-off speed
of the grasshopper that will suffice. (Air resistance is negligible.)
P39∗ A straight uniform rigid hair lies on a smooth table; at each end
of the hair sits a flea. Show that if the mass M of the hair is not too great
relative to that m of each of the fleas, they can, by simultaneous jumps with
the same speed and angle of take-off, exchange ends without colliding in
mid-air.
P40 A fountain consists of a small hemispherical rose (sprayer) which
lies on the surface of the water in a basin, as illustrated in the figure. The
rose has many evenly distributed small holes in it, through which water
spurts at the same speed in all directions.
What is the shape of the water ‘bell’ formed by the jets?

P41 A particle of mass m carries an electric charge Q and is subject to
the combined action of gravity and a uniform horizontal electric field of
strength E. It is projected with speed v in the vertical plane parallel to the
field and at an angle θ to the horizontal. What is the maximum distance the
particle can travel horizontally before it is next level with its starting point?
P42∗∗ A uniform rod of mass m and length ' is supported horizontally
Problems 9
at its ends by my two forefingers. Whilst I am slowly bringing my fingers

together to meet under the centre of the rod, it slides on either one or other
of them. How much work do I have to do during the process if the coefficient
of static friction is µstat , and that of kinetic friction is µkin (µkin ≤ µstat )?
P43 Four identical bricks are placed on top of each other at the edge of
a table. Is it possible to slide them horizontally across each other in such a
way that the projection of the topmost one is completely outside the table?
What is the theoretical limit to the displacement of the topmost brick if the
number of bricks is arbitrarily increased?
P44 A plate, bent at right angles along its centre line, is placed onto a
horizontal fixed cylinder of radius R as shown in the figure.
How large does the coefficient of static friction between the cylinder and
the plate need to be if the plate is not to slip off the cylinder?
P45 Two elastic balls of masses m1 and m2 are placed on top of each other
(with a small gap between them) and then dropped onto the ground. What
is the ratio m1 /m2 , for which the upper ball ultimately receives the largest
possible fraction of the total energy? What ratio of masses is necessary if
the upper ball is to bounce as high as possible?
m1
m2
P46 An executive toy consists of three suspended steel balls of masses

M, µ and m arranged in that order with their centres in a horizontal line.
The ball of mass M is drawn aside in their common plane until its centre
has been raised by h and is then released. If M = m and all collisions are
elastic, how must µ be chosen so that the ball of mass m rises to the greatest
possible height? What is this height? (Neglect multiple collisions.)
P47 Two identical dumb-bells move towards each other on a horizontal
air-cushioned table, as shown in the figure. Each can be considered as two
point masses m joined by a weightless rod of length 2'. Initially, they are not
rotating. Describe the motion of the dumb-bells after their elastic collision.
Plot the speeds of the centres of mass of the dumb-bells as a function of
time.
m
v
2F m
m v
2F
P48 Two small identical smooth blocks A and B are free to slide on √ a
frozen lake. They are joined together by a light elastic rope of length 2L
which has the property that it stretches very little when the rope becomes
taut. At time t = 0, A is at rest at x = y = 0 and B is at x = L, y = 0
moving in the positive y-direction with speed V. Determine the positions and
velocities of A and B at times (i) t = 2L/V and (ii) t = 100L/V.
P49∗ After a tap above an empty rectangular basin has been opened, the
basin fills with water in a time T1 . After the tap has been closed, opening a
plug-hole at the bottom of the basin empties it in a time T2 . What happens
if both the tap and the plug-hole are open? What ratio of T1 /T2 can cause
the basin to overflow? As a specific case, let T1 = 3 minutes and T2 = 2
minutes.
P50 A cylindrical vessel of height h and radius a is two-thirds filled with
liquid. It is rotated with constant angular velocity ω about its axis, which
is vertical. Neglecting any surface tension effects, find an expression for the
greatest angular velocity of rotation Ω for which the liquid does not spill
over the edge of the vessel.
P51 Peter, who was standing by a racetrack, calculated that as one of
the cars, in accelerating from rest to a speed of 100 km h−1 , used up x litres
of fuel, it could increase its speed to 200 km h−1 , by using a further 3x litres
of fuel. Peter, who has learned in physics that kinetic energy is proportional
to the square of the speed, assumed that the energy content of the fuel
was mainly converted into kinetic energy, i.e. he neglected air resistance and
other types of friction.
A railway runs by the racetrack. Paul, who also knows some physics,
saw the start of the race from the window of a train travelling at a speed
of 100 km h−1 in the opposite direction to that of the car. He reasoned as
Problems 11
follows: since the car’s speed increased from 100 to 200 km h−1 during the
first stage, when the car accelerates from 200 to 300 km h−1 in the second
stage, it will need (3002 − 2002 )/(2002 − 1002 ) x = (5/3)x litres of fuel.
Who is right, Peter or Paul?
P52 The distance between a screen and a light source lined up on an
optical bench is 120 cm. When a lens is moved along the line joining them,
sharp images of the source can be obtained at two lens positions; the (linear)
size ratio of these two images is 1 : 9. What is the focal length of the lens?
Which image is the brighter? Determine the ratio of the brightness values
of the two images.
P53 A short-sighted person takes off his glasses and observes a fixed
object through them, while moving the glasses away from his eyes. He is
surprised to see that at first, the object looks smaller and smaller, but then
becomes larger and larger. What is the reason for this?
P54 A glass prism whose cross-section is an isosceles triangle stands with
its (horizontal) base in water; the angles that its two equal sides make with
the base are each θ.
h h
Water
An incident ray of light, above and parallel to the water surface and
perpendicular to the prism’s axis, is internally reflected at the glass–water
interface and subsequently re-emerges into the air. Taking the refractive
indices of glass and water to be 32 and 43 , respectively, show that θ must be
at least 25.9◦ .
P55 A glass prism in the shape of a quarter-cylinder lies on a horizontal
table. A uniform, horizontal light beam falls on its vertical plane surface, as
Light
R
n
If the radius of the cylinder is R = 5 cm and the refractive index of the

glass is n = 1.5, where, on the table beyond the cylinder, will a patch of light
be found?
P56 How much brighter is sunlight than moonlight? The albedo (reflec-
tivity) of the Moon is α = 0.07.
P57 Annie and her very tall boyfriend Andy like jogging together. They
notice that when running they move at more or less the same speed, but
that Andy is always faster when they are walking. How can this difference
between running and walking be explained using physical arguments?
P58 A simple pendulum and a homogeneous rod pivoted at its end are
released from horizontal positions. What is the ratio of their periods of
swing if their lengths are identical?
F F
P59∗ A helicopter can hover when the power output of its engine is P .
A second helicopter is an exact copy of the first one, but its linear dimensions
are half those of the original. What power output is needed to enable this
second helicopter to hover?
P60∗ A uniform rod is placed with one end on the edge of a table in
a nearly vertical position and is then released from rest. Find the angle
it makes with the vertical at the moment it loses contact with the table.
Investigate the following two extreme cases:
(a)
(b)
Problems 13
(i) The edge of the table is smooth (friction is negligible) but has a
small, single-step groove as shown in figure (a).
(ii) The edge of the table is rough (friction is large) and very sharp, which
means that the radius of curvature of the edge is much smaller than
the flat end-face of the rod. Half of the end-face protrudes beyond
the table edge (see figure (b)), with the result that when it is released
from rest the rod ‘pivots’ about the edge. The rod is much longer
than its diameter.
P61∗∗ A pencil is placed vertically on a table with its point downwards.

It is then released and tumbles over. How does the direction in which the
point moves, relative to that in which the pencil falls, depend upon the
coefficient of friction? Will the pencil point lose contact with the table (other
than when the ‘shoulder’ of the pencil ultimately comes into contact with
the table)?
? ?
P62 Two soap bubbles of radii R1 and R2 are joined by a straw. Air
goes from one bubble to the other (which one?) and a single bubble of
radius R3 is formed. What is the surface tension of the soap solution if the
atmospheric pressure is p0 ? Is measuring three such radii a suitable method
for determining the surface tension of liquids?
P63 Water, which wets glass, is stuck between two parallel glass plates.
The distance between the plates is d, and the diameter of the trapped water
‘disc’ is D d.
D
What is the force acting between the plates?

P64 A spider has fastened one end of a ‘super-elastic’ silk thread of length
1 m to a vertical wall. A small caterpillar is sitting somewhere on the thread.
v0
The hungry spider, whilst not moving from its original position, starts
pulling in the other end of the thread with uniform speed, v0 = 1 cm s−1 .
Meanwhile, the caterpillar starts fleeing towards the wall with a uniform
speed of 1 mm s−1 with respect to the moving thread. Will the caterpillar
reach the wall?
P65∗ How does the solution to the previous problem change if the spider
does not sit in one place, but moves (away from the wall) taking with it the
end of the thread?
P66 Nails are driven horizontally into a vertically placed drawing-board.
As shown in the figure, a small steel ball is dropped from point A and reaches
point B by bouncing elastically on the protruding nails (which are not shown
in the figure).
2m
A
1m
Is it possible to arrange the nails so that:

(i) The ball gets from point A to point B more quickly than if it had
slid without friction down the shortest path, i.e. along the straight
line AB?
(ii) The ball reaches point B in less than 0.4 s?
P67 One end of a rope is fixed to a vertical wall and the other end pulled
by a horizontal force of 20 N. The shape of the flexible rope is shown in the
figure. Find its mass.
20 N
Problems 15
P68 Find the angle to which a pair of compasses should be opened

in order to have the pivot as elevated as possible when the compasses are
suspended from a string attached to one of the points, as shown in the figure.
Assume that the lengths of the compass arms are equal.
P69∗ Threads of lengths h1 , h2 and h3 are fastened to the vertices of a

homogeneous triangular plate of weight W. The other ends of the threads
are fastened to a common point, as shown in the figure.
h2
h1 h3
What is the tension in each thread, expressed in terms of the lengths of

the threads and the weight of the plate?
P70∗ A tanker full of liquid is parked at rest on a horizontal road. The
brake has not been applied, and it may be supposed that the tanker can
move without friction.
In which direction will the tanker move after the tap on the vertical outlet
pipe, which is situated at the rear of the tanker, has been opened? Will the
tanker continue to move in this direction?
P71 Two small beads slide without friction, one on each of two long,
horizontal, parallel, fixed rods set a distance d apart. The masses of the beads
are m and M, and they carry respective charges of q and Q. Initially, the
larger mass M is at rest and the other one is far away approaching it at
speed v0 .
v0
m, q v=0 d
M, Q
Describe the subsequent motion of the beads.

P72∗ Beads of equal mass are strung at equal distances on a long,
horizontal wire. The beads are initially at rest but can move without friction.
F d
m m m m m m
One of the beads is continuously accelerated (towards the right) by a

constant force F. What are the speeds of the accelerated bead and the front
of the ‘shock wave’, after a long time, if the collisions of the beads are:
(i) completely inelastic,
(ii) perfectly elastic?
P73∗ A table and a large jug are placed on the platform of a weighing
machine and a barrel of beer is placed on the table with its tap above the
jug. Describe how the reading of the machine varies with time after the tap
has been opened and the beer runs into the jug.
P74 A jet of water strikes a horizontal gutter of semicircular cross-
section obliquely, as shown in the figure. The jet lies in the vertical plane
that contains the centre-line of the gutter.
Water jet
Calculate the ratio of the quantities of water flowing out at the two ends
of the gutter as a function of the angle of incidence α of the jet.
Problems 17
P75∗ An open-topped vertical tube of diameter D is filled with water up

to a height h. The narrow bottom-end of the tube, of diameter d, is closed
by a stop as shown in the figure.
When the stop is removed, the water starts flowing out through the bottom
√
orifice with approximate speed v = 2gh. However, this speed is reached by
the liquid only after a certain time τ. Obtain an estimate of the order of
magnitude of τ. What is the acceleration of the lowest layer of water at the
moment when the stop is removed? Ignore viscous effects.
P76∗ Obtain a reasoned estimate of the time it takes for the sand to run
down through an egg-timer. Use realistic data.
H
d
P77 A small bob joins two light unstretched, identical springs, anchored
at their far ends and arranged along a straight line, as shown in the figure.
F0 F0
The bob is displaced in a direction perpendicular to the line of the springs

by 1 cm and then released. The period of the ensuing vibration of the bob is
2 s. Find the period of the vibration if the bob were displaced by 2 cm before
release. The unstretched length of the springs is '0 1 cm, and gravity is
to be ignored.
P78∗ One end of a light, weak spring, of unstretched length L and force
constant k, is fixed to a pivot, and a body of mass m is attached to its other
end. The spring is released from an unstretched, horizontal position, as in
the figure.
L m
What is the length of the spring when it reaches a vertical position?

(Describing a spring as weak implies that mg kL, and that the tension in
the spring is directly proportional to its extension at all times.)
P79∗ A heavy body of mass m hangs on a flexible thread in a railway
carriage which moves at speed v0 on a train-safety test track, as shown in
the figure.
?
v0
The carriage is brought to rest by a strong but uniform braking. Can the
pendulum travel through 180◦ , so that the taut thread reaches the vertical?
P80∗∗ A glass partially filled with water is fastened to a wedge that
slides, without friction, down a large plane inclined at an angle α as shown
in the figure. The mass of the inclined plane is M, the combined mass of the
wedge, the glass and the water is m.
M
a
If there were no motion the water surface would be horizontal. What

angle will it ultimately make with the inclined plane if
(i) the inclined plane is fixed,
(ii) the inclined plane can move freely in the horizontal direction?
Examine also the case in which m M. What happens if the handle of the
inclined plane is shaken in a periodic manner, but one that is such that it
does not cause the wedge to rise off the plane?
P81∗∗ If someone found a motionless string reaching vertically up into
the sky and hanging down nearly to the ground, should that person consider
Problems 19
it as an evidence for UFOs, or could there be an ‘Earthly’ explanation in

agreement with the well-known laws of physics? How long would the string
need to be?
P82 There is a parabolic-shaped bridge across a river of width 100 m.
The highest point of the bridge is 5 m above the level of the banks. A car
of mass 1000 kg is crossing the bridge at a constant speed of 20 m s−1 .
v m
h
Using the notation indicated in the figure, find the force exerted on the
bridge by the car when it is:
(i) at the highest point of the bridge,
(ii) three-quarters of the way across.
(Ignore air resistance and take g as 10 m s−2 .)
P83 A point mass of 0.5 kg moving with a constant speed of 5 m s−1
on an elliptical track experiences an outward force of 10 N when at either
endpoint of the major axis and a similar force of 1.25 N at each end of the
minor axis. How long are the axes of the ellipse?
P84∗ A boatman sets off from one bank of a straight, uniform canal for
a mark directly opposite the starting point. The speed of the water flowing
in the canal is v everywhere. The boatman rows steadily at such a rate that,
were there no current, the boat’s speed would also be v. He always sets
the boat’s course in the direction of the mark, but the water carries him
downstream. Fortunately he never tires! How far downstream does the water
carry the boat? What trajectory does it follow with respect to the bank?
P85∗∗ Two children stand on a large, sloping hillside that can be con-
sidered as a plane. The ground is just sufficiently icy that a child would fall
and slide downhill with a uniform speed as the result of receiving even the
slightest impulse.
v0
a
For fun, one of the children (leaning against a tree) pushes the other with
a horizontal initial speed v0 = 1 m s−1 . The latter slides down the slope with
a velocity that changes in both magnitude and direction. What will be the
child’s final speed if air resistance is negligible and the frictional force is
independent of the speed?
P86∗ Smugglers set off in a ship in a direction perpendicular to a straight
shore and move at constant speed v. The coastguard’s cutter is a distance a
from the smugglers’ ship and leaves the shore at the same time. The cutter
always moves at a constant speed in the direction of the smugglers’ ship
and catches up with the criminals when at a distance a from the shore. How
many times greater is the speed of the coastguard’s cutter than that of the
smugglers’ ship?
P87 Point-masses of mass m are at rest at the corners of a regular n-gon,
as illustrated in the figure for n = 6.
n 1
How does the system move if only gravitation acts between the bodies?
How much time elapses before the bodies collide if n = 2, 3 and 10? Examine
the limiting case when n 1 and m = M0 /n, where M0 is a given total mass.
P88 A rocket is launched from and returns to a spherical planet of radius
R in such a way that its velocity vector on return is parallel to its launch
vector. The angular separation at the centre of the planet between the launch
and arrival points is θ. How long does the flight of the rocket take, if the
period of a satellite flying around the planet just above its surface is T0 ?
What is the maximum distance of the rocket above the surface of the planet?
Consider whether your analysis also applies to the limiting case of θ → 0.
P89∗∗ Two identical small magnets of moment µ are glued to opposite
ends of a wooden rod of length L, one labelled C, parallel to the rod, and
the other labelled D, perpendicular to it.
L
C
D
Problems 21
(i) Show that the couples that the magnets exert on each other are not
equal and opposite.
(ii) Ignoring the Earth’s magnetic field, explain quantitatively what would
happen if the system were freely suspended at its centre of gravity.
P90 A point-like body of mass m and charge q is held above and close
to a large metallic fixed plane and released when a distance d from it. How
much time will it take for the body to reach the plane? Ignore gravity.
P91∗ A plastic ball, of diameter 1 cm and carrying a uniform charge
of 10−8 C, is suspended by an insulating string with its lowest point 1 cm
above a large container of brine (salted water). As a result, the surface of
the water below the ball wells up a little.
How large is the rise in water level immediately below the ball? Ignore
the effect of surface tension, and take the density of salted water to be
1000 kg m−3 .
P92 A point charge is at rest inside a thin metallic spherical shell, but is
not at its centre. What is the force acting on the charge?
P93 Boron atoms of mass number A = 10 and a beam of unidentified
particles, moving in opposite directions with the same (non-relativistic) speed,
are made to collide inside an ion accelerator. The maximum scattering angle
of the boron atoms is found to be 30◦ . What kind of atoms does the particle
beam consist of?
P94 A billiard ball rolling without slipping hits an identical, stationary
billiard ball in a head-on collision. Describe the motion of the balls after
the collision. Prove that the final state does not depend on the coefficient of
sliding friction between the balls and the billiard table. (Rolling friction is
negligible.)
P95∗ A long slipway, inclined at an angle α to the horizontal, is fitted with
many identical rollers, consecutive ones being a distance d apart. The rollers
have horizontal axles and consist of rubber-covered solid steel cylinders each
of mass m and radius r. Planks of mass M, and length much greater than d,
are released at the top of the slipway.
d m
r a
Find the terminal speed vmax of the planks. Ignore air resistance and
friction at the pivots of the rollers.
P96 A tablecloth covers a horizontal table and a steel ball lies on top of
it. The tablecloth is pulled from under the ball, and friction causes the ball
to move and roll.
What is the ball’s speed on the table when it reaches a state of rolling
without slipping? (Assume that the table is so large that the ball does not
fall off it.)
P97∗ If the law were changed so that traffic in Great Britain travelled
on the right-hand side of the road (instead of on the left), would the length
of the day increase, decrease, or be unaltered?
P98 In a physics stunt, two balls of equal density, and radii r and R = 2r,
are placed with the centre of the larger one at the middle of a cart of mass
M = 6 kg and length L = 2 m. The mass of the smaller ball is m = 1 kg. The
balls are made to roll, without slipping, in such a way that the larger ball
rests on the cart, and a straight line connecting their centres remains at a
constant angle φ = 60◦ to the horizontal. The cart is pulled by a horizontal
force in the direction shown in the figure.
F
M
(i) Find the magnitude of the force F.

(ii) How much time elapses before the balls fall off the cart?
P99∗∗ The following equipment can be seen in the Science Museum in
Canberra, Australia. A disc of radius R has been cut from the centre of a
horizontal table, and then replaced into its original place mounted on a axle.
Problems 23
As illustrated in the figure, the disc is spun and a solid rubber ball is
rolled onto the table. When it reaches the spinning disc, the ball leaves its
straight-line course and follows a curve. On leaving the disc, it continues its
original course, rolling without slipping, along a straight line. The final speed
of the ball is the same as it was before it reached the disc.
What are the conservation principles underlying this motion?
P100 A thin ring of radius R is made of material of density ρ and
Young’s modulus E. It is spun in its own plane, about an axis through its
centre, with angular velocity ω. Determine the amount (assumed small) by
which its circumference increases.
P101∗ A light, inelastic thread is stretched round one-half of the circum-
ference of a fixed cylinder as shown in the figure.
A FA
FB
As a result of friction, the thread does not slip on the cylinder when the
magnitudes of the forces acting on its ends fulfil the inequality
1
FA ≤ FB ≤ 2FA .
2
Determine the coefficient of friction between the thread and the cylinder.
P102∗∗ Charlie is a first-year student at university, studying integral
calculus in mathematics. As an exercise, he has to determine the position
C of the centre of mass of a semicircular arc which has radius R and a
homogeneous mass distribution.
C
His younger sister, Jenny, only attends secondary school, but is studying
rotation in physics. She eagerly watches the calculations of her brother, but
as she has never heard of integral calculus, she does not understand much
of it. The only clear thing to her is the problem itself.
After thinking and calculating for a while, she calls out: ‘I have got the
result, and I can determine not only the position of the centre of mass of a
semicircle but also that of any part of a circle or any sector of it!’
How has she done it?
P103∗ A table of height 1 m has a hole in the middle of its surface. A
thin, golden chain necklace, of length 1 m, is placed loosely coiled close to
the hole, as shown in the figure.
1m
One end of the chain is pulled a little way through the hole and then
released. Friction is negligible, and, as a result, the chain runs smoothly
through the hole with increasing speed. After what times will the two ends
of the chain reach the floor?
P104∗ A flexible chain of uniform mass distribution is wrapped tightly
round two cylinders so that its form is that of a stadium running-track, i.e.
it consists of two semicircles joined by two straight sections. The cylinders
are made to rotate and cause the chain to move with speed v.
v
For some reason, the chain suddenly slips off the cylinders and falls
vertically. How does the shape of the chain vary during the fall?
According to Steve, it takes a circular shape because of the centrifugal
force. Bob accepts this point, but he considers that the initially ‘elliptical’
chain will be deformed beyond the circular by this effect and become a
vertical ellipse with its new major axis at right angles to the original one. He
expects that this process will repeat itself and that the chain shape will cycle
Problems 25
between the two ‘ellipses’. Frank guesses that the chain retains its original
shape, but he cannot give any reasons for his guess. Who is right – or are
they perhaps all wrong?
P105∗∗ A heavy, flexible, inelastic chain of length L is placed almost
symmetrically onto a light pulley which can rotate about a fixed axle, as
L/2
What will the speed of the chain be when it leaves the pulley?
P106∗ A long, heavy, flexible rope with mass ρ per unit length is stretched
by a constant force F. A sudden movement causes a circular loop to form at
one end of the rope. In a manner similar to that in which transverse waves
propagate, the loop runs (rolls) along the rope with speed c as shown in the
figure.
x
c
F F
(i) Calculate the speed c of the loop.

(ii) Determine the energy, momentum and angular momentum carried
by a loop of angular frequency ω. What is the relationship between
these quantities?
P107 Sand falls vertically at a rate of 50 kg s−1 onto a horizontal

conveyor belt moving at a speed of 1 m s−1 , as shown in the figure.
....... ..
.
..
...
.
... 1 m s −1
...
......
.. .. . .. . . .. . . . .... .. . . ..
What is the minimum power output of the engine which drives the belt?
How is the work done by the engine accounted for?
P108∗∗ A fire hose of mass M and length L is coiled into a roll of radius
R (R L). The hose is sent rolling across level ground with initial speed
v0 (angular velocity v0 /R), while the free end of the hose is held at a fixed
point on the ground. The hose unrolls and becomes straight.
M
(i) How much time does it take for the hose to completely unroll?
(ii) The speed of the roll continually increases and its acceleration a is
clearly a vector pointing in the same direction as its velocity. On
the other hand, the vector resultant of the horizontal external forces
(frictional force plus the restraining force at the fixed end of the hose)
points in the opposite direction. How are these two facts consistent
with Newton’s second law?
(To simplify the analysis, suppose the initial kinetic energy of the roll to be
√
much higher than its potential energy (v0 gR), thus allowing the effect
of gravity to be neglected. Assume further that the hose can be considered
as arbitrarily flexible, and that the work necessary for its deformation, air
resistance and rolling resistance can all be neglected.)
P109 Where is gravitational acceleration greater, on the surface of the
Earth, or 100 km underground? Take the Earth as spherically symmetrical.
The average density of the Earth is 5500 kg m−3 , and that of its crust is
3000 kg m−3 . (The depth of the crust may be assumed to be at least 100 km.)
P110∗ The Examining Institute for Cosmic Accidents (EXINCA) sent
the following short report to one of its experts:
A spaceship of titanium-devouring little green people has found a perfectly
spherical asteroid. A narrow trial shaft was bored from point A on its surface
to the centre O of the asteroid. This confirmed that the whole asteroid is
made of homogeneous titanium. At that point, an accident occurred when
one of the little green men fell off the surface of the asteroid into the trial
shaft. He fell, without any braking, until he reached O, where he died on
impact. However, work continued and the little green men started secret
Problems 27
excavation of the titanium, in the course of which they formed a spherical

cavity of diameter AO inside the asteroid, as illustrated in the figure.
A A
O O
Then a second accident occurred – another little green man similarly fell
from point A to point O, and died.
EXINCA asked the expert to calculate the ratio of the impact speeds and
the ratio of the times taken to fall from A to O by the two unfortunate little
men. What figures did the expert give in her reply?
P111∗ The titanium-devouring little green people of the previous prob-
lem continued their excavating. As a result of their environmentally de-
structive activity, half of the asteroid was soon used up and, as shown in
the figure, only a regular hemisphere remained. The excavated material was
carried away from the asteroid.
g0
g ?
What is the gravitational acceleration at the centre of the circular face of

the remaining hemisphere if the gravitational acceleration at the surface of
the original (spherical) asteroid was g0 = 9.81 cm s−2 ?
P112∗ The little green titanium-devouring people found another titanium
asteroid with a radius of 10 km and a homogeneous mass distribution. They
started to excavate and to convey the material of the asteroid to the surface.
The excavation of the metal was effected by boring shafts along a strip
1 m wide round the equator of the asteroid until they had cut the asteroid
completely in two. Then the accident happened; the props separating the
two hemispheres broke and the asteroid collapsed.
The experts from EXINCA need to calculate the total force exerted on the
props just before they collapsed. Please help them.
P113∗ A metal sphere, of radius R and cut in two along a plane whose
minimum distance from the sphere’s centre is h, is uniformly charged by a
total electric charge Q. What force is necessary to hold the two parts of the
sphere together?
P114 A small positively charged ball of mass m is suspended by an
insulating thread of negligible mass. Another positively charged small ball is
moved very slowly from a large distance until it is in the original position of
the first ball. As a result, the first ball rises by h. How much work has been
done?
h
m
P115∗∗ Hydrogen gas is stored at high pressure in a small, spherical

container. The gas is introduced into a light balloon and its pressure becomes
equal to the external atmospheric pressure. Is it possible that the balloon
could lift the container in its final state? Assume that the temperature of the
gas remains constant.
P116 In olden times, people used to think that the Earth was flat. Imagine
that the Earth is indeed not a sphere of radius R, but an infinite plate of
thickness H. What value of H is needed to allow the same gravitational
acceleration to be experienced as on the surface of the actual Earth? (Assume
that the Earth’s density is uniform and equal in the two models.)
P117∗ Electrical charges are evenly distributed along a long, thin insu-
lating rod AB.
E
A + + + + + + B
Show that at an arbitrary point C (see figure), the electric field due to the
rod points in the direction of the bisector of angle ACB.
Problems 29
P118 Using the result of the previous problem, determine the direction
and magnitude of the electric field in a plane which is perpendicular to a
long, charged rod, and contains one of the rod’s endpoints.
P119 At the beginning of nineteenth century the magnetic field of wires
carrying currents was the focus of investigations in physics, both experimen-
tally and theoretically. A particularly interesting case is that of a very long
wire, carrying a constant current I, which has been bent into the form of a
‘V’, with opening angle 2θ.
P h
d
According to Ampère’s computations, the magnitude B of the magnetic

field at a point P lying outside the ‘V’, but on its axis of symmetry and
at a distance d from its vertex, is proportional to tan(θ/2). However, for
the same situation, Biot and Savart suggested that the magnetic field at P
might be proportional to θ. In fact they attempted to decide between the
two possibilities by measuring the oscillation period of a magnetic needle as
a function of the ‘V’ opening angle. However, for a range of θ values, the
predicted differences were too small to be measured.
(i) Which formula might be correct?
(ii) Find the proportionality factor in this formula and guess the most
likely factor appearing in the other one.
P120∗∗ A direct current flows in a solenoid of length L and radius R,

(L R), producing a magnetic field of magnitude B0 inside the solenoid.
...
P B0
R
...
(i) What is the strength of the magnetic field at the end of the coil, i.e.
at the point P shown in the figure?
(ii) What is the magnetic flux at the end of the coil, i.e. through a virtual
disc of radius R centred on P ?
(iii) Sketch the magnetic field lines in the vicinity of P .
P121 The inner surfaces of two close parallel insulating plates are each
given a uniform charge of +Q. What force is required to hold the plates
together?
P122 Two parallel plate capacitors differ only in the spacing between
their (very thin) plates; one, AB, has a spacing of 5 mm and a capacitance of
20 pF, the other, CD, has a spacing of 2 mm. Plates A and C carry charges of
+1 nC, whilst B and D each carry −1 nC. What are the potential differences
VAB and VCD after the capacitor CD is slid centrally between and parallel
to the plates of AB without touching them? Would it make any difference if
CD were not centrally placed between A and B?
P123∗ The distance between the plates of a plane capacitor is d and the
area of each plate is A. As shown in the figure, both plates of the capacitor
are earthed and a small body carrying charge Q is placed between them, at
a distance x from one plate.
A
d Q
x
What charge will accumulate on each plate?

P124∗ A point-like electric dipole is placed between the earthed plates of
the plane capacitor discussed in the previous problem. Its dipole momentum
vector p is perpendicular to the plates and the distances of the dipole from
the plates are x and d − x, respectively.
How does the charge which accumulates on each of the plates depend on
x? (Ignore edge effects.)
P125∗ The refractive index of the medium within a certain region, x > 0,
y > 0, changes with y. A thin light ray travelling in the x-direction strikes
the medium at right angles and moves through the medium along a circular
arc.
Problems 31
How does the refractive index depend on y? What is the maximum possible
angular size of the arc?
P126 A compact disc (CD) contains approximately 650 MB of informa-
tion. Estimate the size of one bit on a CD using an ordinary ruler. Confirm
your estimate using a laser beam. Can you suggest the shape of one unit of
information?
P127 When a particular line spectrum is examined using a diffraction
grating of 300 lines mm−1 with the light at normal incidence, it is found
that a line at 24.46◦ contains both red (640–750 nm) and blue/violet (360–
490 nm) components. Are there any other angles at which the same thing
would be observed?
P128∗ A parallel, thin, monochromatic laser beam falls on a diffraction
grating at normal incidence. How does the interference pattern it produces
on a viewing screen change if the grating is rotated through an angle φ < 90◦
around an axis, which is
(i) parallel to the lines of the grating; or
(ii) perpendicular to the lines of the grating?
P129 Two floating objects are attracted to each other as the result of
surface tension effects, irrespective of whether they are floating on water or
on mercury. Explain why this is so.
P130∗ Water in a clean aquarium forms a meniscus, as illustrated in the
figure.
Calculate the difference in height h between the centre and the edge of the
meniscus. The surface tension of water is γ = 0.073 N m−1 .
P131∗ Is it possible to have a (spherical) drop of water that could
evaporate without taking up heat or losing internal (thermal) energy?
P132∗∗ Small liquid drops of various sizes are in a closed container, to
whose walls the liquid does not adhere. Over a sufficiently long time, the
size of the smallest drops is found to decrease whilst that of the larger ones
increases, until finally only one large drop remains in the container. What is
the explanation for this phenomenon?
P133 A horizontal frictionless piston, of negligible mass and heat capac-

ity, divides a vertical insulated cylinder into two halves. Each half of the
cylinder contains 1 mole of air at standard temperature and pressure p0 .
A load of weight W is now suspended from the piston, as shown in the

figure. It pulls the piston down and comes to rest after a few oscillations.
How large a volume does the compressed air in the lower part of the cylinder
ultimately occupy if W is very large?
P134∗ How high could the tallest mountain on Earth be? And on Mars?
P135∗∗ The sealed lower half of a straight glass tube, of height 152 cm,
is filled with air. The top half contains mercury and the top of the tube
is left open. The air is slowly heated. How much heat has been trans-
ferred to the air by the time all the mercury has been pushed out of the
tube?
Make a plot showing how the molar heat of the enclosed air changes with
its volume during the process. (Atmospheric pressure is 760 mm Hg.)
P136 Vulcanism is very common in Iceland, but glaciers cover 11 per cent
of its surface area. This is why volcanic eruptions quite often occur under
glaciers, as one did in October 1996 under Vatnajokull, Europe’s largest
glacier. At the site of the eruption the glacier was 500 m thick and more or
less smooth and flat. After a day’s activity the visible sign of the eruption
was a deep crater-like depression on the surface of the ice cap, in the
form of a upside-down cone with a depth of 100 m and a diameter of
1 km. Explain the formation of the depression. What would have been
found under the ice crater at this time? Try to predict the subsequent
events.
P137∗ The most famous geyser in Yellowstone National Park is Old
Faithful. This geyser can be considered as a large underground cavity with
a narrow flue leading to the surface.
Problems 33
The surrounding earth is warm as a result of residual volcanic activity

and boils the water in the cavity. After coming to the boil, the water in the
flue is expelled, and approximately 44 tons of steam leave the geyser in 4
minutes. After the eruption, underground springs refill the cavity and the
flue to ground level in 20–30 minutes, and the process then repeats itself. An
eruption occurs every 90 minutes.
Geological experiments show that the underground temperature in this
area increases by 1 ◦ C for each metre of depth. Determine the minimum
distance below the surface at which the cavity is situated. If the cavity is
assumed to be located at this minimum depth, what is its volume?
P138 The air above a large lake is at −2 ◦ C, whilst the water of the lake
is at 0 ◦ C. Assuming that only thermal conduction is important, and using
relevant data selected from that given below, estimate how long it would
take for a layer of ice 10 cm thick to form on the lake’s surface.
Data:
Thermal conductivity of water, λw = 0.56 W m−1 K−1
Thermal conductivity of ice, λi = 2.3 W m−1 K−1
Specific latent heat of fusion of ice, Li = 3.3 × 105 J kg−1
Density of water, ρw = 1000 kg m−3
Density of ice, ρi = 920 kg m−3
P139 If it takes two days to defrost a frozen 5-kg turkey, estimate how
long it would take to defrost an 8-tonne Siberian mammoth.
P140∗ A 0.6-kg block of ice at −10 ◦ C is placed into a closed empty
1 m3 container, also at a temperature of −10 ◦ C. The temperature of the
container is then increased to 100 ◦ C. How much greater is the heat required
than that necessary to raise the empty container alone to that temperature?
P141∗ A strong-walled container is half-filled with water. The other half
contains air, initially at standard temperature and pressure. The container
is closed and slowly heated. When does the water in the container start
boiling? In what state(s) does the water exist, as the temperature rises?
P142 Two cobwebs each of length ' and under a tension F are contained
in a glass case at temperature T . Because they are struck by air molecules
they undergo random vibrations. What is the ratio of the amplitudes of these
motions if cobweb A has twice the mass of cobweb B?
P143 Outdoors at night, water vapour often condenses on cobwebs, on
which we can find periodical lines of very small identical water drops. Find
the minimum distance between these drops.
P144∗ Imagine a cylindrical body that can move without friction along
a straight wire parallel to its axis of symmetry, as illustrated in the figure.
v
0
v0
Tiny particles moving horizontally at speed v0 bombard the body uniformly

from both left and right. Collisions with the right end of the cylinder are
perfectly elastic whilst those with the left are perfectly inelastic, though the
particles do not stick to the cylinder after the collision. What is the speed of
the cylinder
(i) after a long time, and
(ii) after a very long time?
P145 A totally black spherical space probe is very far from the solar
system. As a result of heating by a nuclear energy source of strength I
inside the probe, its surface temperature is T . The probe is now enclosed
within a thin thermal protection shield, which is black on both sides and
attached to the probe’s surface by a few insulating rods. Find the new
surface temperature of the probe. Determine also the surface temperature
which would result from using N such shields.
Problems 35
P146∗ Two thermally insulated containers hold identical masses of water.

The water is at temperature T1 in one of them, but at temperature T2 ,
(T2 > T1 ), in the other. What is the maximum work that this system can
do if it is used as a heat engine? Take the specific heat of water as constant
over the working range.
P147 What is the change of entropy that occurs when two moles
of helium and three moles of oxygen, both at s.t.p. (T ≈ 273 K and
P ≈ 1.01 × 105 Pa) and in adjacent volumes, are allowed to mix by removing
the partition between them?
P148 By slowly pumping air into a 10-litre container, its pressure is
increased to ten times atmospheric pressure. How much work is done during
this process if the displacement of the piston in the pump is 1 litre? The
walls of the container and pump are all good heat conductors and so the
temperature can be taken as constant.
P149 A distant planet is at a very high electric potential compared with
the Earth. A metal space ship is sent from Earth for the purpose of making
a landing on the planet. Is this mission dangerous? What happens when the
astronauts open the door of the space ship and step onto the surface of the
planet?
P150 By what percentage does the capacitance of a spherical capacitor
change when its surface is dented in such a way that its volume decreases
by 3 per cent?
3%
P151∗ A closed body, whose surface F is made of metal foil, has an

electrical capacitance C with respect to an ‘infinitely distant’ point. The foil
is now dented in such a way that the new surface F ∗ is entirely inside or on
the original surface, as shown in the figure.
F*
Prove that the capacitance of the deformed body is less than C.

P152 The plates of a parallel-plate capacitor have surface area A, and

are initially separated by d. They are connected to a voltage V0 . What work
is required to pull the plates apart to a separation of 2d? How does the
energy of the capacitor change during the process?
P153∗ What is the change in length of the spiral spring shown in the
figure, which has N turns, radius R, length x0 and spring constant k, when
a current I0 is made to flow through it?
2R
I0
x0 N
P154∗ A very short magnet A of mass m is suspended horizontally by

a string of length ' = 1 m. Another very short magnet B is slowly brought
closer to A in such a way that the axes of the magnets are always on the
same horizontal level as each other. When the distance between the magnets
is d = 4 cm, and magnet A is s = 1 cm away from its initial position, A
spontaneously moves to attach itself to B.
A B
s d
(i) The dependence on distance of the interaction force between the

magnets is given by the relation Fmagnet (x) = ±K/xn , the sign de-
pending on the relative orientation of the magnets. Using the given
data, find the value of exponent n.
(ii) Magnet B is placed in a vertical glass tube, which is closed at the
bottom. Magnet A is then placed above it in the tube in such an
Problems 37
orientation that the magnets repel each other. Although magnet A

may tend to reverse its direction within the tube, it is constrained by
the tube and cannot do so. Find the distance apart of the magnets
in static equilibrium.
P155 A battery consists of N identical cells, each of e.m.f. E. Is it true

that the energy wasted when using the battery to charge a capacitor through
a resistor can be reduced by charging it in N stages? That is by connecting
it first across a single cell, and then across two cells, and so on, rather than
across the whole battery in a single step.
P156 An ‘energy-generating device’ consists of a parallel-plate capacitor
with nearly all the space between the plates filled with an oil of relative
permittivity ε > 1. Calculate the stored energy in the capacitor when its
plates are given charges of ±Q. The oil, which cannot come into direct
contact with the plates, is now poured out and replaced by air; calculate the
new stored energy and show that it has increased. Explain the catch in this
world-beater!
P157 An insulating sheet of relative permittivity εr is slowly slid between
the plates of a parallel-plate capacitor, completely filling the space between
the plates. What force acts on the sheet if (i) the charge, or (ii) the voltage
of the capacitor is kept constant during the process?
How does the insulating sheet affect the energy of the capacitor in cases
(i) and (ii)?
P158 Each element in the finite chain of resistors shown in the figure is
1 Ω. A current of 1 A flows through the final element.
V 1A
What is the potential difference V across the input terminals of the chain?
What is the equivalent resistance of the chain? How does the equivalent
resistance change if one or two more resistors are connected to it? Compare
this result with the equivalent resistance of an ‘infinite’ chain.
P159 All the elements in the ‘infinite’ grid shown in the figure are
of the same resistance R. What is the equivalent resistance between two
neighbouring grid points?
What would be the equivalent capacitance between two neighbouring grid

points if all the elements in the grid were capacitors with capacitance C?
What would be the equivalent inductance if the elements were inductors of
inductance L?
P160∗ A grid in the shape of a regular polyhedron (tetrahedron, cube,
dodecahedron, etc.) is made up of identical, say 1-Ω, resistors. What is the
equivalent resistance between two neighbouring grid points?
P161 The previous two problems were calculations about electrical net-
works consisting of identical resistors (an infinite grid or a regular poly-
hedron). Find the equivalent resistance of a grid between two neighbouring
grid points, if the resistor joining them is removed.
P162∗ A plane divides space into two halves. One half is filled with a
homogeneous conducting medium and physicists work in the other. They
mark the outline of a square of side a on the plane and let a current I0 in
and out at two of its neighbouring corners using fine electrodes. Meanwhile,
they measure the p.d. V between the two other corners. This is illustrated in
the figure.
B I0
a
A
C
I
0
D V
How can they calculate the resistivity of the homogeneous medium using
this data?
P163∗ You are given a large complex electrical circuit containing a lot of
resistors and other passive elements and wish to determine the resistance of
Problems 39
a particular resistor in the circuit without unsoldering it (i.e. without taking

it out of the circuit). A battery, an ammeter and a voltmeter, all of high
quality, are provided. How would you carry out the measurement?
P164∗ All the sides of a cube are made of 1-Ω resistors. What is the
equivalent resistance of the cube between the two endpoints of one of its
body diagonals?
Examine one-, two- and four-dimensional ‘cubes’ as well. Find a general
formula for the n-dimensional case.
P165 A current of 1 mA flows through a wire made of a piece of copper
and a piece of iron of identical cross-sections welded end-to-end as shown
in the figure.
1 mA
Cu Fe
How much electric charge accumulates at the boundary between the two
metals? How many elementary charges does this correspond to?
P166 The Earth’s magnetic field approximates that of a dipole with a
field of 6 × 10−5 T at the North Pole. Over London, the magnetic flux density
is 5 × 10−5 T and the angle of dip is 66◦ .
The wing span of a jumbo jet is 80 m, its length 60 m, and its depth 8 m.
Estimate the potential differences that could be detected over the surface of
the jet when it flies horizontally at 720 km h−1 :
(i) over the North Pole,
(ii) northwards over the Equator,
(iii) eastwards along the Equator,
(iv) northwest over London.
P167 A homogeneous field of magnetic induction B is perpendicular to
a track of gauge ' which is inclined at an angle α to the horizontal. A
frictionless conducting rod of mass m straddles the two rails of the track as
L
C
S
R B
m
a F
How does the rod move, after being released from rest, if the circuit formed
by the rod and the track is closed by:
(i) a resistor of resistance R,
(ii) a capacitor of capacitance C, or
(iii) a coil of inductance L?
P168∗ One end of a horizontal track of gauge ' and negligible resis-
tance, is connected to a capacitor of capacitance C charged to voltage V0 .
The inductance of the assembly is negligible. The system is placed in a
homogeneous, vertical magnetic field of induction B, as shown in the figure.
S B
R F
C
m
V0
A frictionless conducting rod of mass m and resistance R is placed per-

pendicularly onto the track. The polarity of the capacitor is such that the
rod is repelled from the capacitor when the switch is turned over.
(i) What is the maximum velocity of the rod?
(ii) Under what conditions is the efficiency of this ‘electromagnetic gun’
maximal?
P169 A resistor and an inductor in series are connected to a battery
through a switch.
L R
After the switch has been closed:

(i) What is the magnitude of the current flowing when the rate of the
increase of magnetic energy stored in the coil is at a maximum?
(ii) When will the Joule heat dissipated in the resistor change at the
fastest rate?
P170∗ (i) Sketch qualitatively, as a function of x = ω/ω0 , the magnitude
of the current drawn from the source by the two circuits shown in the figure;
here ω0 = (LC)−1/2 .
Problems 41
L L
L C
V0 cos x t V0 cos x t
C L
C C
(a) (b)
(ii) Using three or more of the components shown in figure (a), construct
five new circuits, each of which shows current resonance (maximum current
drawn from the source at some frequency), but all at different frequencies.
P171∗ The circuit shown in the figure – consisting of three identical lamps
and two coils – is connected to a direct current source. The ohmic resistance
of the coils is negligible.
After some time, switch S is opened. What are the relative brightnesses of
the three lamps immediately afterwards?
P172 The turns of a solenoid, designed to provide a given magnetic flux
density along its axis, are wound to fill the space between two concentric
cylinders of fixed radii. How should the diameter d of the wire used be
chosen so as to minimise the heat dissipated in the windings?
P173∗ A solid metal cylinder rotates with angular velocity ω about its
axis of symmetry. The cylinder is in a homogeneous magnetic field B parallel
to its axis. What is the resultant charge distribution inside the cylinder? Is
there an angular velocity for which the charge density is everywhere zero?
P174∗ Consider the result of the previous problem using a rotating frame
of reference, fixed to the cylinder. Describe the electric and magnetic fields
in this rotating (non-inertial) frame of reference.
(Assume that the angular velocity of rotation is much smaller than the
cyclotron frequency, ω0 = eB/m, where e and m are the elementary charge
and mass of the electron, respectively.)
P175∗ Jack and Jill have been set a similar task to that in P173. They
have to calculate what charge distribution is formed in a metal bicycle
spoke, rather than a metal cylinder, when it is rotated in a homogeneous
magnetic field. The spoke rotates about a perpendicular axis at one end of
it.
Jill knows the solution to P173, and she simply adopts it. Ignoring the
electron mass she concludes that the charge density is ρ = 2ε0 Bω. Jack’s
solution is based on the fact that a bicycle spoke is a thin metal rod; and so
he considers the problem to be one-dimensional. The induced electric field is
E(r) = rBω at a distance r from the rotational axis.
Applying Gauss’s law to a short section of the spoke of length ∆r, Jack
finds the charge density: (ρ/ε0 )A∆r = ∆EA = Bω∆r × A, where A is the
cross-sectional area of the spoke. From this equation he derives: ρ = ε0 Bω,
which is only half of Jill’s value.
Comment on these differing results.
P176 A circular metal ring of radius of r = 0.1 m rotates about a
vertical diameter with constant angular velocity. As shown in the figure, a
small magnetic needle that can turn freely about a vertical axis sits in the
middle of the ring.
When the ring is stationary, the needle points in the direction of the
horizontal component of the Earth’s magnetic field. However, when it rotates
at the rate of ten turns per second, the magnet deviates by an average of 2◦
from this position.
What is the electrical resistance R of the ring?
P177 A uniform thin wire of length 2πa and resistance r has its ends
joined to form a circle. A small voltmeter of resistance R is connected by
tight leads of negligible resistance to two points on the circumference of the
circle at angular separation θ, as shown in the figures.
Problems 43
R V
a a
h h
R V
r r
(a) (b)
A uniform magnetic flux density perpendicular to the plane of the circle

is changing at a rate Ḃ. What will the reading of the voltmeter be if the
voltmeter is positioned:
(a) at the centre of the circle, and
(b) on the chord joining the two points of attachment?
P178∗ A ‘twisted’ circular band (called a Moebius strip) is made from a
strip of paper of length L and width d. A wire running along the edge of
the strip is connected to a voltmeter, as shown in the figure.
L
What does the voltmeter register when the strip is placed in a homogeneous
magnetic field which is perpendicular to the plane of the strip and changes
uniformly with time, i.e. B(t) = kt?
P179 A long solenoid contains another coaxial solenoid (whose radius
R is half of its own). Their coils have the same number of turns per unit
length and initially both carry no current. At the same instant currents start
increasing linearly with time in both solenoids. At any moment the current
flowing in the inner coil is twice as large as that in the outer one and their
directions are the same. As a result of the increasing currents a charged
particle, initially at rest between the solenoids, starts moving along a circular
trajectory (see figure). What is the radius r of the circle?
I
2I
2R R r
P180 Charge Q is uniformly distributed on a thin insulating ring of

mass m which is initially at rest. To what angular velocity will the ring be
accelerated when a magnetic field B, perpendicular to the plane of the ring,
is switched on?
P181∗ A metal disc of radius r can rotate with negligible friction inside
a long, straight coil, about a shaft parallel to the axis of symmetry of the
coil. One end of the coil wire is connected to the edge of the disc and the
other to the shaft. The coil has ohmic resistance R and contains n turns per
unit length. It is placed so that its axis is parallel to the Earth’s magnetic
field vector B0 .
B
x 0
A
What current flows through the ammeter shown in the figure if the disc
rotates with angular frequency ω? Plot the current as a function of ω for
both directions of rotation.
Prove that the power needed to rotate the disc is equal to the rate of Joule
heating generated by the ohmic resistance of the coil.
P182∗ A thin superconducting (zero resistance) ring is held above a ver-
tical, cylindrical magnetic rod, as shown in the figure. The axis of symmetry
of the ring is the same to that of the rod. The cylindrically symmetrical mag-
netic field around the ring can be described approximately in terms of the
vertical and radial components of the magnetic field vector as Bz = B0 (1−αz)
and Br = B0 βr, where B0 , α and β are constants, and z and r are the vertical
and radial position coordinates, respectively.
Bz
Br
Problems 45
Initially, the ring has no current flowing in it. When released, it starts to
move downwards with its axis still vertical. From the data below, determine
how the ring moves subsequently? What current flows in the ring?
Data:
Properties of the ring: mass m = 50 mg
radius r0 = 0.5 cm
inductance L = 1.3 × 10−8 H
Initial coordinates of
the centre of the ring: z=0
r=0
Magnetic field constants: B0 = 0.01 T
α = 2 m−1
β = 32 m−1
P183∗ A small, electrically charged bead can slide on a circular, friction-
less, insulating string. A point-like electric dipole is fixed at the centre of the
circle with the dipole’s axis lying in the plane of the circle. Initially the bead
is on the plane of symmetry of the dipole, as shown in the figure.
How does the bead move after it is released? Find the normal force
exerted by the string on the bead. Where will the bead first stop after being
released? How would the bead move in the absence of the string? Ignore the
effect of gravity, assuming that the electric forces are much greater than the
gravitational ones.
P184∗ A point-like body of mass m and charge q, initially at rest, is
released in a homogeneous gravitational field. What path does the body
follow if it is also acted upon by a homogeneous horizontal magnetic field?
P185∗ A long, thin, vertical glass tube is surrounded by a much wider
coaxial glass tube of outer radius r. Wound on the wider tube there are
many separate circular conducting loops, each of resistance R and spaced a

distance h apart.
m l
h
R
v
0
If a small magnet bar of mass m and magnetic moment µ is dropped into

the thin tube, after a relatively short time it reaches a constant terminal
velocity v0 , with which it then sinks uniformly.
In the course of each later investigation only one the five quantities
mentioned above (m, µ, h, R, r) is doubled, whilst the other four remain at
their original values. By what factor does the terminal velocity of the magnet
change in each case? Ignore mechanical friction and air resistance, as well
as the self- and mutual inductance of the conducting loops.
P186∗ In a vacuum chamber a current of 10 A is flowing in a long,
straight wire, which has a very high conductivity. Electrons with an initial
velocity v0 start moving perpendicularly towards the wire from a point which
is a radial distance r0 away from the wire. Given that they cannot approach
any closer to the wire than r0 /2, determine v0 . Ignore the effect of the Earth’s
magnetic field.
P187∗ The distance between the plates of an initially uncharged capac-
itor is d. Parallel to its plates, there is a magnetic field of strength B, as
B V
d v er
Problems 47
What voltage does the voltmeter connected to the plates of the capacitor
register when an electrically neutral liquid of relative dielectric constant εr
flows between the plates with velocity v?
P188 The energy released by the fission of uranium nuclei would be

higher if the uranium nucleus split into three parts rather than into two.
Despite this, the fission of uranium only produces two nuclei. Why is
this?
P189∗ 7 Be is a radioactive element with a half-life of 53.37 days. When

isotope 7 of beryllium is heated to a few thousand degrees, its half-life
changes. What is the explanation for this?
P190∗ Part of the series of isotopes produced by the decay of thorium-

232, together with the corresponding half-lives, is given below:
232
90 Th −→ 88 Ra −→ 89 Ac −→ 90 Th −→ 88 Ra −→ 86 Rn −→
228 228 228 224 220
···
1.41×1010 y 5.7 y 6.13 h 1.91 y 3.64 d 56 s
Thorium-232 and thorium-228 in equilibrium are extracted from an
ore and purified by a chemical process. Sketch the form of the variation
in the number of atoms of radon-220 you would expect to be present
in 10−3 kg of this material over a (logarithmic) range from 10−3 to 103
years.
P191 Through what voltage must protons be accelerated if they are to be

able to produce proton–antiproton pairs when they collide with stationary
protons? The rest-mass energy of a proton is approximately 1 GeV.
P192∗ How does a positron move in a Faraday cage if it is ‘dropped’

with no initial speed? Consider the positron as a classical particle, acted on
by electrical forces and the gravitational field of the Earth, as indicated in
the figure.
e+
mg
P193∗ Two positrons are at opposite corners of a square of side a = 1 cm.

The other two corners of the square are each occupied by a proton, as shown
in the figure.
e+ p
p e+
Initially the particles are held in these positions, but all four are released at
the same time. What will their speeds be when they are a significant distance
apart? The particles can be considered as classical point masses moving in
each other’s electric fields. Gravity can be ignored.
P194 In an experiment on Compton scattering, stationary electrons are
bombarded by photons whose energy is equal to the rest energy of an
electron. For events in which the scattered photon and the recoil electron
have momenta of the same magnitude, find the angle between them. What
is the speed of the recoil electron in this case?
P195 X-ray photons are scattered through an angle of 90◦ by electrons
initially at rest. What is the change in the wavelength of the photons?
P196 Imagine a ‘classical electron’ as a small, spherical ball. What is its
minimum radius, if its electrostatic energy is not to be greater than its total
rest energy, mc2 ? What is its angular velocity if its angular momentum is
h/(4π)? To what ‘equatorial speed’ does this correspond, if the whole of the
electron’s rest energy is provided by the electrostatic field?
P197∗ An electron is enclosed in a large rectangular box. Estimate the
order of magnitude of the thickness of the layer (at the bottom of the box)
which, as a result of gravitational effects, is occupied by the electron.
P198∗ Classically, the Coulomb field of an atomic nucleus could confine
an electron to that nucleus. However, the Heisenberg uncertainty principle
prescribes such a high kinetic energy for an electron enclosed in such a small
space that it would escape from the nucleus in any case. What would be the
atomic number of a transuranic element able to confine an electron within
its nucleus for a significant time, if only the element itself were sufficiently
stable?
P199∗ Show how the size of water molecules can be estimated using the
speed of water surface (capillary) waves and the speed of sound waves in
Problems 49
water? The speed of propagation of surface waves of wavelength 1 cm is

approximately 10 000 times smaller than that of sound in water.
P200 Congratulations to the reader! You have reached the last problem
in the book and the proper manner in which to congratulate you would be
to drink your health in champagne. Unfortunately, this kind of recognition
is not really practical – though we can at least make the last problem one
about champagne.
The bubbles in champagne are familiar. They form almost exclusively at

particular points in the champagne glass, and from these points they rise
faster and faster. Why do the bubbles in champagne accelerate?
Hints
H1 Resolve the velocity vectors of the snails into suitable components.

There is more than one way to do this, and they lead to different methods of
arriving at the same solution. The equation of the path can be determined
by expressing the velocity in polar coordinates.
H2 Calculate the maximum possible value of the coefficient of friction

if the object is not to stop on the table.
H3 The solution of part (i) is trivial, since the boat is faster than the
river. In part (ii), a suitably chosen vector addition can help determine the
directions in which the boatman could go; the direction corresponding to
the shortest path has still to be chosen.
H4 Although the whole of the moving part of the carpet has unit speed,
its centre of mass has a lower speed. The reason for this is the increasing
mass of the moving part.
H5 Draw the ‘space–time’ world lines of the snails. The result can also
be obtained using the equivalence of different inertial frames of reference
(Galilean symmetry).
H6 Compare the amounts by which the centres of mass of the two worms
are raised.
H7 You can base your solution on the conservation of energy and the
conditions for static equilibrium.
H8 Determine how much additional volume is submerged if the berg is

depressed by a small distance x. Use the flotation condition to relate the
mass M of the berg to its overall dimensions.
50
Hints 51
H9 The sum of the tensions in the suspension springs remains unaltered

by parking on the pavement, and the net torque about any axis must be zero.
H10 The trickiest part of the solution is to appreciate how frictional
forces can balance both Jean Valjean’s weight and the reaction forces of the
walls at the same time.
H11 The distance between the geometrical centre of the composite sphere
and its centre of mass cannot be greater than a certain distance. Find this
limiting distance.
H12 Describe the motion of the ball in terms of its components perpen-
dicular and parallel to the slope.
H13 Because of the acceleration associated with its motion, the hamster
exerts a force on the platform. The torque resulting from this force can
balance the torque about the pivot due to the hamster’s weight.
H14 Get a bicycle and try it.
H15 Use Newton’s law of gravitation and express the mass of the Sun
in terms of its average density.
H16 Compare the field produced by one of the stars at the position of the
other, to that experienced by the Earth as a result of the Sun’s gravitation.
H17 The space probes should be launched from the Equator and directed
eastward.
H18 Examine the energies involved.
H19 The centre of mass of the system, i.e. honey plus steel ball, moves
steadily downwards. The total momentum of the system can be calculated
using the speed of the centre of mass, and the momentum of the honey
obtained by then subtracting that of the ball.
H20 If the wall of the container is at a temperature different from that
of the gas, then collisions of the gas molecules with the wall either take
energy from the wall or give energy to it.
H21 The difference in temperature between the two spheres arises be-
cause their centres of mass are displaced in opposite directions.
H22 No student reading this book should need a hint.
H23 The situation is possible – and using only two resistors!
H24 Examine how the combined centre of gravity of the bucket and
water changes as the water leaks away.
H25 Under what conditions does adding a little more water inevitably
raise the overall centre of gravity?
H26 Our assertion is that water does not flow into the bowl. In order to
prove this, the effects of both the force and the torque exerted by the chain
have to be considered.
H27 The only unusual part of the solution is the calculation of the
buoyancy force.
H28 The average density of the bubble has to be the same as that of air,
since the bubble floats.
H29 At the end of the capillary tube, the pressure of curvature bal-
ances the difference between the pressure inside the liquid and atmospheric
pressure.
H30 The whole system (the current distribution and the electric field)
is spherically symmetrical, and therefore the magnetic field also has to be.
Consider which spherically symmetrical magnetic fields are consistent with
the (experimentally observed) non-existence of magnetic monopoles.
H31 Make use of the symmetry of the charge distribution.
H32 It is not sufficient to simply compare the gravitational accelerations,
since it isn’t clear that the high-jumper would be able to take off with the
same initial speed. The movement of the high-jumper’s centre of mass during
the jump has to analysed.
H33 The time intervals of the motions and the lengths of the paths do
not have to be found exactly; only the inequalities relating them need to be
determined.
H34 Resolving the tension in the string into radial and tangential com-
ponents, the direction of the string can be calculated using the dynamical
conditions for uniform circular motion.
H35 Prove that, at any instant, bodies which started at the same time,
from the same point and slid down frictionless wires in different directions,
all lie on the surface of a common (imaginary) sphere.
H36 The problem can be solved in an elementary way using a rotating
frame of reference fixed to the minute hand.
H37 The stone moves away from the thrower until the component of its
velocity parallel to its position vector has decreased to zero. If this never
occurs, the condition imposed in the problem has been met.
Hints 53
H38 It is false to assume that the trajectory of the grasshopper (with the
minimum take-off speed) just touches the trunk at its topmost point.
H39 They clearly cannot jump directly towards each other without
mishap, so consider jumping in some other direction whilst preserving the
symmetry of the situation. Note that the mass of the hair is given!
H40 The shape of the common surface of the water jets, their envelope,
has to be determined. They start from the same place, have the same initial
speed, and follow parabolic paths. Examine the condition for determining
whether any water jet passes through a given point in space.
H41 Show that
v2 EQ
Range = sin 2θ + (1 − cos 2θ) ,
g mg
and maximise with respect to θ.

H42 The normal reactions exerted on the rod by my fingers are not
equal in general. Thus the maximum static frictional force is smaller on one
side than on the other, and sliding will occur there first. However, because
of the increasing normal reaction at the finger where sliding is taking place,
the kinetic frictional force there increases, and the moment it becomes larger
than the static friction on the other side, slipping will stop at the first finger
and start at the second. During the process there is alternating slipping and
sticking at both fingers. The work done can be calculated from the length of
each stage.
H43 The process should be started from the top! The correct strategy
is to slide the topmost brick as far as possible and then do the same thing
with the two uppermost, considered as a unit, and so on downwards.
H44 Use the balances of forces and torques acting on the plate to find
connections between the frictional forces involved. A graphical method of
treating the linear equations so derived is recommended.
H45 The process has to be considered as a series of consecutive collisions.
H46 Show that in the first collision the fraction of the initial kinetic
energy transferred to the middle ball is 4µM/(µ + M)2 .
H47 During the collision, the momentum, energy and angular momen-
tum of the system are all conserved.
H48 Decompose the motion into that of the centre of mass and that in
the centre of mass system. Show, by considering the conservation of energy
and angular momentum, that when the rope first tightens the centre of mass
velocities of both particles are unchanged in magnitude but turned through
π/2, with the result that they then travel parallel to the x-axis.
H49 The usual reasoning which assumes that one-third of the basin is
filled in 1 minute, and one-half of the basin empties in the same time, (and
hence that 12 − 13 = 16 of the basin becomes empty in 1 minute), is false. Water
flows into the basin uniformly from the tap, but (according to Torricelli’s
law of efflux) it flows out more quickly when the water level in the basin is
higher.
H50 Show that the free surface is part of the paraboloid of revolution
z = ω 2 r 2 /2g, where z is measured from the lowest point of the free surface
and r is the radial distance from the central axis. Consider the volume of
the air above the liquid but still inside the vessel.
H51 In the context of mechanics, the car is not a closed system; it is in
contact with its surroundings, in this case, the Earth.
H52 The focal length can be obtained using the relationship between
the lens formula and the magnification. The ratio of the brightness values
depends not only on the size of the images, but also on the amount of light
reaching the lens.
H53 The apparent magnitude of the virtual image is not determined by
the size of the image itself, but by the angle it subtends at the eye.
H54 Obtain ng sin(θ + φ) ≥ nw , where φ is the angle in the glass between
the ray and the normal to the surface at the point where it enters the prism.
H55 No patch of light can be seen either right next to the quarter-
cylinder, or very far from it. The closer light patch is excluded by total
internal reflection. The distance of the furthest part of the light patch can be
determined by considering the part of the quarter-cylinder close to the table
as a plano-convex lens.
H56 Suppose that the sunlight falling on the Moon is diffusely reflected
with the given coefficient. Calculate how much of it reaches unit area of the
Earth.
H57 The most comfortable walking rate can be related to the period of
the human leg swinging freely like a pendulum. Running can be considered
as a forced oscillation, with its period dependent on the moment of inertia
of the leg and the torque applied by the muscles.
H58 By choosing its length suitably, a simple pendulum can be made to
Hints 55
have the same angular velocity, in any position, as that of the rod pendulum
given in the problem. Compare the periods of swing of this pendulum and
the actual simple pendulum of the problem. The ratio of the periods of two
simple pendulums of different lengths displaced through the same angle can
be deduced using dimensional analysis.
H59 Identify the physical quantities on which the power necessary for
hovering depends.
H60 Use conservation of energy to determine the rod’s angular velocity ω
when its inclination is θ, and relate the components of the reaction between
the table and the rod to the accelerations they produce. In case (i), the
smooth horizontal and vertical walls of the groove can exert only vertical
and horizontal forces on the end of the rod, respectively. In case (ii), the
edge of the table is a very small quarter-circle, so the normal force is always
directed along the rod’s axis.
H61 When the coefficient of friction is small, the point of the pencil
moves ‘backward’. If the coefficient of friction is larger than a certain critical
value (which can be shown to be about 0.37), the pencil moves ‘forward’.
Using the fact that kinetic friction decreases the mechanical energies, it can
be shown that the point of the pencil never loses contact with the table.
H62 Use the ideal gas equation to express the conservation of air mass.
Also note that, after a sufficiently long time, the temperature of the system
will not have changed.
H63 Because of the surface tension (pressure of curvature) of the water,
the pressure inside the trapped water is lower than atmospheric pressure.
H64 Calculate the velocity of the points of the thread at any given
moment.
H65 Consider the (‘elastic’) frame of reference fixed to the thread.
H66 Find a simple – mathematically easy to treat – trajectory, in which the
ball reaches as high a speed as possible, and the time so gained compensates
for the longer path involved.
H67 From the figure you can determine the angle not given in the text.
H68 The centre of mass (CM) of the compasses is directly below the
attachment point. If the angle between the arms were changed, the horizontal
position of the CM would have to remain the same, although the positions
of the CM of the individual arms would change. Use this argument to find
the solution to the problem with a minimum of actual calculation.
H69 Express the condition for equilibrium using vectors.

H70 Note that the system as a whole has no external horizontal forces
acting upon it. As well as keeping the centre of mass of the system fixed, you
will need to use the law of conservation of linear momentum. You should
be able to show that the tanker will initially move forward, but later reverse
its direction of motion.
It may seem surprising that the tanker alters its direction of motion, and it
might help to first consider the following, rather than the original problem:
A poor student and a zealous ticket collector, both of mass m, are in

a stationary, frictionless railway carriage of mass M. When the collector
realises that the student has no ticket, the student runs to the end of the
carriage with the collector, who moves with speed v relative to the carriage,
in pursuit. The student stops at the end of the carriage and jumps out. Find
the velocity of the carriage when the ticket collector reaches the open door,
stops there and watches the student making his escape.
H71 Consider the motion in the frame of reference of the common centre
of mass of the beads.
H72 In the case of the inelastic collisions, after a sufficiently long time,
a growing mass of the coalescing beads reaches a constant velocity. Apply
Newton’s law of motion to this cluster. For elastic collisions, first examine
what would happen if the external force acted only until the first collision
had occurred.
H73 Consider what is happening to the centre of gravity of the table
plus jug plus beer.
H74 The viscosity of the water can be taken to be small and the change in
potential energy of the liquid should be neglected compared with the kinetic
energy. Note that the gutter cannot change the horizontal momentum of the
jet of water and Bernoulli’s equation is applicable (several times!).
H75 Obtain expressions for the changes, over a very short time interval
∆t, in the potential and kinetic energies of the initially stationary liquid, as
it starts to move with (initial) acceleration a.
H76 Experience indicates that the rate at which the sand runs through
the constriction does not depend upon the amount of sand in the upper part
of the egg-timer. The explanation for this is that, due to the friction between
the grains of sand, the average speed of the emerging sand depends only
on its nearby environment, primarily on the diameter of the hole, and not
on effects originating from remote parts. (This is not true for liquids, where
Hints 57
pressure effects are transmitted through large distances; see P49.) Thus the
time that the sand takes to run through the hole has to be proportional
to the cube of the initial height H of the sand. Find the other quantities
on which this time may depend and then apply the method of dimensional
analysis.
H77 For small displacements the net force exerted on the bob is F(x) ≈
−kx3 /'20 , where k is the spring constant. Using dimensional analysis one can
deduce the dependence of the period on the spring constant, the mass of the
bob and the amplitude of its motion.
H78 In the given circumstances, both the horizontal and the vertical
motion of the body can be approximated by harmonic oscillations.
H79 Describe the motion in the (decelerating) frame of reference of the
train.
H80 Examine the motion in the frame of reference fixed to the wedge.
H81 Under what conditions would a long, thin thread move uniformly
above the Equator in a synchronous orbit, i.e. with the same angular velocity
as the Earth?
H82 The normal component of the acceleration of the car is an = v 2 /ρ,
where v is the speed of the car and ρ is the radius of curvature of the
bridge. The latter can be deduced by considering the motion of a projectile;
it follows a trajectory which has the same shape as the surface of the bridge.
H83 Determining the radius of curvature of the track is the core of the
solution (see H82).
H84 In any time interval, the water carries the boat downstream by the
same amount as the remaining distance to the mark has been reduced.
H85 Calculate by how much the speed of the pushed child and its
velocity component down the slope change in unit time. Find a relationship
between the rates of change of these two quantities.
H86 Compare the rate of decrease of the distance between the smugglers’
ship and the coastguard’s cutter, to the speed at which the latter moves away
from the shore.
H87 Because of the symmetry of the problem, the bodies are always at
the corners of an ever-decreasing regular n-gon, and each of them moves
as if only the gravitational attraction of a centrally placed single body (of
a suitably chosen mass Mn ) acted on it. The time taken for the system to
collapse into the centre can be calculated using Kepler’s third law.
H88 The solution needs the application of all three of Kepler’s laws of
planetary motion.
H89 Recall the strengths and directions of the field associated with a
magnetic dipole; B = 2κµ/L3 on its polar axis (A position of Gauss) and
B⊥ = κµ/L3 on its equator (B position of Gauss), where κ has been written
for µ0 /4π.
H90 The force acting on the charge can be found using the so-called
method of image charges. The force – analogous to gravitational attraction –
is inversely proportional to the square of the distance. Therefore, the body’s
behaviour is similar to the motion described by Kepler’s laws for a degenerate
elliptical orbit.
H91 Use the method of image charges to find the value of the electric
field and the induced surface charge density in the region below the ball.
The ‘negative pressure’ due to the electrostatic forces acting on the surface
of the brine below the ball is balanced by the hydrostatic pressure of the
water ‘hump’.
H92 Apply the method of spherical image charges. The basis of this
method is that the electric field produced by two point charges, of opposite
signs and different absolute values, has a sphere as its zero potential surface.
H93 You need to use both the laboratory and the centre of mass reference
frames.
H94 If the sequence of events were re-played in slow motion, it would
be seen that immediately after the collision, the first ball stops and rotates in
a fixed place, whilst the second ball moves on but without rotation. Thus, in
the overall collision, the first ball transfers linear but not angular momentum
to the second ball.
After the collision, friction moves the first ball forward, but slows its
rotation. On the other hand, friction slows the translational motion of the
second ball, whilst increasing its rotation. Thereafter, the angular momentum
of each of the balls about its point of contact with the table remains constant.
H95 Consider energy conservation, but don’t forget to include dissipation
as heat.
H96 Examine the angular momentum of the ball about a point on the
table which lies on the ball’s path, but is otherwise arbitrary.
H97 Convince yourself that it is only the east–west component of the
traffic momentum that matters.
Hints 59
H98 Dynamical equations of motion and connections between transla-

tional and angular accelerations are central to the solution.
H99 From the point of view of energy and momentum, the system is
not closed and, therefore, these quantities are not conserved. Conservation
of angular momentum explains the strange phenomenon described in the
problem.
H100 Consider the forces acting upon a short length of the ring which
subtends an angle ∆θ at the axis of rotation.
H101 Determine the difference, ∆F, in the force stretching the thread
around the surface of the cylinder at two points on the cylinder’s surface
whose azimuthal separation is ∆α. This change in the force is propor-
tional to the force acting normally on the cylinder, and this in turn is
proportional to F. Consider an equivalent phenomenon, in which the rate of
change of some quantity is proportional to the quantity itself, (e.g. radioactive
decay, capacitor discharge, etc.). Using the analogy, relations applying to the
friction of the thread can be obtained.
H102 Jenny suggests that Charlie considers a homogeneous ring rotating
with constant angular velocity about an axis perpendicular to its plane and
passing through its centre. He should determine which forces act on the ring,
and consider how Newton’s second law is satisfied for the centre of mass of
a piece cut from the ring.
H103 The gravitational force both accelerates the hanging part of the
chain and impulsively sets into motion the next link. This means that the
changing mass of the moving chain has to be taken into account.
H104 Gravitation can be ignored in the frame of reference moving with
the centre of mass of the chain; in this frame the chain will be weightless,
but not massless. Examine in which directions forces act on a small piece of
the chain, which has radius of curvature R and moves at a uniform speed
v, and consequently determine how the shape of the chain is deformed (see
also P100, P101 and P102).
We can reveal that Frank’s guess is right – the chain keeps its original
shape.
H105 Calculate the tension in the chain when it leaves the pulley. Use
the principle of conservation of energy (see also P104).
H106 Examine the motion of the loop in the frame of reference moving
at the same speed c as the centre of the loop. In this frame the pieces of
the loop travel with uniform circular motion. The conditions governing the
dynamics of circular motion yield an equation for c.
H107 Examine the change in the horizontal momentum of the sand
falling onto the belt in unit time. Consideration of the energies involved is
also useful.
H108 Apply the law of conservation of energy; then find the force by
using the change in the momentum calculated from the speed of the roll as
a function of its position.
H109 The gravitational field of a thin spherical shell of uniform mass
distribution is zero inside the shell. Outside the shell, it is the same as if the
total mass of the shell were concentrated at its centre.
H110 The gravitational field inside a homogeneous sphere is directly
proportional to the radius of the sphere (see P109). The gravitational field
of the hollowed-out sphere can be found by superimposing the fields of a
homogeneous sphere and a smaller sphere of ‘negative mass density’.
H111 Divide the hemisphere into equally thick hemispherical shells.
Prove that these shells each produce the same gravitational field at the
point in question.
H112 Calculate the force a ‘mythical giant’ would have to exert to pull
the two halves of the asteroid (already cut in two) apart by 1 m.
H113 The electric field exerts a force whose magnitude is proportional
to the surface area exposed by the cut and is in a direction perpendicular to
that surface. Note that this force is similar to that caused by liquid or gas
pressure.
H114 At first sight several parameters seem to be missing. Don’t worry
about it! Find the equilibrium condition and calculate the electrostatic energy
of the system in that situation.
H115 Find the maximum amount of hydrogen that the container could
have contained initially without bursting. The material from which the
container is made may be chosen freely, but only from real materials.
H116 The laws of gravitational and electrostatic fields are very similar.
Make use of this similarity and apply Gauss’s law.
H117 Examine the electric field due to a rod element which subtends an
angle ∆α at the point C.
H118 Consider two very long rods joined end-to-end.
Hints 61
H119 You can distinguish between the correct and false formulae by
considering the case in which θ approaches π. Apply the well-known expres-
sion for the magnetic field of a long straight current-carrying wire to find
the proportionality factors.
H120 Imagine that another identical coil is joined symmetrically to the
original solenoid at point P , and that the same current is also allowed to
flow in this second coil. Apply the principle of superposition.
H121 It is easy to find the force if one imagines changing the positive
charges on one of the plates into negative charges of the same magnitude.
On the other hand the electric field line structure of the positive–positive
plates is very different from that of the positive–negative arrangement.
H122 Remember that the total charge on an isolated plate cannot
change.
H123 The total charge induced on each plate would not change if the
point charge Q were considered to be spread uniformly over a plane a
distance x from the lower plate.
H124 The total electric field outside the plates must be exactly zero. What
are the consequences of this well-known fact for the charge distribution?
H125 Imagine that the medium is sliced into thin layers perpendicular
to the y-direction. The individual layers can be considered as plane-parallel
plates with different refractive indices, and the relation between the refractive
index of a layer and the angle of incidence of the light ray can be determined.
H126 Using simple geometry, you can measure the useful surface area of
a CD. To obtain the required result divide this area by 650 M and also by
8, because 1 byte = 8 bits. You can treat a CD as a reflection grating and
measure its diffraction pattern using a laser beam of known wavelength.
H127 Determine nλ for the composite line and consider possible values
of n, the order of the diffraction spectrum.
H128 In case (i), the optical path difference consists of two parts; one
originates in front of the grating and the other behind it.
In case (ii), instead of considering an optical grating, investigate the
diffraction pattern from a single slit, which is tilted ‘forward’ through an
angle φ.
H129 Draw diagrams showing liquid levels and pressures in the space
between the objects and on either side of them.
H130 Find the horizontal forces acting on the meniscus.
H131 During evaporation, the surface area of the drop shrinks, and
its surface energy decreases. Compare this energy decrease with the energy
needed for evaporation.
H132 The equilibrium saturated vapour pressure is slightly higher near
the surface of a smaller drop than near the surface of a larger one. The
vapour pressure is uniform at the bottom of the container and its value
is therefore higher than the equilibrium value for large drops, but lower
than that for small drops. Consequently, vapour evaporates from the smaller
drops, making them smaller, and condenses onto the large ones, making
them larger.
The relationship between the equilibrium pressure of saturated vapour
and the curvature of the drop can be deduced by considering the pressure
balance in a vessel containing a capillary tube hanging into some of the
liquid.
H133 The increase in internal energy of the air enclosed in the container
is equal to the decrease in potential energy of the load hung from the piston.
H134 If a mountain is very high then its base melts because of high
pressure. Compare the energy needed to melt the bottom layer of a mountain
with the gravitational energy that would be released if the mountain then
sank.
H135 If the state of the enclosed air is plotted on a p–V diagram, a
straight line is obtained. The hidden elegance of the problem is revealed
when the implications of the straight line’s being tangential to an isothermal
or adiabatic curve at certain points is realised.
H136 Your explanation should be based on the interaction between the
molten magma and the ice.
H137 The hydrostatic pressure of the water in the flue increases the
pressure of the water in the cavity, and so it boils at a temperature higher
than the usual 100 ◦ C. The relationship between the pressure and temperature
of the saturated water vapour can be obtained from tables or by using the
approximate law
p = Ae−Lm /(RT ) .
Here p is the pressure of saturated water vapour at its boiling point T , Lm
is the molar heat of vaporisation of water, R is the gas constant and A
is a constant with the dimensions of pressure. When the geyser erupts, the
superheated water in the cavity reaches equilibrium again by boiling until it
cools down to 100 ◦ C.
Hints 63
H138 Consider the heat balance at the base of the layer when the layer
thickness is x.
H139 Establish that the time taken varies as the square of the linear
dimensions for similarly shaped bodies.
H140 The ‘trap’ hidden in this problem relates to the heat of vaporisation.
The heat of vaporisation of water at 100 ◦ C and a pressure of 1 atm (the
standard value of 2256 kJ kg−1 found in tables) takes into account not
only the higher internal energy of the vapour but also the work done by
expansion against atmospheric pressure.
H141 A liquid starts boiling when its saturated vapour pressure reaches
or surpasses the pressure of the gas above the liquid.
H142 Consider how T could be incorporated in a formula for the
amplitude.
H143 Compare the surface energy of a long cylinder of water (assuming
that the cobweb is uniformly covered with water) and the surface energy of
the periodic water drops.
H144 The cylinder keeps accelerating until the net momentum received
per unit time, due to the particles colliding with it from both the left and
the right, becomes zero. However, after a very long time, the cylinder stops
moving, in agreement with the second law of thermodynamics.
H145 Take into account both the emission and the absorption of heat by
the space probe and the consecutive inner and outer surfaces of its protecting
shields.
H146 The entropy of the system cannot decrease during the process.
H147 Consider entropy from the point of view of the number of micro-
states available.
H148 Calculate the change in entropy of the air that is pumped into the
container.
H149 The electric field strength is zero inside the space ship, just as it is
inside a Faraday cage. Examine whether the electric potential of the space
ship changes during the journey.
H150 Examine the change in energy of the spherical capacitor when it
carries a set charge.
H151 Compare the energies of the electrostatic fields of the dented and
undented foils.
H152 The opposite charges on the capacitor plates attract each other,
and therefore work has to be done to pull the plates apart. The capacitance
of the capacitor decreases, and as, at a given voltage V0 , the electrostatic
energy of a capacitor is proportional to its capacitance C, the energy of
the capacitor decreases! The solution to this paradox is that a capacitor
connected to a battery cannot be considered as a closed system.
H153 Because the current flows in the same direction in each turn, the
spring contracts. The force of contraction caused by the current can be found
by considering a superconducting spiral spring (at a very low temperature in
practice). A current can flow in such a superconducting coil even if its ends
are short-circuited. Examine the dependence of the energy of this closed
system on its length.
H154 Find the net force (the sum of the magnetic force, the weight and
the tension in the string) exerted on magnet A as a function F(x) of the
distance x apart of the magnets. Use F(x) to determine the conditions for
equilibrium and stability.
H155 Calculate the total work done by the battery.
H156 Remember that the charge is unchanged when the oil is removed.
H157 The energy (per unit volume) of the electrostatic field is pro-
portional to the square of the electric field strength and to the dielectric
constant of the medium: Wel = 12 ε0 εr E 2 . The dielectric between the plates
of the capacitor decreases the electric field (as a result of its polarisation),
and therefore the energy of the system decreases as well. The force acting on
the dielectric can be calculated from this change in energy (using the work
theorem).
H158 Apply Kirchhoff’s laws, starting from the final element of the
chain. Look for a relationship between the currents flowing through the
consecutive resistors and the terms in the Fibonacci series.
H159 Consider two different cases. In the first case, a current I flows into
a grid point. In the second case, a current I flows out of the neighbouring
grid point. In both cases make use of the symmetry of the system, and then
superimpose the two current and voltage distributions.
H160 Apply the method of superposition as in the previous problem.
Be careful, since with a finite grid the current has to flow out of the circuit
somewhere in order to conserve charge. Solve this difficulty without spoiling
the symmetry of the problem.
Hints 65
H161 The key phrase of the solution is ‘in parallel’.

H162 As in the previous three problems, superposition is a great help.
H163 The battery is to be connected to the terminals of the resistor
through the ammeter. You need to ensure somehow that all the current
measured by the ammeter flows through the particular resistor, and not
partly via other electrical elements.
H164 Establish sets of equipotential points on the cube, when a current
I is flowing in at one end of the diagonal and out again at the other end. The
circuit can then be simplified by notionally connecting together all points at
the same potential.
H165 Apply Gauss’s law.
H166 Use Fleming’s right-hand rule.
H167 The rod starts to accelerate down the slope under gravity. Electro-
magnetic induction causes a current to flow in the rod, which in turn brakes
its motion according to Lenz’s law. The equation of motion of the rod
(written in terms of the current) is the same in all three cases. The different
behaviours are due to the different relationships between the current flowing
in the rod and the induced electromotive force.
H168 The change in velocity of the rod is directly proportional to the
change in charge of the capacitor. The rod accelerates until the induced
e.m.f. balances the remaining voltage across the capacitor.
H169 You can answer question (i) without solving the differential equa-
tion for the circuit. Express the rate of increase of magnetic energy as a
function of the current.
For (ii), note that the time dependence of the current for this circuit is
well known and that the Joule heat is directly proportional to the square
of the current. If you sketch a graph of the square of the current as a
function of time, you can find qualitatively the time at which the rate of
change of dissipation in the resistor is the fastest. Using the result from (i),
the quantitative answer can be found without the need to use calculus.
H170 Consider first a circuit containing only one inductor and one
capacitor connected in series, and show that it has current resonance at a
particular frequency.
H171 According to the law of induction, the current flowing in a coil
cannot change suddenly.
H172 Show that n, the number of turns per unit length of the solenoid,
is proportional to d−2 , and consider the resistance of one turn.
H173 Examine the forces keeping the electrons in the metal in circular
orbits. If the electric field strength is known, Gauss’s law can be used to
determine the corresponding charge distribution.
H174 The electric field can be defined by the force acting on a unit
charge, and the magnetic field can be interpreted with the help of the
Lorentz force exerted on a moving charge.
H175 Jill’s result is correct, and Jack’s answer is wrong. The crucial point
is that the electric field lines within the rotating spoke are not parallel.
H176 The magnetic field of the Earth induces a current in the rotating
ring, which changes the average magnetic field at the centre of the ring. As
a result the magnetic needle moves.
H177 Let the current through the voltmeter be I and that through the
major arc of the ring be i. Then, using consistent conventions for current
directions and circuit traversal, apply Kirchhoff’s laws to two different closed
circuits.
H178 A Moebius strip is a surface that has no associated direction and
the law of induction must be applied only with great caution. Imagine the
wire marking the edge of the band to be laid out on a plane in such a way
that it does not cross itself. Determine the area of the plane enclosed by the
wire and find its equivalent for the Moebius strip.
H179 Consider not only the magnetic field and magnetic forces acting
on the charged particle, but also the effects of the induced electric field.
H180 As a result of the electromagnetic induction an electric field is
established in the charged ring, and its tangential component causes the ring
to experience a torque. One can show that the final angular velocity of the
ring depends only on the final field, and not on the way it is turned on.
H181 An electric field is induced in the rotating disc and this induces a
current in the coil. The total magnetic field is that due the Earth, increased or
decreased by that due to the coil, according to the direction of the rotation.
H182 The total magnetic flux through the superconducting ring (consist-
ing of that due to the external field and its own flux) must not change during
the motion. The flux of the external field changes during the motion, but the
change is balanced by the magnetic flux due to the current induced in the
ring. If the current is known, the Lorentz force can be calculated, and the
Hints 67
net force acting on the ring can be found as a function of its position. This
resulting equation of motion is similar to a well-known mechanics equation.
H183 The electrostatic field of a dipole can be calculated from its po-
tential Φ = K (cos θ/r 2 ), where K is a constant proportional to the strength
of the dipole, r is the distance from the dipole and θ is the polar angle
measured from the dipole’s axis. Calculate first the normal force exerted by
the string on the bead.
H184 Moving with velocity v0 perpendicular to a magnetic field of mag-
nitude B, is equivalent to being in an electric field of magnitude v0 B. If v0
is suitably chosen, this electric field can be made to cancel the gravitational
field acting on the particle.
H185 The changing magnetic field induces eddy currents in the loops,
which brake the fall of the magnet. The terminal speed clearly depends
on the resistance of the conductors. The dependence on other parameters
can be found by applying dimensional analysis. Don’t forget that formulae
involving magnetism usually contain the vacuum permeability µ0 , which has
non-trivial dimensions.
H186 Although it is possible to solve this problem in a reference frame
fixed to the vacuum chamber, the solution is rather complex. It is much
easier to handle the problem using a different frame of reference that moves
with velocity v0 parallel to the wire. In this frame of reference the motion
of electrons is subject to both an electric field and a magnetic field. On the
other hand, in this frame, when the electron is closest to the wire its velocity
is zero and the work–energy theorem can be used to solve the problem.
H187 Describe the phenomenon in the frame of reference of the liquid.
Consider first the transformation of the electric and magnetic fields when
they are viewed in two different frames of reference, one moving at speed v0
relative to the other. Take v0 as being much less than the velocity of light
and ignore relativistic effects.
H188 Consider the initial (activation) energy required for three fission
products rather than two.
H189 7 Be decays via electron capture.
H190 In equilibrium, abundances are proportional to half-lives. Use this
to show that the equilibrium value for the number of radon-220 atoms is
3.3 × 105 . Then consider how this element can be produced from the purified
sample.
H191 Use the relativistic formulae for energy and momentum conserva-
tion.
H192 Examine the motion of the electrons in the wall of the Faraday
cage.
H193 The mass of the proton is much (nearly 2000 times) larger than
that of the positron. For this reason, the positrons move with a much larger
acceleration than the protons and there will be a period in the motion when
the positrons have already moved far from the square, whilst the protons
have hardly moved at all.
H194 Apply the (relativistic) conservation laws of energy and momen-
tum.
H195 In the course of this process (Compton scattering), the total energy
and momentum of the colliding particles (photons + electrons) remains
unchanged. It is convenient to take the rest energy of the electron E0 =
me c2 ≈ 510 keV as the unit of energy for the calculations.
H196 Consider the electron as a spherical capacitor with radius r and
a uniform surface charge distribution. The moment of inertia of a sphere
of mass m and radius r is I = Kmr 2 , where K is a dimensionless constant
depending on the mass distribution. For example, for a homogeneous sphere
K = 25 .
H197 Apply the Heisenberg uncertainty principle and consider the total
energy of the electron.
H198 When an electron is enclosed in a sphere of radius r, the uncer-
tainty principle prescribes a minimum momentum for it of p ≈ h̄/r. Using
approximate relativistic formulae, calculate the total energy (the sum of the
electrostatic and kinetic energies) of the electron as a function of the radius
r and find the minimum of this function.
H199 For small wavelengths, the speed of propagation of surface water
waves is determined by the surface tension. Examine the dependence of the
speed of these (capillary) waves on their wavelength and consider whether
this implies a lower limit for the wavelength.
H200 Take a bottle of champagne and try it. If you do not drink too
little (or too much!) of it, you will almost certainly spot the reason for the
acceleration.
Solutions
S1 Resolve the velocity of snail 2 into a component pointing towards

snail 1 and a component perpendicular to this (see Fig. S1.1). These two
snails approach each other at a relative speed of v + 12 v = 32 v = 7.5 cm min−1 ,
and therefore they meet after a time given by 60 cm/7.5 cm min−1 = 8 min.
In fact, they must all meet after this time and, as they actually travel at a
speed of 5 cm min−1 , they each cover a distance of 40 cm before doing so.
3
1 v 2
v /2
Fig. S1.1
The same result can be obtained if the velocity vector of one of the snails is
resolved as shown in Fig. S1.2 into a component pointing towards the centre
of the triangle formed by the snails, and a component perpendicular to this.
This shows that the snails approach the centre√of the triangle
√ (it is obvious
that this is where they meet) at constant speed ( 3/2)v = (5 3/2) cm min−1 ,
whilst travelling around this point at a tangential speed of 12 v.
3
v /2
1 v 2
3v
2
Fig. S1.2
69
√
It is easy to show that the snails are initially at a distance of 60 ( 3/3) cm
from the centre of the triangle, and that therefore they meet in
√
60 ( 3/3) cm
√ = 8 min.
5 ( 3/2) cm min−1
r∆u v
−∆ r a
∆u
u
Fig. S1.3
Because of the geometrical symmetry of the situation, each snail always

moves so that its direction of motion makes an angle of π/6 with the
line joining its current position to the centre of the triangle. However, it is
worth generalising the problem of calculating the trajectory. Consider the
motion of a body moving at constant speed v around a fixed point with
the angle between the velocity and position vectors equal to a fixed value α,
(0 < α < 90◦ ). If the position vector, of initial length r0 , moves through a
small angle ∆φ and its length changes by −∆r (see Fig. S1.3), then, since α
remains constant,
∆r(φ)
= −r(φ) cot α.
∆φ
This equation is very similar to the radioactive decay equation, dm(t)/dt =
−m(t) λ, the known solution of which is m(t) = m0 e−λ t . Using this analogy,
the equation of a snail’s path (in polar coordinates) is
r(φ) = r0 e−φ cot α .
This is the equation of the so-called logarithmic spiral and implies that the
radius r tends to zero only after turning through an infinite angle, i.e. a
point-like body reaches the centre in finite time and by covering a finite
distance, but only after making an infinite number of turns about the centre.
Note. Nocturnal insects try to follow straight flight paths by keeping a

constant bearing with respect to a distant light source (e.g. the Moon). If a
nearby lamp misleads them, then, according to the solution just found, they
will follow a spiral path to disaster. As neither the insects nor the lamp are
point-like, sooner or later the insects hit the lamp.
Solutions 71
S2 The average speed of the object is 12 m s−1 . Since it decelerates uni-

formly, vaverage = 12 (vinitial + vfinal ), and thus its initial speed cannot be greater
than 1 m s−1 , (since vfinal ≥ 0). It follows that the speed of the body decreases
by a maximum of 1 m s−1 in 2 s. Thus the absolute value of its acceleration
|a| is at most 0.5 m s−2 , i.e. 20 1
times the gravitational acceleration. Therefore,
the coefficient of kinetic friction between the object and the table surface
1
cannot be greater than 20 . This is much smaller than the coefficients of
friction between ordinary materials and therefore it is very likely that the
object does not slide, but that all or part of it rolls.
S3 (i) The shortest path is one perpendicular to the bank and the boat
goes in this direction if the boatman rows in the direction shown in Fig. S3.1.
3 m s −1
−1
5 ms Boat
2 m s−1 a
River
Fig. S3.1
The resultant
√ speed of the boat (in the direction perpendicular to the
bank) is 5 m s ≈ 2.24 m s−1 . The boatman has to row upstream at an
−1
angle α to the bank, where cos α = 23 ; this gives α ≈ 48◦ .

(ii) In this case, the current is so strong that the boat will move down-
stream even if the boatman rows at full speed against the stream. This
means, in contrast to the previous case, he cannot choose his direction with
respect to the bank and, in particular, he cannot travel across in a direction
perpendicular to the bank.
The possible directions he can take may be determined by adding all the
possible still-water velocities of the boat to the velocity of the river. Draw
the velocity vector of the river and, from the endpoint of this vector, draw
velocity vectors in all directions, with a magnitude equal to the speed of the
boat in still water. The endpoints of these vectors will form a circle as shown
in Fig. S3.2.
Boat
3 m s−1
4 m s−1
River
Fig. S3.2
The possible resultant velocities of the boat can be obtained by joining

the starting point of the velocity vector of the river to the points on this
circle. The resultant corresponding to the shortest path will be the one that
makes the greatest angle with the direction of the current, i.e. when the
line of action of the resultant velocity vector is a tangent to the circle (see
Fig. S3.3).
−1 −1
7 ms 3ms
b
−1
4ms
Fig. S3.3
√
Thus, the velocity of the boat with respect to the shore is 7 m s−1 ≈
2.65 m s−1 . Again, the boatman has to row upstream, but this time at an
angle β to the bank, where cos β = 34 , yielding β ≈ 41◦ . The figure also
shows that in this case the distance travelled by the boat will be 43 times the
width of the river.
S4 Let the position of the moving end of the carpet be x as shown in
the figure. It follows that the other end of the moving part is at x/2, and
hence that the coordinate of its centre of mass is 3x/4. Although dx/dt = 1,
the speed of the centre of mass of the moving part is only 34 !
0 x /2 x x
1 2
The linear momentum of the moving part is p = mv, where v = 1 and m

is increasing uniformly with time. The net force acting on the moving part
is thus
dp dm dv dm
F= = v+ m= 1 + 0.
dt dt dt dt
The rate of change of the mass of the moving part can be found with the
help of the following argument. The moving end of the carpet starts from
the origin and the whole carpet will be moving when it reaches x = 2; this
it does after two units of time, i.e. dm/dt = 12 . The corresponding minimal
force (neglecting all dissipative forces) is F = 12 .
Note. (i) The centre of mass of the moving part of the carpet is initially
at the origin and after two units of time at x = 32 , again showing that the
speed of the centre of mass (vCM ) of the moving part is 34 .
Solutions 73
(ii) Notice that the linear momentum of the moving part (p = mv) is not
equal to the product (mvCM) of its mass and the speed of its centre of mass.
(iii) It seems tempting to try to find the minimum force required by using
the conservation of energy, i.e. F × 2L = mv 2 /2, where L is the length of
the carpet, (L = 1). The result would be F = 14 , which is only one-half
of the value calculated earlier. The error in this argument is to ignore the
continuous inelastic collisions which occur when the moving part of the
carpet is jerking the next piece into motion. Half of the work goes into the
kinetic energy of the carpet, but the other half is dissipated as heat.
S5 Solution 1. If space (planar) and time coordinates are established in an

orthogonal frame of reference with axes x, y and t (a space–time diagram),
the ‘world-line’ of a snail travelling with uniform rectilinear motion will
obviously be straight. Encounters occur when two snails are at the same
place at the same time, i.e. when their world-lines intersect as shown in the
figure.
a
A
P
b Q
B
c
d
According to the information given in the problem, the four world-lines

(a, b, c and d) definitely intersect in pairs at five ‘points’. Let us denote these
encounters by A, B, C, P and Q. Points A, B and C determine a plane (the
plane of world lines a, b and c). Since P and Q also lie in this plane, world
line d must do so as well. This means that world lines c and d also cross
each other, implying that the sixth encounter will also take place.
Solution 2. As five encounters have already occurred, there has to be a
snail who has already met all three of his fellows. Let us denote this snail
by α. In our imagination, let us sit on the back of α, i.e. choose a frame of
reference in which α is at rest at the origin.
The three other (moving) snails, (β, γ and δ) have already met α, and
have therefore crossed the origin. Moreover, one of them (β say) has already
met his two other moving fellows (since five encounters have occurred).
This is only possible if β, γ and δ are moving along the same straight line.
Consequently, sooner or later, γ and δ will have to meet as well.
S6 The work done against gravity can be calculated from the increases
in height of the centres of mass. The centre of mass of a worm ‘folded in
two’ is located at the middle of either half, i.e. at a point one-quarter of the
worm’s total length from one end. This is illustrated in the figure.
Centres
of mass
Thus, the centre of mass of the narrow flatworm travels 5 cm up the wall,
whilst that of the broad one moves 7.5 cm. The ratio of the amounts of work
done is therefore 2 : 3.
Note. The centre of mass of the worm is not always in the same position
with respect to the worm; indeed, it need not be at any point of the worm
at all. The centre of mass of the straight worm is obviously at its centre, and
that of the worm folded in two is at its quarter-length point. Thus the centre
of mass of a flexible body does not remain at a fixed point within the body;
its relative position may change. This principle is used by high-jumpers;
when a high-jumper’s body arches over the cross-bar, the body’s centre of
mass remains below it (see also P32).
S7 (i) Let us denote the elastic constant (spring constant) of the rope
by k and its unstretched length by '0 . The maximum length of the rope is
'1 = h − h0 = 23 m, whilst in equilibrium it is '2 = (23 − 8) m = 15 m.
Initially, and at the jumper’s lowest position, the kinetic energy is zero. If we
ignore the mass of the rope and assume that the jumper’s centre of mass is
half-way up his body, we can use conservation of energy to write
1
mgh = k ('1 − '0 )2 .
2
In addition, in equilibrium,
mg = k ('2 − '0 ) .
Dividing the two equations by each other we obtain a quadratic equation
for '0 ,
'20 + 2(h − '1 )'0 + ('21 − 2h'2 ) = '20 + 4'0 − 221 = 0,
which gives '0 = 13 m.
Solutions 75
(ii) When the falling jumper attains his highest speed, his acceleration must
be zero, and so this must occur at the same level as the final equilibrium
position (' = '2 ).
Again applying the law of conservation of energy,
1 2 1
mv + k ('2 − '0 )2 = mg('2 + h0 ),
2 2
where the ratio m/k is the same as that obtained from the equilibrium
condition, namely,
m '2 − ' 0
= .
k g
Substituting this into the energy equation, shows that the maximum speed
of the jumper is v = 18 m s−1 ≈ 65 km h−1 . It is easy to see that his
maximum acceleration occurs at the lowest point of the jump. Since the
largest extension of the rope (10 m) is five times that at the equilibrium
position (2 m), the greatest tension in the rope is 5mg. So the highest net
force exerted on the jumper is 4mg, and his maximum acceleration is 4g.
S8 If the berg has base area A and height H, then M = 13 AHρice . If
the height showing above the surface is h, the flotation condition gives
(H 3 − h3 )ρwater = H 3 ρice . When the berg is depressed by a small amount
x the additional submerged volume is xA(h/H)2 and the upthrust is this
multiplied by ρwater g. This gives that the angular frequency of oscillation ω
is determined by
3h2 ρwater g
ω2 =
ρice H 3
and, on substituting numerical values, that the period of oscillation is about
11 s.
S9 First of all we note that the right front suspension spring will be
further compressed as a result of parking on the pavement. We can measure
both the change of tensions in the suspension springs and the rise of the
car body in centimetres, and will let the sign be positive if the spring is
further compressed. The net torque must be zero about any axis, including,
for example, the diagonals of the rectangle formed by the wheels; so the
changes in tension at the opposite ends of a diagonal must be equal. This is
why the springs of the right front (rf) and left back (lb) wheels are equally
compressed, by an amount x, and the left front (lf) and right back (rb)
suspension springs each lengthen by x. This equality of changes in length
ensures that the net force provided by the springs to support the weight of
the car does not change.
The rises of the car body both at the lf and rb wheels are x, at the lb
wheel the rise is −x and at the rf wheel (on the pavement) it is 8 − x. The
frame of the car is rigid, so, because of the equal movements at the lf and
rb wheels, the midpoint of the chassis also rises by x. Similarly, the other
chassis diagonal remains a straight line, and so the rise at the rf wheel must
be the same as the fall of the body at the lb wheel relative to the midpoint
of the chassis, i.e. (8 − x) − x = x − (−x). From this very simple equation we
get x = 2 cm. We conclude that above the wheel on the pavement the body
of the car rises 6 cm, above the left back wheel it sinks 2 cm and above the
other two wheels it rises 2 cm.
Applying the same calculational technique, it is easy to show that com-
pressions in the suspension springs cannot change when the car is parked
with both right wheels on the pavement. It follows that then the right side
of the car body simply rises 8 cm, the height of the pavement. You can also
show that the result does not depend upon the number and the positions of
the people sitting in the car; this is because we have only investigated the
relative displacement of the body of the car before and after parking on the
pavement.
Note. In the solution above, a slight rotation of the body of the car was
ignored.
S10 Fig. S10.1 shows Jean Valjean’s location on the wall. Figure S10.2 is
a sketch showing his weight (mg), the normal reactions of the walls (N) and
the static frictional forces (Ffr ) acting on his limbs.
h h
Ffr Ffr
N N
mg
Fig. S10.1 Fig. S10.2

Solutions 77
Ffr sin h Ffr sin h
N N
Top view
Fig. S10.3
Let the static frictional forces make a common angle θ with the vertical.
The conditions for static equilibrium (see Fig. S10.3) are
mg = 2Ffr cos θ and N = Ffr sin θ.
From these equations we can express the normal component, N, of the
force exerted by the prisoner on the wall whilst climbing as
N = 12 mg tan θ.
Thus the total force required, F, is given by
mg 2
1 + sin2 θ
F 2 = N 2 + Ffr2 = .
2 cos2 θ
We can also find the minimal force using the inequality
Ffr ≤ µ0 N,
from which it follows that
1 1
sin θ ≥ or tan θ ≥ ,
µ0 µ20 − 1
where µ0 is the coefficient of static friction. Using either of these inequalities
we find the minimal force to be
mg µ20 + 1
Fmin = .
2 µ20 − 1
This expression shows that the coefficient of static friction must be greater
than unity if Jean Valjean is not to fall off the wall. If the coefficient of static
friction approaches infinity, the force on each of his hands is equal to half
of his body weight; this situation corresponds to his being glued to the wall.
S11 If static friction is large enough, the sphere will not slide down
the slope. However, this by itself is not sufficient for equilibrium; it is also
necessary that the sphere does not roll down the inclined plane.
The sphere is made of two hemispheres, implying an inhomogeneous mass

distribution. If the distance between its centre of mass and geometrical
centre is less than 12 r, where r is the radius of the sphere, then, whatever
the orientation of the sphere, its weight will produce a torque about P , the
point of contact with the inclined plane (see Fig. S11.1), which will make the
sphere roll.
r/2
O
o
r
30
P
30o
Fig. S11.1
It will now be shown that this is the situation for any sphere made of two
homogeneous hemispheres – whatever the densities of the two halves.
S
A O
Fig. S11.2
Consider the shaded area in Fig. S11.2. By symmetry, the centre of mass of
this part is obviously at point A, i.e. at a distance 12 r from the centre, O. The
rest of the sphere moves the centre of mass S of the whole even closer to
point O, i.e. OS < 12 r. From our previous considerations, this implies that the
sphere cannot remain in equilibrium on the 30◦ inclined plane. In obtaining
the solution, we have assumed that rolling resistance is small, i.e. no resistant
torque can act at point P. In the case of a surface covered with Velcro, this
is obviously not true, and the sphere may even adhere to a vertical surface.
S12 Viewed within a rectangular coordinate system which has one axis
parallel to the inclined plane, the ball is seen to bounce on a ‘horizontal’
plane in a ‘vertical’ field of gravitational acceleration g = g cos α. It also
experiences an additional constant ‘horizontal’ acceleration (of magnitude
g sin α). The ‘vertical’ motion consists of bounces of identical heights, i.e.
Solutions 79
of identical periods. Meanwhile, since the ‘horizontal’ acceleration is con-

stant, the ball’s average speed between bounces increases uniformly, and so
the distances between two consecutive bounces increase in an arithmetical
progression.
S13 Let the midpoint of the platform be a distance h below the pivot
and the hamster’s distance from that midpoint be x, as shown in the figure.
x
mg
Because of gravity the hamster exerts a torque mgx about the pivot of the
wheel-cage. On the other hand, as the hamster moves it accelerates using
friction with the platform. When its acceleration is a this produces a reaction
force of ma on the platform, directed away from its midpoint. The torque
due to this force is mah. The wheel-cage (and the platform) remains in static
equilibrium if these two torques are equal, i.e.
mgx = mah.
After making due allowance for its direction, the acceleration can thus be
written as a = −(g/h) x. This shows that the required motion of the hamster
is simple harmonic motion with an angular frequency ω = g/h.
S14 (i) The bicycle moves in the direction of the net force (the sum of the
applied backward force and the frictional force directed forwards). In usual
gearings the bike moves backwards, but extremely low gearings can cause
forward displacement. Because the work done is always positive, it follows
that the student’s hand moves backwards relative to the ground. Normally
the gearing N is greater than one, i.e. the rear wheel rotates more rapidly
than the pedals. However, in the unusual case N < r/R < 1 (where R is the
radius of the wheel and r is the length of the pedal arm), the bicycle could
move forwards despite the oppositely applied force.
(ii) The chain-wheel rotates in the same sense as the rear wheel.
(iii) Usually backwards and upwards; the superposition of (i) and (ii), but
with (i) larger because of the gearing and wheel sizes.
Note. It is interesting to note that there is one position of the pedals at
which an arbitrarily large force can be applied without moving the bicycle
either way.
S15 Equate the force of attraction between the Sun and the Earth to the
centripetal force that keeps the Earth in its approximately circular orbit, and
express the angular velocity ω in terms of T , the period of revolution. This
gives
mM 4π2
G 2 = mrω 2 = mr 2 ,
r T
where m and M are the respective masses of the Earth and Sun, and r is the
average distance between them. Divide by m and express M in terms of the
average density ρ and radius R of the Sun as follows:
4
πR 3 ρ 4π2
G3 = r.
r2 T2
This yields
3π r 3
T =
Gρ R
for the period of revolution.
It can be seen that the Earth’s rotation period only depends on the
universal gravitational constant G, the average density of the Sun and the
ratio r/R. Therefore if the density of matter remains constant, any scaling of
the solar system leaves the length of a year unchanged. It can also be seen
that only the density and size of the Sun are relevant; the Earth’s data are
not. Any body that is small in size relative to the Sun would have the same
period and follow the same orbit.
Note. This result can also be obtained using Kepler’s third law T 2 /a3 =
4π 2 /GM, where a is the semi-major axis of the Earth’s elliptical orbit. If
the mass of the Sun is expressed in terms of its average density then it is
clear that a proportional reduction does not change the period of planets
in elliptical orbits.
S16 The gravitational acceleration produced by the Sun at the position of

the Earth is the same as that due to one of the stars at the position of the
other. This is because, according to Newton’s law of gravitation, g (the
gravitational force acting on a unit mass) depends only on the mass at the
centre of attraction and the distance of the second body from it. These
quantities are identical in the two systems.
Thus, the members of the binary star move with the same acceleration as
the Earth but in an orbit of radius only half that of the Earth’s orbit. This
means that, since (a = rω 2 ), the square of their angular velocity has to be
twice as large as that of the Earth. The period of the
√ binary star therefore
equals that of the Earth around the Sun divided by 2, i.e. 8 12 months.
Solutions 81
S17 (i) The acceleration of a satellite moving at speed v in a circular orbit

of radius R is g = v 2 /R. If R is the radius of the Earth (or more precisely, a
slightly larger value), then g has to be the gravitational acceleration at the
√
Earth’s surface; this defines the ‘first cosmic speed’, v1 = Rg = 7.9 km s−1
for the speed of the satellite.
All of this speed is not strictly necessary for launching a satellite if the
‘initial speed’ provided by the rotation of the Earth is taken into account.
This help is greatest at the Equator, approximately 0.5 km s−1 , and means
that an initial speed of 7.4 km s−1 with respect to the Earth is sufficient, but
only if the satellite is launched eastward from on or near the Equator.
(ii) The angular momentum of satellites in polar orbits (passing over the
poles) is zero with respect to the axis of rotation of the Earth. This condition
has to be fulfilled right from the launch, since the angular momentum will
not change later. The ‘help’ described above cannot therefore be utilised.
Indeed, the rotation of the Earth is a drawback, since, not only must the
satellite receive the speed of 7.9 km s−1 in a north–south direction, but, in
addition, the unacceptable west–east speed due to the rotation of the Earth
has to be cancelled. The latter would not have to be taken into account if the
satellite were launched from the neighbourhood of one of the poles; there
are obvious technical difficulties in doing this!
The initial speed necessary to put a satellite into a polar orbit is therefore
at least 7.9/7.4 ≈ 1.06 times greater, i.e. the necessary kinetic energy is at
least 1.13 times as great. This does not seem a big difference, but in reality
the slightest increase of the initial speed requires an enormous effort. The
reason for this is that the carrier rocket has to be accelerated as well as
the satellite, and the mass to be launched increases exponentially with the
intended final speed.
(iii) In order to escape the attraction of the Earth (i.e. to move far from
the Earth), the√probe has to acquire the escape or ‘second cosmic’ speed,
√
v2 = 2gR = 2v1 ≈ 11.2 km s−1 . The rotation of the Earth can again be
used. Launching eastward from the Equator, a launch speed of 10.7 km s−1
relative to the Earth is sufficient.
(iv) The Earth revolves around the Sun at a speed of approximately
30 km s−1 . In order to reach the Sun, a space probe has to be launched
at an initial speed of 30 km s−1 (or more exactly, 29.5 km s−1 ; smaller by
0.5 km s−1 due to the rotation of the Earth). √ If the aim is to leave the
solar system, the space probe has to reach 2 times the speed of 30 km s−1 ,
but the ‘initial speed’ of the Earth in its orbit can be subtracted from this
value if the launching is well timed and directed, i.e. a speed of 12 km s−1 is
sufficient.
Thus, it is easier to make a space probe leave the solar system than
to send it into the Sun. (The former has been successfully attempted, the
latter we are still waiting for.) The situation is even more favourable if the
possibilities offered by the outer planets (Mars, Jupiter, Saturn, . . .) are taken
into account. A space probe launched at the right time and in the right
direction can be significantly ‘pushed by’ (i.e. receive energy from) these
planets, a phenomenon known as ‘gravitational slingshot’. Designed in this
way, the space probe does not have to propel itself very far, it only has to
reach Mars or Jupiter and the rest happens ‘automatically’.
Note. If the rocket driving the probe can be fired in two stages, one to
induce an elliptical orbit and the other to bring the rocket to rest at the
aphelion of the orbit, less total energy is needed to hit the Sun than that
calculated above.
S18 The rocket has to reach the highest possible total energy. If the zero
level of gravitational potential energy is ‘infinitely’ far away, then the energy
of the rocket standing on the surface of the Earth is negative. The energy
released during the operation of the engines increases the total energy of the
rocket, and the rocket can leave the Earth’s gravitational field if the sum of
its potential and kinetic energies becomes positive.
The energy released in the course of the operation of the principal and
auxiliary engines increases the total energy of the rocket and its ejected
combustion products by a fixed value; this increase is independent of the
moment when the engines are switched on. However, the speed at which
the combustion products fall back to the Earth does depend on the timing
of the rocket’s operations. Indeed, if the auxiliary engine starts working
when the rocket is at a greater height, the combustion products fall further
and their speed and total energy are higher when they hit the ground. This
means that the sooner the auxiliary engine is switched on, the higher the
energy ultimately acquired by the rocket. The same argument is valid for
the principal engine, and if only energy considerations apply, it is best to
operate the engines for the shortest time and at the highest thrust.
S19 The mass distribution, and thus the position of the centre of mass,
changes from moment to moment as the ball sinks. In a time t, the centre
of mass is displaced by
ρb V − ρh V
s = vt ,
M
where M is the total mass of the system, V the volume of the ball, and
ρh and ρb are the respective densities of the honey and the ball. This is so
because when the ball has moved through a distance vt, it can be considered
Solutions 83
to have changed places with a ‘honey ball’ of identical volume. Thus the
total momentum of the system is
Ms
ptotal = = vρb V − vρh V .
t
The first term on the right-hand side is the momentum of the steel ball, and
therefore the second is that of the honey:
phoney = −vρh V = −2 g cm s−1 .
The negative sign shows that the direction of the honey’s momentum is
upwards. Its magnitude is the same as that of a honey ball moving upwards
with a speed equal but opposite to that of the steel ball.
S20 The (average) kinetic energy of the gas molecules is proportional to
the square of their velocity. The internal energy of the gas is proportional
to the temperature. Therefore v 2 ∼ T . If the wall is warmer than the gas
(T1 > T ) then the average speed of the rebounding gas molecules will be
increased by the collision (the wall warms the gas). If the wall is colder than
the gas (T1 < T ) then the situation is reversed; the molecules rebound with
a lower speed (the gas cools down).
From a molecular point of view, gases exert a pressure on the walls of
their container because of the changing momentum of molecules that hit the
wall and rebound from it. For a given initial momentum and collision rate,
the rate of change in the momentum of molecules rebounding from a warm
wall is greater than that of molecules rebounding from a cold one. Thus, the
gas exerts a higher pressure on a warm wall than on a cold one.
Note. This phenomenon explains the unexpected rotation of a radiometer
(‘light wheel’). If one side of each blade of a wheel mounted on a delicate
bearing is black and the other one is shiny, then the wheel starts to turn
when it is illuminated. At first sight, one might be tempted to think that
the pressure associated with the reflection of the light turns the wheel. This,
however, is not true, since experience shows it is, in fact, the shiny side
(the one reflecting rather than absorbing the light, and hence causing the
greater change in photon momentum) of the blades that moves forwards!
The correct explanation is that the black side of the blades warms up more
and therefore the pressure of the air molecules rebounding from that side
is greater than on the colder shiny side.
S21 As a result of thermal expansion, the size of both spheres increases.

The centre of mass of the sphere lying on the plate rises, whilst that of the
sphere hanging on the thread sinks. Thus, the potential energy of the first
sphere increases, whilst that of the second one decreases as shown in the
figure.
According to the first law of thermodynamics, the heat transferred to

the spheres produces not only an increase in internal energy and the small
amount of work done in expanding against the atmospheric pressure (this
is the same for both spheres), but also a change in gravitational potential
energy. The potential energy of the sphere lying on the insulating plate in-
creases a little, therefore its internal energy increases by less than the residual
heat transferred. Conversely, the decrease in potential energy of the hanging
sphere contributes positively to the increase in its internal energy. In sum-
mary, the temperature of the sphere suspended from the thread will be higher.
It is worth giving a numerical estimate. If the temperature of the two iron
balls, each with a radius of 10 cm, is increased by 100 ◦ C, a temperature
difference of ∆T ≈ 5×10−6 ◦ C will result from this effect. This is undetectable
in practice.
S22 For quantitative purposes we assume that the resistances of the bulbs
do not depend upon the voltages across them. This is far from accurate,
but will give the correct qualitative conclusion. If the (r.m.s.) supply voltage
is V , the resistance ri of a bulb is V 2 /wi , where wi is the nominal rating
of the bulb. When the two bulbs are connected in series across the supply,
the (r.m.s.) current drawn is V /(rA + rB ) and the power dissipated in bulb
i (i = A or B) is
V2 V 2
Pi = .
wi (V 2 /wA ) + (V 2 /wB )
According to the original agreement (wA = wB = 100 W), both PA and PB
should be 25 W. Actually, PA = 8 W and PB = 32 W, and so A clearly failed
his examinations. By comparison, student B might be considered a double
winner: he gets 32 W, but pays for only (8 + 32)/2 = 20 W. On the other
hand, 32 W is still a very poor light to study by and B also could well have
failed his examinations.
S23 A simple circuit consisting of two identical resistors connected as
shown in the figure would behave as described.
I II
Solutions 85
S24 When the bucket is full of water, the centre of gravity of the water
is above that of the bucket and therefore the common centre of gravity is
at its highest; correspondingly, the period is at its shortest. As the water
starts leaking out, the common centre of gravity moves downwards and the
period becomes longer. When the bucket is half full, the centre of gravity
of the water is below that of the bucket and the common centre of gravity
has moved even lower, significantly lower than the centre of gravity of the
bucket. Thus the period has increased further. When there is no water left
in the bucket, the centre of gravity coincides with that of the empty bucket,
which is higher than in the previous cases, i.e. the common centre of gravity
stopped moving downwards at some point and started to move upwards
again. In summary: the longest period occurs when the common centre of
gravity is at its lowest position. As is shown in the next problem, this occurs
when the common centre of gravity lies in the water surface.
S25 Clearly, the first water to be added is placed below the centre of
gravity of the empty beaker and therefore lowers the overall centre of gravity.
If at some stage water is added above the current overall centre of gravity,
the latter will be raised. Therefore for maximum stability the overall centre
of gravity must lie in the water surface.
−1
10 m
x
x /2
Thus
2 x 2
10−1 × 10−1 + π 3 × 10−2 103 x = x 10−1 + π 3 × 10−2 103 x ,
2
giving x = 55.9 mm.
S26 Assume first that the centre of mass of the bowl remains at the same
height as originally. Then the bowl only turns about its centre (and perhaps
moves sideways) but does not sink any deeper into the water. Under these
circumstances, the rim of the bowl is lowered to the water surface on the
side opposite the chain, and water flows into the bowl. We will now prove
that this cannot occur.
The upthrust acting on the body in the assumed situation remains the
same as in the initial one, i.e. it equals the total weight of the bowl and
soup. Thus, the chain cannot be exerting any force on the bowl. On the
other hand, the centre of mass of the bowl is not on the line of action of
the upthrust when the edge of the bowl is being raised, and the torques can
only be balanced if the chain does exert a force and pulls the bowl upwards.
The two contradictory conditions show that our initial assumption was
wrong. The geometrical centre of the bowl cannot stay in the same place
but has to rise (since a smaller upthrust is sufficient when the chain exerts
an upward force). This implies that even the lowest point of the rim of the
bowl has to remain above the surface of the water.
The possibility of the soup flowing out into the water has also to be
considered. This could occur if the level of the soup in the bowl were higher
than the water level in the lake, i.e. if the density of the soup were lower
than that of water. Realistically this would not be the case.
S27 The forces acting on the ball are the gravitational force mg, the
buoyancy force of the water and the normal force exerted by the rim of the
hole. When the buoyancy force just equals the weight of the ball, the normal
force becomes zero and the ball leaves the hole.
We first calculate the buoyant force exerted on the ball when the water
depth is h. Denote the volume of the ball immersed in water by V , where
V = V (r, R). Now, imagine the ball to have that part of it which protrudes
through the hole removed and the space under the container filled with
water. The buoyancy force would then be
F1 = ρg V (r, R),
where ρ is the density of water. As in reality there is no water under the

hole, a contribution
F2 = ρgh × πr 2
is missing. Thus the actual buoyancy force exerted on the ball is
F = ρg V (r, R) − ρgh × πr2 .
It is clear that, for sufficiently large h, F can be negative and then the
‘buoyant’ force is directed downwards. Decreasing h causes F to increase,
and, provided the top of the ball is not uncovered, it will rise when F = mg
Solutions 87
and the corresponding water depth is h0 . Thus

V (r, R) m
h0 = − 2 .
πr 2 πr ρ
Straightforward integration, or geometry books, give the volume of a
spherical calotte (truncated sphere) as
π √
V (r, R) = 2R 3 + 2R 2 + r 2 R 2 − r2 ,
3
and using this formula we can express h0 as
2R 3 2R 2 + r 2 √ 2 m
h0 = + R − r2 − 2 .
3r 2 3r 2 r πρ
Don’t forget that the formula above √ is only valid if the top of the ball is
still covered by water, that is h > R + R 2 − r 2 . Otherwise we have to modify
V (r, R); instead of a spherical calotte we have to calculate the volume of a
sphere truncated at both ends of a diameter. This leads to a cubic equation
for the critical h0 ,
√ 3m
h20 3 R 2 − r 2 − h0 = .
πρ
Note. When calculating the buoyancy force F, instead of
√ using the radius
of the hole r we can work with another variable: ' = R 2 − r2 (see Fig.
S27.1).
R
F
Fig. S27.1
Either by applying the previous argument or by integrating the vertical

component of the upthrust over the submerged part of the sphere, we can
calculate the buoyancy force to be
3'2 (h − ') + 2'3 − (h − ')3 πρg, if h ≤ R + ',
3
F(h) =
2 R 3 + '3 − 3 (h − ') R 2 − '2
πρg, if h ≥ R + '.
3
The expression for the buoyancy force F as a function of h consists of
two parts; in the interval 0 ≤ h ≤ R + ' the force is a cubic function
of h, whilst in the range h ≥ R + ' it decreases linearly, as shown in
Fig. S27.2.
F(h)
3
4F pqg/3
mg
h
2F R + F
Fig. S27.2
We can find the critical water depth h0 from the intersection of the graph
of F(h) with the line F = mg. Differentiating F(h) shows that it has a
maximum value of (4'3 πρg/3) at h = 2'. If the weight of the ball is larger
than this value (which is just the weight of a water sphere of radius '), then
the ball will not float out of the hole, whatever depth of water we have.
S28 The soap bubble floats and therefore the combined mass of its wall
and the helium inside it is equal to that of the displaced air. Since the density
of helium is less than half the density of air, the mass of the helium is less
than half of the mass of the displaced air. Thus, the wall of the bubble has
to be heavier than the gas it encloses.
S29 In case (a), it is clear that the water cannot flow out of the tube. If it
could, a perpetuum mobile (perpetual motion machine) could be established
using a paddle rotated ad infinitum by the outflowing water.
Cases (b) and (c) are not so simple. The ends of both tubes are lower than
the water level and the water pressures there are consequently lower than
the atmospheric pressure. In each case, the water wells to such an extent
that the pressure corresponding to its radius of curvature equals the pressure
difference between it and the air. The water surfaces corresponding to cases
(a), (b) and (c) are shown in the figure.
(a) (b) (c)
The greatest curvature (smallest radius of curvature) occurs for the hemi-
spherical shape and corresponds to the pressure of a column of water of
height H, since this is the height reached by the water in a vertical tube. If
the air–water pressure difference is greater than ρgH, then surface tension
cannot hold the water in the tube and it flows out. This is what happens
Solutions 89
in tube (c) (assuming that the figure is to scale and H > H); on the other
hand, water does not flow out of tube (b).
S30 The system described in the problem is spherically symmetrical.
Therefore the magnetic field that is built up has to be spherically symmetrical
as well. A spherically symmetrical vector field has to be radial everywhere
and its magnitude can depend only on the distance from the origin: B(r) =
B(r) r/|r|.
On the other hand, a magnetic field contains no sources (magnetic mono-
poles) and the magnetic flux crossing any closed surface has to be zero at
any given moment. In particular, consider a spherical surface of radius r
around the capacitor. The consequences of sourcelessness can only be met
if B(r) = 0 for any r. This means that the current described in the problem
builds up no magnetic field, either inside or outside the spherical capacitor.
Note. It is worth examining how the basic laws of electrodynamics are

satisfied between the plates of the spherical capacitor. Is it true that a
magnetic field builds up around a current flowing in a conductor, and that
the rotation (curl) of this field is proportional to the current?
S31 The radiation has to be spherically symmetrical, since both the

distribution and the motion of the charges are spherically symmetrical. The
magnetic field is always radial and should have the same magnitude at
any given distance from the sphere, irrespective of direction. This, however,
is impossible, since such a magnetic field (unless of zero magnitude) would
imply the presence of a magnetic charge (magnetic monopole), something
that experimentally is found not to exist in nature. Similar reasoning shows
that the electric field is also spherically symmetrical and that its magnitude
depends only on the total surface charge of the sphere and not on the pul-
sation parameters. Therefore, only the static Coulomb field can be observed
outside the sphere, and the sphere emits no radiation at all !
If individual parts of the sphere are examined, they are found to behave
like dipole antennae and emit radiation. But the radiation from different
parts has to be summed, taking into account phases as well as magnitudes.
The individual radiations from the many dipole antennae cancel each other
out, a result which can be proved by direct (but lengthy) calculation without
referring to spherical symmetry.
S32 No high-jump competition has yet been held on the Moon. However,
here we try to estimate the expected result. The men’s high-jump record on
Earth is over 240 cm. A good male high-jumper is more than 190 cm tall
and has a mass of around 80 kg, with his centre of mass approximately
110 cm above the ground. For a successful jump, all of his body must
rise to the height of the cross-bar, but his centre of mass need not. This
achievement requires a special jumping technique (the Fosbury flop), which
can be studied in slow-motion video recordings. The centre of mass of the
high-jumper remains approximately 20 cm below the cross-bar, even when
he is at the apex of his jump. For the western roll and straddle jumping
techniques, the jumper’s centre of mass has to rise above the bar.
The most difficult part of our estimation is comparing the movement of
a jumper at take-off on the Earth to that of a jumper leaving the ground
on the Moon. Assume that the centre of mass of the high-jumper rises by
s = 40 cm from its lowest point (in the crouch just before the jump) to
the highest point (when he has just left the ground) both on Earth and on
the Moon. Then his muscles must have done enough work to subsequently
carry his centre of mass from 110 to 220 cm, i.e. raise it by h = 110 cm on
Earth. Any effect of the run-up has been ignored, or has been assumed to
be identical in both places.
The basis of our estimate is the assumption that the same work is done
and the same jumping technique is employed in both cases. The work done
is the sum of the kinetic energy of his body and the potential energy of his
centre of mass, W = 12 mv 2 + mgs. His speed when leaving the ground can be
calculated using the relation v 2 = 2gh. Thus, the total work done is
W = mg(h + s) = 80 kg × 10 m s−2 × 1.5 m = 1200 J.
We assume that the work done by the high-jumper on the Moon is the
same, and that the rise s of his centre of mass before leaving the ground
is also unaltered. The gravitational acceleration on the Moon is only one
sixth of that on Earth, i.e. the energy equation of the jump on the Moon is
1200 J = 16 mg (s + h ), which yields h = 8.6 m. This is the vertical height by
which the jumper’s centre of mass rises on the Moon. To this must be added
the initial height of his centre of mass, 110 cm, and the extra 0.2 m resulting
from his special technique, to give an estimated record of 9.9 m ≈ 10 m.
Note. This question is usually answered – recognising only the difference

in the gravitational accelerations – by saying that the world record on the
Moon would be six times that on the Earth, i.e. approximately 14–15 metres.
According to the above analysis this is an optimistic expectation. Even our
estimate is probably over-optimistic, since in our model the high-jumper
has to do the same amount of work in a shorter time, i.e. a greater power
output is needed. It can be shown that the jumper has to increase his power
output by nearly 15 per cent when on the Moon. If, instead, the speed at
which he leaves the ground is taken to be the same as on Earth, a result
Solutions 91
of 8 m is obtained; this will be too low a value. All things considered, the
most probable value seems to be a height of about 9 m.
S33 (i) The vertical acceleration of ball B falling from the table is always
g, which makes it possible to determine the time (approximately half a
second) it takes to fall 1 m. The motion of the bob of the simple pendulum
is rather complicated, as no small amplitude approximation is possible, and
therefore the time during which it is in motion is not easy to determine. What
can be stated with certainty is that, since the thread exerts an upward force
on it, its vertical acceleration is always less than g. Therefore the vertical
motion of ball A takes a longer time than the vertical free fall of ball B. Ball
A stays in motion for longer.
(ii) The bob of the pendulum describes one-quarter of a circle (a path of
approximately 1.5 m). The other ball, B, follows a parabolic path, the length
of which cannot be determined by elementary methods. However, it is easy to
√
see that it hits the ground at a distance of vt = 2 × g × 1 2 × 1/g = 2 m
from the edge of the table. The length of its path is therefore not less than
the shortest
√ distance between the beginning and end points of its motion,
namely 5 m ≈ 2.2 m.
In summary: ball B moves on a longer path, but in a shorter time, than
the bob of the pendulum.
S34 For uniform circular motion the tangential acceleration of the body
is zero, but the radial acceleration is v 2 /R, where v is the speed of the
body and R is the radius of the circle.
Ft
Q F P
Fr
mg
Fig. S34.1
As the radial and tangential components of the gravitational force are

mg sin θ and mg cos θ, respectively, as shown in Fig. S34.1, the force F
exerted by the string P Q on the bob must have components:

mv 2
Ft = mg cos θ and − mg sin θ.
Fr =
R
From these expressions we can deduce the direction of the force F, and
hence the direction of the string, characterised by φ:
Fr v2 1.83
cot φ = = − tan θ = − tan θ.
Ft gR cos θ cos θ
8 7
6
9 5
10 4
3 12 11
11 4 13 2 1 10 3
14 24
12 5 15 23 9 2
16 22
18 20
13 6 8 1
7
14 24
15 23
16 22
17 21
18 19 20
Fig. S34.2
In Fig. S34.2 corresponding positions of the two ends of the string are
shown. The angle θ increases in 15◦ steps and the position of the other end
(Q) of the string has been plotted from point P by marking off the length
of the string in the direction of the calculated angle φ.
S35 Look at the two lines drawn in Fig. S35.1. The acceleration (g) is
greater for the one in the vertical direction, but the path length involved
is longer. The path perpendicular to the inclined plane is the shorter one,
but the corresponding acceleration is less. We can presume that the path of
shortest time lies somewhere between these two lines.
P
Fig. S35.1
Solutions 93
We next prove the following auxiliary theorem: bodies starting at the same
time t = 0, from the same point, and following frictionless slopes in different
directions, all lie on a circle at any subsequent time.
P
a C
Fig. S35.2
As shown in Fig. S35.2, the topmost point of any such circle C is the
starting point P . After time t, a body following a vertical wire and in free
fall will have fallen through d = 12 gt2 , and this must be the diameter of C. A
body moving along a wire at an angle α to the vertical has an acceleration
of g cos α. In the same time t it will have covered a distance, measured from
P , of 12 g cos α t2 = d cos α. But this is precisely the length of the chord of C
cut off by the wire. Thus, independent of α, the second body also lies on
C – and the auxiliary theorem is proved.
The original problem is easily solved using the auxiliary theorem. Bodies
starting at the same time from point P and travelling in different directions,
always form a circle that grows with time and has P as its topmost point.
After some time, the circle will touch the inclined plane, with the plane
tangential to the circle at the contact point P . A body starting from point
P reaches the plane in the shortest time by travelling along the line P P . In
fact, the problem is three-dimensional, and bodies starting from point P lie
on a sphere at any one time. The shortest time direction is found by joining
P to the point of the sphere that first touches the inclined plane. However,
it is sufficient for the question in hand to examine the vertical cross-section
through P parallel to the plane’s line of greatest slope, as we have done so far.
P
a/ 2
a a/ 2
P´
a
Fig. S35.3
It is clear from Fig. S35.3 that in the case of a plane inclined at angle α
to the horizontal, the line P P corresponding to the shortest time makes an
angle α/2 with the vertical, i.e. the optimium direction bisects those of the
two lines mentioned in the first paragraph of the solution.
S36 The solution to the problem is surprisingly easy using a rotating
frame of reference fixed to the minute hand. In this reference frame the
minute hand is at rest, whereas the hour hand is moving ‘anti-clockwise’. The
separation between (the ends of) the minute and the hour hands increases at
the highest rate when the line of action of the velocity vector of the end of
the hour hand passes through the end of the minute hand. In this situation,
the two hands and the line joining their ends form a right-angled triangle,
as shown in the figure.
Since the minute hand is twice as long as the hour hand, the angle between
the hands will be θ = cos−1 (1/2) = π/3. We can now calculate the exact time
after midnight when the angle between the hands is θ. As the minute hand
moves 12 times as fast as the hour hand, the angle φ between the hour hand
and the 12 o’clock position is given by 12φ − φ = θ, i.e. φ = 11
1
θ. Thus, since
midnight, the minute hand has moved through an angle of 11 θ = 11
12 4
π and the
time is just before 11 minutes past midnight. There are several subsequent
times (twice in each hour) when the angle between the hands is the same.
The second occurs when the ends of the hands approach each other at the
fastest rate.
Note. Using calculus this problem can also be solved by brute force. From
an expression for the distance between the ends of the hands of the clock,
their relative velocity can be found, and hence the angular positions at
which its stationary values occur determined.
S37 Using the coordinate system shown in the figure, the motion of the
stone can be described by the following relations:
g
x = v0 t cos α, y = v0 t sin α − t2 ,
2
vx = v0 cos α, vy = v0 sin α − gt.
Solutions 95
v0
a
x
x −v v
vx
The stone is at the greatest distance from the origin when its velocity is
perpendicular to its position vector. The condition for this is
y vx
=− ,
x vy
which yields a quadratic equation for the time t at which this happens:
3v0 sin α 2v 2
t2 − t + 20 = 0.
g g
If this is not to happen, the discriminant of this equation must be negative, i.e.
3v0 sin α 2
2v02
<4 .
g g2
Thus, for the stone to be permanently moving away from the thrower, we
must have sin α < 8/9 = 0.94, i.e. α < 70.5◦ .
S38 The trajectory of the grasshopper is a parabola, which touches the
trunk at two symmetrically placed points, B and B ∗ , on the two sides of the
trunk (at the moment we don’t know anything about these points – they may
or may not coincide at the topmost point E of the trunk). The grasshopper
takes off from point A with an initial speed v1 and at an angle θ with the
horizontal, as shown in the figure. At the tangential points B and B ∗ the
grasshopper’s velocity is v2 , making an angle β with the horizontal.
C
v2
E
B B*
F
R
v1
A D A*
For the sake of simplicity we choose β as the independent variable of the

problem. At point B the vertical component of velocity is
v2 sin β = gt2 ,
where t2 is the time of flight for the BC section of the trajectory (C is the
peak of the parabola). The corresponding horizontal displacement BF is
v2 t2 cos β = R sin β.
Multiplying these equations together we obtain

gR
v22 = .
cos β
Conservation of energy between points A and B of the trajectory gives
1 2 1 2
mv = mv2 + mg(R + R cos β),
2 1 2
and so
v12 = v22 + 2gR(1 + cos β)

gR
= + 2gR(1 + cos β)
cos β
1
= 2gR 1 + cos β + .
2 cos β
We can calculate the minimum value of v1 using differential calculus. How-
ever, there is a less complicated method available which uses the inequality
between arithmetic and geometric means:
√
1 1 1 2
cos β + ≥ cos β = .
2 2 cos β 2 cos β 2
√
So the minimum value of cos β + 1/(2 cos β) is equal to 2 and, therefore,
√
β = 45◦ . The case β = 0 requires a larger initial velocity, since 1.5 > 2; it
follows that the trajectory with the minimum initial speed does not in fact
touch the trunk at its topmost point. The gravitational potential energy of
the grasshopper is greater at the peak of the parabola than at the uppermost
point of the trunk, but its kinetic energy and total energy are smaller than
they would be for a top-touching trajectory.
The numerical value of the minimal initial speed is
√
v1min = 2gR 1 + 2 ≈ 2.2 m s−1 .
Solutions 97
Note. (i) It is not very difficult to show that the part of the parabolic
trajectory above B does not intersect the trunk.
(ii) We can also determine the take-off angle as θ = 3π/8 = 67.5◦ , and
the take-off distance as
√
AD = R(1 + 2/2) ≈ 17 cm.
(iii) It is interesting to note that point F is the focus of the parabola.
S39 The fleas jump in directions making angles 12 (π − θ) with the initial
direction of the hair. During the period in which they are in the air, the hair,
reacting to the impulsive couple it receives, rotates in the opposite direction
through an angle of π − θ, so that both fleas land on the hair but with each
at the opposite end from that at which it started.
2L h
m m
p−h
Fig. S39.1
Let v and α be the take-off speed and angle, respectively, and 2L be

the length of the hair. The time of flight t is, as usual, 2v sin α/g and the
range vt cos α. From geometry (see Fig. S39.1) the range must also equal
2L sin(θ/2). Now, each end of the hair receives an impulse, but only the
horizontal part of the tangential component contributes to the impulsive
couple acting on the hair. Thus
θ 1
2mvL cos α cos = Iω, with I = ML2 .
2 3
It is also necessary that ωt = π − θ.
Eliminating α, t and ω from the equations obtained so far, shows that θ
must satisfy the equation
6m
sin θ + θ = π.
M
p −h
6 m sin h
M
6m > M
6m < M
p h
Fig. S39.2
The function f(θ) = n sin θ + θ has the property f(π) = π, whatever the
value of n. In addition, f (θ) = n cos θ + 1 = 0 has a solution in 0 < θ < π
provided n > 1. Thus, if n is strictly greater than 1, f(θ) has a maximum for
some value of θ strictly less than π. This, combined with our observation
about f(π), shows that f(θ) = π has a solution for some value of θ strictly
less than π provided n > 1. In the context of the question, this condition
becomes m > M/6. This is illustrated in Fig. S39.2.
S40 The water ‘bell’ is cylindrically symmetrical about the vertical and
so it is sufficient to solve the problem by considering a cross-section. Let the
point-like rose be at the origin of an x–y coordinate system. The jets of water
then follow parabolic paths starting from the origin, and our mathematical
task is to find the ‘envelope’ (shown dotted in the figure) to this set of
parabolas.
It is well known that the equation of the path of a body projected with
initial speed v at an angle α to the horizontal is
g
y = x tan α − 2 x2 ,
2v cos2 α
which can also be written as
gx2 2 gx2
u − xu + y + = 0,
2v 2 2v 2
where u = tan α.
Solutions 99
If a point (x, y) is fixed, then the above relation is a quadratic equation

for u, which has a real solution if its discriminant is non-negative, i.e.
gx2 gx2
x2 − 4 y+ ≥ 0,
2v 2 2v 2
i.e.
v2 g
y≤ − 2 x2 .
2g 2v
This inequality divides the x–y plane into two regions separated by a
parabola. Water can reach the points under the parabola (a paraboloid
of revolution in three dimensions), but not those above it. The limiting
parabola is the sought-for envelope.
The water ‘bell’ is therefore a paraboloid of revolution. It is clear from
the equation for the limiting curve that the height of the ‘bell’ is v 2 /(2g),
as one would expect from considering an object thrown vertically upwards.
The water ‘bell’ defines a circle on the surface of the basin water, the radius
of which can be found using the condition y = 0; it is r = v 2 /g. This means
that the diameter of the basin should be at least four times the height of the
water ‘bell’ if no water is to be lost.
S41 There is uniform acceleration in both horizontal and vertical direc-
tions giving, in an obvious notation,
1 EQ 2 1 2
x = vt cos θ + t and y = vt sin θ − gt .
2 m 2
When y = 0, elimination of t results in the equation for the range given in
the hint. It has a maximum value of
v2 mg
EQ + m2 g 2 + E 2 Q2 , when tan 2θ = − .
mg 2 EQ
The negative sign of tan 2θ for positive Q indicates that θ needs to be more
than π/4 to take advantage of the ‘following wind’ provided by the electric
field.
Note. It will be clear that this problem is essentially equivalent to that of
finding the maximum range on an inclined plane of a mass projected with
a given speed.
S42 To satisfy the static equilibrium conditions for the rod (net vertical
force and net torque each equal to zero) the reactions of my fingers at
distances x and y from the centre of mass of the rod are (see figure):
y x
Fx = mg , and Fy = mg .
x+y x+y
Fx
F
CM
x
Assume that the rod first slips on my left finger. At any moment the frictional
force exerted on the finger is
y
Ffr = µkin Fx = µkin mg .
x+y
For slow movement (the horizontal acceleration is negligible) this force is
equal to the static frictional force acting on the right finger, which has a
maximum value of
x
µstat Fy = µstat mg .
x+y
Thus the left finger can slide so long as
µkin y ≤ µstat x, i.e. x ≥ ky,
where k = µkin /µstat ≤ 1.

Initially x0 = y0 = 12 ', so my left finger slides to the position x = x1 = k'/2
whilst working against a continuously changing frictional force. The work
done during this sliding can be found using integral calculus:
x1
W (x0 → x1 ) = − µkin Fx dx
x0
k'/2 '/2 ' 2
= −µkin mg dx = mg µkin ln .
'/2 x + ('/2) 2 k+1
At the second stage my right finger is sliding; while x = x1 and is constant,
y is changing from '/2 to y1 = kx1 = k 2 '/2.
The work done is
k2 y0 x1 ' 1
W (y0 → y1 ) = − µkin mg dy = mg µkin k ln .
y0 x1 + y 2 k
In the same way we can calculate the work during all of the following stages
during which the rod slips alternately on my left and right finger. The total
work done is
' 2 1
W = mg µkin ln + (k + k 2 + k 3 + · · ·) ln
2 1+k k
' 2 k 1
= mg µkin ln + ln .
2 1+k 1−k k
Solutions 101
If µkin µstat (i.e. k 1), the work is done in just one step and its value is
µkin mg
W ≈ ' × ln 2.
2
On the other hand, if µstat ≈ µkin (i.e. k ≈ 1), then
k 1
ln ≈ 1,
1−k k
which can be confirmed either with a calculator, or by writing k = 1 − δ,
using ln(1 − δ) ≈ −δ, and then letting δ → 0. Thus
µkin mg
W ≈ '.
2
S43 Take the length of the bricks to be unity and start the process from
the top. The topmost brick can be displaced until half of it protrudes beyond
the table, then the upper two have to be moved relative to the third one
as shown in the figure. The combined centre of gravity of the upper two
bricks must not be beyond the edge of the third one. Thus the second brick
can only be displaced by 14 . The general strategy is to move each subpile of
bricks until its combined centre of gravity is just above the edge of the brick
below it.
Before the third displacement, the combined centre of gravity of the top
three bricks has to be found. That of the two uppermost bricks is over the
edge of the third, and has to be given a double weighting, i.e. the distance
of 12 has to be divided in the ratio 2 : 1; with the third brick being displaced
by only 16 .
For the following (fourth) brick, the three placed above its edge carry
triple weighting, and the distance of 12 between the centre of gravity of the
fourth brick and its edge has to be divided in the ratio 3 : 1, i.e. the fourth
brick can be displaced by only 18 . Adding up the displacements, the result is
2 + 4 + 6 + 8 = 24 > 1, in other words, the topmost brick can hang beyond
1 1 1 1 25
the table.
In the above solution, critically unstable equilibrium positions have been

considered at each step. In practice, the displacements should be chosen to
be a little smaller than those calculated above, but it is still possible for the
topmost brick to hang outside the table.
The solution can be extended to an unlimited number of bricks. The
distance of 12 between the centre of gravity and the edge of the kth brick
has to be divided in the proportion (k − 1) : 1, since the common centre
of gravity of (k − 1) bricks is situated above its edge. The kth brick can
therefore be displaced by a maximum of 2k 1
units. If a total of n bricks is
available then the displacement of the topmost one can be calculated as
1 1 1 1 1 1
1 + + + + ··· + + ··· + .
2 2 3 4 k n
Since the sum of the reciprocals of natural numbers tends to infinity, as is
shown by considering
1 1 1 1 1 1 1 1
1+ + + + + + + + +···
2 3 4 5 6 7 8 9
1 1 1 1 1 1 1 1 1 1 1
> + + + + + + + + ··· = + + +···,
2 2 4 4 8 8 8 8 2 2 2
an arbitrary displacement can be realised by using a suitable number of
bricks, i.e. there is no limit.
S44 Let the mass of the plate be 2m, and denote the normal reactions
and frictional forces by N1 , N2 and F1 , F2 , respectively, as shown in Fig.
S44.1. The equilibrium equations for the horizontal and vertical forces are
N1 + F2 = 2mg, F1 = N2 .
N1
F1
F2
mg
R N2
mg
Fig. S44.1
Solutions 103
Balancing torques about the corner of the plate gives
(mg + N2 )R = N1 R.
The conditions controlling the frictional forces are
F1 ≤ µN1 and F2 ≤ µN2 .
Using the above equations, the following three relations can be derived.
F1 + F2 = mg,
F2 ≤ µF1 ,
µmg
F1 ≤ µN1 = µ(mg + N2 ) = µmg = µF1 , i.e. F1 ≤ .
1−µ
In deriving the last of these relationships we have assumed that µ < 1.
These three relations can be plotted in an F1 –F2 coordinate system. If
the coefficient of static friction is quite large, the situation is as shown in
Fig. S44.2.
F2
F1 < µmg µ F1
mg 1−µ
A F2 < µF1
B
F1
µ mg mg
1−µ
Fig. S44.2
In this case the problem does not have a unique solution; in the region
represented by the straight line segment between points A and B the static
equilibrium conditions can be satisfied by a range of frictional forces.
F2
F1 < µmg
mg 1−µ
F2 < µF1
mg F1
µmg
1−µ
Fig. S44.3
If the coefficient of static friction is too small, the situation is as shown

in Fig. S44.3 and the problem has no solution at all. The minimal possible
value of the coefficient of static friction can be found by making µ such that
the crossing point of the boundaries of the two inequalities occurs on the
F1 + F2 = mg straight line graph as shown in Fig. S44.4. In this case, instead
of inequalities we can use equalities, and after
√ some calculation we find the
minimal value of static friction to be µ = 2 − 1 ≈ 0.414.
F2
µmg
mg F1 < µF1
1−µ
F2 < µF1
mg F1
µmg
1−µ
Fig. S44.4
The possibility of F2 being negative should also be recognised. When µ is
greater than 0.5 the upper limit for F1 is larger than mg and this makes it
possible for F2 to have negative values, as shown in Fig. S44.5.
µmg
F2 F1 < = 1.5 mg
1−µ F2 < µF1 = 0.6 F1
mg
F1
mg
Fig. S44.5
S45 Air resistance is neglected and the balls are considered as perfectly
elastic. If the balls are dropped from height h, they reach the ground with
√
speed v = 2gh. The bottom ball first hits the ground, and then collides
with the top ball, which receives the largest possible energy if the lower ball
is at rest after the two collisions.
The bottom ball rebounds with speed v and collides with the top ball
moving downwards at speed −v. Since the speed of the ball of mass m2 is
Solutions 105
to be zero after the collision, the equations expressing the conservation of

momentum and energy are
v2 u2
(m2 − m1 )v = m1 u and (m1 + m2 ) = m1 .
2 2
The speed u of the top ball after the collision and the ratio of the masses
can be calculated from these equations, giving u = 2v and m1 /m2 = 13 .
With its speed doubled on rebounding, the upper ball rises to a height
of 4h.
Surprisingly, the top ball could bounce even higher than this. If m2 m1
then the top ball only takes a very small fraction of the total energy after
the collisions, but its speed is 3v and the height of the bounce is, in an ideal
case, 9h. This may sound rather incredible, but it is in agreement with the
principle of conservation of energy.
Readers interested in theoretical problems may generalise the problem to
n balls, while those interested in practical experimentation may try dropping
sets of non-identical balls – they bounce in very amusing ways!
S46 For the first collision, momentum and energy conservation give
M 2gh = MV + µv,
1 1
Mgh = MV 2 + µv 2 .
2 2
√
Eliminating V gives v as 2M 2gh/(M +µ) and the kinetic energy transferred
to the middle ball as 4µM 2 gh/(µ + M)2 . As a fraction of the initial energy of
the first ball, this is 4µM/(µ + M)2 . The fractional energy transfer to the final
ball, is the product of two such expressions using different pairs of masses.
Thus in order to maximise the energy of √ the final ball µ should be chosen to
maximise µ2 /(µ + M)2 (µ + m)2 , i.e. µ√= Mm. With this choice the overall
√
fractional energy transfer√is 16Mm/( M + m)4 and the height attained by
√
the final ball is 16M 2 h/( M + m)4 .
S47 Since the dumb-bells approach each other at identical speeds, the
sum of their momenta is zero in the reference frame of the air-cushioned
table (the same as that of their combined centre of mass). Thus, conservation
of momentum implies that the centres of mass of the two dumb-bells always
move at identical speeds and in opposite directions.
When the dumb-bells collide, both their energy and their angular momen-
tum are conserved, since the collision is perfectly elastic and no external
torque acts on them. The states before and after the collision are shown in
Fig. S47.1.
v
P
v
x
V
V
x
Fig. S47.1
Before the collision the dumb-bells have only translational kinetic energy,
while a rotational term appears after the collision. When writing down the
conservation of energy and angular momentum equations for the dumb-bells,
we calculate the latter with respect to their point of contact, P :
1 1 1
2 2mv 2 = 2 2mV 2 + 2m'2 ω 2 .
2 2 2
Before the collision the dumb-bells only have orbital angular momentum, but
a term describing their spin about their centres of mass has to be included
after the collision, i.e.
4'mv = 4'mV + 4m'2 ω.
The non-trivial solution (V = v, ω = 0) of the above set of equations

is found to be V = 0, and ω = v/'. That is that the centres of mass of
the dumb-bells stop moving after the collision, and that the colliding point
masses change velocities while the non-colliding ones keep their original
velocities. This can be interpreted in the following way: point masses joined
by a rigid but weightless rod are not aware of each other’s presence in the
course of a momentary collision. The rod only exerts a force directly after
the collision, when the dumb-bell rotates about its stationary centre of mass.
The hidden point of interest in the problem is that the dumb-bells collide
again after half a turn of each, i.e. after a time, t = π/ω. Using the previous
results, the resulting motion can be predicted without writing equations. The
rotation of the dumb-bells stops, and they move again with the same speeds
Solutions 107
as before the first collision. Their path is the same straight line but they are
now travelling ‘upside down’. In other words the dumb-bells spend the time
between the two collisions turning round. The speed of the dumb-bells as a
function of time is shown in Fig. S47.2.
v
t
p/x
Fig. S47.2
S48 (i) At this time the two masses are travelling parallel to the x-axis
in the centre of mass system and are both crossing the y-axis, one in each
direction. This situation must be superimposed on that due to the motion
of the centre of mass, which has moved along the line x = L/2 through a
distance L and has a speed of V /2 in the y-direction. Block A is at ( 12 L, 12 L)
with velocity ( 12 V , 12 V ); block B is at ( 12 L, 32 L) with velocity (− 12 V , 12 V ).
(ii) Establish that the centre of mass motion is cyclic with period 8L/V .
Block A is at (L, 50L) with velocity (0, V ); block B is at (0, 50L) at rest.
S49 Let x denote the ratio of the actual water level to the level at the
top of the basin; the same number shows the ratio of the current volume of
water to the maximum possible volume.
During filling, x increases uniformly with time and, since it reaches the
value x = 1 in time T1 ,
dx 1
= .
dt in T1
When water flows out, the speed of efflux – and therefore the rate of
decrease of x – is proportional to the square root of the height of the column
of water, i.e. to the square root of x,
dx √
= −K x. (1)
dt out
The coefficient of proportionality has to be chosen so that x just decreases
from 1 to 0 in time T2 .
Since the equation for the efflux is of the same form as the relation
√ between
the speed and the displacement for uniform acceleration, v = 2ax, it can
be concluded that the liquid level decreases to zero at a uniformly changing
speed. The initial rate of decrease is K, the final rate is zero; therefore the
average rate of decrease of x equals K/2. This deceleration can be expressed
in terms of the total time: K/2 = 1/T2 . The same conclusion can be reached
by integrating the differential equation (1) and applying the initial and final
conditions.
When both the tap and the plug-hole are open, the net rate of change
caused by the water flowing both in and out is
dx dx dx 1 2 √
= + = − x.
dt dt in dt out T1 T2
In a state of equilibrium, the water level does not change. The condition
for this is
T2 2
x = xe = .
2T1
For example, if the basin fills up in the same time as it empties, (T1 = T2 )
then the stationary state obtained by opening the tap and the plug-hole
together corresponds to x = 14 , regardless of the initial conditions. With the
data given in the problem, this ratio is 19 . It can also be seen that overflow
can only be a danger if it takes more than twice as long to empty the basin,
as to fill it (T2 > 2T1 ).
Note. One condition for the validity of Torricelli’s law of efflux is that the
size (diameter) of the orifice be much smaller than the depth of the water.
This condition is certainly not satisfied when the basin is nearly empty, and
therefore our results are only approximate. If the orifice is very small the
viscosity of the water (neglected so far) also plays an important role.
S50 When the vessel is rotating the free surface of the liquid must be
an equipotential surface for the system; for if it were not, the energy of the
system could be lowered by changing the surface profile. The total potential
energy per unit volume at any point (r, z) in cylindrical polar coordinates,
is made up of two parts, the gravitational potential energy ρgz and the
centrifugal potential energy. Since the centrifugal force is ρω 2 r directed
away from the axis, the potential energy at r is the integral of this with
respect to r, i.e. − 12 ω 2 r 2 . Both potentials are relative to arbitrary zeros. With
the origin of z chosen as in the Hint and r = 0 taken as the zero for the
centrifugal potential, the equation of the free surface is ρgz − 12 ρω 2 r 2 = 0.
If Z is the vertical distance of the lowest point of the surface below the
rim of the vessel when the liquid is on the point of overflowing, then both a
and Z lie on the free surface, and the volume of air in the paraboloid above
the liquid but within the vessel is still one-third of the volume of the vessel.
Solutions 109
Z
Integrating to find this volume, 0 πr 2 dz with r 2 = 2gz/ω 2 , gives πgZ 2 /ω 2 .
Thus we have
ω 2 a2 πgZ 2 πa2 h
Z= and = ,
2g ω2 3
leading to Ω = (4gh/3a2 )1/2 .
S51 Solution 1. Whilst accelerating, a car ‘pushes’ the Earth back a little
and changes its rotational angular velocity. This very small effect has to be
considered in order to resolve the paradox.
For the sake of simplicity, consider the car, of mass m, to be travelling on
a body of mass M (M m) that can move freely in the direction of the
motion of the car. In the actual situation, the Earth can rotate freely under
the car. A stationary observer would say that if the car accelerates to some
speed v0 and then subsequently to 2v0 , the body of mass M reaches a speed
u1 = −mv0 /M and then u2 = −2mv0 /M, whilst its kinetic energy increases
from an initial zero to Mu21 /2 and then to Mu22 /2. Since M m, the kinetic
energy of the body of larger mass and the change in its energy can be
neglected. Thus the ratio of the fuel consumption values has to be 1 : 3.
The situation is different for the observer moving with speed v0 . He can
see the speed of the car increasing from v0 to 2v0 and then to 3v0 , while,
in accordance with the law of conservation of momentum, the speed of
the other body changes from the initial −v0 to (1 − m/M)v0 and then to
(1 − 2m/M)v0 . The changes in the kinetic energy of the whole system (car
plus Earth) are, therefore, firstly
1 1 m 2
1 1
m (2v0 )2 − v02 + M 1 − v02 − Mv02 ≈ mv02 ,
2 2 M 2 2
and, then, secondly
1 1 2m 2
1 m 2
3
m (3v0 )2 − (2v0 )2 + M 1 − v02 − M 1 − v02 ≈ mv02 .
2 2 M 2 M 2
It can be seen that the energy (fuel consumption) ratio is 1 : 3 for both
observers.
Solution 2. Friction pushes the ground backwards, i.e. it accelerates the car
forward through its wheels. In the Earth’s frame of reference (Peter’s frame),
work is done only on the car and not on the ground. On the other hand,
in the train’s frame of reference (Paul’s frame), static friction acting on the
ground also does work. Viewed from this frame, the ratio of work done on
the car in the two stages is 3 : 5, since the displacements of the car in time
t are 3v0 t/2 and 5v0 t/2, respectively. The work done on the ground is −2
units, as the displacement of the ground is v0 t. The total work done is thus
3 − 2 = 1 unit in the first stage of acceleration and 5 − 2 = 3 units in the
second; their ratio, when viewed from the train, is therefore again 1 : 3.
S52 Since the object and image distances, respectively u and v, can be
exchanged in the lens formula, the square of their ratio gives the ratio of
the image sizes as (v/u)2 = 9 (or 19 ), implying that v/u = 3 (or 13 ). Thus,
the object distance is 30 cm (or 90 cm) and the image distance is 90 cm
(or 30 cm). The focal length can be calculated from the lens formula as
f = 22.5 cm.
If the same amount of light passed through the lens in both cases, the
nine-times smaller image would be 81-times brighter, as the smaller image
occupies a surface area 81-times smaller on the screen than the larger one.
However, when the lens is placed at the greater distance from the source it
receives only one-ninth of the light reaching it when it is close to the source.
As a result, the small image is only nine-times brighter than the large one.
It can be shown in general, for such pairs of images, that the small image
is always as many times brighter than the large one, as the large one is
bigger in linear dimension than the small one.
S53 The lenses of the glasses of short-sited people are divergent. Let −f
denote the (negative) focal length of a divergent lens, d the distance between
the object and the eye, and O the size of the object (see figure). According
to the lens formula, the distance between the (virtual) image and the lens is
given by
1 1 1 uf
+ = , i.e. v = ,
−v u −f u+f
whilst the size of the image is
v f
I= O= O.
u u+f
d–u+v
O
I
v
f
d–u u
Solutions 111
The apparent size of the image is determined by the angle φ it subtends

at the eye, which, assuming the object is small, is
I f
φ= = O.
d−u+v u(d − u) + fd
This angle is smallest as a function of u when the denominator on the right-
hand side of the above formula is a maximum. This condition is satisfied
when u = d/2. In other words, the apparent size of the object is smallest
when the lens is equidistant from the eye and the object. It is interesting that
this condition is independent of the focal length of the lens.
S54 With φ as defined in the Hint, Snell’s law applied to the initial
entry into the glass gives sin( π2 − θ) = ng sin φ. Straightforward geometry
then determines the angle between the incident ray and the normal to the
glass–water interface at the point where the ray meets the boundary as θ + φ.
p −h
2
u h +u
h h
Water
For total internal reflection to occur this must exceed sin−1 (nw /ng ). These
two conditions can be combined using the formula sin(θ + φ) = sin θ cos φ +
cos θ sin φ to eliminate φ and obtain
n2g − n2w ≥ cos2 θ(n2g + 1 − 2nw ).
Substitution of the given values for the refractive indices yields the stated
result.
S55 Consider the light beam as consisting of parallel light rays. They
cross the vertical plane face of the quarter-cylinder without changing their
direction, and strike the curved surface of the cylinder at various angles of
incidence. The normals at the points of incidence of the rays are radii of the
cylinder.
a R x
The higher the position of a light ray entering the quarter-cylinder, the
larger is its angle of incidence at the cylinder’s curved surface. The angle
of incidence for the ray shown in the figure is the critical angle for total
internal reflection. Therefore only light rays closer to the table than this one
can leave the quarter-cylinder (refracted to different extents). The limiting
case is determined using the figure:
1 2 R
sin α = = and = cos α,
n 3 R+x
which yield x = 1.71 cm. This is the closest to the quarter-cylinder that light
can reach the table.
As the angle of incidence of light rays close to the table top is smaller, they
are deviated less from their original direction by refraction, and therefore
might reach the surface of the table further away. One is inclined to think
that, in principle, the light patch could reach to any distance along the table,
since the direction of a light ray travelling adjacent to the surface of the
table is not altered. This, however, is false; the path of each light ray can be
parameterised (e.g. as a function of the angle of incidence), and it can then
be shown that each ray does not get very far up the table.
Instead of through tedious calculation, the furthest point of the light patch
can be found by means of a simple ‘trick’. Consider the part of the quarter-
cylinder close to the table as a plano-convex lens. The cylinder material
before the lens behaves like a plano-parallel plate and can be ignored. The
focal length of the plano-convex lens can be calculated using the thin lens
formula:
1 n−1
= .
f R
This yields f = 10 cm, and this is the distance from the quarter-cylinder of
the furthest point of the light patch.
S56 If RM is the Moon’s radius and R is the Moon–Earth distance, the
light power diffusely reflected into a solid angle of 2π is απRM 2 E, where E
is the intensity of direct sunlight (on either the Earth or the Moon). The
intensity received on Earth as moonlight is this divided by 2πR 2 . The Moon’s
◦
diameter subtends about 12 , or 9 × 10−3 rad, at the Earth’s surface, and so
the ratio of moonlight intensity to that of direct sunlight is
αRM
2
0.07 1 2
= × × 9 × 10−3 ≈ 10−6 .
2R 2 2 4
Thus sunlight is about one-million-times brighter than moonlight.
Solutions 113
Note. In fact, the reflectivity of the Moon was actually measured by com-
paring the brightnesses of sunlight and moonlight. Moreover, the albedo
of the Earth could be similarly determined by measuring the (very low)
brightness of the dark part of a new moon illuminated by reflected light
from the Earth.
S57 Let us approximate comfortable walking by a model in which the

human leg is a freely swinging pendulum. The period of a freely swinging
body supported at its upper end is
I
T = 2π ,
mgs
where I is the moment of inertia of the body, m is its mass, and s is the
distance between the pivot and the centre of mass of the body. Now introduce
the so-called effective length Leff = I/ms by expressing the period as
Leff
T = 2π .
g
We can assume that the effective length is directly proportional to the actual
length of the leg and, for different people with the same leg length, we will
find only very slight differences in the effective length.
Using this result for the natural period, we can now estimate a person’s
natural gait – the one involving the least muscular effort. To a first approx-
imation, we assume that the length of a stride is proportional to the length
of the leg. The time for a single stride is one-half the period given above,
and the walking speed v depends on the leg length:
L √
vwalk ∝ ∝ L.
T /2
This equation predicts that people with longer legs have a more rapid natural
walking gait. The prediction is made on the basis of an oversimplified
model that assumes minimum energy expenditure and ignores differences
in anything (e.g. shape, strength, etc.) other than leg length. However, the
prediction is borne out by common experience.
When analysing a person running, an important change must be made
to our model. During running, the leg does not swing freely but is subjected
to a torque acting about its pivot. The torque is produced by a force F
supplied by the muscles. This force is roughly proportional to the cross-
sectional area of the muscles involved, and if we assume that, for people
of different size, the relative proportions of the leg are the same, then the
cross-sectional area, and therefore the force F, depends on the square of the
length L. The torque is then proportional to the product of F and L:
τ ∝ FL ∝ L2 × L = L3 .
The moment of inertia I is proportional to the mass and to the square of the
length. Again, we assume that all legs have essentially the same proportions;
that is, width and thickness are proportional to length. Thus the mass varies
as the cube of the length and
I ∝ mL2 ∝ L5 .
It can generally be shown that for a body oscillating about a fixed point
and subject to a periodic torque, the period T depends on the maximum
torque τ and the moment of inertia I of the body about that point, and is
given by
I
T ∝ .
τ
Upon substituting for I and τ, we find
L5
T ∝ ∝ L.
L3
The speed of running is the product of the frequency of taking steps and the
length of a single step, and hence
L L
vrun ∝ f × L ∝ ∝ = 1.
T L
So the model predicts, in accord with Annie and Andy’s experience, that the
speed of running does not depend on leg length. Whilst its predictions are
not, of course, strictly accurate, the model does offer some explanation for
the observation that the ordinary walking rate of people with long legs is
usually greater than that of people with short ones, whereas the speed at
which they can run is often not significantly different.
S58 Consider a simple pendulum of length L and a pendulum consisting
of a uniform rod of length ' pivoted at one end. If both are released from
a horizontal position, what are their angular speeds after they have each
travelled through an angle α?
L m F m
a a
x
x
Solutions 115
The principle of conservation of energy yields
1 2g
mL2 ω 2 = mgL sin α, i.e. ω = sin α
2 L
for the simple pendulum, and
1 m'2 2 ' 3g
ω = mg sin α, i.e. ω = sin α
2 3 2 '
for the rod. If L = 23 ', then the angular velocities of the two motions are
equal for all values of α. It then follows that the two motions are identical
at all times and their periods are equal.
How can the period of this equivalent pendulum be calculated? The
formula T = 2π L/g, valid for small oscillations, cannot be applied as
the amplitude here is large. Exact calculations would require complicated
mathematical analysis, but this is not necessary if, instead of calculating the
period T , we only wish to determine its dependence on L.
The period of swing of the simple pendulum may depend on its length
L, the mass of its bob m, the gravitational acceleration g and the maximum
angle of deviation αmax . If the dimensions of the quantities involved are
taken into consideration, this functional dependence can only be of the form
L
T (L, m, g, αmax ) = f(αmax ) .
g
To justify this assertion, we note the following points. The dimension of mass
is the ‘kilogram’, and since the ‘kilogram’ does not occur in the dimensions
of any of the other quantities, the period (which has dimension ’seconds’)
cannot depend upon the mass of the bob. On the other hand, ‘seconds’ occur
only in g, and therefore the required dimension of ‘seconds’ in T can only
be obtained if T is inversely proportional to the square root of g. Finally,
in order to settle the ‘metre’ dimension, the period has to be proportional to
the square root of L. The form of the function f(αmax ) cannot be determined
via dimensional analysis, since the angle is dimensionless. The only available
information is that for small angles f(αmax ) ≈ 2π.
From the above reasoning, it can be concluded that (with the same initial
displacements) the period of a simple pendulum of length (2/3)' is 2/3
times that of a simple pendulum of length '. Thus, the period of a pivoted
rod of length ' is approximately 82 per cent of that of a simple pendulum
of the same length. This conclusion is valid not only for horizontal release,
but for any common initial starting position.
S59 The required power for the hovering helicopter depends on the
gravitational acceleration g, the linear size of the helicopter L, the average
density of the helicopter ρhel , and the density of air ρair .
It is reasonable to assume that the mechanical power needed depends only
on these quantities and that the dependence is a power relationship:
P ∝ g α × Lβ × ργhel × ρδair .
The dimensions of the left- and right-hand sides must be equal:
α γ δ
kg m2 m kg kg
= × mβ × × ,
s3 s2 m3 m3
which yields
γ + δ = 1,
α + β − 3(γ + δ) = 2,
−2α = −3.
The solution of this system of linear equations is β = 72 , α = 32 and γ = 1 − δ.
It can be seen that the mechanical power needed is proportional to the 72
power of the linear size. Consequently, the second helicopter should have an
engine producing power (1/2)7/2 P = 0.088P .
Note. (i) The efficiency of a mechanical engine can be characterised by the
ratio of the power produced to the mass of the engine. According to the
above result the ‘specific power’
P P √
∝ 3 ∝ L,
m L
i.e. the efficiency required increases as the linear size increases. This means
that the smaller a helicopter is, the more easily it can hover. There are many
small animals (bees, dragonflies, hummingbirds, etc.) that can hover like a
helicopter, but larger birds are unable to do so.
(ii) Using simple dimensional analysis we could find only the sum of the
exponents γ and δ. However, it is clear that P can depend only on the
product of the density of the helicopter and g, because, when the helicopter
is hovering, the relevant quantity is not its inertial mass, but its weight.
Thus γ must be equal to α, i.e. γ = 32 with δ = − 12 . Finally, we get
−1/2 ρhel
P ∝ (gρhel )3/2 × L7/2 × ρair = L3 ρhel g × Lg × .
ρair
Here, on the surface of the Earth, we can change only the size and density
of the helicopter. Nevertheless, for a space mission using robot helicopters,
it could be useful to know how P depends on the gravitational acceleration
and the atmospheric density of the target planet.
Solutions 117
S60 The gravitational potential energy lost as the rod falls through an
angle θ is Mg 2' (1 − cos θ), and this is converted into rotational kinetic energy
about the edge of the table. Either by direct calculation or by using the
parallel axes theorem, the relevant moment of inertia of the rod is found to
be 13 M'2 . Combining these two results gives
3g
ω2 = (1 − cos θ).
'
The centripetal acceleration ac , is '/2 times ω 2 and therefore equal to
2 g(1−cos θ). Using the same moment of inertia and the instantaneous torque
3
of Mg 2' sin θ gives the tangential acceleration of the rod as at = 34 g sin θ.

(i) The smooth (frictionless) horizontal and vertical walls of the groove
can exert only positive vertical and horizontal forces, V and H respectively,
on the end of the rod, (see Fig. S60.1). The rod will lose contact with the
table as soon as one of these falls to zero and is required by the equations
of motion to become negative, i.e. to become a (physically impossible) force
of attraction.
at
ac
H g
Fig. S60.1
Resolving forces and accelerations horizontally and vertically (Fig. S60.1)

gives
H = M(at cos θ − ac sin θ),

Mg − V = M(ac cos θ + at sin θ).
Solving these two equations for H and V we obtain

3
H = Mg sin θ (3 cos θ − 2),
4
1
V = Mg (3 cos θ − 1)2 .
4
The first reaction component to vanish is H and this happens when θ =

cos−1 32 ≈ 48◦ . At larger angles, H would be negative, so the rod really loses
contact with the table because the smooth groove is not able to pull back
the end of the rod.
(ii) As the edge of the table is now a very small quarter-circle, the normal
force N is always directed along the rod’s axis. The static frictional force Ffr
is tangential to this quarter-circle and can have any arbitrary value because
of the rough edge (see Fig. S60.2).
at
Ffr
ac
g
Fig. S60.2
The sum of the component of the rod’s weight along the rod and the
normal force of the table gives the centripetal force:
' 3
Mg cos θ − N = Ma c = M ω 2 = Mg(1 − cos θ).
2 2
We can thus express the normal force as
Mg
N= (5 cos θ − 3).
2
The reaction of the table on the rod becomes zero when N = 0, i.e. when
θ = cos−1 35 ≈ 53◦ . At larger angles the normal force should be negative,
which is impossible, and thus the rod loses contact with the table. Because of
the rough edge, the static frictional force is always large enough to prevent
slipping except when the normal force becomes zero; consequently, it has no
effect on the motion.
Note. In this problem we considered two extreme cases represented by (i)
and (ii). In general, the direction of the normal force is perpendicular to the
common tangent to the table’s edge and the bottom of the rod; this means
that the normal force can act in virtually any direction. This shows that the
motion of the falling rod strongly depends on the geometrical details of the
touching surfaces, as well as on the value of the coefficient of friction (see
also P61).
Solutions 119
S61 Suppose, first, that the surface of the table is very smooth (the
coefficient of friction is very small). Just after the pencil is released, its centre
of mass accelerates in the direction of the fall and acquires both horizontal
and vertical velocities. The horizontal component of the acceleration is
produced by the frictional force between the pencil point and the table, but,
since the surface of the table is smooth, the point soon slips, in the direction
opposite to that of the fall.
If the friction is very large, the pencil does not slip for a relatively
long time. Initially, the horizontal velocity of the centre of mass, which
is moving on a circular path, increases in the direction of the fall, but
later it starts to decrease and, if the pencil continued moving this way
until it was horizontal, it would tend to zero. The sign of the horizon-
tal acceleration of the centre of mass, and thus also that of the fric-
tional force, changes during the motion. If the pencil does not slip during
the first stage, then it can only slip ‘forward’, i.e. in the direction of the
fall.
In the following we are going to prove that the pencil will slip in some
way (‘backward’ or ‘forward’) but the point of the pencil never loses contact
with the table. For sake of mathematical simplicity let us use quantities with
a value of unity for the length and mass of the pencil, as well as for the
gravitational acceleration: ' = M = g = 1. Thus the weight of the pencil
is 1, its centre of mass (CM) is 12 measured from either end, its moment of
1
inertia about its CM is 12 , and its moment of inertia about one of its ends
1
is 3 .
=1
,
1
(1 − cos )
2 at
ac
g=1
F
N
Fig. S61.1
During the first stage of the motion the point of the pencil does not slip,
and so the pencil rotates about its point (Fig. S61.1). We can find the angular
velocity of the pencil with the help of the conservation of energy:

1 1 1
(1 − cos θ) = × ω 2 ,
2 2 3
√
which yields ω = 3(1 − cos θ). The instantaneous torque of 1
2 sin θ gives
the angular acceleration of the pencil:
1 1 3
sin θ = α, i.e. α = sin θ.
2 3 2
Vertically there are two forces acting on the pencil: its weight and the
normal force of the table, N. The vertical component of the centripetal
acceleration of CM is 12 ω 2 cos θ and that of the tangential acceleration is
2 α sin θ. Thus the vertical component of the equation of motion is
1
1 1
1−N = α sin θ + ω 2 cos θ,
2 2
which yields:
3 cos θ − 1 2
N= .
2
It seems that the normal force is never negative, and that the point of the
pencil cannot lose contact with the table during the rotational phase of the
motion. The normal force is zero when θ = cos−1 13 ≈ 70.5◦ , which means
that the frictional force is also zero at this angle, and the pencil will slip
there if it has not done so before.
The horizontal component of the equation of motion is
1 1
F= α cos θ − ω 2 sin θ,
2 2
where F is the frictional force. Substituting the angular acceleration and
velocity into this expression, we get
3
F= sin θ(3 cos θ − 2).
4
The condition for slipping is |F| > µN, where µ is the coefficient of (static)
friction. We can reformulate the condition for slipping with the help of a
function f(θ), defined as the absolute value of the ratio of the forces F/N,
F 3 sin θ(3 cos θ − 2)

f(θ) = = > µ.
N (3 cos θ − 1)2
The function f(θ) is plotted in Fig. S61.2.
Solutions 121
f (h)
0.37
h (deg)
35 48 51 70.5
Fig. S61.2
The frictional force changes sign at θ = cos−1 23 ≈ 48◦ , implying that

‘backward’ slipping can occur in the region 0 < θ < 48◦ . Using numerical
methods it can be shown that in this region f(θ) has a maximum value of
µcrit ≈ 0.37 at an angle of θ ≈ 35◦ . It means that the pencil slips ‘backward’,
if µ < µcrit .
If µ > µcrit then the pencil slips ‘forward’ before it reaches the angle
of θ ≈ 70.5◦ , where f(θ) approaches infinity. (Note that the pencil cannot
start slipping in the range of 35◦ < θ < 51◦ .) The ‘backward’ and ‘forward’
motions of the pencil are shown in Fig. S61.3. In both cases the slipping
point of the pencil can stop again.
l < l crit l > l crit
Fig. S61.3
Finally, it will be shown that the point of the pencil does not lose contact
with the table. Let us consider first the case of ‘forward’ slipping shown in
Fig. S61.4. According to the work–energy theorem and the cosine law for
(vector) triangles,
1 1 1 1 ω 2
1
(1 − cos θ) = Wfr + × ω 2 + + vpoint
2
+ 2vpoint × ω cos θ .
2 2 12 2 2 2
If we neglect the work of friction and the two terms containing the velocity,
vpoint of the point of the pencil, (all three terms are positive), then we obtain
x, a
vpoint
h 1
2
x
vpoint
lN
N
Fig. S61.4
an inequality for the angular velocity: ω 2 < 3(1 − cos θ). The instantaneous
torque about CM gives
1 1
N(sin θ + µ cos θ) = α,
2 12
whilst the vertical component of the equation of motion is
1 1
1−N = α sin θ + ω 2 cos θ.
2 2
(It will be recognised that this equation is the same as for the non-slipping
case. The reason for this is that the point of the pencil experiences only
horizontal acceleration, and so the vertical component of acceleration of
CM remains unaltered, see Fig. S61.5.)
x, a
1 2 1
2
x a
2
ax
Fig. S61.5
From the two equations above we can express the normal force as a
function of θ and ω 2 . However, we have an inequality for ω 2 which yields
1 − (1/2)ω 2 cos θ (3/2)[cos θ − (1/2)]2 + (5/8)
N= > > 0.
1 + 3 sin θ(sin θ + µ cos θ) 1 + 3 sin θ(sin θ + µ cos θ)
Solutions 123
Thus the normal force is always positive, and the point of the pencil does
not lose contact with the table.
x, a
vx
h
vpoint v
lN
N
Fig. S61.6
For the case of ‘backward’ slipping our method is very similar. Figure S61.6
shows the horizontal and vertical components of the velocity of CM, vx and
vy , respectively. As the point of the pencil has zero vertical velocity, there is
a connection between vy and ω, namely vy − 12 ω sin θ = 0. We can again use
the work–energy theorem to give
1 1 1 2
(1 − cos θ) = Wfr + ω + vx2 + vy2 ,
2 2 12
in which we (again) neglect work against friction and another positive term
containing vx to obtain
1 − cos θ
ω2 < .
(1/12) + (1/4) sin2 θ
Considering again the net torque about CM and the vertical component of
the equation of motion, we obtain a further inequality for the normal force,
namely:
1 − (1/2)ω 2 cos θ 1 + 3(cos θ − 1)2
N= > .
1 + 3 sin θ(sin θ − µ cos θ) [1 + 3 sin θ(sin θ − µ cos θ)](1 + 3 sin2 θ)
The numerator is always positive and the denominator is positive (for
0 ≤ θ ≤ 90◦ ), if µ < 43 . However, in the ‘backward’ slipping region µ < 0.37,
so, again, the point of the pencil does not lose contact with the table.
If the pencil point stops slipping at some stage, it cannot lose contact
again, because then ω 2 < 3(1 − cos θ), and thus N > (3 cos θ − 1)2 /4 ≥ 0 (see
the first part of this solution).
S62 The pressure in a soap bubble of radius R is greater than the

atmospheric pressure p0 by ∆p = 4γ/R. The factor of 4 arises because both
directions of curvature and the fact that the film is two-sided have to be
taken into account. Clearly, the pressure is higher in the bubble of smaller
radius and, therefore, the air flows into the larger bubble, as if inflating it to
finally form a single bubble of radius R3 .
The volume of air in the bubbles is proportional to R 3 , and therefore the
ideal gas equation and the conservation of mass require that
4γ 4γ 4γ
p0 + R13 + p0 + R23 = p0 + R33 .
R1 R2 R3
For bubbles of ordinary size, the pressure of curvature is many orders of
magnitude smaller than the external atmospheric pressure. If the pressure of
curvature is neglected the radius of the resulting bubble is
3
R3 ≈ R13 + R23 .
If the radii are measured ‘accurately’, in order to determine the surface
tension, the formula
p0 R33 − R13 − R23
γ=
4 R12 + R22 − R32
is appropriate. In practice, however, this method cannot be applied, as the
numerator is, as shown, almost equal to zero and thus would carry a large
fractional uncertainty as a result of measurement error. Any measured data
are likely to provide only a rough estimate of the surface tension.
S63 The cross-sectional edge of the disc of water is a semicircle of radius
r = 12 d (see figure). Thus, the curvature of the surface of the water is 2/d,
which corresponds to a pressure of curvature of ∆p = 2γ/d, where γ is the
surface tension. (The other component of the curvature is negligible because
D d.)
p
d p
0
The pressure inside the disc is therefore p0 − 2γ/d when the atmospheric
pressure is p0 . This pressure difference acts over a surface area between the
water and each of the glass surfaces of πD 2 /4. This implies that a force,
πD 2 2γ
F=
4 d
‘pulls’ the glass plates together.
Solutions 125
Note. If d is much smaller than D, this force can be quite considerable. It is

in fact very difficult (if not impossible) to separate two parallel glass plates
by pulling them in a direction perpendicular to their common plane when
there is water between them. In order to be separated, they have to be slid
in a direction parallel to that plane.
S64 The velocity of the thread at a distance of x metres from the wall is
obviously proportionately smaller than the velocity of the end of the thread,
i.e. it is xv0 .
v0 x v0
1m x m
If this value is greater than the speed of the caterpillar, then the latter will
move away from the wall. Its situation will become more and more hopeless,
and it will never reach the wall.
On the other hand, if vcaterpillar > xv0 , the net velocity of the caterpillar
is towards the wall and increases as time passes, with the consequence that
the caterpillar will certainly reach the wall. The limiting case corresponds
to x = vcaterpillar /v0 = 0.1 m. Starting at this point, the caterpillar does not
move in either direction.
S65 Imagine signs attached to points on the thread and labelled with
the ratio of the distance from the wall to the total length of the thread.
These figures are precisely the coordinates x mentioned in the solution to the
previous problem: x = 0 corresponding to the position of the wall and x = 1
to that of the spider. Now, however, these ‘stretch’ as the thread stretches.
We first calculate how long it takes the caterpillar to get from a point x to
a nearby point x − ∆x when the spider has been moving for a time t. Since
the distance between the points in question is (1 + v0 t)∆x and the caterpillar
moves at speed c, the relationship
c∆t
∆x =
1 + v0 t
holds. Summing (integrating) this relationship for the whole motion of the
caterpillar, which starts from x0 and reaches the wall in time T , gives
c∆t T c c
x0 = ≈ dt = ln(1 + v0 T ).
1 + v0 t 0 1 + v0 t v0
Since the above integral can be made arbitrarily large by a suitable choice of
T , the perhaps surprising result is that however quickly the spider pulls the
end of the thread (i.e. with an arbitrarily large v0 , an arbitrarily small c and
an arbitrary x0 ), the caterpillar will still reach the wall within a finite time.
S66 (i) Suppose that the ball falls freely for 1 m, then reaches point B
by bouncing practically horizontally along a row of closely spaced nails near
the bottom edge of the drawing-board. The duration of the vertical fall is
t1 = 0.45 s, at the end of which the ball has reached a speed of v1 = 4.4 m s−1 ,
and covers the remaining distance of 2 m in a time t2 = 0.45 s. As the ball
would have reached point B in√a time t3 = 1.01 s by sliding down the straight
line AB (with acceleration g/ 5), the answer to the first question is that the
quickest way for the ball to get from A to B is not by following the shortest
route.
Note. It can be proved by means of complicated mathematics (the calculus
of variations) that the curve along which the transit time is the shortest is
a cycloid.
(ii) A body dropped from rest at point A has maximum vertical velocity
if it is in free fall. Its maximum kinetic energy, and therefore its maximum
speed, is determined solely by the magnitude of its vertical displacement.
Thus, a bouncing ball cannot reach the bottom of the drawing-board faster
than a body in free fall, i.e. in less than t1 = 0.45 s. The answer is therefore
no!
S67 The puzzling aspect of the problem is that insufficient data have been
given in the text. However, the figure can be used as a source of information.
Using a protractor you can measure with sufficient accuracy that the tangent
to the fixed end of the rope makes an angle of 30◦ with the vertical, as
shown in the figure. This means that the tension at the fixed end of the rope
is T = 20 N/ sin 30◦ = 40 N. Similarly, the weight of the rope is equal to the
vertical component of the tension there; mg = T cos 30◦ = 34.6 N, giving
the mass of the rope as m = 3.5 kg.
T − g
20 N
o
30
20 N
P
Solutions 127
Note. Further information can be obtained from the figure. The centre of
gravity of the rope must be vertically above the point P , because the lines
of action of all three forces acting on the rope must meet at a single point.
S68 Using the argument described in the Hint it is possible to prove

that the compasses have to be opened to the extent necessary to make the
lower arm hang horizontally when the compasses are suspended, as shown
in Fig. S68.1.
Fig. S68.1
Starting from that situation, let us imagine for the moment that the upper
arm of the compasses is fixed. If the lower arm is then bent either upwards
or downwards, the horizontal position of the CM of the compasses moves
towards the pivot. After the release of the upper arm the pivot moves
downwards, because that is the only way that the CM of the compasses
can again position itself below the attachment point. So, in either case,
the vertical position of the pivot is lowered, and we can conclude that the
originally described situation is the one required.
Instead of a real pair of compasses let us consider a simplified model of
two identical thin rods joined by a pivot of negligible mass as shown in
Fig. S68.2.
CM 1
Fig. S68.2
Let the angle between the arms of the compasses be 2θ, and the length of
each of the arms be 2 units. It is easy to find congruent angles in Fig. S68.2
and to apply the sine rule to the shaded triangle. Figure S68.3 shows a
magnified version of this triangle.
sin
Fig. S68.3
According to the sine rule it follows that

sin θ sin(90◦ − 2θ)
= .
1 sin θ
After a simple calculation we obtain
1
sin θ = √ ,
3
so θ ≈ 35.3◦ , that is to say the compasses should be opened to 2θ ≈ 70.5◦ .
√
Note. The result 2θ = 2 sin−1 1/ 3 = cos−1 (1/3) is so simple, one might
suppose that a more elegant solution exists. Actually, it is possible to find
the angle 2θ with the help of the theorem of parallel transversals.
C
2h
u
A O
D F G
Fig. S68.4
In Fig. S68.4 two verticals are drawn, one through the overall centre
of mass C of the compasses, and the other through the centre of mass
E of the upper arm alone. By considering the lines forming the angle 2θ,
it can be seen that the equality of OE and EB implies that FG = GO.
Similar consideration of the angle φ, and the fact that DC = CE, shows
that FG = DF. Thus points F and G trisect the section OD, which in turn
is equal to OE, thus implying that OE = 3GO. As EGO is a right-angled
triangle, it follows that 2θ = cos−1 (1/3) ≈ 70.5◦.
S69 Clearly, the centre of mass S of the triangle has to be below the
point of suspension. Denote the vectors pointing from the centre of mass
Solutions 129
S to the vertices of the triangle by r1 , r2 and r3 , and that to the suspension

point by m (see figure). The forces F1 , F2 and F3 exerted on the plate by
the threads can now be expressed in terms of the vectors defining the thread
directions:
Fi = λi (m − ri ), i = 1, 2, 3.
m F
2
F3
r2
F1 S
r3
r1
Since the plate is in equilibrium, the vector sum of the forces acting on it
is zero, i.e.
F1 + F2 + F3 + W = 0.
Making use of the fact that the vector pointing to the centre of mass of the
triangle (the origin of the vector reference frame) is the arithmetic mean of
the vectors pointing to its vertices, we have
r1 + r2 + r3 = 0.
We note that W and m are parallel and, therefore, W = −km. Eliminating
r3 from the above equations gives
(λ3 − λ1 )r1 + (λ3 − λ2 )r2 + (λ1 + λ2 + λ3 − k)m = 0.
Since r1 , r2 and m are not in the same plane, a linear combination of them
can only be zero if the coefficient of each is zero.
Thus λ1 = λ2 = λ3 , which implies that the tensions in the threads are
proportional to their lengths. This deduction would become invalid if one
of the threads were slack, since the plane of the plate would then become
vertical and m would lie in it.
S70 Initially the tanker and the liquid in it are at rest. As the outlet pipe
is at the rear of the tanker, when the tap is opened the centre of mass of the
liquid will move backwards. As the centre of mass of the whole system is fixed,
the tanker itself must move forward. However, the outlet pipe is vertical and
so the emerging liquid will acquire a forward horizontal-velocity component.
This does not contradict the law of conservation of linear momentum,
because the liquid inside the tanker will be moving backwards (relative to
the ground). Nevertheless, the forward direction of motion of the tanker
must subsequently change to backwards, since if it did not, the position
of the centre of mass of the whole system would ultimately start to move
forwards. The dynamical reason for the change in the direction of motion
is the force exerted on the rear of the container by the backward-moving
liquid as it is brought to rest relative to the tanker just before discharge.
Note. (i) It could be that the direction of travel of the tanker changes several
times during the motion, but a detailed analysis is virtually impossible as it
depends on too many parameters.
(ii) Finally, we give the solution to the scenario proposed in the Hint
to this problem. Consider the situation when the student has reached the
end of the carriage and stopped, but the ticket collector is still moving
backwards with speed v relative to the carriage. In accord with linear
momentum conservation, the carriage must be moving forward with speed
u = mv/(M + 2m). So, when the student jumps out, he carries away a
forwards-directed momentum of mu = m2 v/(M + 2m). When the collector
stops, the carriage (with the collector aboard) changes its direction of motion
and moves backwards, having a total linear momentum of −mu. Thus the
final velocity V of the carriage is
m2
V =− v.
(M + 2m)(M + m)
S71 Taking motion to the right as positive, the initial velocities in the
centre of mass frame of the beads (see Fig. S71.1) are
M m
vm = v0 and vM = − v0 .
m+M m+M
Since their centre of mass is at rest in this frame, the ratio of the two
velocities remains constant (at M/m) throughout the motion.
vm
m, q CM M, Q d
− vM
Fig. S71.1
Solutions 131
As the beads approach each other, their speeds will decrease if q and Q
have the same sign, but increase if their charges are opposite. In the latter
case, when they are finally sufficiently far apart, their speeds will return to
their initial values, since their energies are conserved. In the original frame
of reference, after a temporary acceleration, the body of mass m slows until
its speed has the original value v0 , while the body of mass M is finally at
rest having been displaced (to the left) through a certain distance.
If the beads repel each other, a more detailed discussion is required. If
their initial energy is sufficient, the beads pass by each other, and as they
part their speeds return to their original values (as viewed from either the
centre of mass frame or the ‘lab’ frame). If, on the other hand, their initial
kinetic energy is too low for them to approach within a distance d, they ‘turn
back’. In the centre of mass system, the body of mass m then moves to the
left at speed −vm , whilst the body of mass M moves to the right with speed
−vM (Fig. S71.2).
−vm
m, q CM M, Q d
vM
Fig. S71.2
The condition for this to happen is

1 2 1 1 qQ
mvm + MvM
2
< ,
2 2 4πε0 d
i.e.
1 mM 2 1 qQ
v0 < .
2m+M 4πε0 d
The quantity mM/(m + M) is called the reduced mass of the system. The
velocities of the bodies in the laboratory frame can be obtained by adding
those in the centre of mass frame to the relative velocity of the two frames;
the latter is |vM |. Explicitly,
m−M 2M
vm∗ = v0 and ∗
vM = v0 .
m+M m+M
In the limiting case, when the initial kinetic energy is just sufficient to allow
the two beads to approach within a distance d, the two beads stop with respect
to each other (Fig. S71.3), i.e. when viewed in the ‘lab’ frame, they move on
with a common speed vm∗ = vM = mv0 /(m + M). In fact, this theoretical final
situation is unstable, since any small disturbance of the relative positions
will grow as a result of the mutual repulsion between the beads.
CM
m, q Stopped
M, Q Stopped
Fig. S71.3
Note. The three cases discussed above model one-dimensional mechanical

collisions. The limiting case, when the two bodies move on together, models
an inelastic collision. In this case only the mechanical momentum remains
constant; the mechanical energy decreases. The velocities in the case in
which the bodies approach each other and then move away again are the
same as those calculated from the laws of elastic collisions (conservation of
energy and momentum). The motion in which the bodies pass each other (in
essence, they do not collide and keep their original velocities) is obviously
in agreement with the conservation laws. The solution corresponding to this
case is usually not used for mechanical collisions since bodies cannot pass
through each other.
S72 (i) Let v0 denote the asymptotic common speed, d the original
distance between the beads and m the mass of one bead.
In a given time interval ∆t, the cluster collides with v0 ∆t/d further beads,
which increases its mass by ∆m = mv0 ∆t/d and its momentum by ∆p =
v0 ∆m = mv02 ∆t/d. According to Newton’s law of motion,
∆p mv02
F= = ,
∆t d
which yields v0 = Fd/m for the ultimate speed in the case of inelastic
collisions.
(ii) In an elastic collision between two equal mass bodies with one of
them initially at rest, their velocities are exchanged. The initially moving
body stops, whilst the second one moves away with the same velocity as that
initially possessed by the first.
The leftmost bead accelerates uniformly and reaches a speed of
2Fd
v1 = = 2v0
m
before the first (elastic) collision takes place. It then transfers its speed to the
second bead and stops, after which it starts accelerating again as a result of
the external force. What happens to the bead it has set in motion? It moves
Solutions 133
at a constant speed v1 , collides with the third bead and stops. The third and
subsequent beads behave similarly, and a ‘shock wave’ propagates forward
at speed v1 .
Meanwhile, the leftmost bead is again accelerated to speed v1 , collides
with the second bead, which is now at rest, and the process is repeated,
thus starting a new ‘shock wave’. The speed of the accelerated
√ bead varies
uniformly from zero to v1 , its average value is v1 /2 = v0 / 2 = Fd/(2m).
Note. The case of partially elastic collisions is also interesting. In this case
(according to the results of computer simulations) it can be stated that
sooner or later, the interacting beads condense into a single cluster that
behaves like a perfectly inelastic body with a final speed of v0 = Fd/m.
The time necessary for the cluster to condense depends upon the degree of
inelasticity (the coefficient of restitution). The more elastic the elementary
collisions, the longer the time necessary for an inelastic cluster to condense.
S73 At any time, the weight of the beer that is in free fall will not be
registered by the weighing machine, although the momentum destroyed as
the beer is brought to a halt by the jug will be, as will the force experienced
by the tap as the direction of the beer flowing through it is changed; there
are several such effects to consider. However, if the overall system of table
plus jug plus beer is considered these are internal actions and reactions
and the only two external forces (ignoring air resistance) are gravity and
the upward reaction from the weighing machine. The net result of these
two has to be such that the centre of gravity of the system falls, initially
accelerating (until the first beer reaches the jug), then sinking with constant
velocity, and finally (when the beer runs out) decelerating. Consequently the
machine reading, relative to the original, will be: increasingly negative – no
change – increasingly positive – no change.
S74 If the cross-section of the incoming water jet is A and its speed
is v, then the mass of water of density ρ flowing into the gutter in unit
time is of ρAv. This quantity of water has a kinetic energy of ρAv 3 /2 and
a horizontal momentum of ρAv 2 sin α. These quantities cannot change if
viscosity is neglected, and so
1 1 1
ρAv 3 = ρA1 v13 + ρA2 v23 , (1)
2 2 2
ρAv 2 sin α = ρA1 v12 − ρA2 v22 . (2)
The law of conservation of matter has to hold as well, and so we also have
ρAv = ρA1 v1 + ρA2 v2 , (3)
where A1 and A2 are the respective cross-sectional areas of the water flowing
out of the gutter to the right and left, while v1 and v2 are the corresponding
speeds.
Equations (1), (2) and (3) are insufficient to determine the four unknown
quantities (the two cross-sectional areas and the two speeds); a further
relationship has to be found. According to Bernoulli’s law, the quantity
ρv 2 /2 + p + ρgh is constant along a streamline of a non-viscous liquid. Inside
the liquid and far from the initial impact point, the pressure is constant and
equals the atmospheric pressure. If the difference in the heights of the streams
or, more exactly, the change in the energy corresponding to that difference,
is neglected (this is correct for a rapidly flowing liquid), the consequence of
Bernoulli’s equation is that v = v1 = v2 . This means that the liquid leaves the
gutter at the same speed at both ends! This is rather surprising, but correct,
within the accuracy of the above approximation.
The equations for the conservation of mass and momentum therefore take
the forms A = A1 + A2 and A sin α = A1 − A2 , which yield
A1 1 + sin α
= .
A2 1 − sin α
This ratio can be examined experimentally and surprisingly good agreement
with the calculated values found, which suggests that the approximations
made were reasonable.
S75 In a time interval ∆t, the level of the liquid with initial acceleration
a decreases by ∆h = a(∆t)2 /2 and the corresponding mass of liquid which
flows out is ∆m = (D 2 π/4)(∆h)ρ. This is equivalent to a decrease of (∆m)gh
in the potential energy of the liquid as a whole (see figure). Meanwhile, the
whole of the liquid is accelerated to a speed ∆v = a∆t and its kinetic energy
increased by (D 2 π/4)hρ(∆v)2 /2. The speed of the emerging liquid is higher
than this, but its effect can be neglected as the quantity of water involved is
small compared with the total.
D
a _a (∆ t )2
2
d
Solutions 135
According to the law of conservation of energy, the changes in the potential

and kinetic energies are the same:
πD 2 a πD 2 hρ (a∆t)2
(∆t)2 ρgh = ,
4 2 4 2
which yields a = g.
This means that initially some of the water starts to free fall. According to
the mass conservation law, the speed of the emerging water is (D/d)2 -times
higher than that of the water surface. Consequently, its acceleration must be
greater than g by the same factor. For example, if the diameter of the orifice
is one-tenth of that of the tube, then the initial acceleration of the emerging
liquid is 100g!
For how long is it true that the liquid is practically in ‘free fall’? Ac-
√
cording to Torricelli’s law of efflux, the speed of the efflux is 2gh and
√
the rate of decrease of the liquid level is (d/D)2 2gh. The time interval τ
between the start of the efflux and the attaining of (nearly) constant veloc-
ities by the surface and the emerging water can be estimated roughly using
the relationship
d2
gτ ≈ 2gh.
D2
If, for example, h = 20 cm and the ratio of the diameters is 1 : 10 then
τ ≈ 0.002 s, which is negligible for most purposes.
S76 The sand flows through the aperture almost uniformly, and therefore
the total operating time T of the sand-glass is proportional to the volume
H 3 of the sand. (As our aim is to obtain only a rough estimate, the difference
between the volumes of a cone and a cube is ignored.) The time T may also
depend on the gravitational acceleration g, the diameter d of the aperture
and the density ρ of the sand, and so T ≈ H 3 × f(g, d, ρ).
As T is a time and only g contains a time dimension, the function f has to
be proportional to the reciprocal of the square root of g. Similar reasoning
shows that T cannot depend on ρ, but is proportional to d−5/2 ; in summary,
T ≈ H 3 / d5 g. The coefficient of proportionality is a dimensionless number,
and since it does not depend on anything, can be assumed to be of order
1 (though such assumptions are notoriously dangerous in some branches of
physics!).
Consider some realistic data. If, for example, H is a few centimetres and d
is around a millimetre, T is a few minutes, which is indeed the sort of time
for which an egg should be boiled.
Note. In principle, the average diameter of the grains of sand could be
included in the formula. However, this is in general much smaller than

the other dimensions involved, therefore (in the same way as the size of
the atoms constituting the sand is ignored) it does not play a role in the
dimensional reasoning. The trickiest point in all dimensional analysis is that
of choosing the relevant parameters for the phenomenon in hand.
S77 Let the displacement of the bob be x and let us calculate the net
force (F) exerted on it (see Fig. S77.1).
m
F
x
h
2 F0
Fig. S77.1
The length of the extended spring is
x2
'= '20 + x2 ≈ '0 + ,
2'0
and so the tension in it is
x2
Fspring = k .
2'0
The net force acting on the bob (see Fig. S77.2) is
x k
F = −2Fspring sin θ ≈ −2Fspring = − 2 x3 ,
'0 '0
Fspring Fspring
h
F
Fig. S77.2
and the resulting equation of motion is
d2 x k
m = − 2 x3 .
dt 2 '0
This is a differential equation which cannot be solved by elementary methods.
However, to solve the problem as posed it is not necessary to solve the
equation explicitly, only to apply dimensional analysis to it. Writing the
Solutions 137
equation in the form

d2 x k
= − 2 x3 = −Cx3 ,
dt 2 m'0
we can suppose that the period of the motion T , depends only on C and A,
the amplitude of the vibration. The dependence can be written as
T ∝ C α × Aβ ,
implying the dimensional relationship

α
kg 1 1
s = [T ] = [C]α × [A]β = × × × mβ = s−2α × m−2α+β .
s2 kg m2
This equation is satisfied if −2α = 1 and −2α + β = 0. Therefore α = − 12
and β = −1, implying that T ∝ 1/A. Accordingly when the amplitude is
doubled (2 cm), the period is halved (1 s).
Since for dimensional analysis the choice of variables to include appears
more or less arbitrary, thus casting some doubt on the validity of the
conclusions reached, we now give another method for the solution of the
current problem.
The velocity of the bob as a function of its position can be calculated
from the law of conservation of energy. The stored energy in the two springs
when the bob is at rest at its maximum displacement must be equal to the
sum of the kinetic energy of the bob and the stored spring energy for a
general displacement. As the elongation of each spring is
1 x2
∆' = ,
2 '0
the equality may be written as
2 2
kA4 1 A2 1 1 x2
2
=0+2× k max
= 2Espring = mv 2 + 2 × k ,
4'0 2 2'0 2 2 2'0
where A is the maximum displacement (amplitude) of the bob. From this
equation the velocity of the bob can be expressed as
dx 1 k √ 4
=v= A − x4 .
dt '0 2m
After separating the variables we get
2m A dx T /4 T
'0 √ = dt = .
k 0 A4 − x4 0 4
Using y = x/A as a new variable, our final result becomes
32m 1 1 dy
T = '0 .
k A 0 1 − y4
This expression shows directly that the period is inversely proportional to
the amplitude, again leading to the conclusion that doubling the amplitude
of the motion to 2 cm, reduces its period to 1 s.
Note. The definite integral in the final expression for the period can be
evaluated using no more than a programmable calculator:
1
dy
≈ 1.31,
0 1 − y4
thus giving the complete solution to the problem.
S78 Because of the weakness of the spring, the body falls virtually
freely at first. The length of the spring is soon several times larger than
its unstretched length (which can be neglected during subsequent motion).
With this approximation, the body executes simple harmonic motion, both
vertically and horizontally. As it is released with no initial speed, it arrives
vertically under the suspension point after a quarter of the period of the
horizontal motion. Meanwhile, the vertical motion has also completed a
quarter-cycle, and the body has sunk to its equilibrium position at a depth
of mg/k (this is much larger than L).
The motion can be described quantitatively. In the coordinate system
shown in the figure, the equations of motion of the body at point (x, y) are:
x
max = −k x2 + y 2 − L ,
x2 + y 2
y
may = −k x2 + y 2 − L + mg.
x + y2
2
During the first part of the motion, whilst the extension of the spring is not
much larger than L, the force exerted on the spring can be neglected. On the
other hand, when
x2 + y 2 L
the original length of the spring can be neglected and the equations of
motion take the following simple forms:
max = −kx and may = −ky + mg.
These equations describe harmonic oscillations of identical periods, about
Solutions 139
the origin in the x-direction and about the equilibrium position y0 = mg/k
in the y-direction. Incorporating the initial conditions gives the solution:
k mg k
x(t) = L cos t , y(t) = 1 − cos t .
m k m
The body is under the point of suspension when x(t) = 0, and y = y0 = mg/k,
in agreement with our previous conclusion.
L
x
Note. For the early part of the motion t m/k (when the assumed
equations of motion are not strictly valid), the above expressions for x and
y can be approximated by x(t) ≈ L and y(t) ≈ gt2 /2, which are in agreement
with the formulae describing free fall, as is appropriate to that part of the
motion.
S79 If the carriage brakes with deceleration a, then in the carriage

reference frame, a ‘virtual inertial force’ of magnitude ma, in the direction of
the carriage’s motion, will appear to act on the body.
If this inertial force acted permanently, the pendulum could certainly
not reach the vertical, since, if it did, the net work done by the inertial
force would be zero (the net displacement of its point of application would
be perpendicular to its line of action) and the gravitational force would
be negative, implying that the kinetic energy of the pendulum should be
negative. This is impossible.
Consider now the fact that the carriage only brakes for a certain length of
time (until it stops). If it stops when the thread of the pendulum is horizontal,
the work done by the inertial force is W = maR, where R is the length of the
thread. If the pendulum subsequently reaches a vertical position with speed
v then, from the conservation of energy,

mv 2
maR − 2mgR = .
2
For the thread to remain taut, even at the topmost point, requires mv 2 /R >
mg, which, together with the above relation for the velocity, implies that the
deceleration of the railway carriage a > 2.5g. The conclusion is, therefore,
that the taut thread can reach the vertical provided the deceleration is great
enough and v0 is large enough for the pendulum to have time to reach the
horizontal before the carriage has come to a halt.
S80 The forces acting on the wedge are its weight mg and a force K,
perpendicular to the inclined plane; the magnitude of the latter may change
with time. As a result of these two forces, the only component of the
wedge’s acceleration a parallel to the inclined plane is g sin α (as measured
in an inertial frame). Newton’s equations of motion remain valid in an
accelerating frame of reference fixed to the wedge only if an ‘inertial force’
−m a is added to the forces actually causing the motion of a body. Here m
is the mass of the body under examination (e.g. that of a small volume of
water).
g sin
g sin
g
The resultant of the gravitational and inertial forces acting on the mass
m must be perpendicular to the inclined plane as the components parallel
to it cancel each other. The bodies on the wedge (the glass and the water in
it) ‘feel’ as if they were in a gravitational field perpendicular to the inclined
plane, with the consequence that the surface of the water lies parallel to the
plane.
This statement does not depend on the motion of the plane; it can be
fixed or move freely or even – as the result of a small force – be shaken to
and fro. As long as the friction between the inclined plane and the wedge is
negligible and the wedge does not rise off the plane, the shape of the water
surface cannot be other than a plane parallel to the inclined surface.
The case m M deserves an additional comment. In this case, the wedge
‘pushes away’ the inclined plane, and falls nearly freely. The weight of the
bodies on the wedge (including the water) are nearly completely ‘lost’; but
Solutions 141
even so, the small force keeping the water inside the glass is still sufficient to
set the water surface parallel to the inclined plane.
Note. The water surface would only become parallel to the inclined plane
after a long time, and on a correspondingly long plane. This is why this
interesting phenomenon cannot be observed experimentally in normal cir-
cumstances.
S81 Assume that the string is of uniform cross-section and mass distri-
bution, and is free at both ends. It orbits the Earth in such a way that its
position relative to the Earth is always the same. Obviously, if the string is
in a vertical position, the phenomenon could only occur at the Equator.
In the Earth’s reference frame a body of mass m orbiting above the Equator
at a distance r and with angular velocity ω experiences a gravitational force
of −GMm/r 2 and a centrifugal force of mrω 2 . Here M is the mass of the
Earth and G is the gravitational constant. The condition for the equilibrium
of the string is that the net force due to gravitation, which varies with r, is
equal to that due to the centrifugal effect, which also changes from point to
point. This condition can easily be derived using integral calculus, but it can
also be found without using such sophisticated mathematics.
Imagine that the string is pulled down a little by some external force. Since
(in the rotating frame of reference) the string was initially in equilibrium, it
can be displaced from its equilibrium position by an arbitrarily small force
and, to first order, the net work done in the course of the change must be
zero. The displacement of the whole string – from the point of view of the
work done – is equivalent to the slow migration of a small piece of the string,
of mass ∆m, from its top to its bottom. The work done is the sum of two
terms, the change in the gravitational potential energy and the work done
by the average centrifugal force (since the centrifugal force changes linearly).
If the bottom end of the string of length L just touches the Earth’s surface,
the work in question is
1 1 R + (R + L) 2
W = GM∆m − − ∆m ω L = 0,
R L+R 2
where R is the radius of the Earth. This is a quadratic equation in L, which
gives
R 8GM
L= −3 + 1+ ≈ 140 000 km
2 R3 ω2
using known data. This length is several times rs = (GM/ω 2 )1/3 ≈ 42 000 km,
the distance of telecommunications satellites from the centre of the Earth!
What is the maximum stress in the string? It is easy to establish that the
greatest stress σmax occurs at the position r = rs , and satisfies σmax /ρ =
4.8 × 107 N m kg−1 , where ρ is the mass density of the string. This figure for
the ratio of tensile strength to density is much greater than that for known
materials (for steel it is 2.6 × 105 , and for carbon 1.7 × 106 ). Therefore, whilst,
in principle, a ‘hook to the sky’ is consistent with Newton’s laws, at the
present time it is impossible (at least with a string of constant cross-section)
to find suitable materials from which to build it.
S82 (i) At the highest point of the bridge the equation of motion of the
car is
v2
mg − N = m ,
ρ
where N is the normal force acting on the car (and the negative of the
required answer), v = 20 m s−1 and ρ is the radius of curvature of the
bridge there. The most difficult part of the problem is to find this radius of
curvature.
If we could find a motion with this trajectory for which the normal accel-
eration is well known, the radius of curvature could be easily calculated. For
a parabolic trajectory the flight of a projectile offers the required analogue.
Let the projectile have an initial velocity of v0 making an angle α with the
horizontal.
The range (d = 100 m) and height (h = 5 m) of the projectile can be
expressed using the initial data,
2v02 sin α cos α v02 sin2 α
d= and h= .
g 2g
The quotient h/d gives tan α = 4h/d (so α ≈ 11.3◦ ), and the horizontal
component of the initial velocity is
g
vx = v0 cos α = d = 50 m s−1 .
8h
Now the radius of curvature at the highest point can be calculated as
ρ = vx2 /g = 250 m.
So the normal force at the highest point is
v2
N =m g− = 8.40 kN.
ρ
(ii) The force exerted on any other part of the bridge can be calculated in
the same way, i.e. using the radius of curvature. At a point three-quarters of
the way across the bridge, the radius of curvature is approximately 254 m
Solutions 143
and the normal force about 8.37 kN. Away from the centre of the bridge
there is also a tangential (frictional) force; here its value is 995 N, and so
the net force acting on the bridge is approximately 8.43 kN.
S83 We can prove that the radii of curvature of the ellipse at the
endpoints of its axes are b2 /a and a2 /b, where 2a and 2b are the lengths
of the major and minor axes, respectively. This geometrical result can be
deduced using calculus or by considering one of a number of physical
situations; what follows is one possibility.
Consider a planet orbiting the Sun in an ellipse. Newton’s second law of
motion applied at the endpoint of the major axis, a distance r from the
Sun, gives
M v2
G = ,
r2 R
where R is the radius of curvature at the endpoint and M is the mass of the
Sun. According to Kepler’s third law the period of the orbit is 2π a3 /GM
and the radius vector sweeps out area at a constant rate. The area of the
ellipse is πab, and so equating two expressions for that constant rate when
the planet is at the endpoint of the major axis, we obtain
vr ab GM
= .
2 2 a3
Comparing the above two equations we conclude that R = b2 /a. For this
argument we utilised the fact that the foci of the ellipse are on the major
axes; we cannot therefore apply the same proof at the endpoints of the
minor axis. However, in respect of their corresponding radii of curvature,
the two axes are symmetrical.
The uniformly moving point mass of the problem obeys the equation of
motion F = mv 2 /R, where R is the appropriate radius of curvature. Using the
data given we obtain: b2 /a = 1.25 m; a2 /b = 10 m and, hence, 2a = 10 m;
2b = 5 m.
Note. The radii of curvature of the ellipse could also be calculated using well-
known formulae from SHM. Consider the point mass moving in the x–y
plane around an ellipse with semi-axes a and b according to the equations
x = a cos ωt and y = b sin ωt.
At t = 0 the mass is moving at the end of the major axis with velocity
v = bω and acceleration A = aω 2 . On the other hand, the acceleration is
A = v 2 /R; so the radius of curvature is R = b2 /a. Similarly, we find the
radius of the curvature at the end of the minor axis to be a2 /b.
S84 Denote the width of the canal by d and draw a straight line perpen-
dicular to its banks a distance d downstream from the boat’s starting point
A (see figure).
v
A F
v
d v
The boat is initially at distance d both from the mark F on the opposite
bank and from this straight line. As both the speed of the water and that of
the boat with respect to the water are v, the water takes the boat downstream
by the same distance as is covered by the boat in the direction of F.
This means that the boat is always equally far from point F and the
straight line. The path of the boat is therefore a parabola, with F as its
focus and the straight line as its directrix. After a very long time, the boat
approaches the opposite bank at a point d/2 from F. Because the speed of
the current equals that of the boat, the boatman cannot land closer than this.
S85 If, after the slightest of pushes, the child would slide (straight)
downhill at a steady speed, the component F of its weight parallel to the
inclined plane must have the same magnitude as the frictional retarding force
S , i.e. F = S .
The force of kinetic friction – the direction of which is always opposite
to that of the instantaneous velocity – causes the speed to decrease, while
the force F increases the component of the velocity parallel to the inclined
plane. These two effects are of course present together and result, in general,
in a rather complicated motion (on a curved path and with a changing
acceleration). Despite this, the final speed can be determined without the
need for a detailed description of the motion.
The figure shows a coordinate system for the general situation in which
the child’s trajectory is not straight downhill. Denote the magnitude of the
instantaneous velocity of the sliding child by v, and its component in the
y-direction by vy . We first calculate the change in these two quantities in a
short time interval ∆t. According to Newton’s second law:
m∆v = (−S + F cos α)∆t, m∆vy = (F − S cos α)∆t.

Solutions 145
x
tree S
v v
F
Adding these two equations and using F = S gives
∆v + ∆vy = ∆(v + vy ) = 0,
i.e.
v + vy = constant.
From the initial conditions, the value of this constant is v0 = 1 m s−1 .

The final speed vmax of the sliding child is directed down the slope, and its
magnitude is determined by the above ‘conservation law’ with v = vy = vmax ,
i.e.
v0
vmax = = 0.5 m s−1 .
2
S86 Let kv denote the speed of the coastguard’s cutter, i.e. k is the
required ratio of the speeds of the two vessels.
v
a
a
kv
a
At a general time t, as shown in the figure, the distance d between the

ships (initially a) decreases by
∆d = kv∆t − v sin α ∆t (1)
in time ∆t. Meanwhile, the distance of the cutter from the shore increases by
∆y = kv sin α ∆t, (2)

where α is the angle the instantaneous velocity of the cutter makes with the
shore.
We now sum the small displacements occurring in equations (1) and (2),
knowing that the three sums ∆d, ∆y and v∆t must all equal a. The
potentially awkward angle α can be eliminated prior to the summations and
a surprisingly simple quadratic equation is obtained for k,
√
1+ 5
k − k − 1 = 0,
2
with k= ≈ 1.618
2
as its positive root. This figure is the famous ‘golden mean’ associated with
the Fibonacci series. In the current situation, it is the ratio of the speeds
of the coastguard’s cutter and the smugglers’ ship if they are to meet as
described in the problem.
S87 From the symmetry of the layout and initial conditions, we deduce
that all the bodies fall towards the centre of the n-gon with the same
non-uniform acceleration. The formation keeps its original shape, but the
distance r from the centre decreases at a non-uniformly accelerating rate.
The resultant force acting on one (say the nth) body when it is at distance r
from the centre is
n−1
m2 1
F(r) = G .
r2 k=1
4 sin(πk/n)
This force, made up of the gravitational forces exerted by all the other
bodies, or, more precisely, of those components of these forces which are
directed towards the centre, is identical to the gravitational attraction of a
fixed body situated at the centre, and of mass
m n−1 1
Mn = .
4 k=1 sin(πk/n)
The values of the masses Mn (in units of m) can be calculated numerically

for all values of n as
M2 = 0.25, M3 = 0.58, M4 = 0.96, ... , M10 = 3.86, ...
The time T of the collapse from an initial distance R onto a central

mass M can be considered as half of the period Te for a severely flattened
(degenerate) elliptical orbit of major semi-axis R/2. The period Tc of a
circular orbit of radius R can be calculated directly from the dynamical
Solutions 147
equation for circular motion,
Mm 2π 2
R3
G = mR , giving Tc = 2π .
R 2 Tc GM
But, according to Kepler’s third law,
3
Te 2
R/2
= .
Tc R
Thus, finally, we obtain T = π R 3 /8GMn for the time.

Note. The limiting case n 1 is interesting. As the number n of bodies
increases, Mn increases even if the total mass of the system is fixed at
M0 , i.e. m = M0 /n. The more finely a given amount of matter is spread
around a circle, the shorter the time it takes for it to collapse under
its own gravitational attraction. However, there is no point in examining
a continuous matter distribution spread along an arbitrarily thin line;
the extent of the matter in the transverse, i.e. radial, direction cannot be
neglected.
S88 According to Kepler’s first law the orbit of the rocket is an ellipse
with one of its foci at the centre of the planet. The launch and return
velocities are parallel to each other (though in opposite directions) if the
launch and return points are at the ends of the minor axis of the ellipse.
But, for an ellipse, the distance from a focus to either end of the minor
axis is equal to the length a of its major semi-axis; consequently a = R (see
Fig. S88.1).
c a
a h c
Fig. S88.1
From Kepler’s third law, satellites in orbits having different eccentricities,

but the same lengths of major axis, have equal periods, and so in our case
the period for a full orbit would be the given T0 . The rocket, however, covers
only one-half of the ellipse. The time required for this is not half of the full
period, but proportional to the fractional area swept by the radius vector
joining the rocket to the focus (Kepler’s second law). The area of the whole
ellipse is
θ
A0 = πab = πa2 sin .
2
The swept area for the half orbit is
πab 1 1 θ θ θ
A1 = + × 2bc = a2 π sin + a2 sin cos .
2 2 2 2 2 2
So the flight time is
A1 1 1 θ
T1 = T0 = T0 + cos .
A0 2 π 2
The maximum distance above the surface of the planet is
θ
2a − a − (a − c) = c = R cos ≤ R.
2
If the angle between the launch and arrival points is allowed to approach
zero (θ → 0), the calculated flight time approaches a maximum value of
1 1
T0 + ,
2 π
and the maximum height achieved approaches the radius of the planet
(c → R). But, in fact, if the take-off and landing sites are the same (θ = 0),
the rocket can reach any arbitrary height, large or small. This implies that
the period and maximum height are not continuous functions of θ at the
point θ = 0.
v1 v2
Fig. S88.2
If the launch speed is sufficiently great (equal to or larger than the first
√
cosmic speed, v = Rg) and the initial velocity is tangential to the surface
Solutions 149
of the planet, then the orbit shown in Fig. S88.2 is possible. Again the return
velocity is parallel to the launch one, but this time in the same direction. The
maximum height achieved can be anything, but the period must be at least
T0 . These are the orbits corresponding to the special case θ = 0.
S89 (i) Writing κ for µ0 /4π, the couple on C due to D is κµ2 /L3 anti-
clockwise; that of C on D is 2κµ2 /L3 , also anti-clockwise.
3 jl 2
L4
jl2 D 2jl2
L3 C L3
B
3jl 2
L4
(ii) Couples are not the only potential result of the magnetic fields; forces
will result if a dipole is positioned in a non-uniform magnetic field. The force
on C due to D is non-zero as the strength of B⊥ is slightly less at the position
of one of the poles of C than at the other. The magnitude of the net force is
µ × ∂B⊥ /∂r where the derivative is evaluated at r = L, and is thus 3κµ2 /L4 ;
its direction is the same as that of B⊥ . This force and its reaction on D
produce a couple on the rod which has magnitude 3κµ2 /L3 and acts in the
clockwise sense. It thus exactly cancels the other two couples acting on the
rod and when the system is suspended, nothing at all happens – something
that must be clear on the grounds of symmetry and the impossibility of free
perpetual motion!
Note. This is an unusual example of a non-central force and its reaction,
which act along parallel, but not identical, lines.
S90 The charge distribution induced on the plane by the charge q,

produces (in the region above the plane) an electrostatic field identical to
that of a charge −q situated below the plane at the point which is the mirror
image of the body’s position, if the plane were considered as a mirror. This
is the principle of image charges. Thus, the force of attraction acting on the
body (moving non-relativistically) can be calculated using Coulomb’s law as
q2
F(x) = k ,
4x2
where k is the Coulomb constant and x is the distance between the plane and
the body at any given instant (initially the body is a distance 2d from the
image charge). This expression for the force can be treated as the analogue
of the gravitational force exerted by a body of mass M = kq 2 /(4Gm) on
another body of mass m situated a distance x from it, i.e. F = GMm/x2 .
In this analogy, Kepler’s third law gives the relationship between the period
of revolution T of a planet and the major semi-axis a of its elliptical orbit,
namely,
T2 4π2
= .
a 3 GM
If the charged body is released at a distance d from the metal plane then
its orbit can be considered as a very flat (degenerate) ellipse with major
semi-axis a = d/2. The time at which the charge hits the plane Th , is half of
the period of the degenerate orbit, and can be calculated by substituting the
corresponding variables:
T π md3
Th = = .
2 q 2k
S91 Brine is a good conductor, because positive and negative ions can
move easily within it. When the charged plastic ball is placed close to the
surface of the water, opposing charges are induced in the surface, whilst like
charges are repelled from it. The resulting electric field lines above the water
surface will be perpendicular to it, whilst beneath it the net electric field
vanishes.
The charged ball attracts the water below it, and the surface wells up in
a hump. The electrical forces exerted on the hump are balanced mainly by
gravity and the effect of surface tension can be ignored. We don’t know the
shape of the hump exactly, but can be sure that the rise in water level will
be small and there will be only a slight deviation from a plane surface; this
is why we can use the so-called method of image charges. It will be sufficient
to consider the maximum effect and find the rise at the point P shown in
Fig. S91.1.
Solutions 151
r
3r
P
Fig. S91.1
At P the electric field due to the charge Q is

1 Q
E1 = .
4πε0 (3r)2
The effect of the unknown surface charge distribution can be replaced by
that due to an image charge of −Q situated at a depth below the surface of
3r (see Fig. S91.2). The electric field at P due to the image charge has the
same magnitude and direction as E1 , and so the net electric field is
1 Q
E = 2E1 = .
2πε0 (3r)2
3r
P
3r
−Q
Fig. S91.2
According to Gauss’s law the surface charge density at P is

1 Q
σ = ε0 E = .
2π (3r)2
At the water surface the force exerted on a unit area is the product of the
surface charge density σ and the electric field E1 due to the ball:
F
= σ E1 .
A
This is the upward force at P per unit area and is balanced by the hydrostatic
pressure associated with the maximum rise h in water level:
F
= ρgh.
A
Substituting for the electric field and surface charge gives an expression
for h as
1 1 Q 1 Q
h= .
ρg 2π (3r)2 4πε0 (3r)2
Inserting the numerical data into this equation, yields h ≈ 0.29 mm, which
is very small compared to the diameter of the ball and justifies our treating
the water surface as being close to flat.
S92 The method of spherical image charges can be applied. Let two
point charges of opposite signs be +Q1 and −Q2 . In the field produced by
them the locus of points of zero potential is given by
Q1 Q2
k −k = 0,
r1 r2
where r1 and r2 are the distances from the two charges of a general point
on the locus. A straightforward rearrangement gives Q1 /Q2 = r1 /r2 , i.e. the
distance ratio r1 /r2 is constant. According to Apollonios’s theorem, points
with this property lie on a sphere (the Apollonios sphere). Therefore the zero
potential surface is a sphere.
Fig. S92.1
If the spherical metal shell mentioned in the problem is earthed, then it is

at zero potential. The point charge +Q inside it induces an inhomogeneous
charge distribution on the inner surface of the shell as shown in Fig. S92.1.
The electric field inside the shell, due to the actual charge and the induced
charge distribution, is the same as if it were caused by the actual charge
and a negative point charge outside the spherical shell. The latter is what is
called a spherical image charge.
The charge distribution inside the spherical metal shell is independent
of the potential of the shell. If the shell were not earthed then charge +Q
would appear uniformly distributed over its outer surface, regardless of the
inner charge distribution. This is because the electric field strength is zero
everywhere inside the material of the shell and therefore the charge on the
outer surface is not aware of the presence of that on the inner one. The
Solutions 153
external electric field makes it appear as if the enclosed charge were at the
centre of the sphere.
B R +Q A − nQ
d
x
Fig. S92.2
The force acting on the charge Q inside the shell is equal to the Coulomb
force acting between the charge and the corresponding image charge. Using
the notation in Fig. S92.2, the charge +Q is a distance d from the centre of
the sphere of radius R, while the image charge −nQ is a distance x from
the shell. The ratio of the charges is therefore n, an expression for which
can easily be found using the two points A and B in which the straight line
connecting the charges intersects the spherical shell:
x x + 2R
n= = .
R−d R+d
This yields for n and x that n = R/d and x = R(R − d)/d. The force acting
on the charge inside the spherical metal shell is therefore
nQ2 Rd
F = −k = −kQ2 2 .
(x + R − d)2 (R − d2 )2
It is clear that this force is zero when d = 0 and tends to infinity as d → R.
The negative sign shows that it is directed towards the position of the
(imaginary) image charge.
Note. The electric charge distribution on the inner surface of the spherical
metal shell can be calculated using Gauss’s law. The magnitude of the
surface charge density is proportional to the electric field strength obtained
by superimposing the fields of the real and image charges.
S93 Denote the mass of the boron atoms (actually boron ions) by M
and that of the unknown colliding particles by m.
LAB
CM m
M
V0 V0
u
Fig. S93.1
Before the collision the particles have opposing velocities of the same
magnitude V0 as measured in the laboratory (LAB) frame of reference.
We can easily transform to the centre of mass (CM) frame of the colliding
particles shown in Fig. S93.1. The total linear momentum of the two particles
in the LAB frame is MV0 − mV0 , thus the velocity of CM in this frame is
M−m
u= V0 .
M+m
In the CM frame the total momentum is always zero and the two particles
must always move in opposite directions with linear momenta of equal
magnitudes. However, in accord with conservation of energy, the magnitude
of the momentum and therefore the velocity of each particle must be the
same before and after the collision – only their directions can change. The
speed of the boron atoms in the CM frame before the scattering is
M−m 2m
V = V0 − u = V 0 − V0 = V0 ,
M+m M +m
and so it must also have this value after the collision (Fig. S93.2).
CM
V
M V v m
Fig. S93.2
We can return to the LAB frame by adding u, the relative velocity of the
frames, to the CM velocity vectors. In the LAB frame the velocity of the
boron atom after scattering u + V, is a vector pointing to some point on the
circle shown in Fig. S93.3. The maximum angle between the final and initial
velocities of the boron atom occurs if u + V is tangential to this circle of
radius V, i.e. V is perpendicular to u + V.
LAB
M o
V
30
u CM
Fig. S93.3
Solutions 155
In this case
2m 1 M −m
|V| = sin 30◦ |u|, i.e. V0 = V0 .
M+m 2 M +m
This gives m = 15 M. Thus the unknown particle has a mass number of A = 2,

and is actually not unidentified any longer; the particle is the deuteron, the
nucleus of deuterium.
S94 Friction between the two balls is negligible, and so, during the
collision, they can only exert forces normal to their surfaces. Thus, the first
ball stops after the collision, while the second acquires the first’s initial speed
v0 . The rotation of the balls, however, does not change, and so, immediately
after the collision, the first ball rotates on one spot and the second slides
without rotation at speed v0 .
I II v
F
fr
− Ffr
Fig. S94.1
The friction between the balls and the table is of course important and
affects the motion of the balls. The first is accelerated forward by the force
of kinetic friction Ffr = µmg, whilst the second is slowed by the same force,
as shown in Fig. S94.1. The rotation of the first ball is reduced by friction,
that of the second one is increased. The part played by the frictional force
lasts until both balls reach the state of rolling without slipping. After that
their motion is unchanged.
It will be shown that the final motion of the balls depends neither on the
frictional coefficient, nor on the possible variation of it with position. After
the collision, the initially moving ball rotates with an angular velocity ω =
v0 /r. Its angular momentum about its axis is therefore Iω = 25 mr 2 (v0 /r) =
5 mv0 r. The angular momentum about the point of contact with the table P ,
2
must be the same, since the centre of mass of the ball is at rest, i.e. the angular
momentum attributable to translation is zero. The angular momentum of
the ball about P cannot be changed by friction any more, as the line of
action of this force runs through P . (The sum of the gravitational force and
the reaction of the table is zero, and so they can produce no net torque
either.)
= v0 /r = v /r
0
I I
v
W
P Ffr
N
Fig. S94.2
The angular momentum of a ball, rolling without slipping at speed v, is the

sum of its own angular momentum 25 mvr and the angular momentum mvr
due to the motion of its centre of mass. Figure S94.2 shows (on the left) the
initially moving ball and the forces acting on it shortly after the collision. On
the right of the figure, the ball is shown ultimately rolling without slipping.
According to the law of conservation of momentum 25 mv0 r = 25 mvr + mvr,
which yields v = 27 v0 . Similar reasoning shows that the final speed of the
other ball has to be 57 v0 , regardless of the magnitude of the coefficient of
friction.
S95 When travelling a distance L, the plank causes L/d rollers to acquire
an angular velocity ωmax = vmax /r. The decrease in potential energy of
the plank is MgL sin α, whilst the kinetic energy of each roller becomes
2 Iωmax = 4 mvmax . Notice that the final tangential surface speed of each
1 2 1 2
roller is equal to the terminal speed of the plank, and the moment of inertia
of each roller is I = 12 mr 2 .
It is false reasoning to suppose that the lost gravitational potential energy
of the plank is simply converted into kinetic energy of the rollers. This would
lead to concluding from the equation
L1 2
MgL sin α = mv (1)
d 4 max
that the terminal speed of the plank is
4dMg sin α
vmax = .
m
However, this result is wrong, because it does not take into account the fact
that the speeding-up of the rollers involves kinetic friction and, consequently,
there is heat dissipation in the process.
Denote by F(t) the kinetic frictional force between a single roller and the
plank. (It is not necessary to assume that this force is constant in time.)
During a short interval ∆t the change in the angular momentum of the
roller is
I∆ω = rF(t)∆t. (2)
Solutions 157
These changes can be summed to give an equation containing the final

angular velocity of the roller
vmax
r F(t) ∆t = Iωmax = I . (3)
r
On the other hand, during time interval ∆t the work done against friction
(heat gain) ∆Q is the product of the kinetic frictional force and the relative
displacement of the surfaces involved:
∆Q = F(t) [vmax − rω(t)] ∆t.
From (2) and (3), the total dissipated energy is
Q = F(t) [vmax − rω(t)] ∆t
= rωmax F(t) ∆t − I ω ∆ω
ωmax
2
ω2
= Iωmax
2
−I = I max .
2 2
In the final line we have used the fact that ω ∆ω = 12 ∆ ω 2 . This result
shows that the dissipated heat is equal in magnitude to the kinetic energy
acquired by the rollers. It is remarkable that the result depends on neither
the magnitude of the frictional force nor its time-dependence. The correct
energy balance is not equation (1), but
L 1 2 L 1 2
MgL sin α = mvmax + Q = 2 mv ,
d 4 d 4 max
which shows that the terminal velocity is
2dMg sin α
vmax = .
m
S96 Treating the problem as two-dimensional, choose a point P on the

surface of the table and examine the angular momentum of the ball about
this point. The line of action of the frictional force passes through this point,
and so there is no frictional torque about P . The gravitational force and the
supporting reaction of the table nullify each other. No other force acts on
the ball, which therefore has constant angular momentum about the chosen
point. As the ball is initially at rest, the value of that angular momentum is
zero.
x
mv
r
P
When the tablecloth is pulled out from under it, the ball starts sliding and
rolling. Following the notation in the figure, its angular momentum J can
be written as the sum of two terms:
J = Iω + r × (mv).
Here I is the moment of inertia and m the mass of the ball. The first term
in the expression is the internal spin, and the second the orbital angular
momentum due to the linear motion of the centre of mass. Taking into
account the directions of the vectors involved, the magnitude of the angular
momentum can be written as J = Iω + mvR, where R is the radius of the
ball.
It is easy to see that, when any ball is rolling without slipping, the sign of
its orbital angular momentum has to be the same as that of its spin. On the
other hand, here the sum of the two has to be zero at all times. These two
conditions can only be fulfilled at the same time if the body has stopped.
The reader can check this experimentally.
The final state depends neither on the size of the frictional force, nor
on how the tablecloth is pulled out. (It can be pulled out evenly, with a
uniform acceleration, or by means of several sudden movements.) However,
it is important that air resistance and rolling resistance are negligible since
their effects can change the angular momentum about P.
S97 Taking the Earth’s actual direction of rotation (from west to east)
as positive, the angular momentum of the traffic about the axis of rotation
would increase if the change were made. This is because the traffic that is
travelling eastward would move to a greater distance from the Earth’s axis
thus increasing its (positive) contribution to the total angular momentum;
conversely the westward-bound traffic would reduce its negative contribution.
Assuming equal amounts of east–west and west–east traffic, the moment of
inertia of the system is unchanged and, since the total angular momentum
of the system cannot change, the Earth’s rate of rotation must decrease. The
length of the day would therefore increase – but you would hardly notice it!
Note. One can also arrive at the same conclusion in a different way. In Great
Britain there are a lot of traffic roundabouts. Any change in the direction in
which these were negotiated would cause a change in the angular momentum
of the traffic, which in turn would cause a small change in the rotation of
the Earth. The whole traffic system can be considered as a series of many
roundabouts.
S98 Let the angular acceleration of the smaller ball be α1 , that of the
larger one α2 , their common horizontal acceleration a1 and the acceleration
Solutions 159
of the cart a2 . As the balls are rolling without slipping, we have

Rα2 = a2 − a1 and Rα2 = rα1 ,
and, because R = 2r,
a2 − a1
α1 = 2α2 = .
r
N1 a1
2
g
N2 N2
a1 N1
8 mg
N
Ffr Ffr a2
F
The moment of inertia of the smaller ball is 25 mr 2 , while that of the larger
one with the same density is 25 × 8m × (2r)2 = 645 mr . Using the notation of
2
the figure, we can write the following dynamical equations of motion:

F − Ffr = Ma2 ,
8mg + N1 − N = 0, Ffr − N2 = 8ma1 ,
mg − N1 = 0, N2 = ma1 ,
2 2
N1 r cos φ − N2 r sin φ = mr α1 ,
5
64 2
2rFfr + 2rN2 r sin φ − 2rN1 r cos φ =mr α2 .
5
From these equations we can express the force F as
7 cos φ
F = 9m + M g ≈ 79 N.
2 1 + sin φ
The acceleration of the balls relative to the cart is
5 cos φ
∆a = a2 − a1 = g.
2 1 + sin φ
At the time t when the balls fall from the cart, the distance they have moved
relative to the cart is L/2. As their initial velocities are zero,
L
t= = 0.55 s.
∆a
Note. It is interesting that this stunt can also be performed with the smaller
ball in the horizontal position, φ = 0. In this situation the frictional force
between the balls balances the entire weight of the smaller ball. What is
more, it is even possible for φ to be negative, if the coefficient of friction
between the balls is sufficiently large!
S99 Consider the motion of the ball with respect to an arbitrary point
P in the plane of the table. Since the forces acting on the ball due to
gravity and the normal reaction of the table are equal and opposite, and the
frictional force acts in the plane of the table, their total moment about P
has no horizontal component. Consequently, the horizontal component of
the angular momentum of the ball about P must remain constant.
When the ball is rolling without slipping, the horizontal component of its
angular momentum is perpendicular to its velocity v and its magnitude is
directly proportional to |v|. (For a solid ball of radius R, the total horizontal
component of its angular momentum is J = IΩ + mvR = 75 mvR see also P96
and S96.) At the end of the motion, when the ball again rolls on the table
without slipping, this component of the angular momentum must therefore
be the same as it was at the beginning. It then follows that the direction and
magnitude of the ball’s velocity v are also unchanged.
Thus, the ball leaves the table at the same speed, and with the same
momentum vector, as it had originally. Further, its total kinetic energy is
unchanged. This last fact is especially strange, as both when the ball arrives
at the disc and when it leaves it, friction does work on the ball and changes
its kinetic energy. However, the algebraic sum of the work done on the ball
is zero; this is not a consequence of the conservation of either momentum
or energy, but of the conservation of angular momentum.
Note. Rather more complicated calculations (using the equations of trans-
lational and rotational motion) show that when the ball is on the rotating
disc it moves uniformly along a circular path, as viewed from the reference
frame of the table. (To obtain this result it is assumed that the coefficient
of kinetic friction is large enough that we can ignore the period during
which the ball slips on the disc. For the solid rubber ball used in the Science
Museum, this assumption is quite reasonable.) The circle, however, is not
centred on the axis of the disc, and the motion’s angular velocity is different
from that of the disc, being 27 times smaller. If the disc rotates steadily, the
Solutions 161
ball ultimately continues along the extrapolation of its original track. This
is not true if the disc does not rotate uniformly, or the period of slippage
is not negligible; in either case, the magnitude of the ball’s momentum is
unchanged, but its track deviates sideways as a result of its encounter with
the disc.
S100 Let the tension in the ring be T . Its resolved component acting
along the radius towards the centre of rotation is 2T sin(∆θ/2) ≈ T ∆θ and
this must balance the centripetal force of R∆θAρRω 2 (see figure).
T ∆ h/2 T T
T
∆h
R 2T sin ∆ h/2
It follows that the longitudinal stress in the ring, T /A, is ρR 2 ω 2 ; the strain
ε is E −1 times this. Finally, the increase in circumference, given by 2πRε, is
2πρR 3 ω 2 /E.
S101 When one end of the thread is pulled by a force F0 , let the maximum
force with which the other end can be pulled without the thread slipping
on the cylinder be Fmax . Specify a general point of the thread in contact
with the cylinder by the angle α, which the radius of the cylinder at that
point makes with a fixed radius. When the thread wound onto the cylinder
is tightened, it exerts a normal force on the cylinder resulting in a frictional
force which opposes any relative motion of the thread and the cylinder. The
tension in the thread increases as α increases, but the excess tension at one
end of a piece of the thread is balanced by the frictional force acting on that
piece.
R ∆a
F F + ∆F
∆a
R
Fig. S101.1
Consider a small length of thread that subtends an angle ∆α at the centre

of the cylinder. If, as shown in Fig. S101.1, the tension at one end of the
small piece is F whilst it is F + ∆F at the other, then the excess force ∆F is

balanced by a frictional force, which can be calculated as
∆F = µN, (1)
where N is the force exerted by the thread normal to the surface of the
cylinder and µ is the required coefficient of friction.
F F + ∆F
∆a ∆a
Fig. S101.2
The normal force can be determined as the vector resultant of the forces
F and F + ∆F ≈ F, shown in Fig. S101.2. This is
∆α
N = 2F sin ≈ F∆α. (2)
2
Substituting this into equation (1) shows the relationship between F and the
angle α to be
∆F(α) = µF(α) ∆α.
This relationship is formally similar to the equation governing radioactive

decay,
∆m(t) = −λ m(t) ∆t,
where m(t) is the mass of radioactive material, t the elapsed time, and λ the
decay constant. As is well known, the mass of radioactive material decreases
exponentially with time, i.e.
m(t) = m0 e−λt .
Thus, using the established correspondence, with −λ replaced by µ, the law

of ‘thread friction’ can be expressed as
F(α) = F0 eµα . (3)
Both of the inequalities stated in the problem are equivalent to

1
2F(0) = F(π) = F(0)eµπ , which yields µ = ln 2 ≈ 0.22.
π
Solutions 163
Note. The force exerted on the thread increases exponentially with the angle
α. The ratio of the forces at the two ends of the thread can reach a large
order of magnitude after only a few turns. Climbers make use of this
interesting fact when they anchor the ropes that prevent them from falling.
Sailors use the same technique to stop large boats with their bare hands!
S102 Jenny considers a homogeneous ring of radius R and linear density

ρ, rotating with constant angular velocity ω about an axis perpendicular
to its plane and passing through its centre. The centripetal acceleration of
points on the ring is Rω 2 , and so unit length of it experiences a centripetal
force of p = ρRω 2 . If a cylindrical container of height 1 m were surrounded
by a gas at pressure p then the force exerted by the gas on the wall of
the cylinder would be exactly the force required to sustain the rotation (see
Fig. S102.1).
p
F q F
Fig. S102.1
In reality, the elements of the ring are not kept in their circular orbit by
some imaginary external pressure, but by the ring’s own internal tension,
whose magnitude is
F = ρR 2 ω 2 .
This can be proved by referring to P100 or by examining, for example, a
1-m length of a container with a semicircular base, surrounded by a medium
at pressure p (see Fig. S102.2). A force of 2Rp acts on the rectangle of area
2R, and has to be balanced by a force of magnitude 2F acting tangentially
within the wall.
p
p
F F
p
2 Rp
Fig. S102.2
Now examine an arc of the ring which subtends an angle at its centre of
2α, and see how Newton’s second law of motion is fulfilled, continues Jenny.
The resultant of the forces acting on a body of mass m = 2R ρ α is 2F sin α,
and the acceleration of the centre of mass is a = s ω 2 , where s denotes the
distance between the centre of mass of the arc and the centre of the ring
(see Fig. S102.3).
F s F
a
Fig. S102.3
According to Newton’s law,

2F sin α = m a,
which, using the previous formulae, yields s = (R sin α)/α. For a semicircle,
as in Charlie’s original problem, α = π/2 and, therefore, s = 2R/π.
If the centre of mass of a sector has to be determined, it can be divided
into thin arcs and each one replaced by a point-like body of appropriate
mass positioned at the centre of mass of the arc. The same procedure could
be adopted to find the centre of mass of a triangle made up of thin stripes.
This implies that the centre of mass of a sector has to be in exactly the
same place as that of a symmetrical triangle of height smax = (R sin α)/α
(see Fig. S102.4), i.e. at a distance of 23 (R sin α)/α from its vertex. With this
deduction Jenny concludes her display of extraordinary logic.
smax
Fig. S102.4
S103 Let us denote the common height of the table and total chain
length by L (= 1 m) and the length and mass of the vertically moving part
of the chain by x and m. The equation of motion (taking into account the
changing mass of the moving part) is
d(mv) dv dm
mg = =m + v,
dt dt dt
Solutions 165
which on rearrangement gives

dv dm
m = mg − v.
dt dt
The left-hand side is the product of the instantaneous mass and acceleration
a, whilst the right-hand side can be converted and simplified using dm =
(m/x)dx and dx/dt = v to yield
v2
a=g− .
x
This result shows that the acceleration of the chain is less than g. The second
term on the right-hand side can be simplified further, since v 2 /x = 2f in
the case of rectilinear motion with constant acceleration f and zero initial
velocity. This means that in the current problem the acceleration of the chain
is constant and satisfies
g
a = g − 2a, yielding a = .
3
As the chain runs down from the table during a time of t1 , its first link
falls a distance L with acceleration g/3. Consequently,
2L 6L
t1 = = = 0.78 s.
a g
When the lower end of the chain reaches the ground, the whole chain is
vertical and its velocity is
g 6L 2Lg
v1 = at1 = = = 2.56 m s−1 .
3 g 3
From this moment on, the chain goes into free fall. Its last link has an initial
velocity v1 , accelerates with g, and covers a distance L in time t2 . Thus
1
L = v1 t2 + gt22 .
2
From this equation we obtain
2L t1
t2 = = = 0.26 s.
3g 3
So the final link of the chain reaches the floor at a time
4
t1 + t2 = t1 = 1.04 s
3
after the start of the process.
Note. (i) Attempting to apply the law of conservation of energy leads to

false results. For example, setting
1 2 L
mv1 = mg ,
2 2
where L/2 is the loss of height of the centre of mass of the chain, and then
substituting v1 = 2Lg/3 leads to the contradictory result
2
L 1 1 2 L
mg = mv12 = m Lg = mg .
2 2 2 3 3
It would appear that one-third of the energy has disappeared. This is, in fact,
accounted for by the energy dissipated in the series of inelastic collisions
occurring when the chain jerks the successive links into motion.
(ii) The problem can also be solved in a different way. Let M be the total
mass of the chain. When the hanging part of the chain of mass m = (M/L)x
causes the next piece, of mass (M/L)∆x = (M/L)v∆t, to move, it accelerates
the piece from rest to a velocity v in a time interval of ∆t. This acceleration
needs a force of
(M/L)v∆t v M 2
= v .
∆t L
The corresponding reaction decelerates the hanging part of the chain, so we
can write
M
ma = mg − v 2 .
L
Inserting m = (M/L)x into this equation, we recover the earlier solution.
(iii) Assuming that the chain consists of n links with an otherwise uncon-
strained separation of ε = L/n between links, leads to the correct answer in
the limit n → ∞.
(iv) It is possible, using calculus techniques, to solve the (non-linear)
differential equation
v2
a=g− ,
x
subject to the initial conditions v = 0, x = x0 (x0 L) at t = 0. The
solution
x
2 x0 3 dx
v(x) = gx 1 − , t=
3 x x0 v(x)
approaches our more heuristic result in the limit x0 → 0.
S104 It will be shown that a chain (flexible rope), moving at a uniform

speed along a closed curve of arbitrary shape, continues moving in the same
way even if no constraints (e.g. pulleys, cylinders, etc.) are placed on it.
Solutions 167
F
F
F
F
F
Fig. S104.1
Consider the chain (shown in Fig. S104.1) moving steadily along some
closed, winding space-curve at speed v. The force stretching the chain has to
have the same magnitude, F, everywhere, as the tangential acceleration of its
links is zero. If the radius of curvature of the chain, of mass per unit length
ρ, is R at some point (R can vary from place to place), then the mass of a
piece of length R∆α is ∆m = ρR∆α, whilst its acceleration is v 2 /R, as shown
in Fig. S104.2. Its equation of motion is
v2
ρR∆α = F∆α,
R
which leads to the relation F = ρv 2 .
F F F F
∆a
∆a
R F ∆a
Fig. S104.2
Notice that R does not occur in this equation, i.e. F is independent of the
radius of curvature. The resultant of the tangential forces, F, is just the right
force to make the chain curve as it does at the given place. If the chain is
straight, the resultant force acting on a small piece of it is zero. The more
curved the chain, the greater the resultant force acting on each small piece.
The direction of the resultant force is also just what is needed.
This means that the chain falls keeping its original shape and speed, just
as Frank guessed.
S105 As shown in the solution to the previous problem, a flexible chain,
or rope, of mass ρ per unit length moving with a speed v has an internal
tension of
F = ρv 2
if it is not contact with any other body. It should be noted that this result is
independent of the radius of curvature R of the arc formed; it also applies
to chains or ropes moving in a straight line when R can formally be taken as
infinite. In the present case, the chain becomes detached from the pulley, as
a result of its accelerating motion, when this freedom condition is satisfied.
Denote the displacement of the chain by x and the acceleration of the
right-hand side of the chain by a. The equations of motion for the two sides
are
L L
F −ρ −x g =ρ − x a,
2 2
L L
ρ +x g−F = ρ + x a.
2 2
The speed of the chain for any x can be determined, without using inte-
gration, from the conservation of energy. The decrease in potential energy
relative to that of the original situation (for which negligible speed is as-
sumed) is the same as if a piece of the chain of length x had been lowered
by x (see figure). Hence
1
ρx2 g = ρLv 2 .
2
L/2
x
Solutions 169
The displacement and speed when the chain leaves the pulley can be
calculated by eliminating F and a from the above equations:
√
1 Lg
x = √ L ≈ 0.35 L, v= .
2 2 2
Thus, the chain becomes detached from the pulley when 15 per cent of its
length is still moving upwards and its speed is less than one would obtain by
naı̈vely substituting x = L/2 into the conservation of energy equation. (This
√
false value would be Lg/2 ≈ 0.71 Lg.)
Note. The subsequent motion of the chain is also interesting. After the
chain has become detached from the pulley, it is in the form of two vertical
sections of unequal length joined by a semicircular arc, with the left-hand
section, whose links move upwards, getting shorter at an increasingly rapid
rate, whilst the right hand section correspondingly accelerates downwards.
If the radius of the pulley is negligible, the speed of the links in its left-hand
vertical section, which has decreasing mass, tends, in principle, to infinity.
In reality, the finite pulley radius and link size, together with air resistance,
place an upper limit on the speed. The kinetic energy of the piece of chain
moving upwards remains finite despite its rapidly increasing speed because
the decrease in the mass of the relevant part of the chain is more rapid than
the increase in its speed. The same phenomenon can be observed when a
whip cracks; a section (of decreasing length) of the whip moves at an ever
increasing speed, and when it reaches the speed of sound, it causes a sharp
supersonic bang.
S106 (i) Examine the motion of the loop in the frame of reference, which
moves with the loop at speed c. In this system, the pieces of a circular loop
of radius R rotate uniformly. The acceleration of a piece of the rope, which
subtends central angle ∆α and has mass ρR ∆α, is c2 /R, whilst the net force
due to the tension in the rope is F ∆α (see S105). The Newtonian law of
motion yields
c2
F ∆α = ρR ∆α ,
R
i.e. the ‘loop-wave’ moves with speed c = F/ρ, – identical to the speed at
which small transverse waves would propagate along the same rope.
(ii) The angular frequency of the rolling loop of radius R is ω = c/R,
and the loop has mass, m = 2πRρ. The energy carried by the loop can be
expressed in terms of these quantities as
1
E = Etransl + Erot = (2πRρ) (c2 + R 2 ω 2 ),
2
which can be written in the form

1 1
E(ω) = 2πFc =K .
ω ω
The quantity K = 2πFc is a constant, characteristic of the rope and the
tension in it. Of course, several (n) loops can be simultaneously excited and
their total energy is
1
En (ω) = nK (n = 0, 1, 2, 3, . . .).
ω
The momentum of the loop(s) can be calculated in a similar way,
1
Pn (ω) = n(2πRρ)c = nK (n = 0, 1, 2, 3, . . .),
cω
as can the angular momentum
1
Jn (ω) = ±n (2πRρ) Rc = ±nK (n = 0, 1, 2, 3, . . .).
ω2
In this latter formula, the two signs correspond to loops moving ‘above’ and
‘below’ the rope, i.e. to the two possible ‘polarisations’.
It can be seen that – provided only circular excitations of a single fre-
quency are allowed – the energy, momentum and angular momentum can
only assume discrete values, those that can be written as the product of an
integer and a basic ‘quantum’. The following relationships are valid between
these (frequency-dependent) ‘quanta’:
E(ω) = cP (ω) = ωJ(ω).
It is not difficult to recognise that the same relationships are valid for photons
Ephoton = h̄ω; Pphoton = h̄ω/c; Jphoton = ±h̄.
Naturally, this formal analogy must not be taken too seriously, e.g. by
thinking of a photon as equivalent to a loop. However, the similarity can
be used to show that, even in classical physics, there are objects more
complicated than a point mass which are easy to understand, but which can
still be excited to many discrete energy, momentum and angular momentum
levels.
S107 A volume of sand of mass ∆m = 50 kg reaches a speed of v =
1 m s−1 in time ∆t = 1 s. The change in its horizontal momentum is therefore
∆p = ∆m v = 50 kg m s−1 . This means that a force
∆p v ∆m
F= = = 50 N
∆t ∆t
Solutions 171
accelerates the sand. The work done by the engine – taking only the acceler-
ation of the sand into account – is 50 J s−1 , i.e. its power output is 50 W.
The sand loses its vertical momentum when it lands on the conveyor belt.
(It hits the belt vertically with a force greater than its weight.) The sand
on the belt is slowed down vertically and then accelerated horizontally; the
belt’s kinetic energy is increased. The kinetic energy of the sand increases by
∆m v 2 /2 = 25 J in ∆t = 1 s. This means that one-half of the power of the
engine (25 W) is converted into kinetic energy of the sand; the rest is the
work done against friction and converted into heat.
Note. The average speed of the sand during the acceleration is v/2. Therefore
the power of the frictional force (F = 50 N) is Fv/2 = 25 W. The belt
experiences a force −F, the power of which is −Fv = −50 W. Thus, exactly
one-half of the power of the engine is used to accelerate the sand.
S108 (i) First, we determine the speed of the roll as a function of

the distance it has covered. The mass of the moving part of the hose
after travelling a distance, x, is m(x) = M 1 − x/L . Its speed v(x) can be
determined using the conservation of energy:
1 1 1 v0 2
1 1 1 2 v 2
Mv02 + MR 2 = mv 2 + mr .
2 2 2 R 2 2 2 r
The change in potential energy and the small vertical speed acquired as
a result of the decrease of the radius of the roll have been neglected. Using
the known variation of the mass with distance, the velocity v is found to be
given by
v0
v(x) = .
1 − x/L
As x increases, so does the velocity, i.e. the roll accelerates as it unrolls.
The total time taken to unroll can be obtained by integrating the reciprocal
of the function v(x) = dx/dt:
T L dx 1 L L 1√ 2 L
T = dt = = 1 − x/L dx = 1 − u du = .
0 0 v(x) v0 0 v0 0 3 v0
Since the hose accelerates, the time taken to unroll is obviously shorter than
if the hose had unrolled with uniform velocity. In fact, it is two-thirds of
that figure.
(ii) The system consisting of a roll of decreasing mass and increasing speed,
and of a motionless horizontal part of increasing length can obviously not
be considered as a point mass! Therefore the basic law of dynamics cannot
be applied to it in the simple form F = ma, but must be used in the more
general form
d
Fexternal = ptotal ,
dt
where ptotal is the momentum of the system as a whole.
The total momentum of the moving roll (and of the whole system) is
v0
p(x) = m(x) v(x) = M(1 − x/L) = Mv0 1 − x/L.
1 − x/L
Clearly, as x increases, p(x) decreases – reflecting the fact that the mass of the
piece in motion decreases faster than the rate at which its speed increases.
The direction of the resultant force K(x) acting on the system is, therefore,
opposite to the direction of the motion, with
dp dp(x) Mv02 1
K(x) = = v(x) = − .
dt dx L 2(1 − x/L)
S109 The gravitational field due to a thin spherical shell of uniform mass
distribution is zero inside the shell. Outside it is as if the total mass of the
shell were concentrated at its centre. At a distance of 100 km below the
surface of the Earth, two factors affect the gravitational field. On the one
hand, the mass of the part of the Earth still ‘underneath’ is smaller than the
total mass of the Earth, meaning that the gravitational acceleration will be
reduced. On the other hand, the centre of the Earth is closer, which will tend
to increase g. Which effect is the stronger?
The shell of thickness 100 km corresponds to 4.6 per cent of the total
volume of the Earth, which has a radius of 6400 km, but its mass is only
2.5 per cent of the total mass of the Earth. Gravitational acceleration can be
calculated as g = GM/r 2 , where M is the mass inside the radius r. At a depth
of 100 km below the surface, the effective mass of the Earth (without its
crust) is M = 0.975 M, with radius r = (6300/6400) r = 0.984 r. Substituting
this data into the above equation, we find that 100 km below the surface of
the Earth, the gravitational acceleration is 0.7 per cent greater than at the
surface!
More generally, it can be proved that g increases as the centre is ap-
proached if the density of the crust is not greater than two-thirds of the
average density.
S110 First consider the trial bore, which is of negligible volume compared
with that of the whole asteroid. Let the density of the asteroid be ρ and its
radius R. The gravitational acceleration at radius r is the same as if only the
Solutions 173
sphere of radius r below it were present:

m(r) (4/3)πr 3 ρ 4πGρ
g(r) = G = G = r.
r 2 r 2 3
Thus the gravitational acceleration is directly proportional to the distance
from the centre of the asteroid and always points towards it.
This means that the first unfortunate little green man approached his fate
whilst executing a harmonic oscillation of amplitude R, and reached it after
the first quarter-period of the oscillation. The coefficient of r in the above
expression corresponds to ω 2 , i.e. to the square of the angular frequency.
Therefore the duration of the fall was
T 1 3π
T1 = = .
4 4 Gρ
The speed at which he collided with the bottom of the hole can be obtained
as the product of his amplitude and angular frequency,
πGρ
v1 = R ω = 2R .
3
By the time of the second accident, the little green people had already
excavated one-eighth of the material in the asteroid. The gravitational field
in the spherical cavity, extending from the surface of the planet to its centre,
has to be determined. This can be done using a ‘cunning’ application of the
principle of superposition: imagine the cavity to be filled with a mixture of
‘normal’ titanium and ‘negative density’ titanium.
The gravitational fields of the complete asteroid and of the sphere of
‘negative titanium’ have to be added. The vector r pointing to an arbitrary
point P in the cavity, the vector c pointing from the centre of the cavity to
the centre of the asteroid and the vector r + c pointing from the centre of
the cavity to point P are shown in Fig. S110.1.
Titanium
Hole P
r+c
r
c
Fig. S110.1
The gravitational accelerations (of the homogeneous asteroid, of the cavity

and of their resultant) are proportional to the position vectors, the coefficient
of proportionality being the constant previously denoted by −ω 2 . The ‘lack
of matter’ in the cavity is represented by a negative density.
The resultant acceleration (see Fig. S110.2) is
g = −ω 2 r + ω 2 (r + c) = ω 2 c.
( r + c) x2
P
2
c cx
2
− rx
O
Fig. S110.2
This gravitational acceleration is a constant vector, regardless of the

position of point P inside the cavity. This means that there is a homogeneous
gravitational field in the cavity, with a magnitude −ω 2 R/2, the gravitational
acceleration in the middle of the trial bore (i.e. at the centre of the cavity).
Note. It can similarly be shown that there is a homogeneous field in any
spherical cavity; it has a magnitude equal to the gravitational acceleration
at the position inside the solid sphere on which the cavity is to be centred.
The duration T2 of the fall and the speed of impact v2 of the second
unfortunate little green man can be calculated by applying the equations for
uniform acceleration:
2 3 πGρ
T2 = = and v2 = 2R .
ω πGρ 3
The ratios reported by the expert were therefore
T1 π v1
= and = 1.
T2 4 v2
It can be seen that the second unfortunate little green man took a slightly
longer time to fall than the first one, but that they both hit the centre of the
asteroid with the same speed.
The fact that the speeds are the same – and, by implication, so are their
kinetic energies – is not a coincidence! In the first case, because the gravita-
tional potential is lower at the centre of the asteroid than at its surface, the
Solutions 175
first unfortunate man is accelerated and gains the corresponding amount of

kinetic energy. The second case is different in that the gravitational potential
of the ‘negative density’ cavity has to be taken into account as well. However,
the man fell from one edge of the cavity to the other, and so his gravitational
potential with respect to the cavity did not change. Applying the principle
of superposition to the gravitational potentials gives the final result, namely
that the total change in the potentials is the same in both cases, and so the
speeds acquired by the two victims are the same.
S111 Imagine the hemisphere to be divided into many concentric hemi-
spherical shells of identical thickness. What is the force exerted by these
shells on a probe of unit mass sitting in the centre of the sphere? Since the
mass of a shell is directly proportional to the square of its radius and the
force exerted by a given-mass shell on the probe is inversely proportional to
the square of its radius, the gravitational accelerations at the centre due to
the different hemispherical shells are all equal (see Fig. S111.1).
R R/n
Fig. S111.1
If there are n shells, then the mass of the outermost shell is 2πR 2 (R/n)ρ,
where R is the radius of the asteroid and ρ is its density. The total gravita-
tional field due to the hemisphere is n times that due to its outermost shell,
i.e. it is the same as if the surface of the hemisphere had mass M = 2πR 3 ρ.
This mass is, in fact, three times the actual mass of the hemisphere.
What force is exerted by this hemispherical shell of mass M on the probe?
Considering only its magnitude, the force is the same as that exerted by the
unit-mass probe on the shell, i.e. p = G(M/2πR 2 ) (1/R 2 ) = Gρ/R per unit
surface area. To integrate this effect over the whole shell, the situation can
be compared with that of finding the force exerted by a liquid at pressure p
on a similar hemispherical shell (see Fig. S111.2).
Fig. S111.2
Since the resultant force of a liquid acting on a complete hemisphere would

be zero, the total force acting on its curved surface would be the same in
magnitude as the force acting on its plane surface. Thus g = pπR 2 = GρRπ
(see Fig. S111.3).
Fig. S111.3
Now consider the initial gravitational acceleration,

4πR 3 1
g0 = G ρ 2,
3 R
at the surface of the asteroid. The value calculated above is therefore g =
−2
4 g0 = 7.36 cm s . Knowing the density of titanium, the (original) radius of
3
the asteroid can be determined as R ≈ 78 km.

S112 If we could calculate the work W necessary to pull the two halves
apart by d = 1 m, the attractive force F could be found from W = Fd. But
how can W be determined? Certainly not directly, as the product force ×
displacement, since that would lead us back to the original problem. Some
other method has to be found.
The gravitational potential energy of the system increases when the hemi-
spheres are pulled apart. W has the same value as this increase. The increase
can be calculated by determining the work done by the little green people
when carrying to the surface the titanium originally in the disc of radius
R and thickness d. The extracted metal can be considered as having been
carried to the surface by means of frictionless devices and then spread evenly
over the asteroid’s surface.
∆x
x
R
d
Solutions 177
If g denotes the gravitational acceleration at the surface of the asteroid

of mass M, then the gravitational acceleration at a distance x from the
centre is g(x) = xg/R. The density of the asteroid, and hence of titanium,
is ρ = 3M/(4πR 3 ). Consider the metal, of volume ∆V = d 2πx ∆x and mass
∆m = ρ∆V , which was originally situated between radii x and x + ∆x (see
figure). The metal has all to be brought up from the same depth and can,
therefore, be considered together when calculating the work done. A force,
∆mgx/R acts on this metal at its original position, whilst at the surface, the
corresponding force is ∆mg. Since the gravitational field increases uniformly
when moving from inside the asteroid to its surface, the arithmetic mean of
the initial and final forces can be used. The total displacement is R − x, and
the work done is therefore
1 + (x/R) R 2 − x2 3 Mgd 2
∆W = ∆mg (R − x) = ∆mg = (R − x2 )x ∆x.
2 2R 4 R4
The total work involved is the sum of that for bringing the titanium up from
the different depths, explicitly
3Mgd
W = ∆W = (R 2 − x2 )x ∆x.
4R 4
In the limit of layers of vanishingly small thickness, the above sum becomes
an integral with value
R R4
(R 2 − x2 )x ∆x ⇒ (R 2 − x2 )x dx = .
0 4
Therefore the total work done is W = 3
16 Mgd.
Note. The same result can be obtained without using integral calculus.
Introduce a new variable u = x2 /R 2 , instead of x, (0 ≤ u ≤ 1), and replace
the summation over the layers at different depths by a summation over
the terms corresponding to different values of u. Using the relationships
u = x2 /R 2 and u + ∆u = (x + ∆x)2 /R 2 (and neglecting terms containing the
square of the small quantity ∆x), the work done can be expressed as
3
W = Mgd (1 − u) ∆u.
8
This sum can be readily evaluated since 1 − u changes uniformly from 1 to
0, and can therefore be taken to be 12 on average, whilst the sum of the
terms in ∆u is 1. In the end, the previous expression for the work done is
obtained.
A ‘mythical giant’ does work W = 16 3

Mgd when separating the hemi-
spheres by d, and the force required to do this is therefore F = 3Mg/16.
This is the force of attraction between the hemispheres and is the force the
props should have borne. Knowing the radius of the asteroid and the density
of titanium, this force can be determined numerically as
2
3 GM 3 4πR 3 ρ G GR 4 ρ2 π2
F= M 2 = = ≈ 4.5 × 1013 N.
16 R 16 3 R2 3
In order to get a feel for the order of magnitude of this force, we calculate the
average pressure exerted on unit surface area as p = F/(πR 2 ) = 1.4 × 105 Pa
(i.e. one and a half times the atmospheric pressure on Earth). This is the
weight, on Earth, of 14 tons of matter; sufficiently strong props should bear
such a load.
Note. The result for the total force can also be obtained in the form
of a double integral using a pure calculus approach. Taking a volume
element dV = 2π r sin θ r dθ dr in the right-hand hemisphere with the axis
of symmetry of the same hemisphere as the polar axis, the force dF acting
on it, and directed towards the centre of the sphere, is ρ dV G (Mr3 /R 3 )/r2 .
The component of dF towards the left is dF cos θ, and the total force pulling
the right hemisphere to the left is the double integral of this: dr from 0 to
R; dθ from 0 to π/2. Then, using GM/R 2 = g and ρ = 3M/(4πR 3) gives
the same answer as above.
S113 At the surface of the charged sphere, whether it consists of a single

piece or two pieces close together, the electric field strength is
1 Q
E= .
4πε0 R 2
The electric charge per unit surface area is
Q
σ= .
4πR 2
This electric field exerts a force ∆F = 12 E∆Q on the charge ∆Q = σ∆A which
resides on a surface area ∆A, as illustrated in the figure. The reason for the
factor of 12 is that the electric field strength is E at the outer surface of the
sphere and zero inside; its average value is therefore E/2.
Solutions 179
The force per unit area exerted by the charges on the pieces of the sphere
is therefore
∆F Q2
= = p.
∆A 32 π2 ε0 R 4
The required force can be compared with the force with which a liquid at
pressure p would push apart the two pieces of the sphere. As this force is
also the product of p and the cross-sectional area of the intersection of the
plane and sphere, i.e. pπ(R 2 − h2 ), it follows that the two parts of the sphere
can be held together by a force
Q2
F= (R 2 − h2 ).
32 πε0 R 4
S114 Using the notation in the figure, the equilibrium condition for the
first ball is
mg '
= ,
F x
where
F = kqQ/x2
is the Coulomb force acting on the first ball and x is the distance between
the balls carrying charges q and Q.
C
F
F
qQ
D q k
x2
h x B
E
A
Q
mg
It is clear that the triangles ABD and CAE are similar, and that conse-
quently
x
: ' = h : x.
2
From the three equations above we can calculate the separation of the
charges and the electrostatic energy of the system:
qQ qQ
x=k and Eelectro = k = 2mgh.
2mgh x
The work done is the sum of the changes in electrostatic and gravitational
potential energy,
W = 2mgh + mgh = 3mgh.
It is perhaps surprising that the work done does not depend on either the
magnitudes of the charges or the length of the thread.
S115 Hydrogen of pressure p is enclosed in a spherical container of
radius R with walls of thickness d. Let ρ be the density of the wall material
and σ its tensile strength, as illustrated in the figure.
air
p
0
p
d
R
We first calculate the maximum pressure the container can sustain without
bursting. If the container were cut in two, the gas would push the two pieces
apart with a force of pπR 2 (the external pressure is negligible compared
with p). This force must not exceed the product of the surface area of the
cut, 2πRd, and the tensile strength, i.e.
p πR 2 < 2πR σ d. (1)
If the gas is released into a large balloon, its pressure decreases to p0 , the
external atmospheric pressure. Its final volume is therefore
4πR 3 p
V = .
3 p0
This volume of gas certainly cannot lift a load any greater than V ρair g (where
ρair is the density of air), as the weight of the balloon and the hydrogen have
still to be subtracted from the upthrust.
The weight of the container is 4πR 2 dρg. In accordance with the above
reasoning, this cannot be greater than the force available to lift it, i.e.
4πR 3 p
4πR 2 dρg < ρair g. (2)
3 p0
Inequalities (1) and (2) give a relationship between the properties of the
Solutions 181
container material and the pressure and density of the external air,
σ 3p0
> . (3)
ρ 2ρair
It is interesting that neither the radius, nor the thickness of the walls of
the container occur in (3)! (A thicker-walled container can sustain a higher
pressure but has a greater weight.) Despite leafing through several tables of
physical properties, we could find no material that would fulfil condition
(3). None of the materials known today are strong enough, relative to their
densities, to be lifted by the upthrust available from releasing a gas stored
inside them.
S116 Gravitational acceleration at the surface of the spherical Earth, of
radius R, mass M and density ρ, can be written as
M 4πR 3 ρ 4π
g=G = G = GRρ.
R2 3R 2 3
In order to solve the given problem, we have to find the magnitude of the
gravitational acceleration on the surface of a very large disc of thickness H
and density ρ, at a point far from the edge of the disc. The result is relatively
easy to obtain using an analogy between the two sets of laws governing
electrostatic and gravitational interactions.
We draw analogies between a mass m (the ‘gravitational charge’) and
an electric charge q, the gravitational constant G and 1/(4πε0 ), and the
gravitational acceleration g = F/m and the electric field strength E = F/q.
In both cases, F is the force experienced by the ‘test charge’. We next
determine the electric field strength (outside the disc) of an infinitely large
disc carrying a homogeneous charge distribution, and then substitute the
analogous quantities to obtain the gravitational acceleration for a mass
distribution of similar geometry.
E
A
H
The electrostatic field strength can be calculated by applying Gauss’s

theorem to a disc of area A,
1
ΦE = q,
ε0
where ΦE = 2AE is the electric flux (see figure). If the electric charge density
is ρq then the total charge surrounded by the closed surface is q = ρq AH.

Therefore
1
2EA = ρq AH,
ε0
which yields
1 ρq H
E=
ε0 2
for the electric field strength. Substituting the relevant analogues, the gravi-
tational acceleration is found to be g = 2πGρH. This has to be equal to the
gravitational acceleration measured at the surface of the Earth, i.e.
4πRρ
G = gsphere = gdisc = 2πGρH,
3
which we can write as
2 2
H = R = 6370 km = 4250 km
3 3
for the thickness of the ‘flat Earth’.
S117 It follows from the similarity of triangles FHC and GKH in
Fig. S117.1, that for the short length of rod KH determined by the small
angle ∆α, the following equalities hold:
r r r2
∆x = GH = r∆α = ∆α.
h h h
C
∆a
h
r
K H F
∆x
Fig. S117.1
There is a charge ∆Q = σ∆x on KH, where σ is the charge per unit
length, and the magnitude of the electric field strength corresponding to this
charge is
1 ∆Q 1 σ
∆E = = ∆α.
4πε0 r 2 4πε0 h
This quantity is independent of the value of angle α itself. It only depends
on the angle ∆α which the piece of rod subtends at C. Thus the electric field
Solutions 183
vectors due to the lengths of charged rod situated symmetrically with respect
to the bisector of angle C are of the same magnitude and their resultant
points in the direction of the bisector (see Fig. S117.2).
∆ ∆
A B
Fig. S117.2
For any given point C there is some particular value of α such that
AĈB = 2α, and by superimposing the results for matched pairs of elementary
lengths of the rod, the stated result is established.
S118 Using the result of the previous problem, it can be stated that
the direction of the electric field at a point on the plane, and a distance
h from the end of the infinitely long rod, makes an angle of 45◦ with the
rod.
The magnitude E of the electric field strength can be found using the
following ‘trick’. Imagine two very long, uniformly charged rods joined end
to end. The resultant field strength will be the vector sum of the field
strengths of the two ‘half-rods’ (see Fig. S118.1).
E E
o o
45 45
Fig. S118.1
The direction of the resultant will obviously √be perpendicular to the rod –
in view of the symmetry – and its magnitude, 2 times the field strength E
for an individual rod, can be found using Gauss’s law for electric field lines.
Enclose a section of length ' of the infinitely long rod in a notional cylinder
of radius h (see Fig. S118.2).
2E
Fig. S118.2
There is a charge Q = 'σ inside the cylinder, where σ is the charge per
unit length, and the number of field lines crossing the cylinder (the electric
flux) is
√
ΦE = 2E 2πh'.
According to Gauss’s law,

√
2 σ
ε0 ΦE = Q, giving E = .
4πε0 h
S119 (i) When the angular opening approaches 2π, point P is in between
the wires and arbitrarily close to them, and so both halves of the current-
carrying wire produce very large magnetic fields at P , and in the same
direction. Thus in this case the magnitude of the net magnetic field at P
approaches infinity. As tan(θ/2) also approaches infinity when θ approaches
π, Ampère’s formula may be correct. However, the expression given by Biot
and Savart must be wrong, because it gives a finite value for B(P ). In fact,
Ampère’s result was later embodied in Maxwell’s electromagnetic theory,
and is now universally accepted.
(ii) When the angular opening is 2θ = π, the ‘V’ becomes a straight
infinite wire. For this case, the magnitude of the field B(P ) is known to be
B = µ0 I/(2πd). Since tan(θ/2) = tan(π/4) = 1, the proportionality factor in
Ampère’s formula is µ0 I/(2πd).
Biot and Savart chose their formula in such a way that it agreed with the
expression for the magnetic field due to a straight infinite current-carrying
wire already generally accepted. Thus they had as their proportionality factor
µ0 I/(π2 d).
Note. In the region θ < π/2 the difference between the two predictions
is relatively small. The ratio of the predicted values for B, 2θ/π tan(θ/2),
shows the greatest difference from unity when θ → 0 and has a maximum
value of 4/π.
Solutions 185
S120 (i) If the magnetic field strength at point P is denoted by B1 then

a symmetrically placed second coil, as shown in Fig. S120.1, would also
produce a field of strength B1 .
B1 B1 B0
Fig. S120.1
Since the resultant magnitude of the superimposed fields is clearly B0 , it
follows that B1 = B0 /2.
(ii) Similar reasoning shows that the horizontal component of the magnetic
field vector through P is B0 /2; this would be true for any P whose distance
from the axis is less than R (Fig. S120.2).
B1 P
Fig. S120.2
Therefore a total magnetic flux of πR 2 (B0 /2) crosses the end of the
solenoid. This is exactly half of the flux inside the solenoid; what happens
to the other half?
(iii) A qualitative sketch of the field lines can be seen in Fig. S120.3.
R
2
Fig. S120.3
The field line crossing the endmost turn of the solenoid (point Q) continues
travelling perpendicularly to the solenoid. Half of the flux travels to the left
of this, the rest leaves the coil between the turns. Deep inside the solenoid,
the√field lines on the boundary between these two halves are at a distance
R/ 2 from its axis.
S121 In theory it is possible to calculate the force between the plates
by dividing the charges on the plates into many small point charges and
summing the Coulomb forces of all point-charge pairs. If we imagine the
positive charges on one of the plates changed into negative ones of the same
magnitude, then the magnitudes of the Coulomb forces remain the same, but
their signs are reversed. Instead of repulsion, we get attraction between the
plates, as for parallel plate capacitors. The energy stored in such a capacitor
with plate separation x is E = Q2 /2C, where the capacitance C is given by
C = ε0 A/x. Hence E = Q2 x/2ε0 A, and the force F is the rate at which this
changes as x is changed, i.e. F = dE/dx. Hence the force required to hold
together plates of area A is
Q2
F= .
2ε0 A
S122 The net charges on the plates cannot change, but the charges on
the plates on either side of any of the spaces must be equal and oppo-
site. Consequently, the charges on C and D must be −1 nC and +1 nC,
respectively, on their outside surfaces and +2 nC and −2 nC, respectively,
on their inside surfaces. The capacitance of any pair of plates is inversely
proportional to their separation, with 5 mm corresponding to 20 pF.
2 mm
x 3−x
A C D B
Thus if AC is x mm and DB is (3 − x) mm, the capacitances of the three

successive capacitors are 100/x, 50 and 100/(3 − x) pF. The voltage VCD is
therefore 40 V, and VAB is 10x + 40 + 10(3 − x) = 70 V, independent of the
value of x.
Solutions 187
S123 The charge between the capacitor plates could be notionally divided
into two parts, which are then moved away from each other in a direction
parallel to the plates. The (induced) charges on the capacitor plates would
also move but their totals on each plate would remain unaltered (using the
principle of superposition). Continuing in the same manner, the charge Q
could be further subdivided until it was ‘spread’ uniformly on a plane of the
same size as, and parallel to, the capacitor plates.
We can now consider two plane capacitors connected in parallel, the
distances between their plates being x and d − x. Since one plate of each
of the capacitors is earthed, and the other is common, the voltages between
the pairs of plates are identical. Their electric field strengths are therefore
inversely proportional to their plate separations, i.e. E1 /E2 = (d − x)/x. The
electric flux emanating from the charge Q is divided in the same proportion
(Gauss’s law), and the ratio of the charges on the earthed plates is the same
as well. Since a total charge of −Q accumulates on the plates, respective
charges of
d−x x
Q1 = −Q and Q2 = −Q
d d
accumulate on the two separate plates.
S124 Very far from the capacitor the electric field is determined by the
total electric charge of the system. The electric field outside the capacitor
is zero; therefore the total electric charge (the sum of the charges on the
plates) must also be zero. That means Q1 = −Q2 = Q.
The long distance behaviour of the electric field of an electrically neutral
system is determined by its total electric dipole. In our case, taking the
component of this dipole moment normal to the plates gives
p − Qd = 0,
which gives the charges on the plates as
|p|
Q= .
d
Thus the charges on the plates do not depend upon the position of the
dipole.
Note. (i) The same method can also be applied in P123. The relevant
equations are:
Q1 + Q2 + Q = 0,
Q1 x − Q2 (d − x) = 0,
and their solution is
Q1 = −Q (d − x)/d, Q2 = −Q x/d.
(ii) Alternatively, we can use the result from the case of a single charge
by placing two charges of opposite signs into the parallel plate capacitor
and applying the superposition principle. This is a simple way to find the
solution of the general problem, in which the dipole momentum is not
necessarily perpendicular to the plates.
S125 Figure S125.1 shows a light ray passing through succesive plano-
parallel plates of different refractive indices.
a1 n1
a2
a2 n2
a3
a3 n 3
a4 n4
Fig. S125.1
According to Snell’s law,

sin α2 n1 sin α3 n2
= , = , ···
sin α1 n2 sin α2 n3
It can be seen that the product of the sine of the angle of incidence and the
absolute refractive index has the same value at all interfaces, i.e. n sin α is
constant along the light ray’s trajectory.
R
R−
a
a
x
Fig. S125.2
Solutions 189
This relationship is also valid for a medium whose refractive index con-
tinuously changes in one direction, since the medium can be considered as
consisting of thin plano-parallel plates. Place the origin of the coordinate
system at the point where the light ray enters the medium. In this case,
the angle of incidence for the first ‘plate’ starting at y = 0 is 90◦ and the
refractive index is n0 , which gives the above constant as n(y) sin α = n0 .
The light travels along a circular arc of radius R and we first examine its
relationship to coordinate y. From Fig. S125.2 it is clear that
R−y
n0 = n sin α = n(y) .
R
This gives the space-dependence of the refractive index as
R
n(y) = n0 .
R−y
The material with the greatest known refractive index is diamond, but even
the refractive index of this material does not reach the value nmax = 2.5. It
is this limit that sets the maximum angular size of the arc the light ray can
cover. If the refractive index changes from n0 = 1 to nmax = 2.5 then the
maximum value of y is 35 R, corresponding to an arc of angular size 66.4◦ .
In practice, it is difficult to constrain light to a circular arc. However, it
is possible to make up solutions in which the concentration of solute, and
therefore the refractive index, shows a continuous vertical change. For such
a medium, the light ray does not propagate along a circular arc but follows
some other continuous curve.
Note. Why does the light ray entering along the x-axis start to bend at
all? The reason is that there is no such thing as an infinitely thin light
ray; a ‘ray’ always implies a beam of finite width, with the refractive index
and, hence, the speed of propagation varying across the beam profile. As a
consequence, the wavefront becomes non-planar and the beam bends.
S126 Using a ruler you can check that on a CD the inner diameter of
the area for storing information is approximately 4.4 cm, whilst the outer
diameter is approximately 11 cm. This means that the useful surface area
is about 80 cm2 . As the unit of information is 1 bit (1 byte = 8 bits) the

surface area per bit is
80 cm2
A= = 1.54 × 10−8 cm2 .
650 × 106 × 8
Assuming that the ‘shape’ of a single unit of information is a square (A = a2 ),

the linear size of a bit is a = 1.24 × 10−4 cm = 1.24 m.
The information on a CD is stored in a very long spiral starting near
the centre and working outwards; this is the reverse of what happens on
traditional LPs. A small piece of a CD can be used as a reflection grating
and you can measure its diffraction pattern using a laser beam of known
wavelength. The effective grating spacing is the width of the grooves and it
is simplest to use normal incidence and measure the distance between the
two first-order interference maxima on a screen. The typical wavelength of
an optical laser is 670 nm and it is convenient to choose 1 m as the distance
between the CD and the screen.
The condition for the first-order intensity maximum in the interference
pattern of a grating is d sin θ1 = λ, where d is the grating spacing. Experi-
mentally, θ1 ≈ 25◦ , and inserting the data into the grating equation gives the
grating spacing as d ≈ 1570 nm = 1.57 m. Although close, this estimate
is somewhat larger than our previous one. The difference between the two
results is not a measurement error, but a consequence of the separating
walls between neighbouring grooves. The effective width of the information
is only about 0.5 m. This means that the total effective area of information
is only some (80/3) cm2 ≈ 27 cm2 , and that 1 bit has a rectangular shape
approximately 1 m long.
Note. The information density on CDs is uniform, but the rotation rate
changes according to the position of the reading head.
S127 nλ = d sin θ = (10−3 m/300) sin 24.46◦ = 1380 nm and the only
possible values for n and λ to put the red and blue/violet light into the
appropriate parts of the spectrum are nR = 2, λR = 690 nm and nBV = 3,
λBV = 460 nm. In all physically possible cases
nλ ≤ d sin 90◦ = 3333 nm,
and the only other pair of integers which are in the ratio 3m : 2m, with m
less than (3333/1380) = 2.4, is 6 and 4. Thus there is only one more angle
Solutions 191
at which a two-component line will be observed; i.e. at

(6 × 460)
sin−1 = 55.9◦ .
3333
S128 (i) When the monochromatic laser beam falls on the diffraction
grating at normal incidence, the positions θ of maxima of the interference
pattern are given by
d sin θ = mλ, (m = 0, ±1, ±2, ±3, . . .),
where d is the grating spacing and λ is the wavelength of the laser beam.
If the grating is rotated through an angle φ around an axis parallel to the
lines of the grating, we have to modify the above equation. It is enough
to consider the interference from only two slits of the grating, as shown in
Fig. S128.1.
h –u
u
d h
Fig. S128.1
The optical path difference consists of two parts:
∆1 + ∆2 = d sin φ + d sin(θ − φ).
Thus the modified equation for the interference pattern is
d[sin φ + sin(θ − φ)] = mλ.
The principal result is that the pattern becomes asymmetrical, with only the
position of the zeroth-order maximum remaining unchanged. If φ is counted
as positive when the grating is rotated anticlockwise (and θ is counted as
positive in the same sense), then the density of interference maxima becomes
larger at positive angles, whilst it decreases for negative ones. If the lines of
the grating are vertical, all the interference maxima lie along a horizontal
straight line; this is to be contrasted with what happens in (ii).
Note. The naı̈ve idea that the effect of rotating the grating is to decrease the
‘effective’ size of the grating spacing is false.
(ii) Since the diffraction grating is an array of a large number of identical

parallel slits, as a first step it is enough to investigate the diffraction from
a single slit. If the beam falls on a very narrow slit at normal incidence,
the diffraction pattern is a weak straight line. If the slit is a little wider,
interference minima along this line can be identified.
When the slit is tilted ‘forward’ (i.e. rotated around a horizontal axis which
is perpendicular to the slit as well as to the direct beam), the zeroth-order
maximum remains unchanged. Though the incoming beam strikes different
parts of the slit with different phases, beyond the slit, in the direction of the
direct beam, there are no phase differences (Fig. S128.2) and no change in
the pattern results.
c
Laser
beam
u
Fig. S128.2
It may be easier to consider the slit as a series of closely spaced very small
holes. Now, we need only recognise that, if the wavelets originating from the
holes produce constructive interference in the direction of the direct beam,
then the same must be true for all directions (in three dimensions), which
make the same angle with the direction of the (rotated) slit. If the slit is
tilted by an angle φ, then the angle between the slit and the direct beam is
γ = 90◦ − φ. The same angle γ between the slit and a diffracted ray occurs
for any ray that lies on the cone which has semi-angle γ and the direction of
the slit as its axis. On the screen we would see a plane section of this cone
(see Fig. S128.3).
Direction of
the rotated
grating
c
Laser beam c Conic section
Rotated grating Direction of
direct beam
Fig. S128.3
Solutions 193
The apex of the cone is the midpoint of the slit, and its semi-angle is
γ. The conic section can be ellipse, parabola or hyperbola. A parabola is
obtained if γ = φ = 45◦ .
Returning to the original problem, we conclude that the interference
pattern of the tilted grating consists of bright spots lying on a conic section.
S129 Because of the surface tension of the liquid, its height between the
two objects is not the same as it is outside the objects; in the case of (a)
water it is higher, whilst for (b) mercury (which has a negative angle of
contact) it is lower.
p0
p0 p0
p0
h h
p0 p0
p0 − g p0 + g
(a) Water (b) Mercury
Just above the liquid surface between the objects the pressure has to be
atmospheric in both cases and, correspondingly, just below the surface it
has to be less than atmospheric for water and more than atmospheric for
mercury. As can be seen from the figure this leads to net inward forces
(acting on the shaded areas) in both cases and a tendency for the objects to
move towards each other.
S130 The pressure of the water changes linearly with the increase in
height. At the bottom of the meniscus it is equal to the external atmospheric
pressure p0 , and at the top to p0 − ρgh. The average pressure exerted on the
wall is paverage = p0 − ρgh/2. The force corresponding to this value, for an
aquarium with side walls of length ', is F1 = 'paverage h.
F
F h 2
1
F
3
Consider the horizontal forces acting on the volume of water enclosed by

the dashed lines in the figure. The wall pushes it to the right with force F1 ,
the external air pushes it to the left with force F2 = 'p0 h, and the surface
tension of the rest of the water pulls it to the right with a force F3 = 'γ.
The resultant of these forces has to be zero, since the volume itself is at rest.
This means that
1
p0 − ρgh 'h − p0 'h + 'γ = 0,
2
2γ 2 × 0.073
h= = = 0.0038 m.
ρg 1000 × 10
Water rises by approximately 4 mm up the wall of the aquarium.

S131 When the radius R of a drop of water with surface tension γ (also
equal to the energy per unit surface area) decreases by ∆R, the energy of the
surface decreases. It changes by
∆Esurface = 4πγ R 2 − (R − ∆R)2 ≈ −8πRγ∆R.
At the same time, the volume of the drop decreases by 4πR 2 ∆R. For this
quantity of water to evaporate, energy
∆Eevaporation = 4πLρR 2 ∆R
has to be supplied. Here ρ is the density of water and L its latent heat of
evaporation.
The decrease in the surface energy could provide the evaporation energy
of the drop if |∆Esurface | > ∆Eevaporation , i.e. if
2γ
R< ≈ 7 × 10−11 m.
ρL
Since this radius is of the same order of magnitude as the size of one water
molecule, a drop of water with this radius cannot exist. Therefore, there is
no water drop that can evaporate without absorbing heat, or losing internal
energy. However, the above reasoning can be used to estimate molecular
sizes using macroscopic properties.
S132 Consider a closed vessel containing a volume of liquid, with satu-
rated vapour of this liquid filling the rest of the vessel. As illustrated in the
figure, let a capillary tube of radius r be immersed in the liquid, which does
not wet its walls.
Solutions 195
. . . . . .
. . . . .
.
. . . .
. . . .
. . .
. . . v .
. . .
. . . .
. . . . . . . . . . .
.
. h
l
In such circumstances, the level in the capillary tube falls to a depth

h below the liquid surface. The magnitude of h can be found using the
equilibrium relationship between hydrostatic force and the surface tension
of the liquid 2πrγ = ρl gh πr 2 , where γ is the surface tension of the liquid and
ρl is its density. This gives h = 2γ/(ρl gr).
The liquid is in equilibrium with its saturated vapour, both in the capillary
tube and at the plane surface of the liquid. In the capillary tube, however,
the pressure of the vapour is a little higher at the interface. The difference
is caused by the pressure of the vapour column, of height h and density ρv ,
above it. It follows that
2γ ρv
∆pv = ρv gh = .
r ρl
As ρv ρl , this difference between the pressures is much smaller than the
pressure of curvature corresponding to the radius r, but, given long enough,
it is sufficient to bring about the phenomenon described.
S133 Let A denote the cross-sectional area of the piston and y the vertical
displacement between its initial and final equilibrium positions (see figure).
h A
The decrease in potential energy of the weight W increases the internal

energy of the air inside the cylinder. Conservation of energy between the
initial and the final states gives

5
Wy = [p1 A(h − y) + p2 A(h + y) − 2p0 Ah] , (1a)
2
where p1 is the final pressure in the lower part of the cylinder and p2 that in
the upper part. The internal energy of a gas made up of diatomic molecules
has been written in the form 52 pV . If W is very large, the decrease in its
potential energy (and the corresponding increase in the internal energies of
the gases) is very large, and the initial internal energy of the air can be
neglected. Thus
5
Wy = [p1 A(h − y) + p2 A(h + y)] . (1b)
2
When the load is finally at rest,
(p1 − p2 )A = W . (2)
The temperatures and the masses of the gases in the two halves are identical,
and so their internal energies must be equal:
5 5
p1 A(h − y) = p2 A(h + y). (3)
2 2
Equations (1b), (2) and (3) yield y = 5/7h for the displacement of the
piston, i.e. the gas in the lower part is compressed to 1 − 5/7 ≈ 15 per cent
of its original volume.
Note. The surprising result is that the volume of air in the lower part does
not tend to zero, however large the weight is, even though gases are supposed
to be compressible! The large load increases the internal energy, and hence
the temperature, of the enclosed gas. This causes considerable increases in
not only the absolute pressures, but also in the difference between the upper
and lower pressures.
In practice, the applied weight is limited by the mechanical load-bearing
capacity of the structure and the melting points of the materials used. We
should also consider whether the increased temperature is one at which air
can still be treated as an ideal diatomic gas.
If the initial internal energy of the air is not neglected then the result is
y 35W 2 + 25p20 A2 − 5p0 A

= .
h 7W
S134 Consider a ‘box-shaped’ mountain of average density ρ, base area
A and height h. In order to melt its bottom layer of thickness d and
specific latent heat L, energy AdρL would be required. The total mass of the
mountain is approximately Ahρ, and the energy released if it sank a distance
Solutions 197
d would therefore be Ahρgd. The base of the mountain does not melt under
its load if
L
AdρL > Ahρgd, i.e. h < .
g
Approximating the required latent heat by the latent heat of melting of
metals (200 − 300 kJ kg−1 ), we estimate the maximum possible height of
mountains on Earth to be 20–30 km. This is of the right order of magnitude.
Allowing for the fact that the base of the mountain does not actually have
to melt, but rather that the size of mountains is limited by the yield strength
of their constituent materials, the estimated height of the highest mountains
on Earth is surprisingly accurate.
Gravitational acceleration is significantly smaller on Mars than on Earth
(gMars ≈ 4 m s−2 ). Therefore mountains, consisting of similar rocks, could
be higher on Mars than on Earth. Indeed, the highest mountain on Mars,
Mons Olympus, is 26 km high!
S135 In order to simplify the calculation, choose a system of units in
which the initial volume and the external pressure are unity, and the units
of the number of moles and the gas constant have a product that is also
unity. This reduces the usual ideal gas equation pV = nRT to pV = T . In
this system, the molar heat has to be multiplied by the gas constant to yield
the normal molar heat capacity.
The initial pressure of the air is given by the sum of the pressure of
the 76 cm-high column of mercury and the atmospheric pressure, a total
of 2 units. Any increase in length of the air column implies a corre-
sponding decrease in the mercury one. Therefore as the air expands from
1 to 2 units, its pressure decreases linearly from 2 to 1 units as shown in
Fig. S135.1.
p T = 9/4
Adiabatic
2
T= 2
V
1 2
Fig. S135.1
The temperature of the enclosed air is initially 2 units, and is still 2 units
at the end of the straight section shown in the figure, since both points
lie on the isothermal pV = 2. In our representation, the isothermals are
symmetrical about the bisector of the axes (i.e. about the line p = V ). Thus
the highest temperature reached corresponds to the isothermal to which the
straight-line segment is tangential. As shown in Fig. S135.1, this occurs at the
middle of the process, when p = 32 and V = 32 . The maximum temperature
is therefore 94 .
The equation of the straight line representing the process is p = 3 − V .
Applying the first law of thermodynamics to a typical section of the line gives
5
∆T = C ∆T − p ∆V ,
2
where the left-hand side of the equation is the change in the internal energy
of the air (diatomic molecules have, CV = 52 R), and on the right-hand side,
C is the molar heat in question. Expressing p as p = 3 − V in the equation
pV = T , gives (3 − V )V = T . From this it follows that small changes in
V and T are connected by (3 − 2V )∆V = ∆T . The change in the internal
energy of the air then becomes
5 3−V
∆T = C ∆T − ∆T ,
2 3 − 2V
which simplifies to
21 − 12V
C= .
6 − 4V
V
1 2
−3
Fig. S135.2
Figure S135.2 shows this result for the molar heat plotted against volume.
The curve is singular (the molar heat approaches infinity) at V = 3/2 because
Solutions 199
at this point the straight line is a good approximation to the isothermal

corresponding to the maximum temperature. During an isothermal process,
heat is transferred but the temperature does not change, and this is why
the molar heat tends to infinity. Beyond this point, the molar heat becomes
negative, which means that in spite of the positive heat transfer, the internal
energy of the air decreases because the work done by it in expanding is
greater than the heat transferred.
The most interesting part of the process occurs when V = 74 , i.e. when
only a quarter of the mercury is left in the tube. At this point the molar
heat is zero and implies that the process is adiabatic, since there is a change
of temperature without any heat transfer; at this point, the straight line
describing the process is tangential to an adiabatic curve. Beyond this point,
the temperature decreases further but the molar heat becomes positive; this
can only correspond to the system being cooled.
If more heat is given to the system, the representative point no longer
follows the straight line but continues along the adiabatic one. The tem-
perature of the air does decrease, but not by as much as the straight line
would suggest. The decrease in the internal energy of the air is equal to the
work done against atmospheric pressure (whilst lifting the mercury) and in
accelerating the ejected mercury.
Returning to the original question, that of the heat transfer necessary to
push the mercury out of the tube, it can be seen that it is only necessary to
transfer heat whilst V is in the range V = 1 to V = 74 . Simple arithmetic
shows that the work done by the air up to this point is (in this system
of units) 39 15
32 , whilst the increase in internal energy is 32 . The required heat
transfer is therefore Q = 2716 . In standard units, this means that if the column
of mercury originally enclosed n moles of air, then the heat transfer required
16 nR.
to remove it is 27
S136 The molten magma came into contact with the ice, and huge
volumes were melted at the base of the 500-m-thick ice cap. As the density
of water is greater than that of ice (i.e. the volume of meltwater is smaller
than that of the ice from which it was formed), and moreover some part of
the water could flow away, a huge conical hole would have formed under the
ice. But the extremely heavy ice above the hole sank, leaving the depression at
the surface. Under the ice crater we could have found the recently solidified
magma intrusion, the meltwater in a conical cavity and the ice cover (see
figure). The amount of meltwater depends on the quantity of magma, but
the shape of the ice crater is determined by the hydrostatic pressure of the
ice and water.
Crater
Ice Ice Ice
Water
Magma
Ground
The eruption broke through the ice cap on the second day and a hot,
black ash-cloud, 500 m high, was formed. This was carried up to an altitude
of 3000 m by the buoyancy force of the cold air. At the end of two weeks
the cloud column had become white and reached a height of 10 km.
Altogether the eruption melted 3 km3 of ice in two weeks. The meltwater
flowed under the ice of the glacier into a lake situated within a nearby
volcanic depression, the Grimsvotn caldera. A deep depression, 8 km long
and a few hundred metres wide was formed in the ice surface at the eruption
site. The rate of melting was extremely high, 0.5 km3 per day for the first four
days. At the same time a new mountain, 0.6 km long and 150–300 metres
high, was built under the ice by the eruption.
The meltwater was held in Grimsvotn caldera under the glacier for five
weeks, before it escaped. A gigantic wave swept across part of the south-east
lowlands, destroying everything (roads, bridges, etc.) in its path.
S137 Assume that the cavity is located at a depth h below the surface.
When the flue is filled with water, the pressure in the cavity is p = p0 + ρgh,
where p0 is the atmospheric pressure and ρ is the density of water. Boiling
starts at that depth when the temperature is such that the pressure of
saturated water vapour (Ae−Lm /(RT ) ) is equal to p. The value of the molar heat
of vaporisation Lm occurring in the formula is approximately 40 kJ mol−1
and that of the molar gas constant R is 8.3 J mol−1 K−1 . Using these data,
the constant A can be determined from the boiling point of water (373 K)
at atmospheric pressure; the result is A = 4.1 × 1010 Pa.
For the purposes of the calculations, the temperature of the ground at the
surface can be assumed to be T0 = 290 K and, as it increases by one degree
per metre, the temperature at depth h is given by T = T0 + h. The following
equation can now be formulated, using the equality of the pressures when
Solutions 201
the water starts to boil:

− R(TLm+h)
p0 + ρgh = Ae 0 .
This transcendental equation can be solved by numerical methods, i.e. by
carefully considered guesses! The solution is h = 197(≈ 200) m. The pressure
at such a depth is approximately 20 atm and the temperature 487 K = 214 ◦ C.
When boiling starts, the first steam bubbles rise and push the water out of
the flue, the pressure decreases to 1 atm and the superheated water at 214 ◦ C
starts to boil violently, producing as much steam as is necessary to reduce
its temperature to 100 ◦ C. The temperature drop of ∆T = 114 ◦ C implies an
excess internal energy of cm∆T , where c is the specific heat of water and m
is the mass of water in the cavity. This surplus energy is sufficient to produce
ms = 44 tons of steam, implying
cm∆T = Lms ,
where L is the heat of vaporisation of water. After substituting the relevant
data, we find the mass of the water in the cavity to be m = 207 tons. Since
the density of water is only around 850 kg m−3 at such a high temperature,
this gives an approximate value of 244 m3 for the volume of the cavity.
S138 The heat conducted away in a given time must equal (minus) the
latent heat of fusion of the additional ice formed in that time. Thus for an
area A of the lake
∆T d
λi A = Li (ρi Ax).
x dt
Simple integration gives that t = 12 x2 B, where B = (ρi Li )/(λi ∆T ). Inserting
the relevant numerical values gives the time as about 90 hours.
S139 According to the law of heat conduction the heat transferred is
directly proportional to the thermal gradient (temperature difference divided
by distance), the area and the time. So
Heat capacity L3
Time ∝ ∝ 2 −1 = L2 .
Area × Temperature gradient L L
Since M ∝ L3 , the time is proportional to M 2/3 and a mammoth should take
2(8000/5)2/3 days, i.e. about nine months, assuming that the turkey and the
mammoth both start to thaw from the same temperature and are defrosting
in similar environments. (In fact, the Siberian summer is too short to defrost
a mammoth.)
S140 Consider first the state of the ice and the pressure in the container at
100 ◦ C. Since the density of saturated water vapour is 0.5977(≈ 0.6) kg m−3
and its pressure is 1 atm, the 0.6 kg of ice is completely transformed into
vapour at 100 ◦ C, indeed into saturated vapour at a pressure of 1 atm.
How does ice, at a temperature of −10 ◦ C, turn into saturated water
vapour at 100 ◦ C? If the temperature is increased very slowly, the system
passes through a number of equilibrium states. First, the ice sublimes and
the ice phase is in equilibrium with the vapour phase. This lasts until the
temperature and pressure of the triple point (0.01 ◦ C, 610 Pa) is reached.
At the triple point, a liquid state appears alongside the ice and water
vapour. Further heating makes the solid phase disappear, and only water
and saturated water vapour remain in the container. It is interesting to note
that subsequently the water boils steadily at 100 ◦ C, until all the water has
been transformed into vapour.
From the point of view of heat absorption, only the initial and the final
states are important. The heat Q absorbed by the system, as it passes with
increasing internal energy through the ‘ice–water–vapour’ states, is indepen-
dent of the intermediate states. For calculational purposes, the heating of
0.6 kg of ice should be divided into four stages (warming the ice, melting
the ice, warming the water and boiling the water) to give:
Q = ci m∆T1 + Lf m + cw m∆T2 + Lv m,
where ci = 2.1 kJ kg−1 ◦ C−1 is the specific heat of ice, ∆T1 = 10 ◦ C, Lf =

334 kJ kg−1 is the heat of fusion of ice, cw = 4.2 kJ kg−1 ◦ C−1 is the specific
heat of water, ∆T2 = 100 ◦ C and Lv is the specific heat of vaporisation of
water.
The specific heats of ice and water are slightly dependent on temperature
and pressure, and the heat of fusion of the ice also depends a little on
pressure. But these variations are small and can be neglected. However, the
situation is different for the heat of vaporisation of water. It is true that the
heat of vaporisation only varies a little with temperature, but its dependence
on pressure is significant.
The value usually given in tables, Lv = 2256 kJ kg−1 , covers, not only
the higher internal energy of the vapour, but also the work done against
atmospheric pressure. In the present problem, this work was done when
the container was evacuated, i.e. the heat to be transferred to the system is
smaller by this amount, which is −p∆V = −101.3 kJ. As this figure is that
for 0.6 kg of water,
101.3
Lv = 2256 − = 2087 ≈ 2090 kJ kg−1
0.6
Solutions 203
is the more accurate value to be used for the present calculation. After
substitution of the data, the heat transfer is found to be Q = 1720 kJ.
Neglecting the work done against atmospheric pressure would have intro-
duced an error of nearly 6 per cent, while ignoring the slight dependence
of the other coefficients on temperature and pressure causes an inaccuracy
of approximately 1 per cent. The main reason for this is that the normal
change in the volume in the course of a water–vapour transition is very
significant (a factor of about 1600). The volumes of the water and ice are
negligible compared with that of the vapour, and they are tacitly neglected
in the solution.
S141 Water continues to vapourise in the closed container until the space
above it is saturated. The total pressure above the liquid, therefore, is the
sum of the pressures of the saturated vapour and the air enclosed in the
container, i.e. it is always higher than the pressure of the vapour alone. This
means that the water cannot start boiling at any temperature.
Water
Ice
T Vapour
T
o o
365 C 374.2 C
If the container is strong enough, it can withstand a large rise of the

temperature. We can ask the question: up to what temperature could liquid
water still be found in the container? Data tables for saturated water vapour
tell us that the critical temperature of water is 374.2 ◦ C. Its critical density
is 326.2 kg m−3 , which means it is not possible for all the water to evaporate
before the critical point C is reached. If it did, the density of water in
the container would have to be around 500 kg m−3 . Indeed, a strange
change of state occurs at lower temperatures! The density of water is
500 kg m−3 at approximately 365 ◦ C, but the density of saturated vapour is
160 kg m−3 at this temperature. This means that at 365 ◦ C, liquid water fills
the whole container. The explanation for this is that, as the temperature rises
in this range, the density of water decreases more rapidly than the density
of water vapour increases.
At temperatures above 365 ◦ C, the water in the container remains in liquid
state, but its pressure increases. As shown in the figure, the system departs
from the boiling curve and continues above it, in the region of the phase
diagram corresponding to the liquid phase. The terms ‘liquid’ and ‘vapour’
lose their meaning above the critical temperature; rigorous application of
the terminology would imply that water is then in a gaseous state.
In what state is the air in the container? In the relevant temperature range
(360–380 ◦C), its pressure is around 200–300 atm. At such a pressure, part of
the air is dissolved in the water, and the rest is compressed into very small
bubbles filling only 0.1–0.2 per cent of the container. Thus, the air has no
noticeable effect on the behaviour of the water.
S142 Only the energy of the air molecules can be relevant and so T
appears in the combination kT . Dimensional analysis using ', F, m and kT
shows that the amplitude must depend on combinations of variables of the
form (kT /F)q '1−q , where q can take any value; m does not appear and so
the amplitude is independent of the mass of the cobweb.
S143 Let us compare the surface energy of a cylindrical water thread
on the cobweb with that of the periodic water drops formed from the
thread. Denote the initial radius of the water thread by r, the ‘wavelength’
(separation) of the drops by λ and the radius of the drops by R, all as shown
in the figure. We can ignore all energies (including gravitational) other than
the surface energy of γ per unit area.
2r
2R
Initially the surface energy of the cylindrical water thread of length L is

E1 = 2πrLγ.
Ultimately L/λ drops, each of radius R, are formed, and their surface energy
is given by
L
E2 = 4πR 2 γ,
λ
where we have ignored the thickness and surface area of the cobweb threads
themselves.
Solutions 205
The radius R is determined from the conservation of matter.

4πR 3 L
πr 2 L = .
3 λ
During drop formation the surface energy must decrease, E2 < E1 . Elimi-
nating R from these equations, we obtain λ > 92 r. This result shows that the
‘wavelength’ of the drops must be larger than a certain critical value λcrit ,
which is proportional to the initial radius of the water thread.
Note. (i) A Belgian physicist, Joseph A. F. Plateau (1801–1883) first showed
that the critical ‘wavelength’ is larger than stated above. He found that
λcrit = 2πr by investigating the constriction of the water thread caused by
surface tension. He assumed a periodical change in the diameter of the
thread and considered the effect of pressure differences associated with the
curvature of the liquid surface.
(ii) The Nobel-laureate English physicist, John W. S. Rayleigh (1842–
1919) investigated the stability of the water thread and, according to his
very careful and detailed calculations on drop formation, the ‘winning
wavelength’ is λwin = 9.02r.
S144 (i) Early on, particles coming from the right, and rebounding
elastically, transmit a greater momentum to the body than those colliding
inelastically from the left. For this reason, a resultant force acting to the left
accelerates the body. The faster the body moves to the left, the lower the rate
at which particles collide with it from the right, and the lower their relative
velocity when they do. For those impinging from the left the converse is true,
and the net force acting on the body decreases with time.
After a sufficiently long time, the body moves at uniform speed v1 . The
condition for this equilibrium situation is that, in unit time, the particles
impart the same momentum from the right as from the left.
The particles from the right reach the body with a relative speed v0 − v1
and rebound with the same relative speed. In a short time interval ∆t, the
particles coming from within a distance of (v0 − v1 )∆t reach the cylinder,
and each imparts an impulse proportional to 2(v0 − v1 ) to the body. Thus
the force from the right is proportional to 2(v0 − v1 )2 . Similarly, particles
inelastically colliding from the left at a relative speed of (v1 + v0 ) produce a
force proportional to (v0 + v1 )2 . The condition for constant speed is
√
2−1
2(v0 − v1 ) = (v0 + v1 ) ,
2 2
v1 = √ v0 ≈ 0.17 v0 .
2+1
(ii) Assume that the colliding particles are the molecules of a gas at a
certain temperature. If the body continued to move uniformly, even after a
very long time, it would imply that a perpetuum mobile (perpetual motion
machine) of the second kind were possible, i.e. a heat engine could be
built that could continually extract energy from a single heat reservoir. All
experience indicates that such an engine cannot be made.
Where is the mistake in the above reasoning? The heat produced by
the inelastic collisions has not yet been taken into account! The particles
inelastically bombarding the left end of the cylinder heat it up. If the heat so
produced is continuously removed (i.e. the body is cooled), then the motion
described in part (i) is sustainable. The result is a normal heat engine working
between two heat reservoirs, the gas of bombarding particles and the cooling
medium.
However, if the body is not cooled, sooner or later it warms up. The
molecules forming the sides of the warm body vibrate at the average speed
corresponding to its temperature, and the gas particles rebound from it,
sometimes at a greater speed, and sometimes at a lower speed, than they
would from a colder body. The result is that finally the body cannot ab-
sorb any more heat from the gas at either end (or it would warm up
further). Then the collisions are effectively elastic on both ends – taking
time averages – and so the impulsive forces are equal. Thus, after a very
long time, when thermal equilibrium has been reached, the body has to
stop.
S145 As the space probe is very far from the solar system we may neglect
the solar and cosmic background radiation. Without any protecting shields,
the heat production of the nuclear energy source is radiated away by the
surface of the space probe according to the Stefan–Boltzmann law:
I = σAT 4 ,
where σ is the Stefan-Boltzmann constant, A is the surface area of the

space probe and T is its surface temperature. When a thin protecting shield
encloses the space probe, the same radiation process occurs at the outer
surface of the shield, and so the temperature of the shield must be T .
However, the shield also emits inwards, and consequently the surface of the
probe absorbs an amount of radiation equivalent to that radiated into space
(see figure). This means that the surface of the probe must re-radiate a total
received intensity of 2I at a new temperature T1 , where
2I = σAT 41 .
√
It follows that T1 = 4
2T.
Solutions 207
I
I T
2I
T1
For N protecting shields, the net radiation through them will still be I.
Repeated application of our previous argument shows that the space probe
radiates (N + 1)I and implies that the temperature of the surface of the
√
probe is TN = 4 (N + 1) T .
Note. This result cannot hold for very large N, because we have ignored the
increase in surface area of successive shields.
S146 When a body at absolute temperature T absorbs a quantity of heat

∆Q, the change in entropy ∆S ≥ ∆Q/T . The equality holds when the process
is reversible.
Let ∆T1 and ∆T2 denote small changes in temperatures of the bodies at
temperatures T1 and T2 , respectively, as a result of receiving quantities of
heat ∆Q1 and ∆Q2 by a reversible process. The change in the entropy of the
whole system is
∆Q1 ∆Q2
∆S1 + ∆S2 = + ≥ 0,
T1 T2
∆Q1 T2 + ∆Q2 T1 ≥ 0.
The ratio of the heat transfer to the change of the temperature is the same for
both bodies, since their masses are equal. The above equation can therefore
be written as
(∆T1 )T2 + (∆T2 )T1 = ∆(T1 T2 ) ≥ 0.
This means that the geometric mean of the temperatures of the two bodies
cannot decrease during the process (though it may increase).
This relationship is valid throughout the process, and hence for the initial
√
and final states, i.e. the common final temperature has to be at least T1 T2 .
If no energy were taken out of the system the common final temperature
would be the arithmetic mean of the two initial temperatures. Thus the
maximum energy that can be taken out is
T1 + T2
− T1 T2 mc,
2
where m is the total mass of the water and c is its specific heat.
S147 The number of microstates available to a system of N objects

confined to a space of volume V is proportional, both classically and quantum
mechanically, to V N . In the current case, initially there are 2NA molecules of
helium occupying 2V0 (V0 ≈ 0.0224 m3 ), and correspondingly for the oxygen.
After the partition is removed there are 5NA molecules occupying 5V0 and
the ratio of the new number of microstates to that before is
W2 (5V )2NA (5V )3NA (5V )5NA
= × = .
W1 (2V )2NA (3V )3NA (2V )2NA (3V )3NA
Thus the change in entropy is
W2
∆S = k ln = kNA (5 ln 5 − 2 ln 2 − 3 ln 3) = 27.9 J K−1 .
W1
Note. If we had investigated the mixing of two identical gases (e.g. two
moles of helium with three moles of helium) both at s.t.p., the calculated
entropy change would seem to be the same. However, it is obvious there
can be no entropy change, because physically nothing happens then. This
is called the Gibbs paradox. The resolution of this paradox depends upon
the indistinguishability of identical particles.
S148 Consider the compression stroke of the pump (shown in the figure)
in two stages. Initially, both valves are closed and the piston isothermally
compresses the air. When the pressure in the pump equals that in the
container, the inner valve opens and the total amount of air is isothermally
compressed. The moment when the inner valve opens becomes later and
later, which makes calculation of the total work done rather complicated.
Fortunately, there is a simpler method!
p
0
Consider the amount of air that is in the container at the end of the
process. This is the 10 litres initially present and a further nine times this
amount, i.e. 90 litres (initially at atmospheric pressure). This amount of air
occupies a volume of 10 litres.
According to the first law of thermodynamics, the sum of the work W
done on the air and the heat Q transferred to it, is equal to the increase in
the internal energy of the air. In the present situation, the temperature of
the air does not change, and therefore its internal energy is unaltered. Thus,
W + Q = 0, i.e. the work done on the air equals the heat −Q it gives out.
Solutions 209
Further, the change in entropy of the air at constant temperature T , in a

reversible process, is ∆S = ∆Q/T . If the change in entropy of the air can be
calculated, the heat given out and the work done can also be determined.
If N molecules are compressed into a ten-times smaller space than they
originally occupied, then the number of possible microstates is 1/10N times
their original number. According to the statistical interpretation of entropy,
the change in entropy of the gas is the logarithm of this number multiplied
by the Boltzmann constant,
1
∆S = k ln = −Nk ln 10.
10N
This gives
−∆Q = −T ∆S = NkT ln 10 = nRT ln 10 = 10p0 V0 ln 10
for the heat given out by the gas. The work done is thus
W = 10 × 105 Pa × 10−2 m3 × ln 10 ≈ 23 kJ.
S149 There has to be an electric field between the Earth and the distant
planet, as there is a potential difference between them. This electric field
causes a charge separation in the wall of the spaceship, and therefore the
electric field inside the spaceship is zero (the Faraday cage effect).
The potential difference between the spaceship and the planet changes in
the course of the flight, being roughly proportional to the distance from the
planet; it is very large at the beginning of the flight but slowly decreases
and becomes zero at the end. Thus, when the spaceship lands on the planet,
its electric potential is exactly equal to that of the planet and the astronauts
can get out safely.
S150 The energy of a capacitor of capacitance C carrying a charge Q is
Q2 /(2C). If the change in energy of the capacitor can be found, the change
in its capacitance can also be calculated.
The energy of the capacitor is higher when it is dented, since the surface
charges have been moved in a direction opposite to that in which their
mutual repulsion acts. Further, an electrostatic field E has an energy εE 2 /2
per unit volume, and an alternative view is that, when the capacitor is dented,
the electric field exists in a volume where it was not previously present.
If the surface of the capacitor is only changed a little, the electric field
near the surface can be taken as the same as the original one. Thus, the
change in energy depends purely on the change in volume and not on the
actual shape of the indentation.
Imagine that the original capacitor is hammered so that its volume de-
creases by 3 per cent, but its shape remains spherical. Its radius is thus
reduced by 1 per cent (as the volume of the sphere is proportional to the
cube of its radius). The ratio of the energy of such a reduced spherical
capacitor to that of the original sphere is the same as the ratio of the energy
of the dented capacitor to the original one. Thus, the relative changes in
their capacitances are the same as well.
Finally, the capacitance of a spherical capacitor is proportional to its
radius. The capacitance of the reduced capacitor is therefore 1 per cent
smaller, and the capacitance of the dented capacitor in the original question
must have decreased by the same amount.
S151 It can be proved that if equal amounts of charge are carried by F
and F ∗ then the electrostatic energy of the configuration belonging to F is
lower.
Let us start from the new surface F ∗ . Imagine that the charges on it are
‘fastened’ to the surface, and that the surface is then hammered in such a
way as to displace the charges perpendicular to the original surface. The
charges have then moved in the direction of the force acting on them. As the
surface was originally an equipotential, the field direction was perpendicular
to the surface. Thus, the energy of the system decreases in the course of the
deformation. (An outward force acts on the surface charges, regardless of
their signs – a field directed outwards emanates from the positive charges,
while an inwardly directed field is produced by the negative ones.)
The new surface will not be equipotential, but, if the fixed charges are
‘released’, they migrate, warming the metal up a little whilst doing so. The
electrostatic energy of the system is therefore lower in the new equilibrium
position.
This process can be repeated until surface F ∗ is transformed into surface
F. The electrostatic energy decreases all the time, while the total charge Q
on the metal does not change. Since the electrostatic energy depends on the
capacitance, C as Q2 /2C, the capacitance of the surface F has to be greater
than that of the surface F ∗.
S152 The capacitance of the plane capacitor of surface area A is C =
ε0 (A/d). The energy of the capacitor, when connected to a voltage V , can be
expressed as
1 AV 2
W = CV 2 = ε0 .
2 2d
When the distance between the plates is increased from d to 2d, both the
capacitance and the energy of the capacitor decrease to half of their original
Solutions 211
values. This may be a surprising result, as pulling apart the plates of the
capacitor requires positive work. However, when charge ∆Q leaves the plates
of the capacitor it increases the energy of the battery (charges the battery) by
∆QV . Since ∆Q = ∆CV, the increase in the energy of the battery is exactly
twice the decrease in the energy of the capacitor. The increase in the energy
of the battery is due half to the decrease in energy of the capacitor and half
to the work done in pulling the plates apart.
Note. The above statement can be verified by direct calculation of the work
done in pulling the plates apart. The force of attraction acting between
the plates of a plane capacitor can be calculated using the relation F =
Q2 /(2ε0 A). Substituting for the charge in terms of voltage and capacitance
gives
C 2V 2 ε0 AV 2
F= = .
2ε0 A 2d2
If the distance between the plates is denoted by x, the work done can be
calculated as
2d
ε0 AV 2 2d
dx ε0 AV 2 1 ε0 AV 2
W = F(x) dx = = = .
d 2 d x2 2 2d 4d
So the work done in pulling the plates apart is equal to the decrease in
energy of the capacitor, and these two quantities together increase the energy
of the battery.
S153 Imagine that the current I0 flows in a superconducting, short-

circuited coil of N turns, with cross-section A = πR 2 and length x0 . The
magnetic field strength inside the coil is then B = µ0 IN/x0 and the total
magnetic flux Φ = BAN = µ0 I0 N 2 A/x0 . Even if the length of the coil
changed for some reason, the magnetic flux would not change, as to do
so would induce a voltage and hence an ‘infinitely’ large current in the
coil of zero resistance. Thus the current must vary with the length x as
I(x) = I0 x/x0 . The inductance of a coil of length x is L(x) = µ0 N 2 A/x, and
the magnetic energy of a coil with current I flowing through it is
1 2 N 2 I02 A
Wm = LI = µ0 x.
2 2x20
Thus the energy of the coil is proportional to x, i.e. Wm (x) ∝ x. The

proportionality constant F0 is the magnetic force of contraction, since work
W = F0 x is needed to re-stretch the coil by length x.
The coil is in equilibrium when the magnetic force of contraction balances
the elastic force F(x) = k(x0 − x), i.e. when the change of spring length is
F0 πI 2 N 2 R 2
∆x = x0 − x = ≈ µ0 0 2 .
k 2kx0
S154 (i) The horizontal force (measured in the direction of increasing x,

i.e. to the left in the figure in the problem) exerted on magnet A is
K (d + s − x)
F(x) = − n
+ mg , (1)
x '
where d + s is the distance from magnet B to the unperturbed position of
magnet A. The condition of equilibrium is
F(x)|x=d = 0. (2)
The stability of equilibrium depends on the behaviour of the function F(x)
close to x = d. The function can be monotonically increasing or decreasing
(Fig. S154.1).
F(x)
Stable Unstable
Fig. S154.1
For stable equilibrium, its derivative has to be negative. In the unstable

case the derivative is positive, and in the limiting case of neutral equilibrium
it is zero, i.e.
F (x)|x=d = 0. (3)
Using the expression in (1) for the force F(x), condition (3) can be formulated
as
nK mg
n+1
− = 0. (4)
d '
But condition (2) can be written in the form
K mg
− + s = 0. (5)
dn '
From these equations we get
d s
n= =4 and K = mgd4 . (6)
s '
Solutions 213
(ii) In the vertical tube (Fig. S154.2) the repulsive force acting on magnet
A is Fvert (x) = +K/x4 − mg.
K
h4
Fig. S154.2
Equilibrium levitation occurs with the magnets a distance h apart, where

Fvert (x)|x=h = 0.
Taking into account result (6) we get
K 1/4
s 1/4
1 1/4
h= =d = 4 cm × ≈ 1.3 cm.
mg ' 100
S155 The energy ultimately stored in the capacitor is 12 CN 2 E2 in both

cases. Direct connection results in an equal amount of wasted energy. For
charging by stages, the total work done by the battery is
1
∆Q (∆V ) = CE(E + 2E + · · · + NE) = CE2 N(N + 1).
2
The wasted energy is thus
1 1
CE2 [N(N + 1) − N 2 ] = CE2 N,
2 2
i.e. only 1/N of the original loss.
Note. You can obtain the same result by considering the following graph.
C A
B
O Q
The triangle OAB represents the stored energy of the capacitor, and the
triangle OAC corresponds to the energy wasted by charging it in a single
step. The small shaded triangles show the energy wasted when using multiple
(e.g. N = 4) steps. The total area of the shaded triangles is 1/N of that of
OAC.
S156 The stored energy is Q2 /2C in both cases and Coil = εCair . Thus
the energy is increased by a factor of ε when the oil is poured out. The catch
is that the oil is attracted to where the field is strongest and work has to
be done to extract it from there; the work needed is at least as great as the
increase in stored electrical energy.
S157 (i) Let the distance between the capacitor plates of area A be d,
and the charge on them be Q. Examine the situation when a length x of
the insulating sheet has already been slid between the plates of length ' as
A
r
d
x − x
The capacitance of the plane capacitor can be calculated as that of two

capacitors connected in parallel, one filled with dielectric and the other
empty:
Ax A '−x ε0 A
C = ε0 εr + ε0 = [(εr − 1)x + '] .
d ' d ' 'd
The total capacitance can be seen to increase linearly with x from ε0 A/d to
εr times its initial value. As the charge is constant in the present case the
energy of the capacitor Wcap can be written as
Q2 Q2 'd
Wcap = = .
2C 2ε0 A [(εr − 1)x + ']
It is clear that the energy of the capacitor decreases as x increases! This
means that the work done in sliding the insulator in is negative, i.e. the plates
effectively suck in the insulator.
The size of the force involved can be calculated by equating the work
corresponding to a small change in x with the resulting change in energy
(F∆x = ∆Wcap ), and differentiating the expression for W ,
dW (εr − 1)Q2 'd
F= =− .
dx 2ε0 A [(εr − 1)x + ']2
Solutions 215
The magnitude of the force is greatest at x = 0, since the denominator in

the above expression increases with x. It is easy to show that when x = ',
the magnitude of the force has decreased by a factor ε2r .
Note. How does the electric field exert a force on the dielectric parallel to
the plates? If the electric field between the plates were homogeneous and
perpendicular to the plates, and zero outside them (i.e. the usual picture of
capacitors), then no force could act on the insulating sheet. The phenomenon
is explained by the curvature of the electric field that is inevitably present
at the edge of the plates.
(ii) The force acting on the dielectric cannot depend on the interaction
of the capacitor with its surroundings. Therefore the previous result would
be valid in the constant voltage case if it were not for the fact that the
charge on the capacitor changes in accordance with Q = CV as soon as
the insulating sheet is placed between the plates. Substituting for the charge
(which depends on x) into the above expression gives
ε0 A(εr − 1) V 2
F=− ,
2'd
i.e. in this case the force acting on the insulating sheet is constant.
An interesting conclusion can be drawn from the expression for the energy
of the capacitor at constant voltage,
CV 2 ε0 A [(εr − 1)x + '] V 2
Wcap = = .
2 2'd
This shows that, in this case, the energy of the capacitor increases linearly
with x, and that the change of energy ∆Wcap corresponding to a small
displacement ∆x of the insulator is
∆Wcap ε0 A(εr − 1)V 2
= .
∆x 2'd
Apart from the sign, this formula is exactly the same as that for the force
acting on the insulating sheet. These results can be summarised in the
following way: when the dielectric is slid between the plates, the system does
work on it (i.e. pulls it in), while the energy of the capacitor increases by the
same amount. This is possible because the energy of the battery decreases
by twice this amount during the process. The decrease in stored battery
energy occurs because the capacitance of the capacitor (and therefore its
charge) increases, and the battery has to supply the additional charge. The
calculation of the work done by the battery is left to you the reader.
S158 Number the resistors, starting with the last element in the chain.
As a current of 1 A flows through the first resistor, a current of 1 A has
to flow through the second one as well, thus there is a potential difference
(p.d.) of 1 V across each resistor. As a consequence, the p.d. across the third
resistor is (1 + 1) = 2 V, and the current flowing through it must be 2 A.
The current flowing through the next resistor is (1 + 2) = 3 A. The current
in the fifth resistor can be determined using the p.d. (2 + 3) = 5 V across the
resistors with currents of 2 and 3 A, respectively, flowing through them, and
so on, as shown in Fig. S158.1.
21 A 8A 3A 1A
V 13 A 5A 2A 1A
Fig. S158.1
Consider the chain of resistors to be built starting with the last element and
then connecting the consecutive elements, alternately in series and parallell,
throughout the chain. The sum of the currents flowing through the two
previous resistors flows through the following series resistor (Kirchhoff’s
first law). The next element connected in parallel creates a new loop in the
chain, and therefore the p.d. across this resistor equals the sum of the p.d.s
across the two previous ones (Kirchhoff’s second law). Since the numerical
values of the p.d. and the current are identical for a 1-Ω resistor, the sum
of the currents of the two previous resistors is the same as the current of
the new resistor connected in parallel. Thus, in this so-called ladder circuit,
Kirchhoff’s laws are satisfied in such a way that the current flowing through
each resistor (and the p.d. across it) is equal to the sum of the corresponding
quantities for the two previous elements.
Notice that the numerical values of the currents (or p.d.s) are the terms
of the Fibonacci series: 1, 1, 2, 3, 5, 8, 13, 21. The p.d. across the two last
resistors is (21 + 13) = 34 V, and this must also be the p.d. across the circuit
input. As a total of 21 A flow as the result of applying a p.d. of 34 V, the
equivalent resistance of the circuit is of 34/21 = 1.619 05 Ω.
If one more element is connected to the chain (in parallel) then the
p.d. across it is unchanged at 34 V, but the total current increases to
(21 + 34) = 55 A. In this case the equivalent resistance is 34/55 = 0.618 18 Ω.
If yet another element is connected to the circuit, a current of 55 A flows
through it and the input p.d. increases to (34 + 55) = 89 V. The total
resistance of the chain is then 89/55 = 1.618 18 Ω.
If the ladder circuit is extended further and further, an ‘infinite’ chain is
obtained. The equivalent resistance of this chain can be calculated using the
Solutions 217
fact that adding two more elements does not change its resistance. Thus,
the whole chain can be replaced by a single resistor of resistance R, which
is such that if two 1-Ω resistors are connected to it, one in parallel and the
other in series with the combination, the equivalent resistance of the new
circuit will also be R (see Fig. S158.2).
1Ω
1Ω R
Fig. S158.2
The condition for this is

1
R =1 Ω+ ,
(1/1 Ω) + (1/R)
which yields the following quadratic equation for the numerical value of R,
R 2 − R − 1 = 0.
The positive root of this equation gives the equivalent resistance for an
‘infinite’ chain as,
√
1+ 5
R= ≈ 1.618 03 Ω.
2
We see that the equivalent resistance of a chain of eight–ten elements
approximates very well that of the ‘infinite’ chain. Hence a ladder circuit
with relatively few elements can be considered as infinite.
Note. (i) In practice, the neutral wires of overhead electric supply networks
can be considered as ladder circuits; the neutral wires are fastened to poles
and earthed at, say, every tenth pole. Such a ladder circuit consists of two
types of resistors, but the equivalent resistance of the ‘infinite’ chain can be
calculated using the above method.
(ii) It is of interest to note that the above quadratic equation is the
√ golden
ratio equation, the solution of which is the golden mean, (1 + 5)/2 =
1.618 03 . . . . As shown, this is the same as the numerical value of the
equivalent resistance of the infinite ladder circuit. Furthermore, the ratio
of consecutive elements of the Fibonacci series was shown to approach the
golden mean surprisingly quickly. It is also easy to prove that dividing the
even elements by the previous odd ones, the golden mean is approached
from below, while dividing the odd elements by the previous even ones, it
is approached from above.
Finally, for the sake of the aesthetic pleasure produced by multiple-level

fractions, it is worth expressing with 1s the equivalent resistance of the infinite
chain made up of 1- Ω resistors. This time, the elements are considered in
order, not from the end but from the beginning, i.e. starting at the left hand
end of the chain:
1
R =1+ .
1
1+
1
1+
1
1+
1
1+
1
1+
1 + ···
If the · · · at the end of the formula is replaced by a 1, then the equivalent
resistance of the original eight-element chain is obtained.
S159 We discuss first the grid of ohmic resistors and reason as follows.
Consider first a grid point at which a current I enters and then flows away
‘to infinity’. From symmetry, identical currents of magnitude I/4 will start
from this point and travel along the four possible directions as shown in
Fig. S159.1.
I/ 4
I
Fig. S159.1
Now consider the neighbouring grid point and let current I flow out of
it (independently of the previous reasoning). Again, identical currents of
magnitude 14 I flow through the four identical resistors adjacent to the point
(see Fig. S159.2).
I/ 4
I
Fig. S159.2
Solutions 219
Now consider superposing these two cases. Because of the linearity of

the circuit equations, scalar quantities (e.g. currents and potentials) simply
add. A voltage (p.d.) V appears between the two neighbouring grid points,
current I goes into one of them and current I flows out of the other. The
currents flowing elsewhere in the grid cannot be easily determined, but the
current flowing between these two neighbouring points can. The two currents
discussed above add in this resistor, i.e. a total current of 12 I flows through
it (see Fig. S159.3).
I/ 2
I I
Fig. S159.3
But if current 12 I flows through a resistance R then a p.d. of V = RI /2

appears across it. The equivalent resistance between the two neighbouring
points is therefore Re = V /I = R/2.
The equivalent values for infinite grids of capacitors and self-inducting
coils can be calculated similarly as Ce = 2C and Le = 12 L.
S160 Denote the number of grid points of the polyhedron by c and
the number of lines meeting at a grid point by n (e.g. in the case of a
dodecahedron, c = 20 and n = 3). If a current of 1 amp flows in at one grid
point and 1/(c − 1) amps are taken out at all other grid points, then, by
symmetry, the current flowing in each of the lines which meet at the point
where the current enters is I = 1/n amps.
Superimpose on the previous current distribution, that produced by a
current of 1 amp flowing out of the neighbouring grid point and currents of
1/(c − 1) amps flowing into all the other points (including the original point).
The current of I = 2/n amps flowing through the ‘direct’ resistor (i.e. the one
directly joining the two neighbouring grid points) causes a p.d. of V = 2/n
volts to appear across it. Since the current flowing through the whole circuit
is I = 1 + 1/(c − 1) = c/(c − 1) amps, the equivalent resistance is
Re = 2(c − 1)/nc.
This figure is 19/30 Ω for the dodecahedron and 1/2 Ω for the infinite square
grid. It can be calculated similarly for any regular shape.
S161 Let us denote the equivalent resistances of the original grid by Reorig
and of the truncated circuit by Retrun . We can consider the original grid as a
circuit of two resistors connected in parallel; the two resistors are Retrun and
R, the latter being the removed resistor. It is then easy to write an equation
which involves Retrun , namely,
R × Retrun R × Reorig
Reorig = or Retrun = .
R + Retrun R − Reorig
For example, in the case of the infinite two-dimensional square grid the
equivalent resistance between two neighbouring grid points is Reorig = R/2.
It follows that the equivalent resistance of the truncated circuit is Retrun = R.
S162 The ‘trick’ of superposition is again employed to combine separate
discussions of currents flowing in and out. Let A denote the corner of the
square where the current I0 flows in and B the neighbouring corner where it
flows out. The p.d. V is measured between the other two corners (C and D)
as shown in the figure.
I0 I0
a
A B
a a
a
D C
If the current I0 is introduced at point A (and flows towards points at zero

potential infinitely far away), it is distributed (hemi)spherically symmetrically
in the half-space containing the matter, i.e. at a distance r from point A, the
current density j(r) is j(r) = I0 /2πr 2 . This relation is not valid for very small
values of r, i.e. when r is not much greater than the size of the electrode
at A.
The local version of Ohm’s law is the differential law that expresses the
current density j in terms of the electric field strength E and the local
resistivity ρ, namely j(r) = E(r)/ρ.
The magnitude of the electric field strength in the half-space can be
determined using this relation as
I0 ρ
E(r) = .
2πr 2
The potential function (and hence the p.d.) could be determined from the
field strength by integration. In this case, however, it can be obtained using a
simple analogy. The electric field of a point charge Q is inversely proportional
Solutions 221
to r 2 , its potential is inversely proportional to r and the proportionality

coefficient is the same in both cases (E = kQ/r 2 and V = kQ/r, respectively).
This means that the potential field corresponding to the electric field strength
determined above is V = I0 ρ/(2πr).
The nearer a point is to the electrode where the current flows in, the higher
The potential VD = I0 ρ/(2πa) of point D is therefore√higher than
its potential. √
VC = I0 ρ/(2 2πa). The p.d. between the two points is I0 ρ(2 − 2)/(4πa).
We next discuss the flow of the current I0 out through point B. Everything
is the same as in the previous case, except that the signs of the quantities
(current, current density, field strength and potential) are reversed. The
potential in the half-space under investigation is described by the function
V = −I0 ρ/(2πr ), where r is the distance from point B. The potential of
point C is lower than that of point D, i.e. point D is again more positive
than C. The p.d. between the two points is the same as previously.
If the two previous cases are now superposed we return to the original
problem, and the p.d. between points √ C and D is exactly twice either of
the above p.d.s, i.e. V0 = I0 ρ(2 − 2)/(2πa). Apart from given data, this
expression contains only the resistivity ρ. Therefore the solution of the
problem is
√
ρ = (2 + 2)πaV0 /I0 .
Note. This method is widely used in ‘real life’, e.g. to determine the average
resistivity of rocks. The measurements are, naturally, not made on infinite
half-spaces, but on volumes and planes with linear sizes much larger than
the side a of the square.
S163 First connect the battery to the terminals of the resistor through
the ammeter, and then connect the voltmeter across the same terminals, as
shown in Fig. S163.1.
A
R
+ —
Fig. S163.1
It is naı̈ve to assume that the quotient of the measured potential difference

and current gives the resistance, because we cannot be sure that at junction
A all the current coming from the ammeter flows through the resistor R, and
that some does not flow towards other parts of the circuit.
The problem can be readily solved with the help of short circuits. Using
zero-resistance wires we connect all of A’s neighbouring junctions to the
same terminal of the battery as that to which the ammeter is connected, as
shown in Fig. S163.2.
V
C A
B R
+ —
Fig. S163.2
As the internal resistance of the ammeter is negligible, the junctions
(B, C, D, . . .) and A are equipotential points. Consequently, there is no current
flowing between them and the current from the ammeter must all flow
through the resistor R. It is possible that the battery has to provide additional
currents flowing towards junctions B, C, D, . . . , but these have no influence
on our measurement.
Note. If resistor R is connected in parallel with other resistors, then there
is no way to measure the resistance of these resistors separately – only their
equivalent resistance can be found.
S164 Let current I flow into the cube at one point and flow out at the
point diagonally opposite to it (see Fig. S164.1).
I/3
B I
X
O X
I/3 I/3
I/3 O
I/3 X
A O
I I/3
Fig. S164.1
Symmetry prescribes that the current flowing through the three resistors
that meet at each endpoint of the diagonal is I/3, and hence that the voltages
across these resistors are identical. Thus, the sets of points denoted by O and
Solutions 223
X in the figure are each equipotentials, i.e. the members of each set can be
joined together without disturbing the system.
When the equipotential points are joined, the circuit can be redrawn
as shown in Fig. S164.2, and the equivalent resistance can be calculated
mentally. If all the resistors have resistances of 1 Ω, then the equivalent
resistance is 5/6 Ω.
B
A
Fig. S164.2
A one-dimensional ‘cube’ is simply a straight section, its resultant resistance
is itself, i.e. 1 Ω. The two-dimensional ‘cube’ is the square. Two resistors
emanate from one end of a diagonal and two resistors converge at the
opposite end of the same diagonal. If current flows through the square,
the endpoints of the other diagonal are equipotential. Therefore two lots of
two parallel resistors are connected in series. For a square the equivalent
resistance is again 1 Ω. As we have already seen, for a three-dimensional
cube, three resistors start from each end of the diagonal and the remaining
six resistors join equipotential surfaces.
A four-dimensional ‘cube’ can be obtained by a parallel displacement of a
three-dimensional cube in the direction of the fourth dimension, followed by
the joining of corresponding points. The four-dimensional ‘cube’ therefore
has 12 + 12 + 8 = 32 edges (12 for each normal cube plus eight to join
corresponding corners). A distorted projection of such a ‘cube’ can be made
for investigative purposes, as shown in Fig. S164.3, replacing the displacement
by a magnification. Figure S164.3 shows one diagonal, AB, from the eight
possible ones.
A
Fig. S164.3
For a four-dimensional ‘cube’, four lines lead out of any one vertex and
the endpoints of these are equipotentials. The same is true at the other end
of the diagonal. The shortest way from one of these equipotential surfaces to
the other is through any two resistors of the remaining 24. This means that
the ‘inner’ points of the 24 resistors are also equipotential, i.e. two lots of 12
resistors are connected in parallel and these are then connected in series. The
redrawn circuit is shown in Fig. S164.4; the equivalent resistance is 2/3 Ω.
... 12 ...
... 12 ...
Fig. S164.4
Note. The problem can be generalised to n-dimensional ‘cubes’ if, after

connecting equipotential points, the circuit can be decomposed into sets of
identical resistors connected in parallel with the sets then connected in series.
In general, the equivalent resistance across a diagonal of an n-dimensional
‘cube’ made of 1-Ω resistors is Rn Ω, where Rn is given by
1 1 1 1
Rn = n + n + n + ··· + n .
1 1 2 2 3 3 n n
S165 Let I denote the current flowing in the wire, A the cross-section of
the wire, and ρ1 and ρ2 the resistivities of the metals. Ohm’s law for a wire
of length ' gives V = Iρ'/A, which yields E = V /' = ρI/A for the electric
field strength in the wire.
+
Cu + Fe
+
+
The resistivity of copper is lower than that of iron, and therefore the
electric field strength has to be smaller in the copper than in the iron.
Solutions 225
According to Gauss’s law, the difference in the electric field strengths implies
an accumulation of charge at the boundary of the two metals (see figure).
The net accumulated charge is
Q = ε0 A(EFe − ECu ) = ε0 I(ρFe − ρCu ).
It is interesting that this quantity depends purely on the current and material
constants, and not on the cross-section of the wire.
Substituting the known data, the charge is found to be Q ≈ 5 × 10−21 C,
1
which is only 30 of an elementary charge! Though a measurable macroscopic
current flows through the wire, the accumulated charge is only a small
proportion of the microscopic elementary charge. This strange result shows
that classical electrodynamics (imagining charge carriers as small balls)
cannot always correctly describe electrical phenomena. Only the application
of the more sophisticated laws of quantum theory and statistical physics can
give an accurate description.
S166 In SI units the jet’s speed is 200 m s−1 .
(i) The magnetic field is vertically downwards, and the induced voltage is
80 × 6 × 10−5 × 200 = 960 mV, with the starboard (right) wing tip at the
higher potential.
(ii) The motion is parallel to the field, and so no potential difference is
developed.
(iii) Since the field is a dipole its strength at the Equator is one-half that at
the Pole. The field is horizontal, and the induced voltage is 8×3×10−5 ×200 =
48 mV, with the bottom of the jet at the higher potential.
(iv) The vertical field√ is 5 × 10−5 × sin 66◦ T and the northward component
of the velocity is 200/ 2 m s−1 , leading to a 520 mV potential difference
across the wings with the starboard tip at the higher potential. In the same
way, a p.d. of 23 mV appears between the top (higher potential) and the
bottom of the jet’s body.
S167 Let that the rod move with speed v and acceleration a along the
inclined plane, while current I flows in it. The magnetic field brakes the rod
in accordance with Lenz’s law, and its equation of motion is
ma = mg sin α − B'I.
This equation is the same in all three cases. The differences result from the
different relationships between the induced voltage and the current flowing
in the rod.
(i) If the circuit is closed by an ohmic resistor R, the current I and the
induced voltage V = B'v are connected by the relationship I = V /R =
B'v/R. This shows that the braking force increases in proportion to the
speed, with the result that the rod experiences decreasing acceleration and
ultimately travels with uniform speed. This final maximum speed vmax can
be found from the equation of motion by setting a = 0,
mgR sin α
vmax = .
B 2 '2
(ii) If the circuit is closed by a capacitor of capacitance C, the relationship
between the induced voltage and the current is different. The charge on the
capacitor is determined by the induced voltage, and given by
Q = CV = CB'v.
Note that the current flowing through the rod is equal to the time derivative
of this, i.e.
dQ dV dv
I= =C = CB' = CB'a.
dt dt dt
In other words, the current flowing in the rod is directly proportional to the
acceleration of the rod. If the above expression for the current is substituted
into the equation of motion, the rod is found to move on the rails with
uniform acceleration
mg sin α
a= .
m + B 2 '2 C
Induction decreases the acceleration caused by gravity by, in effect, increasing
the inertial mass of the rod. The speed of the rod and the charge on the
capacitor are both directly proportional to the time elapsed.
(iii) If the circuit is closed by a coil of inductance L, the relationship
between the induced voltage and the current is
dI dx
L = B'v = B' .
dt dt
We note that, since I = 0 and x = 0 at the start of the motion, the
above formula implies that the current is proportional to the x-coordinate,
LI = B'x. Substituting for the current, from this relationship into the
equation of motion, gives
B 2 '2
ma = mg sin α − x.
L
The force acting on the rod is therefore the sum of a constant term and
a negative term proportional to the displacement. This is the same as the
Solutions 227
equation of motion of a body hung on a spring and then released. Thus, the
rod makes harmonic oscillations about an equilibrium position
mgL sin α
x0 = .
B 2 '2
The amplitude of the oscillation is A = x0 , and the dependence of the
displacement of the rod on time is
x(t) = A(1 − cos ωt),
B 2 '2
where ω 2 = .
mL
S168 (i) At the instant when the capacitor is connected, a current I =
V0 /R starts flowing in the rod, which experiences a force F = B'I and
an initial acceleration a = B'V0 /mR. In accordance with Lenz’s law, the
voltage induced in the moving rod causes the current flowing in the rod to
decrease. The charge Q on the capacitor decreases and consequently so does
the voltage across it. Meanwhile the voltage induced in the rod increases,
until the two voltages cancel each other out. The rod then continues with its
maximum velocity given by
Qmin
B'vmax = . (1)
C
The equation of motion of the rod is
dv dQ
m = ma = B'I = −B' , (2)
dt dt
where the acceleration and the current have been expressed as the rates of
change in velocity and charge, respectively. The proportionality between the
two rates of change holds throughout. The speed of the rod increases from
zero to vmax , whilst the charge on the capacitor decreases from Q0 = CV0 to
Qmin . Equation (2) can therefore be rewritten as
mvmax = B'(Q0 − Qmin ).
The final velocity and the residual charge on the capacitor can be calculated
using equations (1) and (2),
B'CV0 B 2 ' 2 C 2 V0
vmax = and Qmin = .
m + B 2 '2 C m + B 2 '2 C
(ii) The above relations show that the maximum velocity of the rod is
proportional to the initial voltage V0 across the capacitor. Thus, the final
kinetic energy of the rod is proportional to V02 (for given values of C
and m), i.e. proportional to the initial energy of the system. The coefficient
of proportionality can be regarded as the efficiency η of the apparatus

(considered as an electromagnetic gun), and can be written in the form
2 mvmax
1 2
1
η= = √ √ .
2 CV0
1 2 2
m B' C
√ + √
B' C m
The product of the two terms in the brackets is 1, and from the inequality
between arithmetic and geometric means, it follows that their sum is at least
2. This means that the efficiency of the electromagnetic gun cannot be more
than 25 per cent.
Note. If the condition for maximum efficiency m = CB 2 '2 is substituted
into the expression for the final charge on the capacitor we find that
Qmin = V0 C/2, i.e. only half of the initial charge on the capacitor is left.
Thus, only one-quarter of the initial energy of the capacitor is left; one-
quarter of it is transformed into the kinetic energy of the rod, and the other
half is dissipated in the rod as Joule heat.
S169 (i) The rate of increase of magnetic energy (Emagn = LI 2 /2) is the
difference between the power output of the battery and the power dissipated
in the resistor,
dEmagn V 2
V2 V2
= V I − RI 2 = −R I − + ≤ .
dt 2R 4R 4R
It is clear that the rate of increase is maximal when I = V /(2R).
(ii) After the switch has been closed, we can write Kirchhoff’s law for the
circuit as
dI
V = IR + L ,
dt
which gives the current–time relationship (see Fig. S169.1)
V R
I= 1 − e− L t .
R
Fig. S169.1
Solutions 229
The power dissipated in the resistor by Joule heating is

V2 R 2
P = RI 2 = 1 − e− L t .
R
2
P~ I
Fig. S169.2
Figure 169.2 is a sketch of the power–time graph, and shows that the
power increases monotonically. Whilst the rate of change of power initially
increases, it later reaches a maximum, and beyond this point decreases
monotonically to zero as shown in Fig. S169.3.
dP
dt
Fig. S169.3
The fastest rate of change of energy occurs, for both the inductor and
resistor, when I 2 is changing most rapidly. In part (i) we found that this
happens when I = V /2R. Substituting this value for I into the above
expression for the power, shows that it occurs when e−Rt/L = 12 , and hence
that t = (L/R) ln 2 ≈ 0.69 (L/R).
Note. (i) The rate of energy loss from the battery is VI, and so is proportional
to the current, which increases monotonically.
(ii) The fastest increase in the current takes place at t = 0, but that in I 2
(which is proportional to the magnetic energy of the coil) occurs later.
(iii) Just for fun you may wish to solve the twin of this problem, in which
the inductor is replaced by a capacitor.
S170 (i) These circuits are tricky to analyse using differential equa-
tions, but become straightforward if complex impedances are employed. The
impedance of an inductor of inductance L at angular frequency ω is iωL,

whilst that of a capacitor of capacitance C is 1/iωC. Here i is the square
root of −1, i.e. i2 = −1. Combining impedances is governed by the same
rules as those which apply to resistances R1 and R2 in series [R = R1 + R2 ]
and in parallel [R = R1 R2 /(R1 + R2 )].
Circuit (a) thus has total complex impedance
iωL × iωL (1/iωC) × (1/iωC) iωL 1 1 − ω 2 LC
Z= + = + = .
iωL + iωL (1/iωC) + (1/iωC) 2 2iωC 2iωC
The magnitude of the current drawn from the source is therefore
|V0 | V0 |2iωC| 2x
|I| = = = CV0 ω0 ,
|Z | |1 − ω LC|
2 |1 − x2 |
when LC is written as ω0−2 and ω/ω0 is written as x. This is plotted
in Fig. S170.1(a). It will be seen that, theoretically, the current increases
without limit as ω approaches ω0 ; in practice the source will be unable
to supply such a current. In any case, real inductors and connectors have
non-zero resistance, and the calculated peak in Fig. S170.1(a) is then of finite
amplitude.
In a similar way the impedance of circuit (b) is
iωL × (1/iωC) iωL × (1/iωC)
Z= + ,
iωL + (1/iωC) iωL + (1/iωC)
leading to
2ωL CV0 ω0 |1 − x2 |
|Z | = and |I| = .
|1 − ω 2 LC| 2x
This is plotted in Fig. S170.1(b). It will be noticed that at ω = ω0 the circuit
has an infinite impedance and no current is drawn. It is also of interest that
the x-dependences in Fig. S170.1(a) and (b) are reciprocal functions.
I I
1 x 1 x
(a) (b)
Fig. S170.1
Solutions 231
(ii) Inductances of L, 2L, L/2, and capacitances of C, 2C, C/2, constructed

from identical components in series or √ parallel
√ as needed, are used to make
circuits resonant at ω0 , ω0 /2, 2ω0 , ω0 / 2, 2ω0 . As shown in Fig. S170.2,
for two of these there are two alternatives.
or or
0 /2 0/ 2 0 2 0 2 0
Fig. S170.2
S171 When the switch is closed, currents, as shown in Fig. S171.1, flow
round the circuit.
2I I
S
I I I
Fig. S171.1
In the period immediately after opening the switch, the current flowing in
each coil is practically unchanged; if this were not the case, there would be
a rapid change in its magnetic flux which would induce a very high voltage
in the coil. Currents of 2I and I therefore continue to flow in the coils, and
these determine the currents flowing through the lamps (see Fig. S171.2).
2I I
S
2I I I
Fig. S171.2
This means that the lamp closest to the switch suddenly flashes, but the
brightness of the two other lamps does not change. This all takes place in a
very brief period and later all three lamps fade and go out.
S172 The cross-sectional area of the space to be filled is fixed, whilst that
of the wire varies as d2 . Thus n ∝ d−2 . The resistance of one turn is inversely
proportional to the cross-sectional area of the wire, i.e. varies as d−2 , and
hence the resistance per unit length of the solenoid is R ∝ nd−2 ∝ d−4 . The
flux density B is ∝ nI and therefore the required current I ∝ n−1 ∝ d2 . The
2
heat dissipated per unit length is RI 2 , which is ∝ d−4 d2 , i.e. independent
of d. Thus (within limits) it does not matter what diameter wire is chosen so
far as the heating effect is concerned.
S173 The equation describing the forces that keep the free electron
moving on a circular track (inside the cylinder) is
eE ± erωB = mrω 2 ,
where e is the charge of the electron, m is its mass, r is its distance from the
axis of rotation and E is the strength of the electrostatic field produced in
the cylinder by the charge distribution. The ± sign shows that the Lorentz
magnetic force can be directed either inwards or outwards, depending on the
sense of rotation of the cylinder. Re-arranging the equation of motion,
mω 2
E= ± ωB r ≡ Kr,
e
shows the field strength to be directly proportional to the radius.
r ∆r
E
L
Using Gauss’s law, the electric charge density in the cylinder can be found.
Consider the thin cylindrical shell shown in the figure and denote the electric
charge density at distance r from the axis by ρ(r). Electric flux of magnitude
of 2πrLE (r) enters the shell and a flux of 2π (r + ∆r)LE (r + ∆r) leaves it.
Solutions 233
According to Gauss’s law

K(r + ∆r) 2π (r + ∆r) L − Kr2πrL = (1/ε0 )ρ 2πr∆rL.
From this we obtain
2ε0 ωm eB
ρ= ω±
e m
for the charge density within the cylinder, noting that it is independent of r.
The density can be positive, negative or zero – depending on the directions
and magnitudes of the magnetic field and the angular velocity. It is zero,
if ω = |e| B/m. For this situation, positive and negative charges are not
separated inside the cylinder as the centripetal force provided by the Lorentz
effect is just what is needed to sustain circular motion.
Note. In order to get an idea of the orders of magnitude involved, assume
that the magnetic field is comparable with the Earth’s magnetic field near
to the Equator, i.e. B = 3 × 10−5 T. Zero charge density corresponds to an
angular velocity ω = eB/m = 5.3 × 106 s−1 , which is more than 50 million
revolutions per minute! Such rapid rotation cannot be realised in practice,
since no material could stand it.
S174 The electric charge distribution at any point is the same in the
rotating frame K as in the laboratory (inertial) frame K, because the
charge density is proportional to the number of electrons in unit volume,
and both the number and the volume are clearly invariant. It thus follows
that the electric charge density is homogeneous in the rotating frame and
equal to
mω 2
ρ = ρ = 2ε0 ±Bω + ≈ ±2ε0 Bω.
e
The force F acting on a charged particle must be the same in both frames of
reference (as demonstrated by the fact that the elongation of a spring which
measures the force is independent of the frame of reference). Thus F = F.
In the frame rotating with the cylinder the free electrons of the metal are
at rest, and thus the net force exerted on them must be zero (otherwise they
would move); if the centrifugal force is neglected this implies that the electric
field must also be zero (E = 0). Further,
Q(E + v × B) = F = F = Q[E + (v + vrel ) × B ],
where vrel is the relative velocity of the two frames, which is different at
different points in the cylinder. Recalling that E = ωBr and noting that
vrel × B is directed radially with magnitude ωB r, we conclude that the
magnetic field is (vectorially) the same in both frames, i.e. B = B. It follows

that in a rotating frame of reference electric charges can exist without an
associated electric field, as illustrated in Fig. S174.1.
E(r) = 0 +
+ + +
+
+ + +
+ +
+ +
+
+
+ + +
+
q=0 +
+
+
Fig. S174.1
Similarly, it can be proved that in the frame of reference of an observer

spinning in a homogeneous magnetic field, an inhomogeneous electric field
exists, although there are no electric charges present (Fig. S174.2).
E(r) = 0
q= 0
Fig. S174.2
This means that Gauss’s law (the connection between electric flux and the
charges responsible for it) is not obeyed in rotating frames of reference. This
surprising result has to be allowed for, even in low-speed (non-relativistic)
motion.
S175 It is a mistake to consider this question as a one-dimensional
problem. The magnitude of the induced electric field calculated by Jack is
correct [E(r) = rBω], but the electric field vector within the rotating spoke
is not parallel to it; it is radial as shown in Fig. S175.1. This means that
the electric flux is non-zero over the curved surface of the cylindrical rod.
Solutions 235
Moreover, as shown below, for an elementary cylinder the total flux across
this surface is equal in magnitude to the flux across its outer end. It is
perhaps surprising that half of the electric flux escapes through the curved
surface of the cylinder.
E
x
Fig. S175.1
The correct result can be calculated using Gauss’s law. The appropriate
Gaussian surface is shown in Fig. S175.2. The net electric flux is
Ψ = E(r + ∆r)h(r + ∆r)θ − E(r)hrθ
= Bωhθ (r + ∆r)2 − r 2 ≈ 2Bhωθr∆r.
B E
x E
h
h
r ∆r
Fig. S175.2
We can relate this to the charge density using

1 ρhrθ ∆r
Ψ= ρ ∆V = ,
ε0 ε0
which gives a value for ρ identical to Jill’s.
Note. The sign of the charge density can be either positive or negative,
depending on the directions of both the rotation and the magnetic field.
The charge distribution inside the spoke is homogeneous, but the net charge
on the spoke is zero as there are charges of opposite sign on the spoke’s
surface. The surface distribution is complicated and cannot be found by
elementary methods.
S176 Resolve the magnetic field of the Earth into its horizontal and
vertical components. The vertical component induces no current in the ring,

since its flux through the ring is always zero. Let the horizontal component
of the magnetic field be B and the angular velocity of the ring be ω. The
magnetic flux threading the ring is
Φ = πr 2 B cos ωt,
and the induced voltage is V = −(dΦ/dt) = πr 2 Bω sin ωt. The current,
V r 2 πBω
I= = sin ωt,
R R
flowing in the ring induces a magnetic field at the centre of the ring of
magnitude
I πrBω
BI = µ0 = µ0 sin ωt.
2r 2R
The direction of the magnetic field BI is perpendicular to the plane
of the ring and rotates with it. Resolve the vector BI into a component
parallel to B and a component perpendicular to it. The parallel component
is proportional to cos ωt × sin ωt = 12 sin 2ωt, which averages to zero over
time. The perpendicular component can be written as
πrBω πrBω
B⊥ = µ0 sin2 ωt = µ0 (1 − cos 2ωt).
2R 4R
This expression consists of a term varying (relatively rapidly) with time and
which, on average, is zero, and a constant term that causes the magnetic
needle to deviate by α = 2◦ from its original (north–south) direction. Since
the needle aligns itself with the direction of the (average) resultant field,
B⊥ πrω
tan α = = µ0 .
B 4R
The needle will make small oscillations about the above position, with an
amplitude determined by the mechanical and magnetic characteristics of the
needle and by the damping forces.
It is interesting to note that the angle of deviation of the magnetic needle
does not depend on the magnitude of the Earth’s magnetic field. The only
important thing is that its horizontal component is non-zero. The resistance
of the ring can be calculated from the above formula and is found to be
1.78 × 10−4 Ω.
S177 With I and i as defined in the hint, the required voltmeter reading
Solutions 237
is given by RI. In both cases, applying Kirchhoff’s laws yields the equations:
θ 1
RI + r(I + i) = a2 Ḃλ,
2π 2
θ
ri + rI = πa2 Ḃ.
2π
In case (a), λ = θ and solution of the simultaneous equations shows that I,
and hence the voltmeter reading, is zero.
In case (b), λ = θ − sin θ and straightforward but slightly lengthy algebra
shows that the voltmeter reading will be
2π2 Ra2 Ḃ sin θ
|V | = .
4π2 R + rθ(2π − θ)
S178 Consider two touching discs each of radius R = L/2π as shown in

Fig. S178.1.
Fig. S178.1
If the discs are placed with their plane perpendicular to a homogeneous

magnetic field whose strength changes uniformly with time, the voltage
induced in the piece of wire wrapped around their edge is
∆B
V = 2πR 2 .
∆t
V V V
(a) (b) (c)
Fig. S178.2
Now twist the disc on the right by 180◦ about its symmetry axis e
(Fig. S178.2(a)). Its top (dark) side then becomes its bottom side (Fig.
S178.2(b)). Turn the same disc again by 180◦ , but this time about axis t
(Fig. S178.2(c)). At the end of this, the dark sides of both discs are on top
and the perimeter is exactly the same as that of the Moebius strip mentioned
in the problem.
Thus, with the strip in a uniformly changing magnetic field, the voltmeter
reads
∆B kL2
V = 2πR 2 = .
∆t 2π
This value is much higher than what one would naı̈vely expect, if reasoning
from the area of the paper band. The area of the (one-sided) surface covering
the Moebius strip is not the same as the area of the paper band, and for
narrow strips it is, in fact, much larger!
The induced voltage can also be calculated by cutting the wire at the
‘twist’ into two ‘coils’ of one turn each, and adding up the algebraic values of
the voltages induced in each turn (taking account of their directions). In the
present case, the directions of the two turns are the same, and therefore the
voltage V0 = kπR 2 in one turn is doubled to give a total voltage of V = 2V0 .
S179 The current at time t is I = kt in the outer coil, and 2I = 2kt in
the inner one, where k is a constant. Because of these currents the magnetic
field in the outer coil is B = µ0 nkt, whilst in the inner one it is 3B, where n
is the number of turns per unit length. The magnetic flux enclosed by the
particle’s trajectory of radius r is
Φ = πR 2 × 2B + πr 2 × B = 2R 2 + r 2 πµ0 nkt.
The (constant) magnitude of the induced electric field E can be calculated

from the rate of change of magnetic flux with time:
dΦ
E × 2πr = = 2R 2 + r 2 πµ0 nk,
dt
and so
2R 2 + r 2 µ0 nk
E= .
r 2
The charged particle is held in its circular orbit by the magnetic field, and
so, from the zero net radial component of the force acting on it, we obtain
mv 2
= qvB. (1)
r
The particle is accelerated along its circular orbit by the tangential compo-
nent of the net force according to mat = qE, where m is the mass and q the
electric charge of the particle.
Solutions 239
As the magnitude of the electric field is constant, the speed of the particle
increases uniformly with time,
qE 2R 2 + r 2 µ0 nk q
v = at t = t= t.
m r 2 m
Inserting this and the value of B into equation (1), we get
m 2R + r µ0 nk q
2 2
t = qµ0 nkt,
r r 2 m
which is satisfied if
2R 2 + r 2 √
= 1, i.e. r = 2R.
2r 2
S180 The changing magnetic field induces an electric field in the ring.
Let us imagine the ring divided into small sections each of length ∆s and
denote the tangential component of the induced electric field by Et (in the
general case Et can vary from point to point). The charge on a small section
of the ring is
∆s
∆Q = Q ,
2πr
where r is the radius of the ring. The force exerted on it is
∆Ft = ∆Q Et
and the resultant torque is
∆τ = r ∆Ft .
The total torque experienced by the ring is thus
∆s Q
τ= ∆τ = rQ Et = Et ∆s.
2πr 2π
Identifying the expression Et ∆s as the induced electromotive force along
the ring, which is directly proportional to the rate of change in the magnetic
flux, we have
∆Φ ∆B
Et ∆s = − = −πr 2 .
∆t ∆t
As a result of the torque, the ring, which has a moment of inertia I = mr 2 ,
starts to spin with angular acceleration α. During a time interval ∆t its
angular velocity changes by
τ Q ∆B 1 Q
∆ω = α ∆t = ∆t = −πr 2 ∆t = − ∆B.
I 2π ∆t mr 2 2m
Since the magnetic field strength increases from zero to B, the final angular
velocity of the ring will be
QB
ω=− .
2m
Note: (i) The negative sign shows that the direction of the angular velocity
vector is opposite to the magnetic induction if Q is positive.
(ii) It is interesting to note that the final angular velocity does not
depend on the radius of the ring, the time over which the magnetic flux
changes, or even on how the magnetic flux increases with time.
(iii) In our calculation we ignored the magnetic field produced by the
rotating ring.
(iv) Except in the case of a cylindrically symmetric uniform field, it is
not possible to find the actual value of the induced electric field within the
ring because the geometrical structure of the magnetic field is unknown
and we do not know the position of the ring in the magnetic field. We can
determine the total induced electromotive force, but not the electric field
itself.
S181 The resultant magnetic field B is the sum of the magnetic fields of
the Earth and the coil, B0 and B respectively, i.e.
B = B0 ± B. (1)
The current flowing through the coil is determined by the induced voltage
Vi and the resistance R,
Vi r2 ω 1
I= = B , (2)
R 2 R
where the induced voltage has been calculated from the rate at which a disc
radius cuts the field’s magnetic flux. The magnetic field produced by the coil
itself is
B = µ0 nI. (3)
From the three equations above, B, B and I can easily be determined.
The two signs occurring in equation (1) allow for both positive and negative
values of the angular frequency. The value of ω is taken to be positive if
the magnetic field of the coil acts in the same direction as that of the Earth.
The following results are obtained for the resultant magnetic field and the
current:
2RB0 B0 r 2 ω
B = , and I = .
2R − µ0 nr 2 ω 2R − µ0 nr 2 ω
As expected, when the disc is at rest, the current is zero and the resultant
magnetic field inside the coil is simply B0 , the magnetic field of the Earth.
Solutions 241
When the direction of rotation is such that the field in the coil opposes
the external magnetic field (ω < 0) the resultant decreases asymptotically to
zero as the (negative) angular frequency of the disc increases. At such high
rotation speeds, the current flowing in the coil tends to −B0 /µ0 n (the value
needed to cancel the magnetic field of the Earth).
Rotation of the disc in the opposite direction, (ω > 0), causes the resultant
magnetic field to increase. This leads to a higher voltage being induced
and a larger current flowing, which in turn leads to a further increase
in the magnetic field. Under these positive feedback conditions, both the
magnetic field and the current tend to infinity as a particular ‘critical’ angular
frequency, ωcrit = 2R/(µ0 nr 2 ), is approached, as shown in Fig. S181.1. Such
a state is obviously not realised in practice. If the angular frequency is
increased too much, the current and the heat given out by the coil increase
until the wires burn away!
B
0
crit
(a)
I
crit
B0
µ0 n
(b)
Fig. S181.1
The strange behaviour of the system can be more easily understood if
the relationship between the current in the coil and the resultant magnetic
field is represented graphically as in Fig. S181.2. According to equation (2),
I is proportional to B in such a way that the coefficient of proportionality
depends on ω.
This is represented by a straight line through the origin, with a gradient
proportional to ω. Equations (1) and (3) show that B = B0 + µ0 nI; this is
also a linear relationship, but its graph does not pass through the origin.
I x > x crit
x = x crit
x <0
0 < x < x crit
B´
B0
B0
l 0n
Fig. S181.2
The gradient of the latter is 1/(µ0 n) and independent of ω. The intersection

of these two straight lines determines the actual current and the resultant
magnetic field. If ω = ωcrit , then the gradients of the two straight lines are
the same and the equations have no solution. In fact, the critical angular
frequency is so high that the corresponding state cannot even be approached
in practice.
The Joule heat given out by the coil has to be equal to the mechanical
work done in rotating the disc. The electrical power is Pel = I 2 R, while
the mechanical work done per second is the product of the torque and
the angular frequency, Pmech = Mω = B Ir 2 ω/2. (The torque M has been
calculated as the product of the force B Ir and the average perpendicular
distance of its line of action from the axis, r/2.) Using the relationship
between B and I, it can be verified directly that Pel = Pmech .
The strange device described in this problem is called a unipolar dynamo.
S182 The total magnetic flux at the position of the ring is made up of
that due to the external magnetic field and the effects of self-induction,
Φ = Bz πr02 + LI.
Any change in magnetic flux induces a current in the ring in accordance

with
∆Φ
RI = .
∆t
However, this has to be zero since the ohmic resistance of a superconducting
ring is zero. Accordingly, the magnetic flux through the ring has to be
constant, i.e.
Φ = B0 (1 − αz) πr02 + LI = constant.

Solutions 243
From the initial conditions (z = 0, I = 0), the value of the constant is

Φ = B0 πr02 .
The current in the ring can be determined using the above equations which
give
1
I= B0 α πr02 z.
L
The Lorentz force acting on the ring (which can only be vertical, because of
the symmetry of the assembly) can be expressed as
1
Fz = −Br I(z) 2πr0 = −B0 β r0 B0 α πr02 2πr0 z = −kz.
L
The Lorentz force is thus directly proportional to the vertical displace-
ment of the ring, with the coefficient of proportionality calculable from the
given data. (This result is only valid for small displacements, since the mag-
netic induction is not adequately described by the given formulae for large
ones.)
The equation of motion of the ring is
maz = Fz − mg = −kz − mg.
This means that the ring makes harmonic oscillations about the equilibrium
position z0 = −mg/k with
z(t) − z0 = A cos ωt,
where ω = k/m. From the initial conditions it follows that A = −z0 , and
so
g
z(t) = (cos ωt − 1) .
ω2
The vertical z-coordinate is never positive, and it follows that the Lorentz
force always points upwards, being zero at the topmost point of the oscilla-
tion. The current always flows in the same direction around the ring.
Substituting the numerical data gives ω = 31.2 s−1 and A = 1 cm. The
time dependence of the current flowing in the ring can be expressed in terms
of z(t) as
1 1
I= B0 α πr02 z(t) = B0 α πr02 A(cos ωt − 1).
L L
The maximum value of the current, which flows at the bottom of the
oscillation, is Imax = 39 A.
S183 Let us apply the law of conservation of energy for a particle of

mass m and charge Q:
1 2 cos θ 1 QK cos(π/2)
mv + QK 2 = mv02 + = 0.
2 r 2 r2
We can then express the velocity of the bead at angle θ as
QK
v= −2 cos θ, ( π2 ≤ θ ≤ π).
mr 2
− QEr
v
The circular motion needs a radial force component of mv 2 /r. The radial
component of the force on a unit charge due to the dipole (i.e. the effect of
the radial component of the electric field) can be calculated as minus the
derivative of the electric potential with respect to r,
∂Φ K cos θ
Er = − =2 .
∂r r3
Using the earlier expression for the velocity, we notice that QEr is just equal
to −mv 2 /r, the required centripetal force. Thus the string does not need to
exert any force on the bead to sustain circular motion. If the string were
not there, the bead would move along a circular path until it reached the
point opposite its starting position. The bead would stop there, and then
repeatedly retrace its path executing a periodic motion.
Note. The time dependence of this motion is just the same as that of a
simple pendulum subject to gravity after release from a displacement of 90◦ .
S184 Imagine that you are sitting in a frame of reference moving hor-
izontally at constant speed v perpendicular to the magnetic field lines, and
you are carrying a charge q. In a frame of reference fixed to the ground, the
Solutions 245
charge moving with velocity v is observed to experience a force of magni-

tude qvB (pointing upward or downward, depending on the direction of the
motion).
In the moving coordinate system, the charge is at rest and the Lorentz force
does not act on it, though it still ‘feels’ the force. (The presence of a force
caused by some interaction cannot depend on the coordinate system in which
it is described.) This seeming contradiction can be resolved by transforming
the electric and magnetic fields when changing from a stationary coordinate
system to a moving one. In the present case, there is only a magnetic field in
the stationary frame of reference and no electric field. In a frame that moves
(at a speed much less than the speed of light), this same magnetic field can
be observed, but an electric field of field strength E = vB is also present.
This provides the force F = qE = qvB that would otherwise be missing in
this frame.
B
q
Let the velocity of the moving frame of reference be such that the electric
force described above is the same as the gravitational force mentioned in the
problem, i.e.
qvB = qE = F = mg.
We now describe the motion of the observed body in this coordinate

system. Since the effects of the electric and gravitational fields cancel each
other out and the body moves at velocity −v in this system, a force of −qvB
(the magnetic Lorentz force) acts on the body. This makes it move on a
circular track spiralling downward according to qvB = mv 2 /r, where r is
the radius of the circle and is given by r = mv/qB. Using the values of the
velocity and the radius, the time in which the body makes one orbit is found
to be T = 2πm/qB, independent of the velocity v.
In the coordinate system fixed to the ground, a uniform rectilinear motion
is superimposed on this uniform circular motion. The particle therefore
follows a cycloidal path as shown in the figure, falling down to 2r = 2mv/qB
and then rising again to reach its initial height in time T . The corresponding
horizontal displacement is T v = 2πr. At this point the particle stops for a
moment, before starting out on a new cycloid curve.
S185 When the magnet is falling with the constant terminal speed and
it covers a distance L (L h), its potential energy loss mgL is converted
to Joule heat by the currents induced in L/h loops. Denoting by Q the heat
produced in one particular loop, we can write:
L
mgL = Q, i.e. Q = mgh.
h
On what quantities can Q depend? Since the power dissipated in a resistor
by a given voltage is inversely proportional to its resistance, Q ∝ R −1 , when
the magnet is moving with velocity v0 we can express the heat function in
the form
1
Q = f(v0 , µ, r).
R
Thus it is a general function of the terminal velocity v0 , the magnetic moment
µ and the radius r of the circular loop; moreover the formula could involve
the vacuum permeability µ0 .
Now consider the units of the individual quantities. The units of RQ are
[RQ] = kg2 m4 s−5 A−2 ,
whilst
[v0 ] = m s−1 , [µ] = A m2 , [r] = m, [µ0 ] = kg m s−2 A−2 .
Thus it must be that
RQ(v0 , µ, r, µ0 ) ∝ v0 (µ0 µ)2 r −3 .
But we know that Q = mgh, and so
mghRr 3
v0 ∝ .
µ2
4,
1
It follows that the terminal speed ratios in question are: 2, 2, 2 and 8,
respectively.
S186 The speed of the electrons remains constant in the frame of refer-
ence fixed to the vacuum chamber (S ) because the magnetic field can only
change the direction of the velocity of the moving charge. Consider another
frame of reference (S ) moving parallel to the wire with constant speed v0
with respect to the first one. We can write the Lorentz force experienced by
a particle of charge Q in either frame:
F = F = Q(v × B) = Q(v + v0 ) × B = Q(v × B ) + QE .
We first note that the electric current in both frames of reference must
Solutions 247
be the same. In frame S , there are moving electrons and static positive ions
whilst the wire itself is neutral. In reference frame S , the speed of electrons
and so the current of electrons will be different, but the current of the moving
positive ions just compensates for this change. The consequence of this is
that there is the same magnetic field in both frames (B = B). The equation
above shows that in frame S we have, in addition to the unaltered magnetic
field, an electric field given by v0 × B. (Note that the transformation of a
single magnetic field from S to S does not generally alter the magnetic field
if the speed of the moving system is much less than the speed of light.)
Let us now describe an electron’s motion in reference frame S . In this
frame there is an electric field (perpendicular to the wire) of strength
µ0 v0 I 1
E(r) = v0 B(r) = ,
2π r
where r is the distance from the wire. Using the analogy of the cylindrical
capacitor, we can find an appropriate electric potential function for this field
(i.e. one that is such that minus its derivative with respect to r gives the
field), namely,
µ0 v0 I
U(r) = − ln r.
2π
In the√frame S , the initial speed of the electron (at a distance r from the
wire) is 2v0 , and the electron just stops when it has approached within r0 /2
of the wire. We can apply the work–energy theorem to the motion of the
electron as follows,
1 √ 2 µ0 v0 IQ r0 µ0 v0 IQ
m 2v0 =− ln − ln r0 = ln 2.
2 2π 2 2π
Inserting numerical values into this expression, gives the initial speed of the
electron as
µ0 Q
v0 = I × ln 2 ≈ 2.46 × 105 m s−1 ≈ 250 km s−1 .
2mπ
Note. (i) The initial speed of the electron is very large on a macroscopic
scale, but is very small for electrons; an electron attains this speed if it
is accelerated through a potential difference of as little as 0.2 V. Since
250 km s−1 is much less than c, we justifiably ignored the relativistic mass
increase of the electron in this problem.
(ii) It is interesting to observe that if the electron cannot approach the
wire closer than r0 /2, then the maximum distance of the electron is also
limited – it cannot be further away than 2r0 . More generally, if the minimum
distance is r0 /n, then the maximum distance must be nr0 . This can be proved
using another frame of reference moving with velocity −v0 .
S187 The Lorentz force
F = Q(E + v × B)
acts on a small body of charge Q moving with velocity v in a field of electric

field strength E and magnetic field strength B. In a frame of reference
moving at velocity v0 relative to the original one, the velocity of the particle
is v = v−v0 . If all physical quantities in this new reference frame are denoted
by primes, the expression for its Lorentz force is
F = Q (E + v × B ).
Now compare the two frames of reference. Any force acting on the particle
which can be detected (e.g. through the acceleration it produces) cannot
change, i.e. F = F. The same is true for the electric charge: Q = Q . (If
particle charges depended on their velocity then an initially neutral body
would show a net electric charge when heated – no trace of this is seen in
nature.)
Transforming the velocities yields
Q(E + v × B) = Q(E − v0 × B + v × B ).
This relation has to be satisfied for any v, and specifically for stationary
particles. Therefore
B = B leading to E = E + v0 × B.
It can be seen that electric and magnetic fields are not independent physical
quantities, their values depend on the (velocity of the) frame of reference in
which phenomena are described (see also P184). We now apply these general
relationships (twice) to the problem in hand.
Consider the frame of reference of the liquid. To get to this frame from
the capacitor’s frame requires a transformation with v0 = −v. The magnetic
field in this new frame is unaltered as B = B, but an electric field E of
strength −v×B is also present. The stationary liquid is polarised by this field,
and the electric field strength is consequently reduced by a factor 1/εr to
E = −v × B (1/εr ). Returning to the original (capacitor) frame of reference
requires a further transformation, but this time with v0 set equal to +v.
Again the magnetic field is unchanged, B = B = B, but the electric field is
given by
1
E = E + v × B = −v × B (1/εr ) + v × B = 1 − v × B.
εr
Solutions 249
Consequently, a p.d. of
1
V = vBd 1 −
εr
appears between the plates.
For non-polarisable materials (εr = 1), this voltage is obviously zero, whilst
in the case of ‘easily’ polarisable materials (metals), V = vBd. This latter
equation, describing the Hall effect, can be interpreted in the following way.
The charges in the conductor moving in the magnetic field are displaced
by the Lorentz force. The sideways migration of charges continues until
an electrostatic field strong enough to balance the Lorentz force (which is
proportional to the magnetic field) has been built up. From the condition for
this, QE = QvB, the above relationship, V = Ed = vBd, can be recovered.
S188 It is not correct in quantum mechanics to describe processes in
terms of their various parts occurring in a particular order, nor to describe
classically conserved quantities as being ‘borrowed’. However, the ‘classical’
description of a quantum effect usually gives a qualitatively correct picture
if the ‘borrowing’ of a quantity is interpreted as meaning that the more of it
that is borrowed, the less likely the process is to occur.
The volume of the uranium nucleus is equal to the sum of the volumes of
its fission products, but the total surface area of the fission products is greater
than that of the original uranium nucleus. This means there is a temporary
loss of (binding) energy when the process starts (this is the energy that has
to be ‘credited’ to initiate the fission process), which is later refunded by the
decrease in Coulomb energy of the system when the daughter nuclei move
away from each other. If the uranium nucleus were to split into three fission
products, the energy required at the start of the process would be so much
higher than for two, that, both in theory and in practice, the corresponding
probability would be negligible.
S189 Even if atoms are heated to a few thousand degrees, their thermal
energy is still much smaller than their binding energy per nucleon. For this
reason the rate of nuclear reactions is usually independent of temperature.
However, the isotope 7 Be transforms into 7 Li via K-capture (the capture
of an electron from the innermost K shell). At a temperature of a few
thousand degrees the quantum of thermal energy kT is comparable with the
ionisation energy of the two innermost electrons of the Be atom (consider the
tail of the Maxwell velocity distribution). If some fraction of the beryllium
atoms becomes ionised, the probability of nuclear electron capture occurring
decreases by approximately the same percentage. Thus, at such temperatures,
the radioactivity, and hence the half-life, of beryllium can be affected.
S190 In equilibrium, abundances are proportional to half-lives. In 10−3 kg

of thorium, almost exclusively 232 Th, there will be
6.02 × 1023
= 2.595 × 1021
232
atoms of thorium. Consequently, there will be
56 s
2.595 × 1021 × ≈ 3.3 × 105
1.41 × 1010 y
atoms of radon.
Over the stated time span both isotopes of thorium will contribute to the
radon population. Each contributes an amount which first increases at a rate
governed by the slowest decay rate in the relevant intermediate decay chain,
then remains steady at the equilibrium value, and finally decays at a rate
governed by the decay rate of the parent.
Thus the contribution from the 22890 Th rises with a half-life of 3.64 days
and decays with a half-life of 1.91 years. Similarly, the contribution from the
232
90 Th rises with a half-life of 5.7 years and decays with a half-life which is
essentially infinite. The total 220
86 Rn present is the sum of these two as shown
in the figure.
atoms/ 10 5
2
86 Rn
1
220
−3 −2 −1 2 3
10 10 10 1 10 10 10
Time / years
The initial rise in the curve is due to radon produced from the 22890 Th
present in the initial sample. Most of this has decayed away by the time
that the radon originating from the original 23290 Th becomes significant.
After about eight years the numbers of radon atoms from the two different
ancestors become equal at about 0.9 × 105 each. After that the number again
rises to the equilibrium value of 3.3 × 105 with virtually all atoms present
having 232
90 Th as their ancestor; it would remain at this value for about
1010 years.
Solutions 251
S191 The relativistic energy and momentum of a particle with (rest) mass
m and speed v are
mc2 mv
E= and p= ,
1 − v 2 /c2 1 − v 2 /c2
where c is the velocity of light. These two equations can also be combined
as E = p2 c2 + m2 c4 .
As a result of particle–particle collisions, new (additional) particles, e.g.
proton–antiproton pairs, can be produced. Antiprotons have the same rest
mass as protons. The total energy of the final particles is lowest if they
stay together, moving with negligible speed (in the reference frame of their
centre of mass). For this situation, the four particles (three protons and an
antiproton) can be considered as one particle of mass 4m, whose momentum
is equal to the initial momentum of the accelerated proton. The law of
conservation of energy prescribes that
p2 c2 + m2 c4 + mc2 = (4m)2 c4 + p2 c2 .
√
Squaring both sides of the equation gives p = 48mc and Ep = 7mc2 . The
protons thus have to gain a kinetic energy of Ep − mc2 = 6mc2 ≈ 6 GeV
in the course of the acceleration, which requires an accelerating voltage of
6 × 109 V.
S192 Electrons of mass m and charge e can move freely in the wall of
the Faraday cage to reach a state in which the resultant of the electric and
gravitational forces acting on them is zero. This requires a small condensation
of the electrons at the bottom of the metal wall, leading to a surplus
of positive charge (lack of electrons) at the top. The charge displacement
continues until the magnitude of the resulting vertical electric field is E0 =
mg/|e|.
As a homogeneous electric field is formed in the wall of the cage, the same
field has to be present inside the cage itself. If this were not the case, the
motion of a charged particle vertically upwards or downwards in the wall of
the cage, and then back to its initial position through the inner part of the
cage, would form the basis of a perpetual motion machine – and obviously
this is not possible.
Thus, as a result of charge displacement in the wall, an electrostatic field
is set up which acts on a free electron placed inside the cage with the same
force as gravity does. In other words, the cage shields the gravitational field
as well! This shielding, however, only works for electrons; the condition of
zero net force is not satisfied for particles with a different ratio of charge
to mass (e/m). In the case of positrons, for example, the resultant force is
E0 |e| + mg = 2mg, i.e. these particles start moving vertically downwards with
an acceleration of 2g.
Numerically, E0 = 5.6 × 10−11 V m−1 , a very small value, and therefore
the phenomenon described cannot be seen under ordinary circumstances.
Electrons cannot be placed into a Faraday cage with no initial speed, since
even in the case of the photoelectric effect, in which electrons are pushed out
of the metal with an energy of only 0.1 eV, they still travel with a speed of
200 km s−1 . The relative change in velocity after 1 metre of free fall is only
about one part in 1010 , which is immeasurably small.
Note. So far, we have not considered the effect of the displacement of

the charges in the wall due to the presence of the positron in the cage.
The magnitude of this effect can be calculated using the method of image
charges. A charge e at a small distance d from the wall, experiences a force
caused by the polarised charges in the metal that is the same as the force
Fimage produced by a charge of −e at distance 2d from the first charge (an
‘image charge’), i.e.
(−e)e
Fimage = k .
(2d)2
In the case of electrons or positrons, this force is comparable to their weight

if d ≈ 2 m. Therefore if somebody wanted to verify the conclusions of the
above reasoning using experimental data, the experimenter would need to
build a Faraday cage large enough to allow the particles to be more than a
few metres away from the wall. If it were smaller than this, it would not be
the shielding effect of the cage which was being examined, but the effect of
polarising the charges in the walls.
S193 Let m and M denote the masses of the positron and the proton,
respectively (M ≈ 2000m), and let e denote the elementary charge. Because
of the large mass ratio, the protons will hardly have moved by the time the
positrons are already far away. We equate the energy of the initial state to
that of the one in which the positrons have moved much further than 1 cm
away and are travelling at speed v1 :
e2 4 2 e2 1 1
+√ = √ + 2 mv12 .
4πε0 a 2a 4πε0 2a 2
Substitution of the numerical data yields v1 = 350 m s−1 . Thereafter, the

speed of the positrons does not change significantly.
The protons, on the other hand, although hardly moving whilst the
positrons are close, do repel each other and are accelerated to a speed
Solutions 253
v2 . Conservation of energy can again be applied and now gives

e2 1 1
√ = 2 Mv22 .
4πε0 2a 2
This yields the value v2 ≈ 2.7 m s−1 .
S194 The general equations of conservation of energy and momentum
for Compton scattering are:
hf0 + m0 c2 = hf + mc2 ,
hf0 hf
= cos α + p cos β,
c c
hf
sin α = p sin β,
c
2 2
mc2 = m 0 c2 + p 2 c2 ,
where f0 and f denote the frequencies of the incident and scattered photon
respectively, α and β are the photon and electron scattering angles (see figure)
and m0 is the rest mass of the electron.
hf
c
Photon
hf
0
c a
Photon b
p Electron
For the special case considered,

hf
hf0 = m0 c2 and = p.
c
Solving this system of equations we find that
2
α = β = cos−1 ≈ 48.2◦ ,
3
and so the angle between the scattered photon and recoil electron is α + β ≈
96.4◦ . The momentum of the scattered electron is
3 m0 v
p = m0 c = ,
4 1 − v 2 /c2
yielding the speed of the electron as v = 35 c. This is of the same order of mag-
nitude as c, and consequently it was appropriate to use relativistic formulae.
S195 The energy of a photon of frequency f is hf, while its momentum

is hf/c, where h is Planck’s constant and c the velocity of light. The energy
of the photon decreases to hf after the collision, and its momentum, of
magnitude hf /c, is perpendicular to the original momentum.
Electron
p
p
f Photon x
px
Photon
It follows from momentum conservation that, as shown in the figure, the

components of the electron’s final momentum are px = hf/c and py = hf /c.
Using the principle of conservation of energy in its relativistic form, E 2 =
E02 + p2 c2 , where E0 is the rest-mass energy, we obtain
hf + E0 = hf + E02 + (p2x + p2y )c2 .

The above relationships yield the change in the wavelength of the photon as
c c hc h
∆λ = λ − λ =
− = = ≈ 2.4 × 10−12 m.
f f E0 mc
Note. This interesting quantity with the dimensions of a length – dependent
only on the mass of the electron, the speed of light and Planck’s constant,
i.e. natural constants – is called the Compton wavelength of the electron.
S196 Applying the energy formula for spherical capacitors to the ‘classi-
cal electron’ gives
1 e2
< mc2 ,
2 4πε0 r
1 e2
r> = 1.4 × 10−15 m.
4πε0 2mc2
This value is called the classical electron radius.
The classical electron is considered as a rotating sphere of this radius with
a homogeneous mass distribution, an angular frequency ω and an angular
Solutions 255
momentum corresponding to that of a rotating rigid body. Hence we obtain

5 mr ω = h/(4π). This gives the ‘equatorial speed’ as veq = rω ≈ 350c. Since
2 2
this speed is many times higher than the limiting speed of relativistic physics,
namely the speed of light, the original version of the classical electron model –
worked out by Lorentz and Abraham before the birth of quantum theory in
the early 1900s – had to be abandoned.
A modified version suggested that the electron could be a sphere of radius
r, charged on its surface, which did not rotate but would contain a magnetic
dipole to account for the experimental fact that electrons have a magnetic
momentum. Outside the sphere, a multiple of the vector E × B produced
by the electric and magnetic fields, describes the current of electromagnetic
momentum. This current carries angular momentum and with a suitable
choice of the parameters can be made to equal the measured one.
S197 Assume that the electron occupies a horizontal layer of thickness
H just above the bottom of the box. Its vertical coordinate is then known
with accuracy H. Therefore the uncertainty in its vertical momentum has
to be at least ∆p = h̄/∆x = h̄/H, where the quantity h̄ is 1/(2π) times the
Planck constant. In such circumstances, the electron has a potential energy
of
H
Epotential = mgxaverage ≈ mg ,
2
and a kinetic energy of
∆p2 h̄2
Ekinetic = ≈ .
2m 2mH 2
Thus its total energy is
1
E(H) = Epotential (H) + Ekinetic (H) = AH + B ,
H2
where A and B are constants determined by the above equations. If H is
small, the potential energy is low but the kinetic energy is high. If, on the
other hand, H is large, the kinetic energy is low but the potential energy is
high. The total energy will be at a minimum if Epotential and Ekinetic are of
the same order of magnitude. It can be shown using differential calculus that
the optimum value for the energy ratio is 2 : 1 in favour of the potential
energy. As we only want a rough estimate, the ratio of the energies can be
taken to be unity. This gives
1/3
h̄2
H≈ ≈ 1 mm.
m2 g
This is quite a large value compared with usual microscopic sizes. The reason
for this is that gravitation is very weak compared to the electromagnetic
interaction which determines the binding energies and sizes of molecules.
S198 If a nucleus of atomic number Z restricts an electron to a sphere
of radius r, then the electrostatic energy of the electron is Eel ≈ −kZe2 /r
and its momentum can be estimated to be h̄/r. If Z 1 (i.e. the nucleus
is a heavy one), the kinetic energy of the strongly bound electron can be
calculated using the relativistic formula
Ekin = p2 c2 + m2 c4 − mc2 ≈ pc,
where c is the velocity of light and h̄ is the Planck constant divided by 2π.
The total energy of the electron is
1 Ze2 h̄c
E(r) = Eel (r) + Ekin (r) = − + .
4πε0 r r
Since the fine-structure constant e2 /(4πε0 h̄c) is approximately 1/137, the total
energy of the electron for small values of r is
e2 1
E(r) = (137 − Z ) .
4πε0 r
(The above expression is not valid for large values of r, as a small, non-
relativistic momentum would have to be taken into account.)
According to the above formula, the electron would fall into the nucleus
if Z > 137; or rather, it would be confined to a small volume of nuclear
size. The figure 137 is only an estimate of the critical atomic number; more
accurate calculations, which take the finite size of the nucleus into account,
yield values around 150–160 for Zcrit .
These calculations show that an electron could be confined to the nucleus
of an element with atomic number greater than 150, if transuranic elements
of such high atomic numbers could be made at all.
S199 The speed of propagation of surface water waves depends on the
surface tension γ and the density ρ of water, and on their wavelength, λ. The
dimensions of these quantities are
N kg kg
[γ] = = 2, [ρ] = , [λ] = m.
m s m3
An expression for velocity can only be derived from these quantities if γ and
ρ, appear in the combination γ/ρ (otherwise the velocity, which involves only
Solutions 257
length and time, would have to depend upon the unit of mass). However, as
γ m3
= 2,
ρ s
this expression has to be further divided by the wavelength, and then square-
rooted in order to make the result have the dimensions of speed. In summary,
dimensional analysis dictates that the speed of propagation of capillary waves
is proportional to the reciprocal of the square-root of the wavelength,
γ 1
v∼ ∼√ .
ρλ λ
From this functional dependence (and the given data), we can conclude that
the speed of propagation of surface waves would reach that of sound in
water when their wavelengths are of the order of 10−8 cm.
Since the speed of propagation of surface waves cannot be greater than
that of sound (the molecules cannot transmit a disturbance to each other
faster at the surface than inside the matter), waves of wavelength less than
approximately 10−8 cm have no meaning. This is, in fact, the order of
magnitude of the size of water molecules!
S200 You will have noticed not only the acceleration, but also the
increasing size of the champagne bubbles. As the sparkling champagne is
super-saturated with carbon dioxide, gas is released continuously whilst the
bubbles are rising. This is why the size of the bubbles increases, as does
the buoyancy force provided by the liquid. The upthrust is proportional to
the volume of the bubble, while the viscous drag, which is also increasing,
is only proportional to the surface area of the bubble. Consequently, the
net upward force increases with bubble size. However, increasing speed also
leads to a larger viscous drag, and ultimately the bubble moves under the
influence of a collection of balanced forces.

200 Puzzling Problems

Uploaded by

Copyright:

Available Formats

200 Puzzling Problems

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

200 Puzzling Problems

Uploaded by

Copyright:

Available Formats

What is the intended audience of the book?

What is the intended audience of the book?

What are some of the challenges faced in solving the physics problems?

What are some of the challenges faced in solving the physics problems?

200 Puzzling Physics Problems is aimed at strengthening a student’s grasp of

gyula honyek graduated as a physicist from Eötvös University (ELTE) in Budapest in

Cambridge University Press

This publication is in copyright. Subject to statutory exception

First published 2001

Printed in the United Kingdom at the University Press, Cambridge

Library of Congress Cataloguing in Publication data

ISBN 978-0-521-77480-2 Paperback

Preface page vii

In our experience, an understanding of the laws of physics is best acquired

guided by the suggestions and remarks of our colleagues. In particular, we

Gravitational constant, G 6.673 × 10−11 N m2 kg−2

Some astronomical data

Some physical properties

Densitiesa , ρ (kg m−3 )

Optical Refractive Indicesb , n

P1 Three small snails are each at a vertex of an equilateral triangle of

P5 Four snails travel in uniform, rectilinear motion on a very large plane

P20 A gas of temperature T is enclosed in a container whose walls are

P26 Fish soup is prepared in a hemispherical copper bowl of diameter

A point on the rim of the bowl is pulled upwards through 10 cm, by a

P30 A charged spherical capacitor slowly discharges as a result of the

P34 A small bob is ﬁxed to one end of a string of length 50 cm. As a

What is the shape of the water ‘bell’ formed by the jets?

at its ends by my two foreﬁngers. Whilst I am slowly bringing my ﬁngers

P46 An executive toy consists of three suspended steel balls of masses

If the radius of the cylinder is R = 5 cm and the refractive index of the

P61∗∗ A pencil is placed vertically on a table with its point downwards.

What is the force acting between the plates?

Is it possible to arrange the nails so that:

P68 Find the angle to which a pair of compasses should be opened

P69∗ Threads of lengths h1 , h2 and h3 are fastened to the vertices of a

What is the tension in each thread, expressed in terms of the lengths of

Describe the subsequent motion of the beads.

One of the beads is continuously accelerated (towards the right) by a

P75∗ An open-topped vertical tube of diameter D is ﬁlled with water up

The bob is displaced in a direction perpendicular to the line of the springs

What is the length of the spring when it reaches a vertical position?

If there were no motion the water surface would be horizontal. What

it as an evidence for UFOs, or could there be an ‘Earthly’ explanation in

(i) Find the magnitude of the force F.

(i) Calculate the speed c of the loop.

P107 Sand falls vertically at a rate of 50 kg s−1 onto a horizontal

excavation of the titanium, in the course of which they formed a spherical

What is the gravitational acceleration at the centre of the circular face of

P115∗∗ Hydrogen gas is stored at high pressure in a small, spherical

According to Ampère’s computations, the magnitude B of the magnetic

P120∗∗ A direct current ﬂows in a solenoid of length L and radius R,

What charge will accumulate on each plate?

P133 A horizontal frictionless piston, of negligible mass and heat capac-

A load of weight W is now suspended from the piston, as shown in the

The surrounding earth is warm as a result of residual volcanic activity

Tiny particles moving horizontally at speed v0 bombard the body uniformly

P146∗ Two thermally insulated containers hold identical masses of water.

P151∗ A closed body, whose surface F is made of metal foil, has an

Prove that the capacitance of the deformed body is less than C.