Scholar Advanced Higher Maths Unit 3

SCHOLAR Study Guide
SQA Advanced Higher Mathematics

Unit 3
Jane S Paterson
Heriot-Watt University
Dorothy A Watson
Balerno High School
Heriot-Watt University
Edinburgh EH14 4AS, United Kingdom.
First published 2001 by Heriot-Watt University.
This edition published in 2009 by Heriot-Watt University SCHOLAR.
Copyright © 2009 Heriot-Watt University.
Members of the SCHOLAR Forum may reproduce this publication in whole or in part for
educational purposes within their establishment providing that no profit accrues at any stage,
Any other use of the materials is governed by the general copyright statement that follows.
All rights reserved. No part of this publication may be reproduced, stored in a retrieval system
or transmitted in any form or by any means, without written permission from the publisher.
Heriot-Watt University accepts no responsibility or liability whatsoever with regard to the
information contained in this study guide.
Distributed by Heriot-Watt University.

SCHOLAR Study Guide Unit 3: Advanced Higher Mathematics
1. Advanced Higher Mathematics
ISBN 978-1-906686-05-5
Printed and bound in Great Britain by Graphic and Printing Services, Heriot-Watt University,
Edinburgh.
Acknowledgements
Thanks are due to the members of Heriot-Watt University’s SCHOLAR team who planned and
created these materials, and to the many colleagues who reviewed the content.
We would like to acknowledge the assistance of the education authorities, colleges, teachers
and students who contributed to the SCHOLAR programme and who evaluated these materials.
Grateful acknowledgement is made for permission to use the following material in the
SCHOLAR programme:
The Scottish Qualifications Authority for permission to use Past Papers assessments.
The Scottish Government for financial support.
All brand names, product names, logos and related devices are used for identification purposes
only and are trademarks, registered trademarks or service marks of their respective holders.
i
Contents
1 Vectors 1
1.1 Revision exercise . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
1.2 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
1.3 Direction ratios and cosines . . . . . . . . . . . . . . . . . . . . . . . . . 8
1.4 Basic arithmetic and scalar product of vectors . . . . . . . . . . . . . . . 9
1.5 Vector product - algebraic form . . . . . . . . . . . . . . . . . . . . . . . 16
1.6 Vector product - geometric form . . . . . . . . . . . . . . . . . . . . . . . 19
1.7 Scalar triple product . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
1.8 Lines in three-dimensional space . . . . . . . . . . . . . . . . . . . . . . 24
1.9 Planes in three-dimensional space . . . . . . . . . . . . . . . . . . . . . 34
1.10 Lines and planes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
1.11 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
1.12 Proofs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
1.13 Extended information . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
1.14 Review exercise . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
1.15 Advanced review exercise . . . . . . . . . . . . . . . . . . . . . . . . . . 50
1.16 Set review exercise . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
2 Matrix algebra 53
2.1 Revision exercise . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
2.2 Basic matrix terminology . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
2.3 Matrix operations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56
2.4 Determinants and inverses . . . . . . . . . . . . . . . . . . . . . . . . . 62
2.5 Properties of matrices . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
2.6 Matrices and transformations . . . . . . . . . . . . . . . . . . . . . . . . 76
2.7 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
2.8 Proofs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
2.10 Review exercise . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88
3 Further sequences and series 91

3.1 Maclaurin series . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92
3.2 Iterative Schemes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102
3.3 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 120
3.5 Review exercise . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122
ii CONTENTS
4 Further Ordinary Differential Equations 125

4.1 First order linear differential equations . . . . . . . . . . . . . . . . . . . 126
4.2 Second order, linear, differential equations with constant coefficients . . 137
4.3 Homogeneous, second order, linear, differential equations . . . . . . . . 138
4.4 Non-homogeneous, second order, linear, differential equations . . . . . 146
4.5 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 152
4.6 Proofs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 154
4.8 Review exercise . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 160
5 Further number theory and further methods of proof 163

5.1 Revision . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 164
5.2 Implications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 164
5.3 Further proof by contradiction . . . . . . . . . . . . . . . . . . . . . . . . 171
5.4 Further proof by induction . . . . . . . . . . . . . . . . . . . . . . . . . . 174
5.5 Proof using the contrapositive . . . . . . . . . . . . . . . . . . . . . . . . 177
5.6 Direct methods of proof . . . . . . . . . . . . . . . . . . . . . . . . . . . 178
5.7 The division algorithm and number bases . . . . . . . . . . . . . . . . . 181
5.8 The Euclidean algorithm and its applications . . . . . . . . . . . . . . . 187
5.9 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 193
5.11 Divisibility rules and integer forms . . . . . . . . . . . . . . . . . . . . . 195
5.12 Review exercise . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 196
6 End of unit three tests 199
Glossary 201
Hints for activities 209
Answers to questions and activities 210

1 Vectors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 210
2 Matrix algebra . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 221
3 Further sequences and series . . . . . . . . . . . . . . . . . . . . . . . 230
4 Further Ordinary Differential Equations . . . . . . . . . . . . . . . . . . 241
5 Further number theory and further methods of proof . . . . . . . . . . . 246
© H ERIOT-WATT U NIVERSITY
1
Topic 1
Vectors
Contents
1.1 Revision exercise . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
1.2 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
1.3 Direction ratios and cosines . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
1.4 Basic arithmetic and scalar product of vectors . . . . . . . . . . . . . . . . . . . 9
1.4.1 Arithmetic on vectors in three dimensions . . . . . . . . . . . . . . . . . 9
1.4.2 Scalar product of vectors in three dimensions . . . . . . . . . . . . . . . 13
1.5 Vector product - algebraic form . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
1.6 Vector product - geometric form . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
1.7 Scalar triple product . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
1.8 Lines in three-dimensional space . . . . . . . . . . . . . . . . . . . . . . . . . . 24
1.8.1 The equation of a line . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
1.8.2 Intersection of and the angles between lines . . . . . . . . . . . . . . . 29
1.9 Planes in three-dimensional space . . . . . . . . . . . . . . . . . . . . . . . . . 34
1.9.1 The equation of a plane . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
1.9.2 Intersection of and the angles between two planes . . . . . . . . . . . . 39
1.9.3 Intersection of three planes . . . . . . . . . . . . . . . . . . . . . . . . . 41
1.10 Lines and planes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
1.11 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
1.12 Proofs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
1.13 Extended information . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
1.14 Review exercise . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
1.15 Advanced review exercise . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
1.16 Set review exercise . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
2 TOPIC 1. VECTORS
Learning Objectives
• Use vectors in three dimensions.
Minimum Performance Criteria:
• Calculate a vector product.
• Find the equation of a line in parametric form.
• Find the equation of a plane in Cartesian form given a normal and a point in the
plane.
1.1. REVISION EXERCISE 3
Prerequisites
A sound knowledge of the following subjects is required for this topic:
• Plotting points in three dimensions.

• Simple trigonometry.
• Algebraic manipulation.
• Equations of straight lines.
• Gaussian elimination
1.1 Revision exercise

Learning Objective

ÆIdentify areas that need revision

Revision exercise
If any question in this revision exercise causes difficulty then it may be necessary to
revise the techniques before starting the topic. Ask a tutor for advice. 15 min
Q1: Draw and label a three-dimensional graph with axes x, y and z

Q2: Find the equation of the line which joins the two points (4, 5) and (-1, -10)
Q3: Expand fully the following: (ab + 2c)(ac - b)(c + b)
Hence simplify the expression if a2 = 2, b2 = 1 and c2 = 1
Q4: Use Gaussian elimination to solve the equations
7x - 7y - 2z = 8
2x - y - 3z = -6
2x + 3y - z = -4
1.2 Introduction
Learning Objective

ÆDefine basic vector terms

Quantities that have magnitude only are called scalars.
Weights, areas and volumes are all examples of scalars.
Many quantities are not sufficiently defined by their magnitudes alone. For example,
a movement or displacement from a point P to a point Q requires both the distance
between the points and the direction from P to Q.
4 TOPIC 1. VECTORS
Example If Q is 45 km North West of P then the distance is 45 km and the direction is

North West.
There is a special type of quantity to cope with this. It is called a vector quantity.
Vector quantity
A vector quantity is a quantity which has both direction and magnitude.
Geometrically, a vector can be represented by a directed line segment.
The two lines ab and cd represent two vectors in three dimensions.
Parallel vectors
Parallel vectors have the same direction but the magnitudes are scalar multiples of each
other.
There is a particular type of vector called a position vector.

Position vector
A position vector is a vector which starts at the origin.
1.2. INTRODUCTION 5

A vector from the origin O, for example, to the point P, may be expressed by a small
letter in the form p or or by the directed vector OP. It is customary for textbooks to use
the bold form and this will be the style in this topic.
A position vector can also be expressed as an ordered triple.

For example, the vector in the previous diagram from the origin to the point P with
a
coordinates (a, b, c) can be expressed as the ordered triple
b
c
Note that the vector from the origin to the point P has a magnitude (length) and direction
from the origin to the point (a, b, c)
Length of a vector
a
Let p be the vector then the length of p is defined as p = a2 + b2 + c2
b
c
Examples
2
1. Find the length of a = 3
4
Answer:

The length is given by a = 22 + 32 + 42 = 5.38516
6 TOPIC 1. VECTORS
2. These are all position vectors in 2 dimensions.
These are position vectors in 3 dimensions.
In this topic most of the work will be covered using position vectors. The word ’position’
will be omitted unless confusion is likely.
Unit vector
As the name would suggest a unit vector is a vector of magnitude 1. That is, the length
(or magnitude) of the vector is 1
There are 3 special unit vectors which lie along the x, y and z axes. The point P has
coordinates (1, 0, 0), Q has (0, 1, 0) and R has (0, 0, 1)
1.2. INTRODUCTION 7

Each of the vectors OP, OQ, and OR is one unit long and is denoted by i, j, and k
respectively. i, j and k are called the standard basis vectors.
1
Thus i is the vector 0
0

0
similarly j is the vector
1
0
0
and k is the vector 0
1

p
Every vector can be written in component form as pi + qj + rk Conversely every
q
r
p

vector in component form pi + qj + rk can be written as q
r
3

Example The vector p = -5 = 3i - 5j + 2k
2
Vector length exercise

There is another version of this exercise on the web for you to try if you prefer it.
10 min
Q5: Find the lengths of the position vectors in the diagram.
-2

Q6: Find the length of the vector 2
1
Vector notation exercise

There is another version of this exercise on the web for you to try if you like.
-3
5 min
Q7: Express the vector 3 using the standard basis vectors.

-4
8 TOPIC 1. VECTORS
Q8: Express the vector i + 2j + k as an ordered triple.
1.3 Direction ratios and cosines

Learning Objective

ÆFind direction ratios and direction cosines of a vector

Let p be the vector ai + bj + ck
This can be represented on a diagram as follows.
Notice that the vector p makes an angle of with the x-axis, an angle of with the y-axis
and an angle of with the z-axis.
Thus
cos = a
p
cos = b
p
cos = c
p
a b c
These values p , p and p are called the direction cosines of the vector p
The ratios of a : b : c are called the direction ratios of the vector p
Examples
1. Find the direction ratios of the vector -2i - 3j + 4k

Answer:
The direction ratios are -2 : -3 : 4
2. Find the direction cosines of the vector p = 2i - 2j - k

Answer:
1.4. BASIC ARITHMETIC AND SCALAR PRODUCT OF VECTORS 9
p=3
The direction cosines are 2 /3 , -2 /3 and -1 /3
Direction ratios and direction cosines exercise

There is a web exercise similar to this for you to try if you prefer it.
5 min
Q9: Find the direction ratios of the following vectors:
a) -3i + 2j - k
b) i + j + k
c) 2i - j - k
Q10: Find the direction cosines of the following vectors:
a) 2i + 3j + 6k
b) i - j - k
c) 4i - 4j + 2k
1.4 Basic arithmetic and scalar product of vectors

Learning Objective

ÆAdd, subtract, multiply and find the scalar product of vectors in three dimensions

These techniques are straightforward and the following examples will demonstrate them.
Vectors in 2d addition and subtraction

There is a web animation to review addition and subtraction in 2 dimensions.
10 min
1.4.1

Arithmetic on vectors in three dimensions

and q = y are vectors then p + q is the vector b + y
If p =
a
b
x a+x
c z c+z
The diagram in the following example shows how this is related to the parallelogram.

Example
and b =
Add the vectors a =
3
4
-5
6

2 1
Answer:
The sum is found by adding the corresponding values so

+
3
4
-5
6 = 4 + 6 =
3 + ( - 5)

-2
10

2 1 2+1 3
10 TOPIC 1. VECTORS
If p and q are written in terms of the standard bases then

p = ai + bj + ck, q = xi + yj + zk and p + q = (a + x)i + (b + y)j + (c + z)k
Example Add the vectors -2i + 3j - 4k and -3i - 6j - 3k

Answer:
Adding the corresponding values gives
(-2i + 3j - 4k) + (-3i - 6j - 3k)
= (-2 - 3)i + (3 - 6)j + (-4 - 3)k
= -5i - 3j - 7k

and q = y are vectors then p - q is the vector b - y
If p =
a
b
x a-x
c z c-z
Examples
1 2

1. Evaluate -
4 -2
2 1
Answer:
1
2

-
4 -2
2
1-2
1
= 4 - ( - 2)
-1

2-1
= 6
1
The following diagram shows how this relates to a parallelogram. Note that in
subtraction, the vector -b instead of b forms one of the sides (a + (-b) = a - b)
It is also possible to subtract vectors that are written in terms of the standard bases.
2. Evaluate (i + 4j + 2k) - (2i -2j + k)

Answer:
(i + 4j + 2k) - (2i - 2j + k)
= i + 4j + 2k - 2i + 2j - k
= -i + 6j + k
Vectors can also be multiplied by a scalar.

a
and is a real number, (a scalar) then p =
If p =
a
b b

c c
-2

Example If a = 3 find -2a
-1
12 TOPIC 1. VECTORS
Answer:
-2
-2 -2
4

-2a = - 2 3 = - 2 3 = -6
-1 -2 -1 2
Using the standard bases, the multiplication is just as straightforward.
Example If a = 3i - j + 2k find -4a

Answer:
Each of the terms is multiplied by -4 to give
-4a = -12i + 4j - 8k
This is particularly useful for determining whether two vectors are parallel or not. Recall
that two vectors which are parallel have the same direction but their magnitudes are
scalar multiples of each other.

Example Show that the two vectors a =

-3
5
and b =
6
- 10 are parallel.
3 -6
Answer:
Each component of b is -2 times the corresponding component of b
That is, b = -2a and the vectors are parallel.
Basic skills exercise

There is another exercise on the web for you to try if you prefer it.
10 min
Q11: Add the vectors 2i + 3j - 2k and i - 3j - k
-2
2

Q12: Add 5 and -2 and give the answer as an ordered triple.
-1 2
Q13: Subtract 3i - 2j + 5k from -2i
-7
0

Q14: Subtract -2 from 3 and give the answer as an ordered triple.
4 1
Q15: Show that 2i - j + 3k and -4i + 2j - 6k are parallel vectors.
Q16: Find a vector parallel to 2i - 4j + k with z component equal to 3
1.4.2 Scalar product of vectors in three dimensions

There are two ways of expressing the scalar product. The first is algebraically in
component form.
Scalar product in component form
a d

If p = and q = e
b
c f
then the scalar product is the number p q = ad + be + cf
It is important to note that the scalar product of two vectors is not a vector. It is a scalar.
The scalar product is also known as the dot product.
-2
3

Example Find the scalar product of a = 1 and b = 2
3 0
Answer:
a b = -6 + 2 + 0 = -4
The scalar products of the standard basis vectors are useful to note:
i i = j j = k k = 1 and i j = j k = i k = 0
also
i j = j i
jk=kj
i k = k i

These results can be used to give the scalar product in i, j, k notation of the two vectors
-2 3
a= 1
and b = 2
3 0
Example Find (-2i + j + 3k) (3i + 2j) by multiplying out the brackets and using the
properties of the vector product on standard bases.
Answer:
(-2i + j + 3k) (3i + 2j)
= -6i i - 4i j + 3 j i + 2j j + 9k i + 6k j
= -6 -i j + 2 + 9i k + 6j k
= -6 + 2 = -4
14 TOPIC 1. VECTORS
Algebraic scalar product exercise

There is an exercise similar to this on the web for you to try if you wish.
10 min
2 -3

Q17: Find a b where a = and b =
3 2
4 -3
Q18: Find the scalar product of the vectors 2i + 3j - 2k and i - 3j - k
-2
2

Q19: Find the scalar product of the vectors 5 and -2
-1 2
Algebraic rules of scalar products

There are several useful properties of scalar products.
Property 1
a (b + c) = a b + a c
Details of the proof can be found in the Proof section at Proof 3.
Property 2
a b=ba
The proof of this result can be found in the section headed Proofs at Proof 4.
Property 3
a a = a 2 0
Property 4
a a = 0 if and only if a = 0
Q20: Show that the property a (b + c) = a b + a c holds for the three vectors
a = 2i - k, b = -3i - j + 3k and c = -i + 2j - k
a1
b1

Q21: Using a = a2 and b = b2 prove a b = b a
a3 b3
The second way of expressing the scalar product is geometrically.

Scalar product in geometric form
The scalar product of two vectors a and b is defined as
ab= a b cos where is the angle between a and b, 0 180 Æ
The proof of this is shown at Proof 1 in the section headed Proofs.
Example Find the scalar product of the vectors a and b where the length of a is 5, the
length of b is 4 and the angle between them is 60 Æ
Answer:
ab= a b cos = 5 4 cos 60 Æ
= 10
The geometric form of the scalar product is especially useful to find angles between
vectors.
Example Find the angle between i + 4j and -8i + 2j

Answer:
a b = -8 + 8 = 0

a = 17

b = 68
cos = ab
a b = 0 so = 2
The last example demonstrates an important geometric property of the scalar product.
Property 5
For non-zero vectors, a and b are perpendicular if and only if a b = 0
In this case, a and b are said to be orthogonal vectors.
The proof of this result is shown in the section headed Proofs at Proof 7.
To show that two vectors are perpendicular however, the algebraic form is all that is
required.
Example Show that the two vectors a = -2i + 2j + k and b = 3i + 2j + 2k are

perpendicular.
Answer:
a b = (-2 3) + (2 2) + (1 2) = 0
The vectors are perpendicular.
Geometric scalar product exercise

10 min
Q22: Find the scalar product of two vectors that have lengths 4 and 8 units respectively
and an angle of 45 Æ between them. Give the answer to 3 decimal places.
Q23: Find the scalar product of two vectors whose lengths are 2.5 and 5 and have an
angle of 150 Æ between them. Give the answer to 3 decimal places.
Q24: Determine the angle between two vectors whose scalar product is -6 and whose
lengths are 4 and 3
16 TOPIC 1. VECTORS
Q25: Find the angle between the two vectors 3i - 4j and 12i + 5j. Give the angle in
degrees correct to 2 decimal places.
Q26: Find the angle ACB if A is (0, 1, 6), B is (2, 3, 0) and C is (-1, 3, 4)
Perpendicular vectors exercise

There is an exercise on the web for you to try if you like.
10 min
Q27: Find c so that ci + 2j - k is perpendicular to i - 3k
Q28: Find c so that 3i + cj + 4k is perpendicular to 2i -2j + 3k
Q29: Show that the vectors -4i - 3j + 2k and 2i - 2j + k are perpendicular.
1.5 Vector product - algebraic form

Learning Objective

ÆCalculate the vector product algebraically

Like the scalar product, there are two ways of expressing a vector product. The first is
algebraically in component form.
Vector product in component form
If p = ai + bj + ck and q = di + ej + fk
then the vector product is defined as the vector
p x q = (bf - ec)i - (af - dc)j + (ae - db)k
Hence, unlike the scalar product, the vector product of two vectors is another vector.
i j k

This is actually the determinant of the matrix a b c
d e f
By putting the vectors in this form it is easier to see how the calculation is done and why
the vector product is sometimes called the cross product.
Determinants of matrices will be studied in detail in topic 12 in this course.
Examples
4
-1

1. If a = -5 and b = -2 find a x b
-6 3
Answer:
To use the formula, the vectors a and b have to be expressed in terms of the basis i, j
and k
1.5. VECTOR PRODUCT - ALGEBRAIC FORM 17
a = 4i - 5j - 6k
b = -i - 2j + 3k
a x b = ((-5) 3 - (-2) (-6))i - (4 3 - (-1) (-6))j + (4 (-2) - (-1) (-5))k
i j k

= -27i - 6j - 13kIn matrix form the calculation can be taken from the determinant of
4 -5 -6
-1 -2 3
2. Find p x q where p = -2i + 3j - k and q = 4i + 2j - 2k

Answer:
p x q = (3 (-2) - 2 (-1))i - ((-2) (-2) - 4 (-1))j + ((-2) 2 - 4 3)k
= -4i - 8j - 16k
Algebraic vector product exercise

There is a different exercise on the web.
15 min
Q30: Find a x b for the following pairs of vectors
a) -2i - j + 3k and i + 2j
b) -5i + j + 4k and 2i + j - 3k
c) -i + 2j + 4k and 2i - 4j - 8k
The following table shows the outcome of the vector products of the standard vectors
ixj=k j x i = -k ixi=0
jxk=i k x j = -i jxj=0
kxi=j i x k = -j kxk=0
Properties of vector products
The vector product obeys the distributive laws.
Property 1
a x (b + c) = (a x b) + (a x c) and (a + b) x c = (a x c) + (b x c)
Q31: Show that the distributive property a x (b + c) = (a x b) + (a x c) holds for the

vectors a = 3i - 2j + k, b = i - j - k and c = 2i - 3k
However the vector product does not obey all the laws that the product of real numbers
does. For example, from the table of vector products of standard basis vectors shown
previously
i x j = k = -j x i
Hence a x b need not equal b x a
In fact
Property 2
For any two vectors p and qp x q = -q x p
18 TOPIC 1. VECTORS
The proof is shown here rather than in the section headed Proofs as it is important.
Let p = ai + bj + ck and q = di + ej + fk
q x p = (ec - bf )i - (dc - af )j + (db - ae )k
but p x q = (bf - ec)i - (af - dc)j + (ae - db)k
= -[(ec - bf )i - (dc - af )j + (db - ae )k]
So p x q = -q x p
Example Show that (6i - j + 3k) x (9i + j + 4k) = -(9i + j + 4k) x (6i - j + 3k) by expanding
the brackets and using the properties of the vector product of standard bases.
Answer:
(6i - j + 3k) x (9i + j + 4k)
= 54i x i - 9j x i + 27k x i + 6i x j - j x j + 3k x j + 24i x k - 4j x k + 12k x k
= 9k + 27j + 6k - 3i - 24j - 4i
= -7i + 3j + 15k
and -(9i + j + 4k) (6i - j + 3k)
= - (54i x i + 6j x i + 24k x i - 9i x j - j x j - 4k x j + 27i x k + 3j x k + 12k x k)
= - (7i - 3j - 15k)
= -7i + 3j + 15k as required.
From the result a x b = -b x a, then (a x b) + (b x a) = 0

If in particular a = b then 2(a x a) = 0 then
Property 3
For any vector a a x a = 0
Another law that holds for the product of real numbers is a(bc) = (ab)c. It does not hold
for the vector product.
Property 4
For any vectors a, b and c, it is generally the case that a x (b x c) (a x b) x c

Example Show that the property a x (b x c) (a x b) x c is true for the vectors
2 1 -2
-3 , 0 and 3
1 -1 -2
Answer:
b x c = 3i + 4j + 3k
a x (b x c) = -13i - 3j + 17k
a x b = 3i + 3j + 3k
(a x b) x c = -15i + 15k
1.6. VECTOR PRODUCT - GEOMETRIC FORM 19
They are not the same.
Vector product properties exercise

10 min
Q32: Show that a x a = 0 for a = -3i - j - k
Q33: Show that p x q = -q x p for the vectors p = 2i + 3j + 4k and q = 4i + j + 5k

Q34: Show that the distributive law (a + b)x c = (a x c) + (b x c) holds for
4 -3 1
a= -1 , b = 1 and c = 2
2 -3 -1
1.6 Vector product - geometric form

Learning Objective

ÆCalculate the vector product using the geometric formula

The second way of expressing a vector product is geometrically by specifying its
direction and magnitude.
As shown earlier, the angle between vectors can be found using the scalar product.
It was shown that if a b = 0 then the vectors a and b are perpendicular. This can be
used to explore the direction of a x b
Example Show that a (a x b) = 0 and that b (a x b) = 0

Answer:
Let a = a1 i + a2 j + a3 k and b = b1 i + b2 j + b3 k
Then a x b = (a 2 b3 - b2 a3 )i - (a1 b3 - b1 a3 )j + (a1 b2 - b1 a2 )k
Thus a (a x b) = a1 a2 b3 - a1 b2 a3 - a2 a1 b3 + a2 b1 a3 + a3 a1 b2 - a3 b1 a2 = 0
Also b (a x b) = b1 a2 b3 - b1 b2 a3 - b2 a1 b3 + b2 b1 a3 + b3 a1 b2 - b3 b1 a2 = 0
Since a (a x b) = 0 and b (a x b) = 0 the vector a x b is perpendicular to both a and

b
But a x b = -b x a from the properties of vector products given earlier. So a x b gives a
vector perpendicular to a and b in a positive direction whereas b x a also gives a vector
perpendicular to a and b but in the negative and opposite direction.
This direction can be determined by the right hand rule which states that if the right
hand is used with the thumb pointing in the direction of a and the first finger pointing in
the direction of b, then the middle finger points in the direction of a x b
20 TOPIC 1. VECTORS
The vector a x b has length a b sin ( see Proof 2 in the Proof section).
Thus the direction and length of the vector a x b have been specified.
Vector product in geometric form
The vector product of a and b is defined with
• magnitude of a x b = a b sin , where is the angle between a and b,

0 180Æ
• direction perpendicular to both vectors a and b as determined by the right hand

rule.
Note that a x b = 0 if and only if a and b are parallel.
Example Find the magnitude of the vector product of the two vectors which are at an
angle of 50 Æ to each other and have lengths of 5 and 7 units
Answer:
a x b = 5 7 sin 50 Æ
= 26.81
The length of the vector product a x b has the same formula as the area of a
parallelogram in which the vectors a and b are the sides of the parallelogram as shown
in the diagram.
z
b
c a
q y
Area of parallelogram = a c
But c = b sin
Thus the area = a b sin = a x b
Of course to find a x b the algebraic version of the vector product is used.
Example Find the area of the parallelogram which has adjacent edges
a = 3i + j + 2k and b = 2i - j
1.7. SCALAR TRIPLE PRODUCT 21
Answer:
a x b = 2i + 4j - 5k

Hence a x b = 22 + 42 ( - 5)2 = 6.708 units.
The area of the parallelogram is 6.708 sq units.
Geometric vector product exercise

There is a web exercise for you to try if you prefer it.
15 min
Q35: Find a vector perpendicular to 2i - j + 3k and i + 2j
Q36: Find the length of a x b given that a has length 4 units, b has length 5 units and
the angle between them is 45 Æ . Give the answer to 2 decimal places.
Q37: Find the area of a parallelogram with edges i - 4j + k and 2i + 3j - 2k

Give the answer to 3 decimal places.
Q38: Find the area of a triangle with vertices at (0, 0, 0), (3, 1, 2), (2, -1, 0). Give your
answer to 3 decimal places.
Q39: Find the area of the triangle with vertices (1, -1, 0), (2, 1, -1) and (-1, 1, 2)
Activity
Investigate geometrically (a x b) x c and a x (b x c) by choosing some easy points to
manipulate in three-dimensional space.
1.7 Scalar triple product

Learning Objective

ÆCalculate the scalar triple product of vectors.

To show that a x b is perpendicular to a and to b, the numbers a (a x b) and
b (a x b) are found. If a, b and c are vectors then as b x c is also a vector it is possible
to find the scalar product of a with b x c
Scalar triple product
The scalar triple product a (b x c) is the number given by the scalar product of the
two vectors a and (b x c)
This product can also be found using the determinant of a matrix in a similar way to the
vector product. Suppose that
a = a 1 i + a2 j + a3 k
b = b 1 i + b2 j + b3 k
c = c1 i + c2 j + c3 k
22 TOPIC 1. VECTORS
a1 a2 a3

This is the determinant of the matrix b1 b2 b3
c1 c2 c3
The calculation of a (b x c) gives a 1 (b2 c3 - c2 b3 ) - a2 (b1 c3 - c1 b3 ) + a3 (b1 c2 - c1 b2 )
It is sometimes known as the triple scalar product or the dot cross product.
The calculation is best shown by example.
Example Find a (b x c) when a = i + 2j - 3k, b = 4i - j + 2k and c = 3i - 2j + 5k

Answer:
b x c = -i - 14j - 5k
a (b x c) = -1 -28 +15 = -14
The modulus of the scalar triple product is the volume of a parallelepiped with adjacent
edges a, b and c as shown in the following diagram.
b
bxc
q
c
y
The base is formed by b and c

The area of the base = b x c
The perpendicular height is a cos
But the volume of the parallelepiped = area of base perpendicular height
Thus the volume = b x c a cos = a (b x c)
(Remember that a b = b a)
Example Find the volume of the parallelepiped with edges i + j + 2k, i + 2j + 3k and
3i - j - k
Answer:
This is the modulus of the scalar triple product.
Let a = i + j + 2k, b = i + 2j + 3k and c = 3i - j - k
1.7. SCALAR TRIPLE PRODUCT 23
The volume is a (b x c)
b x c = i + 10j - 7k
Then a (b x c) = (1 1) + (1 10) + (2 (-7)) = 1 + 10 - 14 = -3
The modulus is 3 and the volume of the parallelepiped is 3 cubic units.
Note that the volume of a parallelepiped is zero if and only if a, b and c are coplanar,
that is if a (b x c) = 0
If b and c form the base of the parallelepiped then b x c is perpendicular to the base.
However, if a lies in the base formed by b and c then a is perpendicular to b x c which
means that the scalar product of a and b x c is zero.
Scalar triple product exercise

15 min
Q40: Find a (b x c) if a = i + j + 2k, b = i + k and c = 3i - 2j + 5k
Q41: Find a (b x c) if a = i + 2j + 3k, b = 2i + j + 3k and c = 4i + k
Q42: Find the volume of the parallelepiped with edges

-i + 4j + 7k, 3i - 2j - k and 4i + 2k
Q43: Show that the position vectors to the points (1, 2, 3), (4, -1, 2) and (2, -5, -4) lie in
the same plane.
Properties of the scalar triple product

The volume of a parallelepiped was shown as a (b x c) . By considering the
base of the parallelogram formed from a and b, the volume is similarly computed as
c (a x b) and by taking the base formed from a and c the volume is b (c x a)
It is possible that c (a x b) and a (b x c) may have different signs, but provided that
a is followed by b which in turn is followed by c and back to a they have the same signs.
Thus
Property 1
a (b x c) = b (c x a) = c (a x b)
The proof of this is shown in the section called Proofs at Proof 8.
If the order is different then the answer need not be the same. For example:
a (c x b)
= a (-(b x c)) (using the property of vector products)
= -a (b x c) (using the property of scalar products)
Q44: Check that this property holds for the vectors

a = i - k, b = 2i + j and c = i + j - k
24 TOPIC 1. VECTORS
1.8 Lines in three-dimensional space

Learning Objective

ÆDetermine and use the equations of a line in vector, parametric and symmetric form

In a similar way to lines in the plane, lines in space are specified by either:
• One point and a direction.

• Two points on the line.
1.8.1 The equation of a line

There are three forms in which the equation of a line can be written in three-dimensional
space. These are vector, parametric and symmetric form.
Suppose first that the line is specified by a point and a direction.
Vector form of the equation of a line
The vector equation of a line through the point P with direction d is

r = a + d where a = OP, d is a vector parallel to the required line and is a real
number.
Any position vector r of a point on the required line will satisfy this equation.
Example Find the vector equation of the straight line through (2, -1, 6) and parallel to
the vector i + 2j - 8k
1.8. LINES IN THREE-DIMENSIONAL SPACE 25
Answer:
Using the formula r = a + d with a = 2i - j + 6k and d = i + 2j - 8k gives
r = 2i - j + 6k + (i + 2j - 8k)
If r = xi + yj + zk, a = a1 i + a2 j + a3 k and d = d1 i + d2 j + d3 k then the equation can be

split into component parts. This leads to the second form of the equation of a straight
line.
Parametric form of the equation of a line
The parametric equations of a line through the point P = (a 1 , a2 , a3 ) with direction
d = d1 i + d2 j + d3 k are
x = a1 + d1 , y = a2 + d2 , z = a3 + d3 where is a real number.
Example Find the parametric equations of the line through (3, 4, 5) and parallel to the
vector 2i - 3j - 4k
Answer:
Using the formulae with a 1 = 3, a2 = 4 and a3 = 5
x = 3 + 2, y = 4 - 3, z = 5 - 4
It is easy to obtain the parametric form from the vector form and vice versa.
Examples
1. State the parametric form of the equation of the line which has a vector equation of

r = a + d where P is the point (3, -1, 5), a = OP and d = -2i + 4j - k
Answer:
r = (3i - j + 5k) + (-2i + 4j - k)
Equating each component in turn gives
x = 3 - 2, y = -1 + 4, z = 5 -
2. Find the vector form of the equation of a line with parametric equations
x = -1 + 3, y = , z = 2 + 2
Answer:
x = -1 + 3, y = , z = 2 + 2 gives
x -1
3

=
y 0 + 1
z 2 2
a = -i + 2k and d = 3i + j + 2k
The equation is r = ( -i + 2k) + (3i + j + 2k)
Equations of the parametric form can be rearranged to eliminate the real number
giving the third form of the equation of a straight line.
26 TOPIC 1. VECTORS
Symmetric form of the equation of a line

If x = a1 + d1 , y = a2 + d2 , z = a3 + d3 are parametric equations of a line, the
symmetric equation of this line is
x - a1 y - a2 z - a3
d1 = d2 = d3
These symmetric equations of a line are also known as the Cartesian equations of a
line.
Note that if any of the denominators are zero then the corresponding numerator is also
zero. This means that the vector is parallel to an axis.
It is also important to ensure that the symmetric form of the equation of a line is stated
with each coefficient of x, y and z on the numerator equal to 1. An equation with any
other coefficients for x, y and z is not the symmetric equation of the line.
The direction ratios are the denominators d 1 , d2 and d3 in these symmetric equations.
It is straightforward to obtain the symmetric form from either the parametric or vector
form. The following examples will demonstrate how this is done.
Example Find the parametric and symmetric equations of the line with a vector equation
of
r = (2i - 3j + k) + (-i - 2j + k)
Answer:
a = (2, -3, 1) and d = (-i - 2j + k)
The parametric equations are x = 2 - , y = -3 - 2 and z = 1 +
Solving each component for gives the symmetric equations as
x-2 y+3 z-1
-1 = -2 = 1
Note that even if the denominator is 1, the form of equations requires that it should be
left there.
Examples
1. Find the parametric and symmetric equations of the line which has position vector
i - j + k and is parallel to the vector 3i + 2j + k
Answer:
The parametric equations are x = 1 + 3, y = -1 + 2, z = 1 +
and so
y+1
x-1
3 = 2 = z-1
1 =
Hence by eliminating the symmetric equations are
x-1 y+1 z-1
3 = 2 = 1
2. Find the symmetric equations of the line which has parametric equations
x = 2 - 2, y = -1 + 3, z = -
Answer:
Rearrange to make each equation equal to
y+1
This gives x-2
-2 = 3 = z-0
-1 =
x-2 y+1 z-0
Hence the symmetric equations are -2 = 3 = -1
To convert the symmetric equations of a line into the parametric form, take the equation
x - a1 y - a2 z - a3
d1 = d2 = d3 and call this common value the parameter
Then solve for x, y and z in terms of
Examples
1. Find the parametric form of the equation of a line with symmetric equations
x+3 y-1 z-4
2 = 1 = -3
Answer:
Make each part of the equation equal to and solve for x, y and z
2 = so x = - 3 + 2
x+3
y-1
1 = so y = 1 +
- 3 = so z = 4 - 3
z-4
The equations are x = -3 + 2, y = 1 + , z = 4 - 3
2. Find the parametric and vector forms of the equation of a line which has symmetric
equations x +1 4 = y--23 = z 2- 4
Answer:
y-3
Let x+4
1 = -2 = z-4
2 =
Solving for x, y and z gives: x = -4 + , y = 3 - 2, z = 4 + 2
This gives the parametric equations of the line.
Thus a = -4i + 3j + 4k and d = i - 2j + 2k
Thus r = (-4i + 3j + 4k) + (i - 2j + 2k) is the vector equation of the line.
Equations of a line are not unique. The direction vectors however, are proportional (one
is a scalar multiple of the other).
Example Suppose that the equation of a line is given by the symmetric equations
x-3 y+1 z-1
2 = 3 = 1
then in parametric form the equations are

x = 3 + 2, y = -1 + 3, z = 1 +
Answer:
If = 2, a point on this line is = (7, 5, 3)
28 TOPIC 1. VECTORS
x-9 y-8 z-4

Consider the line with symmetric equations -4 = -6 = -2
This has symmetric equations x = 9 - 4, y = 8 - 6, z = 4 - 2

Substitution of the x, y, and z values of the point (7, 5, 3) into either of the forms of this
second line gives = 1 /2 and so the point is also on this line.
But the direction vector d1 = 2i + 3j + k of the first line is a scalar multiple of the direction
vector d2 = -4i - 6j - 2k of the second line.
Thus the lines are in fact the same line with equations that look very different.
Equation of a line exercise

There is another version of this exercise on the web for you to try if you wish.
10 min
Q45: State the vector equation of the line which passes through the point (3, 4, 5) and
is parallel to the line -2i - j + 3k
Q46: Find the vector equation of a line with parametric equations:

x = 3 + 4, y = -2 - 3 and z = 4 -
Q47: Give the parametric and symmetric equations for the lines:
a) through (-1, 3, 0) in the direction -2i - j + 4k

b) through (2, -1, 0) in the direction -i + 4k
Q48: Find the symmetric equations of the lines with vector equations:
a) r = -j + 2k + (-i + j - k)
b) r = 2i - 2j - 2k + (-i + 3j - k)
Q49: From the vector equation of the line given as (2i + j - 3k) + (3i - j - k) state the
point with y coordinate -1 which also lies on this line.
Q50: Find the parametric equations of the line through the point (3, 4, 5) and parallel to
the line i + j + k
Q51: Find the symmetric equation of the line which has parametric equations of
x = - , y = - 2 + , z = 3 + 2
Line determined by two points

Now suppose that the line is determined by two points P and Q. A direction vector on
this line can be obtained by taking PQ

Let P be the point (a 1 , a2 , a3 ) and Q be the point (b 1 , b2 , b3 ) then a direction vector
is represented by the line segment PQ = d = q - p. It is now possible to obtain vector,
parametric and symmetric equations of the line as before.
Example Find the vector, parametric and symmetric equations of the line through
(1, -2, -1) and (2, 3, 1)
Answer:
If a = the position vector i - 2j - k and b = 2i + 3j + k
then a direction vector is b - a = i + 5j + 2k
The vector form of the equation is r = i - 2j - k + (i + 5j + 2k)
The parametric equations are x = 1 + , y = -2 + 5, z = -1 + 2
x-1 y+2 z+1
The symmetric equations are 1 = 5 = 2
Equation from two points exercise

15 min
Q52: Give parametric and symmetric equations for the line through the points (3, -1, 6)
and (0, -3, -1)
Q53: Find the parametric and symmetric equations of the line joining the points
A (1, 2, -1) and B(-1, 0, 1)
1.8.2 Intersection of and the angles between lines
Learning Objective

ÆFind the intersection of and the angle between two lines

Intersection of two lines in three-dimensional space
Now suppose that L 1 and L2 are two lines in three-dimensional space. The two lines
can:
30 TOPIC 1. VECTORS
• be identical
• be parallel
• intersect
• be skew, that is, not identical, parallel or intersecting
The last case cannot arise in two-dimensional space.

Note that if two lines are parallel then their direction ratios are proportional.
Two lines intersecting at a point
In three dimensions lines normally do not intersect. If however they do meet then by
equating the parametric equations, there will be a unique solution. To show this, take
the lines in parametric form.
Example Suppose that L 1 has parametric equations x = 1 + , y = -1 + , z = 2 +

and L2 has parametric equations x = 2 + , y = -2 + 3, z = -1 + 5
Show that the two lines L 1 and L2 intersect but are not the same line.
Answer:
As their direction vectors are not proportional L 1 and L2 are not the same line.
If the lines intersect then there are solutions to the set of equations obtained by equating,
in turn, x, y and z
1 + = 2 + (this is the x equations).
-1 + = -2 + 3 (this is the y equations).
2 + = -1 + 5 (this is the z equations).
These can be solved to give = 1 and = 2
Since there is a solution, the lines meet. By substituting the values for and the point
of intersection is (3, 1, 4)
32 TOPIC 1. VECTORS
Two skew lines

If the equations of two lines have no solution for the real numbers and the lines are
skew.
y-0
Example Show that the lines L 1 with symmetric equation x 3- 4 = -1 = z-2
-1 and L2 with
parametric equations x = , y = , z = 3 + do not intersect.
Answer:
If L1 and L2 do intersect then there is a solution of the set of equations
1) 4 + 3 =
2) - =
3) 2 - = 3 +
However since 4 + 3= = -
then = -1
Substituting = -1 in equation 2) gives = 1 but this is inconsistent with equation 3)
which, using these values of and gives 3 = 4
There is no solution to these equations and so the lines do not intersect.
Angle between two lines

If two lines intersect then the angle between them is the angle between their direction
vectors. The acute angle is always taken.
Example Show that the two lines L 1 and L2 , which have symmetric equations as shown,
intersect and find the angle between them.
x+4 y-5 z-3
L1 : 2 = -4 = 1
x-0 y-3 z-2
L2 : -1 = -1 = 1
Answer:
First of all find the parametric equations that are
L1 : x = -4 + 2, y = 5 - 4, z = 3 +
L2 : x = - , y = 3 - , z = 2 +
The set of equations are
-4 + 2 = -
5 - 4 = 3 -
3+=2+
These can be solved to give = 1 and = 2
The two lines meet at (-2, 1, 4)
The direction vectors are d1 = 2i - 4j + k for L1 and d2 = -i - j + k for L2
If is the angle between L 1 and L2 then

d1 d2
cos = d½

d2 =
3
21 3
So the angle is 67.79 Æ
Note: If two lines a and b have direction cosines equal to a 1 , a2 , a3 and b1 , b2 , b3 then
the angle can be found from the equation cos = a 1 b1 + a2 b2 + a3 b3
Suppose that each direction cosine could be expressed as
cos = a
a = a1
cos = b
a = a2
cos = c
a = a3 for point (a, b, c)
and similarly
cos = d
b = b1
cos = e
b = b2
cos = f
b = b3 for point (d, e, f)
then a b = a b cos
so cos = ab
a b
ad + be + cf
a b a1 b1 + a2 b2 + a3 b3
34 TOPIC 1. VECTORS
Intersection points and angles exercise

There is another mixed exercise on the web for you to try if you like.
15 min
x-2 y-2 z-3 x-2 y-3 z-4
Q54: Show that the lines 1 = 3 = 1 and 1 = 4 = 2 intersect.
Q55: Determine whether the two lines L 1 and L2 intersect when

x+3 y-2 z-1
L1 has symmetric equations 1 = 1 = -3 and L2 has symmetric equations
x+4 y-1 z-0
-2 = 1 = 1
Q56: Find the intersection of and the angle between the lines L 1 with parametric
equations x = -2 + 2, y = 1 - 3, z = -1 + and the line L 2 which passes through
the point (-3, 4, 0) and is parallel to -i + j - k
Q57: Find the angle between and the point of intersection of the two lines
x-5 y-7 z+3 x-8 y-4 z-5
4 = 4 = -5 and 7 = 1 = 3
Q58: Find the point of intersection of the line through the point (1, 3, -2) and parallel to
4i + k with the line through the point (5, 3, 8) and parallel to -i + 2k
Then find the angle between them.
Q59: Find the intersection of and the angle between the lines L 1 with symmetric
equations x 2- 1 = y 3- 3 = z 1- 2 and L2 with symmetric equations x +2 1 = y 1- 6 = z--17
1.9 Planes in three-dimensional space

Learning Objective

ÆDetermine and use equations of the plane

Planes in space are specified either by:
• One point on the plane and a direction n at right angles (normal) to the plane.
• Three non-collinear points in the plane.
1.9.1 The equation of a plane

There are three common forms in which the equation of a plane can be written. These
are the vector equation, the Cartesian equation and the parametric equation.
1.9. PLANES IN THREE-DIMENSIONAL SPACE 35
Vector equation of a plane

The vector equation of a plane containing the point P and perpendicular to the vector
n is (r - a) n = 0 where a = OP
The equation can also be written as r n = a n
Example Find the vector equation of the plane through (-1, 2, 1) with normal vector
i - 3j + 2k
Answer:
The general equation is (r - a) n = 0 or r n = a n
Here a = -i + 2j + k
and n = i - 3j + 2k
The vector equation is
r (i - 3j + 2k) = -1 - 6 + 2 = -5
If r = xi + yj + zk then evaluating the scalar products r n and a n will give the second
form of the equation of a line.
This is called the Cartesian form of the equation of the plane.
Cartesian equation of a plane
The Cartesian equation of a plane containing the point P (a 1 , a2 , a3 ) and perpendicular
to the vector n = n1 i + n2 j + n3 k is xn1 + yn2 + zn3 = d where d = a1 n1 + a2 n2 + a3 n3
Example Find the Cartesian equation of the plane through (-1, 2, 1) which has normal
vector i - 3j + 2k
Answer:
The vector form of this plane given in the previous example is r (i - 3j + 2k) = -5
If r = xi + yj + zk then evaluating the vector product r n gives x - 3y + 2z = -5
36 TOPIC 1. VECTORS
Alternatively, the Cartesian equation in the last example can be found by immediate
substitution into the definition of the equation in this form.
The Cartesian form of the equation of a plane is the simplification of the vector equation.
It is straightforward to find the vector equation of a plane if the Cartesian form is known.
Example Find the vector equation of a plane whose Cartesian equation is

2x - y + 3z = -4
Answer:
If r = xi + yj + zk then 2x - y + 3z = r (2i - j + 3k)
The vector equation is r (2i - j + 3k) = -4
The examples shown previously have defined the plane using a point on the plane and
a normal to the plane. It is also possible to define the plane from three non-collinear
points.
The vector equation of a plane from three non-collinear points

Suppose that A, B and C are three non-collinear points. The vectors AB and AC lie in a
plane through all three points.

Their vector product AB x AC is perpendicular to this plane. Any multiple of the vector
product forms a normal n to the plane. Since a is the position vector to the point A that
lies in the plane it is straightforward to give the equation of the plane as r n = a n
Example Find the equation of the plane through A(-1, 3, 1), B(1, -3, -3) and C(3, -1, 5)
Answer:

AB = 2i - 6j - 4k

AC = 4i - 4j + 4k
These two vectors lie in the plane and their vector product is normal to the plane.

AB x AC = -40i - 24j + 16k
Let n = -5i - 3j + 2k (n is 1 /8 of the vector product for ease)

and a = -i + 3j + k
Thus the equation of the plane is r n = a n
So -5x - 3y + 2z = 5 - 9 + 2 = -2
The equation is 5x + 3y - 2z = 2
The third form of the equation of the plane is the parametric equation.
Parametric equation of a plane
The parametric equation of the plane is
r = a + b + c where a is a position vector of a point in the plane, b and c are two
non-parallel vectors parallel to the plane and and are real numbers.
Note that the parametric equation of the plane is sometimes referred to as the symmetric
equation of the plane. Do not confuse this with the distinct differences in the equations
of a line in parametric and symmetric form.
Example Find the parametric equation of the plane through (2, 3, -1) and parallel to the
vectors
-2i - 3j + k and -i + 3j + 4k
Answer:
The equation is r = a + b + c where
a = 2i + 3j - k, b = -2i - 3j + k and c = -i + 3j + 4k
The Cartesian form, ax + by + cz = d can be thought of as one linear equation in three

unknowns and solved by putting y = and z = to give the parametric solution.
38 TOPIC 1. VECTORS
Example Find the parametric equation of the plane with Cartesian equation
x + 2y + z = 3
Answer:
To obtain the parametric equation let y = and z = , then x = 3 - 2 +
x 3
-2
1

so the general solution is = 0 +
y 1 + 0
z 0 0 1
3
-2
1

Let a = 0 , b = 1 and c = 0
0 0 1
Then the parametric equation of the plane is r = a + b + c
It is also possible to find the Cartesian form of the equation of a plane from the
parametric form of r = a + b + c
By definition of the parametric form, b and c are vectors parallel to the plane.
Thus b x c is a normal to the plane.
Evaluating r (b x c) = r a gives the Cartesian form of the equation of the plane.
Example Find the Cartesian form of the equation of a plane given the parametric form
of
r = a + b + c where a = i + 2j - k, b = i + j + 2k and c = 2j - k
Answer:
b x c = -5i + j + 2k
This is a normal n and a is the position vector i + 2j - k
The equation of the plane is r n = a n
So -5x + y + 2z = -5 + 2 - 2 = -5
It is worth noting that two parallel or coincident planes have normals that are
proportional.
Equation of a plane exercise

There is an exercise on the web for you to try if you like.
15 min
Q60: Find the equation of the plane which passes through (1, 2, 3) and is perpendicular
to 6i - j + 1 /3 k
Q61: Find the equation of a plane through the points
A = (1, 1, -1), B = (2, 0, 2) and C = (0, -2, 1)
Q62: Find the equation of a plane which passes through the points
A = (1, 2, 1), B = (-1, 0, 3) and C = (0, 5, -1)
Q63: Find the parametric form of the equation of the plane which has Cartesian
equation x - 2y - 6z = -3
1.9.2 Intersection of and the angles between two planes
Learning Objective

ÆFind the intersection of and the angle between two planes

Consider two planes in three-dimensional space.
In relation to each other, the two planes can:
• be coincident
• be parallel
• intersect on a line
Note: parallel or coincident planes have normals that are proportional.
Example : Parallel planes

Determine whether the two planes P 1 and P2 with equations 6x + 2y - 2z = 5 and
-3x - y + z = 2 respectively are parallel, coincident or intersect.
40 TOPIC 1. VECTORS
Answer:
A normal n for P1 is 6i + 2j - 2k
A normal m for P2 is -3i - j + k
As n = -2m, P1 and P2 are either parallel or coincident.
If P1 and P2 did intersect, the two equations 6x + 2y - 2z = 5 and -3x - y + z = 2 would
have a common solution for x, y and z.
This set of two equations have no solution and so P 1 and P2 have no points in common.
Hence, the planes do not intersect and are not coincident.
The planes are parallel.
If two planes intersect, Gaussian elimination can be used to find the vector equation
of the line of intersection. The symmetric form of the equation can also be found more
directly by algebraic manipulation of the equations. The following examples demonstrate
this.
Examples
1. Intersection of two planes in a line by Gaussian elimination

Find the vector equation of the line of intersection of the two planes P 1 with equation
4x + y - 2z = 3 and P2 with equation x + y - z = 1
Answer:
1
1
Using Gaussian elimination on the two equations gives

1
4 1
-1
-2
1
3
0
1
-3
-1
2
1
-1
That is, x + y - z = 1 and -3y + 2z = -1
Let z =
Then by substituting in -3y + 2z = -1 gives y = 1
3 + 23 and
2
+ 13

by substituting in x + y - z = 1 gives x = 3
2
1
The vector equation is r = a + d = + 23

3
3
1
3
0 1
2. Intersection of two planes in a line by algebraic manipulation

Find the symmetric equation of the line of intersection of the two planes P 1 with equation
4x + y - 2z = 3 and P2 with equation x + y - z = 1
Answer:
Subtract 2P2 from P1 to eliminate z and give the equation 2x - y = 1
Subtract P2 from P1 to eliminate y and give the equation 3x - z = 2
y+1 z+2
Solving both for x gives x = 2 = 3 or as an equation of a line in symmetric form this
is x 1- 0 = y +2 1 = z +3 2
Angle between two planes

If two planes intersect, the angle between them is equal to the angle between their
normal vectors. The equation cos = nn m
m will give the angle where n and m are the
normals.
Example Find the angle between the two planes with equations
2x + y - 2z = 5 and 3x - 6y - 2z = 7
Answer:
The normal vectors are n = 2i + j - 2k and m = 3i - 6j - 2k

cos = 3
4
7 = 4
21 so = 79Æ
Intersection and angles exercise

There is another exercise on the web if you want to try something different.
30 min
Q64: Find the equation of the plane that passes through (1, -3, 2) and is parallel to
x - 2y + z = 5
Q65: Ascertain whether the two planes 3x - 6y +9 z = -12 and -x + 2y - 3z = 10 intersect
and give the equation if they do.
Q66: Find the equation of a plane in Cartesian form which passes through the point
(1, -1, 3) and is parallel to the plane 3x + y + z = 7
Q67: Find the symmetric equation of the line of intersection of the two planes
3x + y - z = -3 and x - y - z = 1
Q68: Ascertain whether the two planes 3x - y + 2z = 5 and x + 2y - z = 4 intersect and
if they do give the equation of the line in parametric form.
Q69: Find the angle between the planes x + y - 4z = -1 and 2x - 3y + 4z = -5
Q70: Find the acute angle between the planes 2x - y = 0 and x + y + z = 0
1.9.3 Intersection of three planes
Learning Objective

ÆFind the intersection of and the angle between three planes

Consider three distinct non-parallel planes in three-dimensional space. If no two of the
three planes are parallel or coincident, the three planes can:
42 TOPIC 1. VECTORS
• intersect at a point
• intersect at a line
• not intersect
The technique of Gaussian elimination can be used to find the point or line of intersection
of three planes.
Examples
1. Intersection of three planes at a point

Find the point of intersection between the three planes
P1 : x - 2y - z = 3
P2 : 2x + y - z = -1
P3 : -x + y + 2z = 2
Answer:
Gaussian elimination gives
1 -2 -1 3
1 -2 -1 3
1 -2 -1 3

2 1 -1 -1 =0 5 1 -7 =0 5 1 -7
-1 1 2 2 0 -1 1 5 0 0 6 18
This gives z = 3 and substituting this value back into row 2 gives 5y + 3 = -7
Thus y = -2
A further substitution gives x = 2 and the full solution of x = 2, y = -2 and z = 3
The point of intersection is (2, -2, 3)
2. Intersection of three planes in a line

Find the symmetric equation of the line of intersection of the planes
P1 : x + y - z = 0
P 2 : 2x - y + 4z = -3
P3 : x + 3y - 5 z = 2
Answer:
1 1 -1 0
1 1 -1 0
1 1 -1 0

2 -1 4 -3 =0 -3 6 -3 =0 1 -2 1
1 3 -5 2 0 2 -4 2 0 0 0 0
From row 2: y = 2z + 1
From row 1: x = z - y = -z - 1
Let z = to give the parametric equations x = -1 - , y = 1 + 2, z =
x+1 y-1 z-0
This gives the intersection as a line with symmetric equation -1 = 2 = 1
3. No intersection of three planes

Show that the planes do not intersect where
P1 : x + y + z = 2
P2 : 2x - 2y + z = 5
P3 : 3x - y + 2z = -1
Answer:

Using Gaussian elimination gives
1 1 1 2 1 1 1 2
1 1 1 2

2 -2 1 5 =0 -4 -1 1 =0 -4 -1 1
3 -1 2 -1 0 -4 -1 -7 0 0 0 -8
This gives 0 = -8, which is impossible. The planes do not intersect at a common point
or line.
Intersection of three planes exercise

There is another exercise on the web if you want to try something different.
30 min
Q71: Find the point of intersection of the planes
P1 = 6x - 9y - 8z = 6
P2 = 3x - 2y - 9z = -17
P3 = 6x + 6y - 5z = -15
Q72: P1 = x + y + z = 6, P2 = -x + y + 2z = 5 and P3 = 3x + 2z = 12
Ascertain whether the three planes P 1 , P2 and P3 as shown intersect. If they do, provide
the relevant information on the intersection.
44 TOPIC 1. VECTORS
Q73: If the planes P1 with equation x + y + z = 1, P2 with equation 2x + 2y + z = 3 and

P3 with equation 4x + 4y + 3z = 5 intersect, give details of the intersection.
Q74: Ascertain whether the three planes with equations

4x + y + 5z = 2, -x + y - 2z = 7 and 3x - 3y + 6z = 21 intersect.
If they do provide details of the intersection.
Q75: Find the angle between the planes x + y - 4z = -1 and 2x - 3y + 4z = -5
Q76: Find the acute angle between the planes 2x - y = 0 and x + y + z = 0
1.10 Lines and planes

Learning Objective

ÆFind relationships between lines and planes

The intersection of a line and a plane can be found by substituting the equation of the
line in parametric form into the equation of the plane and solving for
If there is a unique solution then a point of intersection exists. If there are infinite
solutions then the line lies on the plane. If there is no solution then the line is parallel to
the plane.
Example : Intersection of a line and a plane

Find the intersection between the plane x - y - 2z = -15 and the line with symmetric
equation x 1- 1 = y 1- 2 = z 2- 3
Answer:
In parametric form this line is given by the equations
x = 1 + , y = 2 + and z = 3 + 2
Substitute this into the equation of the plane to give
1 + - 2 - - 6 - 4 = -15
i.e. = 2
When = 2, x = 3, y = 4 and z = 7
Hence the point of intersection is (3, 4, 7)
If a line L and a plane P intersect, it is possible to calculate the angle of intersection

between the line and the plane. To do this, find the normal n to the plane and find the
angle between L and L 2 where L2 is in the direction n. The angle of intersection between
the line and the plane is the complement of the angle between the line and the normal
to the plane.
1.10. LINES AND PLANES 45
Example : Angle of intersection between a line and a plane

x-1 y-2 z+3
Find the acute angle between the line 4 = 3 = 2
and the plane -10x + 2y - z = 1

Answer:
Take the direction vector of the line, that is, d = 4i + 3j + 2k
The normal to the plane is n = -10i + 2j - k
Thus cos = nd
n d =
- 36
and so = 130.72 Æ
105 29
Thus the acute angle is 49.28 Æ

The angle between the line and plane is 90 - 49.28 = 40.72 Æ
It is particularly difficult to visualise lines and planes in three dimensions but special
isometric graph paper can help.
Activity
Take some isometric graph paper and for a selection of the lines or planes in this topic,
try some drawings. Note any intersections and confirm the results given such as parallel
planes, skew lines or three planes intersecting at a point.
Interactive lines and planes in space

There is an animation on the web of lines and planes in space to help with the concepts
of intersections, angles and parallel properties. 10 min
46 TOPIC 1. VECTORS
Lines, points and planes exercise

There are some questions on the web for you to try if you wish.
15 min
Q77: Find the point of intersection between the plane 3x + 2y - z = 5 and the line with
symmetric equations x 2- 1 = y -+11 = 3z
Q78: Find the acute angle between

x+1 y z-3
the line 2 = 3 = 6 and the plane 10x + 2y - 11z = 3
Q79: Show that the line of intersection of the planes

x+3 y z-1
x + 2y - 2z = 5 and 5x - 2y - z = 0 is parallel to the line 2 = 3 = 4
Q80: Find a plane through the point (2, 1, -1) and perpendicular to the line of
intersection of the planes 2x + y - z = 3 and x + 2y + z = 2
1.11 Summary
At this stage the following techniques and methods should be familiar:
• Using basic vector arithmetic.
• Calculating scalar, vector and scalar triple products.
• Using direction cosines and direction ratios.
• Finding the equation of a line and solving related problems.
• Finding the equation of a plane and solving related problems.
• Finding the intersections of and angles between lines and planes.
1.12 Proofs
Proof 1: a b = a b cos
a1
b1

Let a = a2 and b = b2
a3 b3
c=b-a
1.12. PROOFS 47
c 2 = b - a 2
= (b1 - a1 )2 + (b2 - a2 )2 + (b3 - a3 )2
= b21 - 2a1 b1 + a21 + b22 - 2a2 b2 + a22 + b23 - 2a3 b3 + a23
Thus = (a21 + a22 + a23 b21 + b22 + b23 - 2(a1 b1 + a2 b2 + a3 b3
a b - 2(a b but
=
c 2 = a 2 + b 2 - 2 a b cos by the cosine rule so
(a b = a b cos
Proof 2: a x b = a b sin
a x b 2 = (a x b) (a x b)
a2 b3 - b2 a3
a2 b3 - b2 a3

= a1 b3 - b1 a3 a1b3 - b1a3
a1 b2 - b1 a2 a1 b2 - b1 a2
= a22 b23 - 2a2 b2 a3 b + a23 b22 + a21 b23 - 2a1 b1 a3 b3 + a23 b21 + a21 b22 - 2a1 b1 a2 b2 + a22 b21
= a21 b21 + a22 b21 + a23 b21 + a21 b22 + a22 b22 + a23 b22 + a21 b23 + a22 b23 + a23 b23
- a21 b21 + 2a1 b1 a2 b2 + a22 b22 + 2a2 b2 a3 b3 + a23 b23 + 2a1 b1 a3 b3
= (a21 + a22 + a23 b21 + b22 + b23 - a1 b1 + a2 b2 + a3 b3 2
= a b - a b
= a b - a b cos2
= a b (1 - cos2
= a b sin2
a x b = a b sin
Note the trick of adding in and then subtracting the terms that are underlined.
Proof 3: a (b + c) = a b + a c
a1
b1
c1

Let a = a2 , b = b2 and c = c2
a3 b3 c3
b1 + c1

b + c = b2 + c2
b3 + c3
a1
b1 + c1

a (b + c) = b2 + c2 = a1b1 + a1c1 a2b2 + a2c2 + a3b3 + a3c3
a2
a3
b3 + c3
a1
b1

ab= a2 b2 = a1b1 + a2b2 + a3b3

a3 b3
a1
c1

ac= a2 c2 = a1c1 + a2c2 + a3c3
a3 c3
Thus
48 TOPIC 1. VECTORS
a b+a c = a1 b1 + a2 b2 + a3 b3 + a1 c1 + a2 c2 + a3 c3
= a1 b1 + a1 c1 + a2 b2 + a2 c2 + a3 b3 + a3 c3
Proof 4: a b = b a
Let a = a1 i + a2 j + a3 k and b = b1 i + b2 j + b3 k
ab
= a1 b1 + a2 b2 + a3 b3
= b1 a1 + b2 a1 + b3 a3 (since ab = ba for any real numbers)
=ba
Proof 5: a a = a 2 0
Recall the definition of length.

If p = ai + bj + ck then the length of p = a2 + b2 + c2 = p
But using the scalar product
p p = (a a) + (b b) + (c c) = a2 + b2 + c2 = p
Since any square of a real number is positive then p p is always greater than or equal
to zero.
Proof 6: a a = 0 if and only if a = 0
a1

Let a = a2
a3
a1
a
then a a = a2 a12 = a1 2 + a2 2 + a3 2
a3 a3
The square of a real number is positive.
So a a = 0 only if a1 2 = a2 2 = a3 2 = 0
Thus a1 = a2 = a3 = 0 a = 0
Proof 7: For non-zero vectors, a and b are perpendicular if and only if a b = 0
ab= a b cos
If a b = 0 then a b cos = 0
cos = 0 since a and b are non-zero vectors.
Thus = 90Æ and the vectors a and b are perpendicular.
Conversely if a and b are perpendicular then a b cos = 0
and so a b = 0
Proof 8: a (b x c) = b (c x a) = c (a x b)
b x c = (b2 c3 - c2 b3 )i - (b1 c3 - c1 b3 j + (b1 c2 - c1 b2 )k
a (b x c) = a1 b2 c3 - a1 c2 b3 - a2 b1 c3 + a2 c1 b3 + a3 b1 c2 - a3 c1 b2
1.13. EXTENDED INFORMATION 49
c x a = (c2 a3 - a2 c3 )i - (c1 a3 - a1 c3 )j + (c1 a2 - a1 c2 )k

b (c x a) = b1 c2 a3 - b1 a2 c3 - b2 c1 a3 + b2 a1 c3 + b3 c1 a2 - b3 a1 c3
a x b = (a2 b3 - b2 a3 )i - (a1 b3 - b1 a3 )j + (a1 b2 - b1 a2 )k
c (a x b) = c1 a2 b3 - c1 b2 a3 - c2 a1 b3 + c2 b1 a3 + c3 a1 b2 - c3 b1 a2
Rearrangement of the terms in each show that they are all equal.
1.13 Extended information

Learning Objective

ÆDisplay a knowledge of the additional information available on this subject

There are links on the web which give a selection of interesting sites to visit. These
sites can lead to an advanced study of the topic but there are many areas that will be of
passing interest.
Stevin
Simon Stevin was a Flemish mathematician of the 16/17th century. He was an
outstanding engineer and used the concept of vector addition on forces. This was
however a long time before vectors were generally accepted.
Hamilton
This Irish mathematician was the first, in 1853, to use the term ’vector’.
Gibbs
In the late 1800s, Josiah Gibbs used vectors in his lectures. He made major
contributions to the work in vector analysis.
Weatherburn
An Australian, Charles Weatherburn published books in 1921 on vector analysis.
There are many more mathematicians, such as those already mentioned in the complex
numbers topic, who played a part in the development of vectors.
1.14 Review exercise

Review exercise in vectors
Choose two questions to complete in a time of about 15 minutes.
15 min
There is a web exercise available for you to try if you prefer it.
Q81: Given the vectors a = 3j - k, b = i + 2j - k and c = -i + 2j + 2k,

calculate a x b and a (b x c)
50 TOPIC 1. VECTORS
Q82: Obtain, in parametric form, an equation for the line which passes through the
points (2, - 3, 1) and (1, -1, 7)
Q83: Find the equation of the plane which has a normal vector n = 2i + j + 3k and
passes through the point (1, 2, 3)
Q84: Find the equation of the line of intersection in parametric form of the planes
2x - 3y + z = 3 and 3x + 2y + z = 2
Q85: Find the point of intersection of the planes

P1 : 3x - 2y - 4z = 3
P2 : x + y + z = 4
P3 : -2x - 2y - 3z = -10
1.15 Advanced review exercise

Advanced review exercise in vectors
There is a similar exercise on the web for you to try if you like.
15 min
Q86: Let u = i - 4j - k and v = 2i - 2j + k.
a) Find u v and u x v
b) Three planes 1 , 2 , 3 are given by the equations
1 : x - 4y - z = 3
2 : 2x - 2y + z = 6
3 : 3x - 11y - 2z = 10
i) Find the acute angle between the planes 1 and 2
ii) By using Gaussian elimination show that the planes 1 , 2 and 3 intersect in a point
Q and obtain the coordinates of Q
iii) Find an equation for the line L in which 1 and 2 intersect, and the point R in which
L intersects the xy-plane.
c) Three non-zero vectors a, b and c are such that
a x b = c and b x c = a
Explain briefly why a, b and c must be mutually perpendicular and why b must be a unit
vector.
(1996 CSYS paper II)
Q87: P is the point (2, -3, 1) and Q is (-1, 2, -1)

a) Find p q and p x q
b) Three planes P 1 , P2 and P3 are given by the equations
P1 : 4x - 4y + z = 8
P2 : 2x + 3y - z = -1
P3 : x + y + z = 0
1.16. SET REVIEW EXERCISE 51
i) By Gaussian elimination, or otherwise, find the point of intersection of the three planes.
ii) Find the equation of the line of intersection of the planes P 2 and P3
iii) State where it meets the yz-plane.
Q88: a) Find the equation of the plane through the points

P (1, -2, 6), Q ( 1, 0, 3) and R (2, 1, 2) in Cartesian form.
b) Find the angle QPR
c) Find the angle between the plane and the line with parametric equations
x = 2 - 3, y = 3, z = 4 + 1
1.16 Set review exercise

Set review exercise
The answers for this exercise are only available on the web by entering the answers
obtained in an exercise called ’set review exercise’. The questions may be structured 15 min
differently but will require the same answers.
Q89: Let p = 3i - j - k and q = -i + 3kFind q x p and p q
Q90: Find the vector, parametric and symmetric forms of the equation of the line which
passes through the points (-2, 1, -1) and (3, 2, -2)
Q91: Give the equation of the plane with normal vector n = i + 2j - 2k and which passes
through the point (1, -3, 1)
52 TOPIC 1. VECTORS
53
Topic 2
Matrix algebra
Contents
2.1 Revision exercise . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
2.2 Basic matrix terminology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
2.3 Matrix operations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56
2.4 Determinants and inverses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
2.4.1 Determinants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
2.4.2 Inverses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
2.5 Properties of matrices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
2.6 Matrices and transformations . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76
2.7 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
2.8 Proofs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
2.10 Review exercise . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88
Learning Objectives
• Use matrix algebra.
• Perform the matrix operations of addition, subtraction and multiplication.
• Calculate the determinant of a 3 3 matrix.
• Find the inverse of a 2 2 matrix.

54 TOPIC 2. MATRIX ALGEBRA
Prerequisites
A sound knowledge of the following subjects is required for this unit:
• Basic matrix concepts.
• Transformations of functions.
2.1 Revision exercise

Learning Objective

ÆIdentify areas that need revision

Revision exercise
20 min revise the techniques before starting this topic. Ask a tutor for advice.
2 7 4

Q1: Name the element a 23 in the matrix

-3 -2 1
Q2: Using elementary row operations, solve the system of equations

2x - y + 2z = 11
x+y+z=4
3x - y - 2z = 1
Q3: Give the coordinates of the point (2, 3) under the following transformations:
1. under reflection in the x-axis.

2. under rotation by 90 Æ anticlockwise.
Q4: Find the vector product of the two vectors a and b

where a = a1 i + a2 j - a3 k and b = b1 i - b2 j - b3 k
2.2 Basic matrix terminology

Learning Objective

ÆKnow the basic terminology used in matrix algebra

The following terms are worth emphasing at this point as they will be used frequently in
this topic.
2.2. BASIC MATRIX TERMINOLOGY 55
Matrix
A matrix is a rectangular array of numbers.
Example
2 -3

1 5 This matrix has 3 rows, 2 columns and 6 elements.
3 -1
The order of this matrix is taken from the number of rows and the number of columns.
In this case the matrix has order 3 x 2
In general terms, an m n matrix will have m rows, n columns and mn elements.

a11
a1n
a12

a

m rows
a21 a22
It can be shown as
2n
am1 am2
amn

n columns

This matrix A is written as A = a ij mn
aij is the element in the i-th row and the j-th column.
Here are examples of matrices:
2 1
, 1
1 -2
, 0
2 5 2 and
0
0 1
3 1 2
Column matrix
A column matrix has only one column but any number of rows.
1
The matrix is called a column matrix.
2
3
Row matrix
A row matrix has only one row but any number of columns.
The matrix
1 2
is called a row matrix.
Square matrix
A square matrix has the same number of rows and columns.
1 0

The matrix is a square matrix. This particular example has a special property
0 1
which will be explained in the section on determinants and inverses.
2.3 Matrix operations

Learning Objective

ÆPerform addition, subtraction and multiplication operations on matrices
It is possible to carry out basic arithmetic operations such as addition and subtraction

with matrices.
In fact, matrix addition has already been demonstrated in a previous topic. Recall the
following example from topic 11 on vectors.

and q = y are vectors then p + q is the vector b + y ’
’If p =
a
b
x a+x
c z c+z
The ordered triple form of the vector is, of course, a 3 x 1 matrix and this example shows
the matrix addition of two 3 x 1 matrices.
Note: only matrices of the same order can be added (or subtracted). The technique is
to add the corresponding ijth elements together.
Matrices which have the same order are said to be conformable for addition (or
subtraction).
Expanding on this, matrix addition can be defined as follows:
Addition of two matrices
If A and B are two matrices of the same order, the matrix A + B is formed by adding
the corresponding elements of each matrix.
Example : Addition of matrices

Add the following matrices
-2 3

1 -1

1. and
-1 -1 2 -2
-1
-2

2. 2 and 3
0 2
3.
-1 0 -1 1
and 1 1 -2 1

Answer:
-1 2

1.
1 -3
-3

2. 5
2
3.
0 1 -3 2

2.3. MATRIX OPERATIONS 57
Matrix adding examples

There are also examples of matrix addition on the web.
10 min
Adding matrices exercise
There is another exercise on the web for you to try if you wish.
-2 -4

0 -1

5 min
Q5: Add and
2 -1 2 3
1
2

Q6: Add -2 and 2
0 1
-1 9 -4
-3 2 1

Q7: Add 2 3 - 2 and 0 -4 3
0 -1 2 -1 1 -2
-2 3

3 -3

Q8: Add and

-1 -1 1 2
Subtraction is just as straightforward as addition. Note that again the matrices must
have the same order.
Subtraction of two matrices
If A and B are two matrices of the same order, the matrix A - B is formed by subtracting
the corresponding elements of matrix B from those in matrix A.
Example : Subtraction of matrices

Subtract the following matrices:
-2 3

1 -1

1. -
-1 -1 2 -2
-1
-2

2. 2 - 3
0 2
3.
-1 0 -1 1
-1 1 -2 1

Answer:
-3 4

1.
-3 1
1

2. -1
-2
3.
-2 -1 1 0

Matrix subtracting examples

There are also examples of matrix subtraction on the web.
10 min
Subtracting matrices exercise
5 min
-3 -4

0 -1

Q9: -
-3 -1 2 1
1 2

Q10: -
2 2
0 -1
-1 4 -2
-3 0 1

Q11: 2 -1 -1 - 0 -2 3
0 3 2 -1 3 -2
-2 3

3 -3

Q12: -
-1 -1 1 2
Matrices can be multiplied by a scalar. Again, in topic 11 on vectors this was

demonstrated. Recall the paragraph:
’Vectors can also be multiplied by a scalar.
a a

If p = and is a real number, (a scalar) then p = b ’
b
c c
This technique is used for a matrix of any order.
Multiplication by a scalar
To multiply an m x n matrix A by a scalar , take each element and multiply it by to
give the matrix A.
Example : Multiplication of a matrix by a scalar

-3 0 1

Multiply the matrix 2 1 -2 by -3
-3 1 1
Answer:
-3 0 1
9 0 -3

-3 2 1 -2 = -6 -3 6
-3 1 1 9 -3 -3
From the previous example it is clear to see that for any matrix A multiplied by a scalar
then A = A
Note also that the new matrix A has the same order as the original matrix A.
Multiplication of a matrix by a scalar exercise

There is a similar exercise on the web for you to try if you prefer.
Q13: Find the matrix formed by A where is -2 and A =

-2 -1 3
5 min
Q14: State the value of when

2 -1 3

-2 1 -3

A= and A =
-1 1 -2 1 -1 2
1 0 -1

Q15: If A = 2 1 -2 find the matrix A which has element 23 = 8
3 -2 4
The final matrix operation to consider in this section is matrix multiplication.

Two matrices can only be multiplied together if the number of columns in the first matrix
is equal to the number of rows in the second matrix.
When this is possible the two matrices are said to be conformable for multiplication.
Example : Matrices conformable for multiplication

times
is possible since the first matrix (of order 3 x 1) has 3 column

and the second matrix (of order 1 x 3) has 3 rows.

However times is impossible since the first matrix has 3

columns but the second matrix only has 2 rows.
It is worth remembering to multiply rows into columns.

The first element of the product of two 3 x 3 matrices is shown on the diagram. This
structure may help to make it easier to remember.
The formal definition of matrix multiplication follows.
Multiplication of matrices
The product of the two matrices A mn and Bnp = Cmp
a11 a12 a1n
b11 b12 b1p

a21 a22 a2n and B =
b b22 b2p

If A = 21
am1 am2 amn bn1 bn2 bnp
then the matrix AB =
a11 b11 + + a1n bn1 a11 b12 + + a1n bn2 a11 b1p + + a1n bnp


am1 b11 + + amn bn1 am1 b12 + + amn bn2 am1 b1p + + amn bnp
Some examples will help to make this clearer.
Example : Matrix multiplication

-1 2

Find AB where A = 0 3 and B = -1 2

3 1
-2
0
-2 1
Answer:
Since matrix A has order 3 x 2 and matrix B has order 2 x 3 the resulting matrix AB will
have order 3 x 3
(-1 -1+2 3) ( - 1 2+2 1) ( - 1 -2+2 0)

AB (0 -1+3 3) (0 2+3 1) (0 -2+3 0)

(-2
7 0 2
-1+1
3) ( - 2 2+1 1) ( - 2 -2+1 0)
= 9 3 0
5 -3 4
Multiplying matrices exercise

10 min
Q16: Multiply the matrices
-1 2

- 3 and
-2
1
-2
-1 0 -2
0 2 -1 1

Q17: Multiply 3 2 -1 and 4 -3 2 -2
1 2 0 2 0 1 1

Q18: For the four matrices A, B, C, D, form all the possible products.
2
1 2
A=
3
1
-2
-1
1
,B= -2 1
1 0 3
3
0 3
1 1 -2

C= 2 -1 -1 and D = 1 1 0 -1
-3 2 -2 2 -2 1 2
-2 1 0 2
0 1 2 0

by

3 -2 1 0 -2 1 -1 3
Q19: Multiply 1 0 1 0 2 -1 1 1
2 1 -1 1 -1 0 0 1
Note carefully:
Two matrices are equal if and only if the following two conditions apply:
• They have the same order.
• Each corresponding element ij is equal.
Example : Equality of matrices

Find the values of x and y which make the two matrices A and B equal when
x+2 4

3 x+3

A= and B =
-1-y 0 1 2x + y
Answer:
Both matrices are 2 x 2
Equate each element in turn
x+2=3
4=x+3
-1 - y = 1
0 = 2x + y
and solve these to give x = 1 and y = -2
3 4

Thus if x = 1 and y = -2 both matrices represent the matrix

1 0
If there are no values of x and y which satisfy all the equations then the matrices are not
equal.
Q20: Are the following pairs of matrices equal:

x 4

2-x x+3

1. and

2-x
y-2 2 4 y
x 4 x+3
2. and

y-2 2 4 y-4
1
1 2
3.
2
3 4
and 3 4

0 0
x-2 4 1 x+1
4. x 2 and 3 y+2
1 y-1 y+1 -1
2.4 Determinants and inverses

Learning Objective

ÆFind and use the determinant and the inverse of a matrix

2.4.1 Determinants
The word ’determinant’ may be familiar. If you have studied vectors consider the vector
product . The work which now follows should clarify the relationship between the vector
product and the determinant.
Determinant of a matrix
The determinant of a matrix is a value representing sums and products of a square
matrix.
a11 a12 a1n

If the matrix A =
an1 an2 ann

then the determinant of A is det A =
a11 a12 a1n

an1 an2 ann
Note that the determinant only exists for a square matrix and that the determinant is a
number, not a matrix.
Determinant value of a 2 x 2 matrix

then the value of det A = = ad - bc
b a b
If A =
c d c d
-1 -2

Example Find the determinant of the matrix A =

2 -3
Answer:
det A = ad - bc = (-1 -3) - (-2 2) = 3 + 4 = 7
2 x 2 determinant exercise
There is another exercise on the web for you to try if you prefer.
5 min
Q21: Find the determinant of the following matrices:
-1 2

a)
-3 1
3 -2

b)
-1
-1 0
2

c)
-3 1
2 2

d)
-1 1
2.4. DETERMINANTS AND INVERSES 63
-2 2

e)
-3 3
The calculation of the determinant of a 3 x 3 matrix is slightly more complicated and

involves the technique previously shown for the 2 x 2 matrices.
There are two methods. The one explained here is called the Laplace expansion.

Determinant value of a 3 x 3 matrix
a b c
If A = d e f then the value of det A =
g h i
d e
d f
a - b + c =
e f
h i g i g h
a(ei - hf) - b(di - gf) + c(dh - ge)
Remember the minus sign for the middle term in this calculation.
Compare this formula with that for the vector product :
’If p = ai + bj + ck and q = di + ej + fk
then p x q = (bf - ec)i - (af - dc)j + (ae - db)k
i j k

This is actually the determinant of the matrix a b c ’
d e f
Important: The same formula is used but the vector product is a vector whereas the
determinant of a matrix is a value.
There is an easy way to remember the calculation for the determinant of a 3 x 3 matrix.
• Make two copies of the matrix side by side.
• For each entry in the top row of the first matrix multiply down the diagonal to the
right.
• For each entry in the top row of the second copy of the matrix, multiply down the
diagonal to the left and add a minus sign.
2 -1 2

Example Find the determinant of -3 0 1
0 1 2
Answer:
The determinant is
2 (0 2 - 1 1) - (-1)(-3 2 - 0 1) + 2(-3 1 - 0 0) = -2 - 6 - 6 = -14
3 x 3 determinant exercise
There is a web version of this exercise for you to try if you prefer.
10 min
Q22: Find the determinant of the following matrices:
-2 1 3

a) 0 -4 2

-1 2 1
3 -2 0
b) 3 -1 1

-2 1 1
1 0 0
c) 0 1 0

0 0 1
-2 1 3
d) 1 -2 1
3 1 -2
-2 8

Consider the matrix

1 -4
This has a determinant of 0 and matrices such as this have a special name.
Singular matrix
A matrix whose determinant is zero is called a singular matrix.
Conversely, any matrix which has a non-zero determinant is said to be non-singular.
2.4.2 Inverses
Although the formula for a determinant can be used to find the vector product, its main
use is in finding the inverse of a matrix.
In basic arithmetic for any number n there is a multiplicative inverse (1 /n ) such that
n 1 /n = 1
A multiplicative inverse also exists for a matrix if it satisfies two conditions.
1. The matrix must be square.
2. The determinant of the matrix must be non-zero.
When these conditions are satisfied the matrix is said be to invertible.
Inverse of a 2 x 2 matrix
a b

Let A be the square non-singular matrix

c d
a b

d -b

1
The inverse of A is denoted A -1 = = ad - bc
c d -c a
The definition clearly shows why the matrix has to be non-singular.

1 1
If the determinant is zero then ad - bc = 0 which is undefined.
When the inverse is calculated it is customary either to leave the fraction outside or, if
taken inside, to leave each element of the matrix as a fraction.
Example : Inverse of a 2 x 2 matrix

-2 2

Find the inverse of the matrix A =

3 4
Answer:
The inverse exists since the matrix A is a square non-singular matrix.
det A = (-2 4) - (3 2) = -14
4 -2

-2 1

1
a-1 = - 14
7
3
7
1
-3 -2 14 7
Inverse example
There is an animation of the construction of an inverse on the web for you to look at if
you wish. 5 min
2 x 2 inverse exercise
There is a web exercise for you to try if you prefer.
10 min
Q23: Find the inverses of the following matrices:
-1 2

a)
-3 1
3 -2

b)
-1
-1 0

c)
-3 1
2 2

d)
-1 1
2 2

e)
-3 3
In the same way that it is possible to check that an integer times its inverse is equal to
one (e.g. 3 1 /3 = 1), it is possible to check that the matrix times its inverse gives the
matrix equivalent of 1. The matrix equivalent of 1 is called the identity matrix.
Identity matrix
An n x n matrix which has only ones on the leading diagonal and zeros elsewhere is
called an identity matrix and takes the form
1 0 0

0 1 0 n

0 0 1 It is denoted by I.
n
The structure of this matrix may look rather familiar.

1 0 0

In vectors the standard basis vectors can be shown as the matrix 0 1 0
0 0 1
This in fact is an identity matrix. It is the 3 x 3 identity matrix.
Q24: What is the 2 x 2 identity matrix?
Example A previous example showed the inverse of A where

-2 2

-2 1

A= as A-1 = 3
7 7
1
3 4

1
14 7
2 1
Checking AA-1 gives
-2 2
3 4
-
3
7 7
1 =
0
0 1
14 7
Activity
Calculate AA-1 and verify that this gives the identity matrix shown. Find A-1 A and make
a conclusion.
In all cases this is a sensible simple check to make when an inverse of any matrix has
been calculated.
The easiest method to use for finding the inverse of a 3 x 3 matrix relies on the
elementary row operations that were covered in topic 5 - Systems of linear equations.
Strategy for finding the inverse of a 3 x 3 matrix

1

• Form an augmented matrix A I ( A 0
0 0

1 0 ) using the given 3 x 3 matrix
0 0 1
(A) at the left and the 3 x 3 identity matrix (I) at the right.
• Perform elementary row operations to reduce the left-hand matrix to the 3 x 3

identity matrix (Gauss-Jordan elimination).
• The effect on the right matrix will be to convert it into the inverse required.
As a brief reminder, here is the definition of elementary row operations.

Elementary row operations
The three ways in which a matrix can be manipulated to solve a system of equations
are called elementary row operations. They are:
• Interchange two rows.
• Multiply one row by a non-zero constant.
• Change one row by adding a multiple of another row.
Elementary row operations

There are also examples of the application of each operation on the web.
It is important to carry out the elementary row operations in such a way that the identity 10 min
matrix is formed on the left hand side.
3 x 3 inverse exercise
There is another version of this exercise on the web for you to try if you like.
15 min
Q25: By using ERO’s find the inverses for the following matrices:
1 -2 0

1. -3 8 -3

2 -5 2
-1 0 0
2. 2 -1 0

5 -4 1
1 0 -2
3. -1 1 0

2 2 1
2 -1 3
4. 1 3 0
1 2 2
1 0 2

5. 1 2 4
3 -2 1
In topic 5, elementary row operations were used to solve a system of three equations in
three unknowns say, x, y and z.
The coefficients of these unknowns were used to form a matrix of the coefficients, say
A.
The system can then be represented by the equation AX = B where X is the column
matrix of the unknowns x, y, and z and B is the column matrix of the constants.
If the augmented matrix A I is subsequently converted by elementary row operations

to I A - 1 this can be used to find a solution to this system of equations. A little bit
of algebra completes the picture.
AX = B (the system of equations)
Multiply both sides by A-1 to give A-1 AX = A-1 B
But A-1 A = I and so X = A-1 B
Thus multiplying the column matrix of constants by the inverse of the matrix of
coefficients gives the column matrix of values for the unknowns.
An example will demonstrate this technique.
Example Find the solution to the system of linear equations using the matrix method,
where the equations are:
x + 2y + z = 1
2x + 5y + 2z = 3
2x + 4y + z = 5
Answer:
Find the matrix of coefficients, say A.
1 2 1

This gives A = 2 5 2
2 4 1
1 2 1 1 0 0

By using EROs on the augmented matrix 2 5 2 0 1 0
3 -2 1
2 4 1 0 0 1
the inverse A-1 = -2 1 0

2 0 -1
x
Let X be the matrix then AX = B where B = 3
y
z 5
This equation (AX = B) represents the system of equations.
so A-1 AX = A-1 B
But A-1 A = I
so X = A-1 B
3 -2 1
1
2

Thus X = -2 1 0 3= 1
2 0 -1 5 -3
The values are x = 2, y = 1 and z = -3
This method is an alternative to the Gaussian elimination method but it is doubtful

whether it is any quicker or easier to use.
The determinant however can indicate if a system of equations has a unique solution or
not.
If an inverse exists, that is, if the matrix of coefficients is square and non-singular then
there is a unique solution to the system.
This determinant check can also be used to determine the nature of an intersection of
three planes in three-dimensional space.
If the determinant exists, then the intersection of the planes is at a point. It can also
indicate, by the same argument, that three lines do intersect in three-dimensional space.
It is important not to confuse the inverse matrix A-1 explained earlier with the transpose
of a matrix.
Transpose of a matrix
The transpose of a matrix A of order m n is found by reflecting the matrix in its main
diagonal.
It is denoted AT or A .
The effect is to interchange the rows of the matrix with the columns and A T has order
n m.
In contrast to an inverse, a transpose exists for any matrix; the matrix does not have to
be square or non-singular.
Example : Transpose of a matrix

2 0

Find AT when A = -1 -4
1 3
Answer:
2 -1 1

AT =
0 -4 3
Tranposing matrices
There are four examples of forming the transpose on the web.
10 min For some square matrices AT = A-1 . These matrices have a special name.
Orthogonal matrix
A square matrix A is orthogonal A T = A-1
AT = A-1 det A = 1
This information will be useful in the section on transformations.
2.5 Properties of matrices

Learning Objective

ÆUse the properties of matrix operations

There are several interesting results which can be found from examining the operations
on matrices. These are rather similar in some cases to the properties of operations on
real numbers. Some of these may seem obvious.
The following properties are assumed:
• AI = A
• AA-1 = I
In each of the properties mentioned A, B and C are matrices with general elements of
aij bij and cij respectively.
Property 1: A + B = B + A
Condition: A and B must have the same order for addition.
For any two real numbers a 1 and b1 the sum a1 + b1 = b1 + a1
The matrix (A + B) will have elements (a + b) ij and (B + A) has elements (b + a) ij
But (a + b) = (b + a) and so each element of A + B and B + A are equal.
Thus A + B = B + A
Example Show that A + B = B + A

-2 1
-3 2

when A = 0 3 and B = 1 0
-4 -2 3 2
Answer:
- 2 + ( - 3) 1+2
-5 3

A+B= 0+1 3+0 = 1 3
-4+3 -2+2 -1 0
2.5. PROPERTIES OF MATRICES 71
- 3 + ( - 2) 2 + 1
-5 3

and B + A = 1+0 0+3 1 3 as required.
3-4 2-2 -1 0
Property 2: AB BA in general - this is the commutative property.

Condition: to form either AB or BA the matrices A and B must be square, otherwise both
products cannot be formed.
Suppose that A and B have order m m and that A and B are not identical. (That is,
each element of A each element in B.)
Consider a general element e rs in each product.
For AB the element e rs = (ar1 b1s + ar2 b2s + ... arm bms )
However for BA the element e rs = (br1 a1s + br2 a2s + ... brm ams )
Since A and B are not identical, it follows that there is at least one element e pq such that
apm bmq bpm amq
Thus AB BA in general.
3

then AB
4 -2 1
Example Show that if A = and B = BA
2 1 0 -1
Answer:
-6 -1

-4 -7

AB = and BA =
-4 1 -2 -1
Property 3: (AB)C = A(BC) - this is the associative property.

Condition: for either product to exist, the orders of the matrices must be
A = m n, B = n p, C = p q
This property follows from the fact that for any three real numbers a, b and c,
(ab)c = a(bc) and so each element of ABC can be expressed in either form. The general
form of this proof is cumbersome and shown, in the section headed Proofs near the end
of this topic, as Proof 1.

1 0 1 -1
1

3 and C =
1 2 0 2
Example If A =
-2 1 1
, B= 2
-2
1 0
-1 1 0
-1 show
1
that (AB)C = A(BC)
Answer:
5 2 1 5

AB = and
-2 0 -1 5

5
1
13
(AB)C =
5 2
-2 0
1
-1 5
2
-1

= 4
1
-1

-1

BC = 7 and A(BC) = 1 2
-2 1 1
0 7 = 134 as required.
-5 -5
Property 4: A(B + C) = AB + AC - this is the distributive property

Condition: B and C must have the same order and the number of columns in A must
equal the number of rows in B and C
The proof is shown for interest in the section headed Proofs near the end of this topic
as Proof 2.
1 -2

Example Let A = -1 3 , B = -3
2
1
-1
and C =
0
-1
2
-1
and show
0 2
that A(B + C) = AB + AC
Answer:
-3 3

B+C= and
1 -2
1 -2

-5 7

A(B + C) = -1 3 -3
1
3
-2
= 6 -9
0 2 2 -4
-7 3
2 4

AB = 9 -4 and AC = -3 -5
4 -2 -2 -2
-7 3
2 4
-5 7

so AB + AC = 9 -4 + -3 -5 = 6 -9 as required.
4 -2 -2 -2 2 -4
Property 5: (A’ )’ = A or (AT )T = A

a11 a12 a1n

Let A be the m n matrix
a21 a22 a2n

am1 am2 amn
a11 a21 an1

a12 a22 an2 . Call this the matrix B
Thus AT =
a1m a2m anm
a11 a12 a1n

a21 a22 a2n

But BT = (AT )T = = A as required.
am1 am2 amn
(AT )T = A
1 -2

Example Verify that (AT )T = A when A = 3 -1
2 0
Answer:

1 -2

AT =
1
-2
3
-1 0
2
and transposing this matrix gives 3 -1
2 0
So (AT )T = A
Property 6: (A + B)T =AT + BT

Condition: A and B have the same order for addition to take place.
Suppose that A and B are m n matrices.
Then (A + B)T =
a11 + b11 a21 + b21 an1 + bn1

a12 + b12 a22 + b22 an2 + bn2

=
a1m + b1m a2m + b2m anm + bnm
a11 a21 an1
b11 b21 bn1

a12 a22 an2 + b12 b22 bn2

a1m a2m anm b1m b2m bnm
= AT + BT
Example Show that (A + B) T = AT + BT

1 0 2
2 1 3

when A = -1 3 -2 and B = -1 -2 -1
-4 -3 0 -1 2 0
Answer:
3 1 5

A+B= -2 1 -3 and
-5 -1 0
3 -2 -5

(A + B)T = 1 1 -1
5 -3 0
1 -1 -4
2 -1 -1

But AT = 0 3 -3 and BT = 1 -2 2
2 -2 0 3 -1 0
1 -1 -4
2 -1 -1
3 -2 -5

so AT + BT = 0 3 -3 + 1 -2 2 =1 1 -1
2 -2 0 3 -1 0 5 -3 0
as required.
Property 7: (AB)T = BT AT
Condition: A and B must be conformable for multiplication for the product AB to exist.
This is another cumbersome proof, which has been placed in the Proof section near the
end of this topic as Proof 3.
1 -2

0 -1

Example If A = and B = show that (AB)T = BT AT

2 -1 2 3
Answer:
-4 -7

-4 -2

AB = so ABT =
-2 -5 -7 -5
0 2

1 2

But BT = and AT =
-1 3 -2 -1
0 2

1 2

-4 -2

which gives B T AT = = as required.

-1 3 -2 -1 -7 -5
Property 8: (AB)-1 = B-1 A-1

Condition: A and B must be square non-singular matrices and conformable for
mutliplication. (They are therefore of the same order.)
(AB)(B-1 A-1 )
= A(BB-1 )A-1
= A IA-1 since BB-1 = I
= AA-1 since AI = A
=I
Thus B-1 A-1 is the inverse of AB, but the inverse of AB is (AB)-1
Therefore (AB)-1 = B-1 A-1
0 2

1 -2

Example For A = and B = show that (AB)-1 = B-1 A-1

-1 3 2 -1
Answer:
4 -2

AB = and det A = 6
5 -1
-1 2

1
Thus (AB)-1 = 6 -5 4
3 -2

-1 2

1 1
det A = 2 and det B = 3 so A -1 = 2 and B-1 = 3
1 0 -2 1
-1 2

3 -2

-1 2

1 1 1
Thus B-1 A-1 = 3 times 2 = 6 as required.
-2 1 1 0 -5 4
Property 9: det (AB) = det A det B

Condition: A and B must be conformable for multiplication.
The proof is shown for 2 x 2 matrices but can be extended to 3 x 3 and larger matrices.
a b

e f

Let A = and B =
c d g h
ae + bg af + bh

AB =
ce + dg cf + dh
det A = ad - bc, det B = eh - fg and
det A det B = (ad - bc)(eh - fg) = adeh - bceh + bcfg - adfg
det (AB) = (ae + bg)(cf + dh) - (ce + dg)(af + bh)
= acef + bcfg + adeh + bdgh - acef - adfg - bceh - bdgh
= bcfg + adeh - bceh - adfg
Thus det A det B = det (AB)
Activity
Verify the property for 3 x 3 matrices.
Properties exercise
There is an alternative exercise on the web for you to try if you wish.
15 min
Q26: Show that A + B = B + A
-2 3
-2 1

when A = 0 -2 and B = 2 1
1 -1 1 2
2
0
then AB
3 -1
Q27: Show that if A = and B = BA
1 1 2 -1

-1
and C =
-1 1 0 -1
Q28: If A =
2 0
-1 2
-1
3
,B= -2 0 1 3

1
-2

show that
-2 1 0 2
1
(AB)C = A(BC)
2 -1

2
Q29: Let A = -3 0 , B = -2
1
-1
-2
and C =
0
1
-1
and show that
1 1
A(B + C) = AB + AC
-1 3

Q30: Verify that (AT )T = A when A = 2 1
4 2
Q31: Show that (A + B) T = AT + BT

2 1 0
-1 2 -2

when A = -2 1 1 and B = 2 0 1
-3 -1 2 -2 3 1
3 -1

3 0

Q32: If A = and B = show that (AB)T = BT AT

1 -2 1 2
1 0

2 -3

Q33: For A = and B = show that (AB)-1 = B-1 A-1

-2 3 -2 -1
2.6 Matrices and transformations

Learning Objective

Æ
Identify the relationship between matrix operations and transformations in the (x, y)
plane
Transformations in the plane, other than translations, can be represented by 2 x 2

matrices.
Any point (x, y) can be mapped to the point (x 1 , y1 ) with a 2 x 2 matrix such that
a b x x1
= In fact a linear transformation maps any line in the x-y
c d y y1

x1

where ad - bc 0
a b x
plane to a line in the plane such that =
c d y y1
Transformations include:
• rotations
• reflections
• dilatations (commonly known as scalings).
Rotations
Consider the effect of taking the point (5, 2) and rotating it by 90 Æ in an anticlockwise
direction about the origin.
After rotation the image of the point is at coordinates Q = (-2, 5)
2.6. MATRICES AND TRANSFORMATIONS 77
The point P (5, 2) can be represented by the 2 x 1 (column) matrix A =

2
0 -1

If this matrix is then multiplied by the 2 x 2 matrix R = the result is

1 0
0
5

1
-1
0
2
=
-2
5
In matrix terms RA = B where B is the matrix which represents the point Q, the
coordinates of the image of P under the rotation.
Q34: Find the simplest matrix which will rotate this point (5, 2) through -90 Æ
The angle of rotation determines the transformation matrix through values of the cosine
and sine of this angle.
Activity
Before continuing further, examine the cos and sin of the angles 90 Æ and -90Æ in
relationship to the two rotation matrices shown earlier. Try to deduce the form of a
general matrix of rotation.
Rotation matrix
cos - sin

The matrix R = acting, by multiplication on the column matrix

x
sin cos
A= , represents the effect of rotation through Æ anticlockwise about the origin
y
on the point (x, y)
Example : Rotation of a point

Find the coordinates of the point (3, 4) under an anticlockwise rotation of 60 Æ about the
origin.
Answer:
cos 60 Æ
- sin 60Æ

3
Let R = and A =
sin 60Æ cos 60Æ 4
3 cos 60Æ - 4 sin 60Æ

- 1 964

Then RA = =
3 sin 60Æ + 4 cos 60Æ 4 598
and the rotated point has its image at Q = (-1.964, 4.598)
cos - sin

Q35: If R = find R-1

sin cos
Rotation of a point
There is a very useful simulation on the web which allows exploration of these rotations
5 min both mathematically and graphically.
Rotation exercise
There is a web exercise for you to try if you like.
10 min
Q36: Find the coordinates of the point (8, 1) rotated through an angle of 30 Æ
anticlockwise.
Q37: The point (2, 1) maps to the point (1.537, 1.624) under a rotation about the origin.
Find the angle through which the point has been rotated to the nearest degree.
Q38: A point is rotated about the origin by -100 Æ and finishes at the coordinates
(-1.796, 1.332). Find the original coordinates to the nearest whole number.
Reflections
It would be useful to find another 2 x 2 matrix which could be used to determine the
coordinates of a point reflected in any line through the origin.
Such a matrix S does exist and SA = B where A is the column matrix representing the
original point and B is the matrix representing the reflected coordinates.
Consider the reflection in the y-axis of the point (2, -3)
-1 0

2

The matrix S = when multiplied to the column matrix

0 1 -3
representing the point (2, -3) has this effect.
-1 0

2

-2

=
0 1 -3 -3
1 0

2

Also the matrix S = acting on the matrix represents a reflection in

0 -1 -3
the x-axis.
The reflection matrix is similar to the rotation matrix but uses the cos and sin of twice
the angle where is the angle that the line of reflection makes with the positive x-axis.
cos 2
Reflection matrix
sin 2

The matrix acting, by multiplication on the column matrix

x
sin 2 - cos 2
, represents the effect of reflection on the point (x, y) in the line through the
y
origin which makes an angle of Æ , -90Æ 90Æ
Examples
1. Reflection of a point
Find the coordinates of the point (-2, 5) under a reflection in the line through the origin
which makes an angle of 60 Æ with the positive x-axis.
Answer:
cos 120 Æ
sin 120Æ

-2

Let S = and A =
sin 120Æ - cos 120Æ

5.330
5
- 2cos 120Æ + 5 sin 120Æ
Then SA = =
- 2sin 120Æ - 5 cos 120Æ 0.768
and the rotated point has its image at Q = (5.33, 0.768)
2. Find the coordinates of the point (-3, 1) when reflected in the line y = 3x
Answer:
First find the angle which the line makes with the positive x-axis. By simple trigonometry,
tan = 1 /3 and so = 18.43 Æ
cos 36.86 Æ
sin 36.86Æ

Let S be the reflection matrix then S =

sin 36.86Æ - cos 36.86Æ
-3

Thus if the point is represented by A = , the image of A is

cos 36.86

Æ
sin 36.86Æ -3 - 1.8
SA = = to one decimal place.
sin 36.86Æ - cos 36.86Æ 1 - 2.6
Reflection of a point
There is a very useful simulation on the web which allows exploration of these reflections
Reflection exercise
10 min
Q39: Find the coordinates, correct to one decimal place, of the point (3, 1) reflected in
a line through the origin at an angle of -30 Æ
Q40: The point (-1, -2) maps to the point (-2, -1) under a reflection in a line at an angle
of to the x-axis . Find the angle to the nearest degree.
Q41: A point is reflected in the line y = - 4x and has its image at the point with
coordinates ( -1, 0). Find the original coordinates to one decimal place.
Dilatations or scalings
It is clear that for both rotations and reflections the values of the elements in the matrix
lie between -1 and 1
However, it is possible to take a point (x, y) and multiply it by a 2 x 2 matrix
whose elements are real numbers outwith this range. This produces another range
of transformations.
Examples
2
4
1. Scaling in the x direction

0
Consider the effect on the point (2, 3) by multiplying the matrix by
3 0 1
Answer:
The point has an image at (8, 3). The x-coordinate has been scaled by 4 and the
y-coordinate has been left alone (scaled by 1).
2
1
2. Scaling in the y direction

0
3 0 3
Answer:
The point has an image at (2, 9) representing a scaling in the x direction of 1 (no real
scaling) and in the y direction by 3
2
4
3. Scaling in both directions

0
3 0 5

Answer:

2

4 0
0 5
3
=
8
15
The point is now (8, 15) = (4 2, 5 3)
This demonstrates scaling in both the x and y directions: the x direction is scaled by 4
and the y direction by 5
A general definition can now be stated.

Dilatation or scaling matrix
0

x
The matrix acting by multiplication on the column matrix represents
0 y
the effect on the point (x, y) thus scaling the x-coordinate by and the y-coordinate by

Scaling of a point
There is a very useful simulation on the web which allows exploration of these scalings
Scaling exercise
10 min
Q42: Find the scaling matrix and the coordinates of the image of the point (3, 1) scaled
in the x direction by -3
Q43: The point (-1, -2) maps to the point (-2, -1) under a scaling in both directions. Find
the scaling matrix.
Q44: A point (3, 4) is scaled to give an image at the point with coordinates (-1, 0). Find
the scaling matrix.
General linear transformations of a line

b
a
In a more general way the matrix will transform sets of points in the x-y plane.
c d
In this case a and d are not equal to zero.
However, it is important to note that the origin remains the origin under any linear
transformation. Thus a translation is not a linear transformation.
Activity
b
a
Experiment with some points on the x-y plane using the matrix with a, b, c, d
c d
as integers. Check sets of points and verify that the origin remains fixed. This is a good
stage at which to experiment with a graphics calculator for these multiplications.
Example : General linear transformation

Find the 2 x 2 matrix which will transform the point (-3, 2) to (-8, -13) and the point (5, 4)
to (6, 7)
Answer:
a b

Let the matrix be

c d
a

Thus
b
c d
-3
2
=
-8
- 13
gives the equations
1) -3a + 2b = -8 and
2) -3c + 2d = -13
a
5
6
Also
b
c d
4
=
7
gives the equations
3) 5a + 4b = 6 and
4) 5c + 4d = 7
Solving the set of equations, 1) and 3) gives a = 2 and b = -1
Similarly the set of equations 2) and 4) gives c = 3 and d = -2
2 -1

The matrix is
3 -2
General transformation exercise

There is an alternative exercise on the web for you to try if you prefer it.
5 min
Q45: Find the 2 x 2 matrix which will transform the point (-3, 2) to (6, 7) and the point
(5, 4) to (-8, -13)
Q46: Find the transformation matrix which takes the point (2, -3 ) to (-8, 16) and the
point (-1, -4) to the point (-7, 3)
Q47: Find the image of the point (-6, 4) under the transformation by the matrix
-1 3
2 -1
Q48: Find the transformation matrix which maps the points (-2, 2) to (2, 2) and (-3, 2)
to (3, 0)
Q49: Find the 2 x 2 transformation matrix which maps the two points (2, 2) and (3, 4)
onto the points (0, 8) and (-2, 13) respectively.
Earlier in this topic orthogonal matrices were briefly mentioned.

Recall the definition: AT = A-1 a matrix is orthogonal. The matrices for rotation and
reflection transformations are orthogonal.
Activity
Verify this definition for rotation and reflection matrices using algebra and the trig.
property that cos2 + sin2 = 1
Having established that rotations and reflections have orthogonal matrices, the next step
is to look at the transpose and the inverse of each more closely.
Rotations
Suppose that the point P is transformed under the rotation matrix R by Æ
and maps to
the point Q.
The inverse of the rotation matrix R, (R -1 ) maps the point Q back to the point P. In other
words it rotates through an angle of - or through in a clockwise direction.
Since R is orthogonal, R -1 = RT
Thus for any clockwise rotations it is a simple task to find the transpose of the matrix of
rotation and multiply the coordinates of the point by this to find the image.
Reflections
Suppose now that the point P is transformed under the reflection matrix S about a line
at Æ through the origin and maps to the point Q.
2.7. SUMMARY 85
The inverse of the reflection matrix S, (S-1 ) will reflect the point Q again in the same line
back to the point P.
Since S is orthogonal, S T = S-1 and so ST will reflect Q back to the point P.
In this case, however, the reflection operation is identical, whether it maps P to Q or Q
to P and this is confirmed by examination of S T
For reflections ST = S.
2.7 Summary
The following points and techniques should be know after studying this topic:
• Performing arithmetic operations on two matrices.
• Finding the inverse and determinant of 2 x 2 and 3 x 3 matrices.
• Deducing and using the properties of matrix operations.
• Relating matrix operations to transformations in the (x, y) plane.
2.8 Proofs
Learning Objective

ÆFollow the proofs given and understand the manipulation of the algebra

Proof 1
Property 3: (AB)C = A(BC)
Only the relevant cells are shown to keep this simple. If time permits, try this proof in
detail.
a11 a1n
b b1p
c c1q

Let A = , B = 11 and C = 11
am1 amn bn1 bnp cp1 cpq
a11 b11 + + a1n bn1 a11 b1p + + a1n bnp

AB =
am1 b11 + + amn bn1 am1 b1p + + amn bnp
a11 b11 c11 + + a1n bnp cp1 a11 b11 c1q + + a1n bnp cpq

(AB)C =
am1 b11 c11 + + amn bnp cp1 am1 b11 c1q + + amn bnp cpq
b11 c11 + + b1p cp1 b11 c1q + + b1p cpq

However BC = and
bn1 c11 + + bnp cp1 bn1 c1q + + bnp cpq
a11 b11 c11 + + a1n bnp cp1 a11 b11 c1q + + a1n bnp cpq

A(BC) =
am1 b11 c11 + + amn bnp cp1 am1 b11 c1q + + amn bnp cpq
(AB)C = A(BC)
Proof 2
Property 4: A(B + C) = AB + AC
a11 a1n
b b1p
c c1p

Let A = , B = 11 and C = 11
am1 amn bn1 bnp cn1 cnp
b11 c11 b1p + c1p

B+C=
bn1 + cn1 bnp + cnp

and A
a11 (b11 + c11 ) + + a1n (bn1 + cn1 ) a11 (b1p + c1p ) + + a1n (bnp + cnp )
(B + C) =
am1 (b11 + c11 ) + + amn (bn1 + cn1 ) am1 (b1p + c1p ) + + amn (bnp + cnp )
a11 b11 + + a1n bn1 a11 b1p + + a1n bnp

But AB = and
a11 c11 + + a1n cn1 a11 c1p + + a1n cnp

AC =
am1 c11 + + amn cn1 am1 c1p + + amn cnp

AB +
a11 (b11 + c11 ) + + a1n (bn1 + cn1 ) a11 (b1p + c1p ) + + a1n (bnp + cnp )
AC =
am1 (b11 + c11 ) + + amn (bn1 + cn1 ) am1 (b1p + c1p ) + + amn (bnp + cnp )
A(B + C) = AB + AC
Proof 3
Property 7: (AB)T = BT AT
a11 a1n
b b1p

Let A = , B = 11
am1 amn bn1 bnp
a11 b11 + + a1n bn1 a11 b1p + + a1n bnp

(AB) = and so
a11 b11 + + a1n bn1 am1 b11 + + amn bn1

(AB)T =
a11 b1p + + a1n bnp am1 b1p + + amn bnp
b11 bn1
a am1

But BT = and AT = 11
b1p bnp a1n amn
a11 b11 + + a1n bn1 am1 b11 + + amn bn1

Thus BT AT = since ab = ba in real
a11 b1p + + a1n bnp am1 b1p + + amn bnp
numbers.
(AB)T = BT AT

Learning Objective

ÆShow an awareness of additional information on this topic

SEKI KOWA
He was a Japanese mathematician who used matrix methods taken from the Chinese
ideas of the centuries before.
LEIBNITZ
He was working on matrices at the same time as Seki Kowa using determinants applied
to systems of three equations in three unknowns.
LAPLACE
The method of expanding the determinant shown in this course is named after him.
CAUCHY
The word determinant was first used as we know it by Cauchy.
SYLVESTER
The term matrix was not used until 1850 when the English mathematician defined it as
an array.
CAYLEY
In the late 1800s Cayley identified many of the algebraic properties of matrices which
are studied today.

Review exercise
There is a web version of this review exercise for you to try if you prefer it.
15 min
-3 2

-4 3

Q50: Let A = ,B= ,

-4 3 -5 -1
2
-1 0 -2
C=
4
-2
3
and D = 3 -4 2
-3 2 5
Find
a) BC
b) A-1
c) A -3B + C
d) det D
-1 4

0 -2

Q51: Let A = ,B= ,

1 2 -3 -5

1 3 3
C=
-2
4
3
-1
and D = 3 -3 2
0 -1 2
Find
a) BC
b) A-1
c) A +2B - C
d) det D

Advanced review exercise
There is another exercise on the web that you may like to try.
15 min
Q52: The n x n matrix A satisfies the equation 2A 2 = A + I where I is the n x n identity
matrix. Show that A is invertible and express A -1 in the form pA + qI.
Q53: If A is a square matrix which satisfies the equation A 2 - 4I = 0 find two solutions
for A.
Q54: The point (2, -3) is transformed first by a rotation through -60 Æ and then by a
scaling of -2 in the x direction and 3 in the y direction. Finally the point is further
transformed by a reflection in the x-axis. Give the coordinates of the image of the point
after these transformations have taken place to one decimal place.
Q55: If A is an orthogonal matrix show that the matrix (AA T )-1 + A(3AT - A-1 ) is 3I where
I is the 2 x 2 identity matrix.

Set review exercise
differently but will require the same answers.
-2 0

-4 2

Q56: Let A = ,B= ,

-2 3 3 1

2 -2 -2

C=
-5
2
-1
3
and D = -2 0 1
-1 2 3
Find
a) BC
b) A-1
c) 2A - B + C
d) det D
2 2

-1 1

Q57: Let A = ,B= ,

4 3 -5 -1

2 3 1

C=
2
-2 0
3
and D = -2 4 1
3 4 1
Find
a) AB
b) C-1
c) 2A - B - 3C
d) det D
91
Topic 3
Further sequences and series
Contents
3.1 Maclaurin series . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92
3.1.1 Power series . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92
3.1.2 Maclaurin series for simple functions . . . . . . . . . . . . . . . . . . . . 93
3.1.3 Maclaurin’s theorem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95
3.1.4 The Maclaurin series for tan-1 x . . . . . . . . . . . . . . . . . . . . . . . 98
3.1.5 Further Maclaurin series . . . . . . . . . . . . . . . . . . . . . . . . . . . 99
3.1.6 Maclaurin series expansion to a given number of terms . . . . . . . . . 100
3.2 Iterative Schemes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102
3.2.1 Recurrence relations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102
3.2.2 Recurrence relations and corresponding equations . . . . . . . . . . . . 105
3.2.3 The iterative scheme xn + 1 = g (xn ) . . . . . . . . . . . . . . . . . . . . . 109
3.2.4 Locating an initial approximation . . . . . . . . . . . . . . . . . . . . . . 112
3.2.5 Conditions for convergence . . . . . . . . . . . . . . . . . . . . . . . . . 114
3.2.6 Order of convergence . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117
3.3 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 120
3.5 Review exercise . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122
Learning Objectives
Understand and use further aspects of sequences and series.
Minimum performance criteria:
• Use the Maclaurin series expansion to find power series for simple functions to a
stated number of terms.
• Find a solution of a simple non-linear equation, to a prescribed degree of accuracy,

using an iterative scheme of the form xn + 1 = g (xn ) where x = g (x) is a
rearrangement of the original equation.
92 TOPIC 3. FURTHER SEQUENCES AND SERIES
Prerequisites
Before attempting this unit you should have a thorough knowledge of the following:
• Geometric series, in particular the formula for the sum to infinity of a geometric
series.
• The Binomial expansion for an expression in the form (a + x) -1
• Higher derivatives of simple functions.
• Linear recurrence relations.
Revision Exercise
Q1: Find the sum to infinity for the geometric series
10 min
16 + 8 + 4 + 2 + 1 + 1 /2 + ...
Q2: Write down the binomial expansion for (1 + x) -1 , -1 x 1
Q3: For f (x) = e2x calculate the derivatives

f (x), f (x) and f (x)
Q4: Find the first five terms and the limit for the recurrence relation
un + 1 = 0.7un + 9, u0 = 5
3.1 Maclaurin series

3.1.1 Power series
Learning Objective

ÆRecognise power series

A power series is an expression of the form
a xn

a0 + a1 x + a2 x2 + a3 x3 + + an xn +
n
n=0
where a0 , a1 , a2 , a3 , ..., an , ... are constants and x is a variable. It is called a power

series as it is made up of a sequence of powers of x with coefficients a 0 , a1 , a2 , a3 , ...,
an , ...
Power series are useful in solving differential equations that occur in physics, including
the equations that describe motion of a simple pendulum, vibrating strings, heat flow,
electrical current, and many other examples.
In numerical analysis, power series can be used to determine how many decimal
places are required in a computation to guarantee a specified accuracy.
Numerical analysis is a branch of mathematics. It can be described as the analysis
and solution of problems which require calculation.
It is also useful if we are able to express some simple functions such as e x , sin x, cos x,
3.1. MACLAURIN SERIES 93
tan-1 x, (1 + x)n and ln (1 + x) in terms of power series.
Examples
1. 1 + x + x2 + x3 + x4 + ... xn + ... is an example of a power series

with 1 = a0 = a1 = a2 = a3 = a4 = ... = an = ...
You should also recognise this as a geometric series.
2. Another important example of a power series is

x2 x3 x4 x5
1
0! + x
1! + 2! +
3! + 4! + 5! +
This time ar = r!1
This power series converges for all values of x
Q5: Substitute x = 1 into the above power series and calculate the sum of the first ten
terms, S10 , to 6 decimal places. Do you recognise this answer?
In this topic we will learn a method that allows us to rewrite a variety of functions as
power series.
3.1.2 Maclaurin series for simple functions
Learning Objective

ÆGenerate Maclaurin series for various functions using the given formula

Suppose we have a function f (x), which we are able to keep differentiating as often as
we want and that there is no problem differentiating when x = 0. Functions like this do
exist and for example ex and sin x are functions that satisfy these conditions. With these
conditions we are able to find a special type of power series called the Maclaurin series.
The Maclaurin series generated by the function f (x) is

xr x x2 x3 xn
f (r) (0) = f (0) + f(1) (0) + f(2) (0) + f(3) (0) + + f(n) (0) +
r! 1! 2! 3! n!
r=0
The following examples show how we can find the Maclaurin Series generated by e x
and sin x
Example Find the Maclaurin series generated by f (x) = e x

Answer
When f (x) = ex is repeatedly differentiated we obtain
f (x) = ex f (0) = e0 = 1
f (1) (x = ex f (1) (0) = e0 = 1
f (2) (x) = ex f (2) (0) = e0 = 1
f (3) (x) = ex f (3) (0) = e0 = 1
f (4) (x = ex f (4) (0) = e0 = 1
f (5) (x) = ex f (5) (0) = e0 = 1
Therefore the Maclaurin series generated by f (x) = e x becomes
xr

x x2 x3 x4 x5
f (r) (0) = f (0) + f (1) (0) + f (2) (0) + f (3) (0) + f (4) (0) + f (5) (0) +
r! 1! 2! 3! 4! 5!
r=0
x x2 x3 x4 x5
+ (1) + (1) + (1) + (1)
= 1 + (1)
1! 2! 3! 4! 5!
x 2 x 3 x4 x 5
= 1+x+ + + + +
2! 3! 4! 5!
You should recognise this from before.
Example Find the Maclaurin series generated by f (x) = sin x

Answer
When f (x) = sin x is repeatedly differentiated we obtain
f (x) = sinx f (0) = sin(0) = 0
f (1) (x) = cosx f (1) (0) = cos(0) = 1
f (2) (x) = - sinx f (2) (0) = - sin(0) = 0
f (3) (x) = - cosx f (3) (0) = - cos(0) = - 1
f (4) (x) = sinx f (4) (0) = sin(0) = 0
f (5) (x) = cosx f (5) (0) = cos(0) = 1
Therefore the Maclaurin series generated by f (x) = sin x becomes
xr

x x2 x3 x4 x5
f (r) (0) = f (0) + f (1) (0) + f (2) (0) + f (3) (0) + f (4) (0) + f(5) (0)
r! 1! 2! 3! 4! 5!
r=0
x x2 x3 x4 x5
= 0 + (1) + (0) + ( - 1) + (0) + (1)
1! 2! 3! 4! 5!
x3 x5 x7 x9 x11
= x- + - + - +
3! 5! 7! 9! 11!
Now try the questions in Exercise 1.
Exercise 1
An on-line assessment is available at this point, which you might find helpful.
15 min
Q6: Find the Maclaurin series generated by the following functions
a) f (x) = cos x
b) f (x) = (1 + x)n
c) f (x) = ln (1 + x)
3.1.3 Maclaurin’s theorem

Learning Objective

Æ
Appreciate that given certain conditions the Maclaurin series generated by a function
is actually equal to that function
What is the relationship between the Maclaurin series of a function and the function
itself?
Suppose we have a function f (x) which we are able to keep differentiating as often as
we want and that there is no problem differentiating when x = 0. Also suppose that it is
possible to write this function as a series expansion so that
f (x) = a0 + a1 x + a2 x2 + a3 x3 + a4 x4 + a5 x5 + ...
Then the following definition applies
Maclaurin’s theorem
Maclaurin’s theorem states that

xr
f (x) = f (r) (0)
r!
r=0
x x2 x3 xn
= f (0) + f (1) (0)
+ f (2) (0) + f (3) (0) + + f (n) (0) +
1! 2! 3! n!
Notice that this theorem claims that the function f (x) is actually equal to its infinite
power series expansion. The theorem is named after the Scottish mathematician
Colin Maclaurin (1698-1746) who first proposed this result in his publication Treatise
of fluxions.
We can see why this is true from the following reasoning.
Let f (x) = a0 + a1 x + a2 x2 + a3 x3 + a4 x4 + a5 x5 + ...
If we substitute x = 0 in the expansion we get f (0) = a 0
Differentiating with respect to x, we obtain
f (x) = a1 + 2a2 x + 3a3 x2 + 4a4 x3 + 5a5 x4 + ......
Now substituting x = 0 into this equation gives us f (0) = a1 = 1!a1
We can repeat this process again.
Differentiating again with respect to x, we obtain
f (x) = (2 1)a2 + (3 2)a3 x + (4 3)a4 x2 + (5 4)a5 x3 + ...
Therefore f
(0) = (2 1)a2 = 2!a2
Differentiating once more with respect to x, we obtain
f (x) = (2 1)a2 + (3 2 1)a3 + (4 3 2)a4 x + (5 4 3)a5 x2 + ...
Therefore f
(0) = (3 2 1)a3 = 3!a3
Remember that for higher derivatives it is often more convenient to replace a series of
dashes with a number.
In other words, for example, f
can be rewritten as f(3)
Q7: Do the next differentation yourself to find f (4) (0)
Rearranging the previous results we can rewrite the coefficients as follows,

a0 = f (0)
f (1) (0)
a1 =
1!
(2)
f (0)
a2 =
2!
f (3) (0)
a3 =
3!
(4)
f (0)
a4 =
4!
Since we assumed that we could keep on differentiating f (x) then we can also say that
f (n) (0)
an = n!
Now substituting these expressions for ar back into the power series
f (x) = a0 + a1 x + a2 x2 + a3 x3 + a4 x4 + a5 x5 ...
gives the Maclaurin series expansion.
2 3
f (x) = f (0) + f (1) (0) 1!x + f (2) (0) x2! + f (3) (0) x3! +
Note that this result depends on us being able to differentiate the infinite series term-by-
term and is only valid within an interval of convergence.
The following example should help you understand more clearly what we mean by this.
Example Consider the power series

1 + x + x2 + x3 + x4 + ... + xn + ...
In Unit 2, Sequences and Series, we learned that for a geometric convergent series
S a + ar + ar2 + ar3 + ar4 + = a
1-r where 1r1
The interval -1 r 1 is known as the interval of convergence for the previous series.
For values of r outside this interval the series does not converge and we say that the
series is a divergent series for r -1 and for r 1
divergent series
A divergent series is one which is not convergent. For example 1 + 2 + 3 + 4 + 5 ... is

a divergent series. The sum of this series will continue increasing the more terms we
add. It will not tend towards a limit.
Therefore comparing the power series
1 + x + x2 + x3 + x4 + ... + xn + ...
to a geometric series we can see that this series is convergent for -1 x 1 and from
the formula for S , taking a = 1 and r = x we can write
1 + x + x2 + x3 + x4 + ... + xn + ... = 1 /1 - x , for -1 x 1
-1 x 1 is the interval of convergence for this series. Outside the interval of
convergence the power series is divergent.
For example, when x = 1 the series is simply 1 + 1 + 1 + 1 + ... + 1 + ... which is
obviously divergent as the sum of the series will continue increasing the more terms we
add. It will not tend towards a limit.
convergent series
A convergent series is one for which the limit of partial sums exists. This limit is called
the sum and is denoted by S
It was also established in Unit 2, Sequences and Series, that we can perform a binomial
expansion on 1 /1 - x which of course can be written as (1 - x)-1 . Remember that when the
binomial expansion is used with negative powers the expansion is infinite and to ensure
that the infinite series converges we need -1 x 1, just as in the geometric series
formula.
The Binomial expansion gives
( - 1) ( - 1)( - 2) ( - 1)( - 2)( - 3)
(1 - x) - 1 = 1 + ( - x) + ( - x)2 + ( - x)3
1! 2! 3!
( - 1)( - 2)( - 3)( - 4)
+ ( - x)4 +
4!
= 1 + x + x2 + x3 + x4 +
which you will notice gives the same series as previously.
It will be useful for you to remember the following information.
We can normally expect a Maclaurin series to converge to its generating function in an

interval about the origin. However, for many functions this is the entire x-axis.
For example, the Maclaurin series for functions such as sin x, cos x, ex all converge for
x
Whereas the Maclaurin series for
ln(1 + x) converges for 1x1
converges for x 1
1
1+x
tan - 1 x converges for x 1
The Maclaurin series for sin x

There is a simulation on the web that allows you to compare the graph for sin x with the
10 min graphs for partial sums of its Maclaurin series.
3.1.4 The Maclaurin series for tan-1 x

We can find the Maclaurin series for tan -1 x in the same way as for the previous functions.
However the working soon becomes a bit complicated as you can see here.
When f (x) = tan-1 x is repeatedly differentiated we obtain
f (x) = tan - 1 x f (0) = 0
1 (1)
f (1) (x) = f (0) = 1
1 + x2
- 2x
f (2) (x) = f (2) (0) = 0
(1 + x2 )2
6x2 - 2
f (3) (x) = f (3) (0) = - 2
(1 + x2 )3
From this we only obtain the first two terms in the Maclaurin series.
xr

tan - 1 x = f (r) (0)

r!
r=0
x x2 x3 x4 x5
= f (0) + f (1) (0)
+ f (2) (0) + f (3) (0) + f (4) (0) + f (5) (0)
1! 2! 3! 4! 5!
x x2 x3
= 0 + (1) + (0) + ( - 2) +
1! 2! 3!
x 3
=x- +
3
You can see that to differentiate further and so obtain more terms in the series soon
becomes quite tedious. The following method gives another way to obtain more terms.
f (x) = tan - 1 x
1 -1
f’ (x) = = 1 + x2
1 + x2
Provided x2 1 we can now use the binomial expansion to give

( - 1) 2 ( - 1)( - 2) 2 2 ( - 1)( - 2)( - 3) 2 3
f’ (x) = 1 + x + x + x
1! 2!
( - 1)( - 2)( - 3)( - 4) 2 4 3!
+ x +
4!
= 1 - x2 + x4 - x6 + x8 -
Integrating this will take us back to f (x) = tan-1 x
tan - 1 x = C + x - 13 x3 + 15 x5 - 17 x7 + 19 x9
C is the integrating factor. However when x = 0, tan -1 x = 0 and therefore C = 0
So we can now write
tan - 1 x = x - 13 x3 + 15 x5 - 17 x7 + 19 x9
The previous series for tan-1 x actually converges for x 1 and we can use it to obtain
an approximation for in the following way.
Let x = 1 then
tan - 1 1 = 4 =1- 1
3 + 1
5 - 1
7 + 1
9 - 1
11 +
and so

4 1- 1
+ 1
- 1
+ 1
- 1
+

3 5 7 9 11
The previous series is known as the Leibniz formula for . However, this series
converges very slowly so in practice it is not used in approximating to many decimal
places. In fact 1000 terms are needed before it gives a value accurate to 4 decimal
places.
There is more information about different methods for calculating in the extended
information chapter.
What happens when you try to obtain the Maclaurin series for the following functions?
Q8: f (x) = ln x
Q9: f (x) =
x
Q10: f (x) = cot x
3.1.5 Further Maclaurin series

Learning Objective

ÆCalculate the Maclaurin series for simple composite functions
-3x

It is also possible to find Maclaurin series for functions such as e and sin 2x. The
following example shows how this can be done.
Example Find the Maclaurin series for e -3x

Answer
When f (x) = e-3x is repeatedly differentiated we obtain
f (x) = e - 3x f (0) = e0 = 1
f (1) (x) = - 3e - 3x f (1) (0) = - 3e0 = - 3
f (2) (x) = ( - 3)2 e - 3x f (2) (x) = ( - 3)2 e0 = ( - 3)2
f (3) (x) = ( - 3)3 e - 3x f (3) (x) = ( - 3)3 e0 = ( - 3)3
f (4) (x) = ( - 3)4 e - 3x f (4) (x) = ( - 3)4 e0 = ( - 3)4
f (5) (x) = ( - 3)5 e - 3x f (5) (x) = ( - 3)5 e0 = ( - 3)5
Therefore the Maclaurin series for f (x) = e-3x becomes
xr

e - 3x
= f (r) (0)
r!
r=0
x (1) x2 x3 x4 x5
= f (0) + f (0) + f (2) (0) + f (3) (0) + f (4) (0) + f (5) (0)
1! 2! 3! 4! 5!
x x 2 x3 x4 x5
= 1 + ( - 3) + ( - 3)2 + ( - 3)3 + ( - 3)4 + ( - 3)5
1! 2! 3! 4! 5!
3x (3x)2 (3x)3 (3x)4 (3x)5
=1- + - + - +
1! 2! 3! 4! 5!
We could have obtained this same result if we had substituted (-3x) for x in the Maclaurin
series for ex . Check this for yourself!
Exercise 2
30 min
Q11: Find the Maclaurin series for the following functions
a) e5x , x
b) sin 2x, x
c) cos 3x, x
d) cos (-x), x
e) ln (1 - 3x), -1 /3 x 1 /3
Q12:
a) For i =
-1 write down the Maclaurin expansion for e ix
b) Can you suggest a connection between e ix , sin x and cos x
3.1.6 Maclaurin series expansion to a given number of terms
Learning Objective

Æ
Determine the Maclaurin series expansion for a given function to a specified number
of terms
Often you may be required to find a specific number of terms in a Maclaurin series
expansion. We proceed as before as you can see in the following example.
Examples
1. Use Maclaurin’s theorem to write down the expansion of (1 + x) -3 as far as the term
in x3
Answer
When (1 + x)-3 is repeatedly differentiated we obtain
f (x) = (1 + x) - 3 f (0) = 1
(1) -4 (1)
f (x) = - 3(1 + x) f (0) = - 3
f (2) (x) = ( - 3)( - 4)(1 + x) - 5 f (2) (0) = ( - 3)( - 4)
f (3) (x) = ( - 3)( - 4)( - 5)(1 + x) - 6 f (3) (0) = ( - 3)( - 4)( - 5)
We may stop here as we are only required to find the series as far as the term in x 3
From this we obtain the Maclaurin series

xr (r)
(1 + x) - 3 = f (0)
r!
r=0
x (1) x2 x3
= f (0) + f (0) + f (2) (0) + f (3) (0) +
1! 2! 3!
x x 2 x3
= 1 + ( - 3) + ( - 3)( - 4) + ( - 3)( - 4)( - 5) +
1! 2! 3!
= 1 - 3x + 6x - 10x +
2 3
Thus the Maclaurin series for (1 + x)-3 , as far as the term in x3 , is 1 - 3x + 6x2 - 10x3
This is also known as the Maclaurin series to third order because the highest power of
x in the expansion is 3
2. Be careful if you are asked to find the Maclaurin series generated by, for example,
(1 + x)5 . You should notice that since this expression has a positive integer power then
it will have a finite number of terms when expanded.
Answer
When (1 + x)5 is repeatedly differentiated we obtain
f (x) = (1 + x)5 f (0) = 1
(1) 4 (1)
f (x) = 5(1 + x) f (0) = 5
f (2) (x) = (5)(4)(1 + x)3 f (2) (0) = 20
f (3) (x) = (5)(4)(3)(1 + x)2 f (3) (0) = 60
f (4) (x) = (5)(4)(3)(2)(1 + x)1 f (4) (0) = 120
f (5) (x) = (5)(4)(3)(2) f (5) (0) = 120
f (6) (x) and further derivatives = 0 f (6) (0) and further derivatives = 0
From this we obtain the following Maclaurin series
xr

(1 + x)5 = f (r) (0)

r!
r=0
x (1) x2 x3 x4 (4) x5
= f (0) + f (0) + f (2) (0) + f (3) (0) + f (0) + f (5) (0)
1! 2! 3! 3! 3!
x x2 x3 x4 x5
= 1 + (5) + (20) + (60) + (120) + (120)
1! 2! 3! 4! 5!
2 3 4 5
= 1 + 5x + 10x + 10x + 5x + x
You should recognise that this is the same as the binomial expansion for (1 + x) 5
Exercise 3
30 min
Q13: Use Maclaurin’s theorem to write down the expansions, as far as the term in x 4 ,
of
a)
(1 - x), x 1
b) (1 + x)-5 , x 1
c) (x + 1)3/2 , x 1
Q14:
Find all the terms in the Maclaurin series generated by
a) (2 + x)4
b) (1 - 2x)3
Q15:
Use Maclaurin’s theorem to write down the expansions, as far as the term in x 3 , of
a) 2x + 3
1
x 12
b) 1
(x - 2)3
, x1
3.2 Iterative Schemes

Iteration
Iteration is the successive repetition of a mathematical process using the result of one
stage as the input for the next.
In this section we use iteration to solve equations.
3.2.1 Recurrence relations

Learning Objective

ÆInvestigate when a first order linear recurrence relation converges to a limit

You will already be familiar with the concept of a recurrence relation. However, as a
reminder here is the definition.
recurrence relation
A recurrence relation describes a sequence of terms u0 , u1 , u2 , u3 , ..., un , un + 1 , ...
where each term is a function of previous terms.
So in this way a recurrence relation is a type of iterative process.
There are various types of recurrence relation.
3.2. ITERATIVE SCHEMES 103
first order recurrence relation

A first order recurrence relation is a recurrence relation where u n + 1 depends only on
un and not on values of u r where r n
Example un + 1 = 2un - 4 is an example of a first order recurrence relation.
second order recurrence relation A second order recurrence relation is a

recurrence relation of the form u n + 2 = aun + 1 + bun + c, with a and b 0. Note
that un + 2 depends on the previous two terms in the sequence.
Example un + 2 = 2un + 1 + 0.5un - 6 is a second order recurrence relation. Note that

un + 2 depends on the previous two terms in the sequence.
Linear first order recurrence relation

A linear first order recurrence relation is a recurrence relation of the form
un + 1 = aun + b, a 0
Example un + 1 = 0.8un + 2 is a linear first order recurrence relation.
Example
However, un + 1 = ln (un ) + 3 is a non-linear first order recurrence relation. It includes
the non-linear term ln (u n )
Previous work has concentrated on first order linear recurrence relations. It will be useful
if we are reminded of some of the properties of these recurrence relations. In particular it
is important to be able to decide from a given recurrence relation whether the sequence
of terms will converge or diverge. Look at these examples.
Examples
1. For the recurrence relation u n + 1 = 0.2un + 4 with u0 = 1 we get the following sequence
of terms.
1, 4.2, 4.84, 4.968, 4.9936, 4.99872, 4.999744, ...
This sequence of terms converges to the limit 5
2. For the recurrence relation u n + 1 = 3un + 2 with u0 = 1 we get the following sequence
of terms.
5, 17, 53, 161, 485, 1457, 4373, 13121, ...
This sequence continues to get bigger and bigger. It diverges.
Now try the following questions in Exercise 4.
Exercise 4
Q16:
20 min Consider the recurrence relation u n + 1 = aun + 4. Investigate whether this recurrence
relation converges or diverges for the following values of a
a) 4
b) 1
c) 0.8
d) 0.5
e) 0.2
f) 0
g) -0.2
h) -0.6
i) -2
Q17:
For which values of a would you expect the recurrence relation to converge to a limit?
Q18: Consider the recurrence relation u n + 1 = 0.2un + b with u0 = 0. Investigate the

effect of substituting b with the following values.
a) 10
b) 0.6
c) -0.2
d) -4
Q19:
What effect does changing b seem to have on the recurrence relation?
Q20: Consider the recurrence relation u n + 1 = 0.2un + 6

Investigate the effect of substituting u0 with the following values.
a) 0
b) 2
c) 10
d) 100
Q21:
What effect does changing u 0 seem to have on the convergence of the recurrence
relation?
3.2.2 Recurrence relations and corresponding equations
Learning Objective

Æ
Equate recurrence relations with their corresponding equations and identify when the
recurrence relation converges to a fixed point
In the previous section we were able to confirm that
For the linear recurrence relation u n + 1 = aun + b, un tends to a limit, L,

when -1 a 1
b
L is given by the formula L = 1-a
We can derive the above formula for L from the following.

In general we can say that if u n + 1 L as n then we also have that
un L as n
Therefore, as n , the formula u n + 1 = aun + b tends to
L = aL + b
L - aL = b
L(1 - a) = b
b
L=
1-a
Now consider the following.
Q22: Solve the following equations

a) x = 4x + 4
b) x = 0.8x + 4
c) x = -0.2x + 4
Compare your answers to these equations to the following recurrence relations which
were already investigated in Exercise 4.
un + 1 = 4un + 4 (Ex 4 Q16 a)
un + 1 = 0.8un + 4 (Ex 4 Q16 c)
un + 1 = - 0.2un + 4 (Ex 4 Q 16g)
Is there a connection?
You should notice that the above recurrence relations have corresponding equations.
So, for example, un + 1 = 0.8un + 4 has corresponding equation x = 0.8x + 4 When the
recurrence relation converges (-1 a 1) then the limit of the recurrence relation is
equal to the solution of the corresponding equation. This solution is also known as the
fixed point of the recurrence relation. Divergent recurrence relations also have a fixed
point but application of the recurrence relation does not provide the fixed point. You will
understand this more clearly once you have done the exercise on Staircase and Cobweb
diagrams later in this section.
fixed point
The recurrence relation u n + 1 = aun + b has fixed point given by b //(1 - a) . The recurrence
relation will only converge to this fixed point if (-1 a 1)
Q23: Write down a recurrence relation which corresponds to each of the following
equations.
a) x = -0.3x + 7
b) x = 2x = 5
c) x = 0.4x + 6
d) 0.2x = 4
e) 1.2x = 3.6
f) -0.5x = 15
Q24: Which of these recurrence relations converge to a limit that is the same as the
solution of the equation?
The following diagrams may help you to understand why this happens.
Staircase and Cobweb Diagrams

You will find interactive versions of the following Staircase and Cobweb diagrams on the
15 min course web site.
The recurrence relation xn + 1 = axn + b with initial value x0 can be represented
geometrically by the straight lines y = x and y = ax + b
The strategy given here can then be used to draw the diagrams.
1. Draw the lines y = x and y = ax + b. From the given value of x 0 draw a vertical
line to meet the line y = ax + b. The coordinates of the point of intersection are
(xn , xn + 1 ) with xn + 1 = axn + b and n = 0
2. Draw a horizontal line from (xn , xn + 1 ) to meet the line y = x. The point of
intersection has coordinates (xn + 1 , xn + 1 )
3. Draw a vertical line from (xn + 1 , xn + 1 ) to meet the line y = ax + b. The point of
intersection has coordinates (xn + 1 , xn + 2 ) with xn + 2 = axn + 1 + b
4. Repeat steps 2, 3 and 4 of this iterative process.
This single process gives two types of diagram.
• A staircase diagram when a 0
• A cobweb diagram when a 0
There follows examples of these two types of diagram.
Example : Staircase Diagram

(a 0)
This diagram shows the straight lines y = 0.4x + 6 and y = x which is a geometrical
illustration of the recurrence relation un + 1 = 0.4un + 6
This iterative process can be written in x notation as xn + 1 = 0.4xn + 6
The initial value in this example is x0 = 1
The iterative process is repeated and the value of x n tends to the fixed point of the
recurrence relation as n . This is also the solution to the equation x = 0.4x + 6
which is x = 10
Notice that the fixed point is where the lines y = x and y = 0.4x + 6 intersect.
Example : Cobweb Diagram

(a 0)
This diagram shows the straight lines y = -0.5x + 6 and y = x which is a geometrical
illustration of the recurrence relation un + 1 = -0.5un + 6
This iterative process can be written in x notation as xn + 1 = -0.5xn + 6
The initial value in this example is x0 = 1.
The iterative process is repeated and the value of x n tends to the fixed point of the
recurrence relation as n . This is also the solution to the equation x = -0.5x + 6
which is x = 4
Notice that the fixed point is where the lines y = x and y = -0.5x = 6 intersect.
Staircase and Cobweb diagrams

Q25: Draw Staircase or Cobweb diagrams for the following recurrence relations. For
each diagram write down the fixed point and determine whether the recurrence relation 15 min
leads to this fixed point or not.
a) xn + 1 = 0.2xn + 4, x0 = 0
b) xn + 1 = -0.5xn + 12, x0 = 1
c) xn + 1 = 1.5xn + 6, x0 = 1
d) xn + 1 = -2xn + 12, x0 = 2
3.2.3 The iterative scheme xn + 1 = g (xn )

If we require to obtain solutions for the equation x 2 + 3x - 2 = 0 we can use the formula

- b b2 - 4ac
x 2a

and so obtain the solutions
x= 1
2 -3 17
However, suppose that we wish to find the solutions of equations such as

x3 + 2x2 + 10x - 20 = 0, xex = 1, or x = 3 sin x then we have no convenient formula
and so we need to use another method. One technique for solving such equations is to
use an iterative method. We rewrite the original equation in the form x = g (x) and then
we set up the corresponding recurrence relation x n + 1 = g (xn ). Taking x0 as an initial
approximation to the solution we are then able to calculate x 1 , x2 , x3 , ... and hopefully
this will lead to a fixed point and therefore a solution to the original equation. However
you should be wary from our previous work on first order linear recurrence relations that
this may not always be the case. The following activity should highlight some of the
difficulties you may encounter.
Testing Convergence
Learning Objective

Æ
25 min
Evaluate different iterative processes.
Consider the cubic equation

f (x) = x3 - 6x + 3
Using a graphic calculator it is possible to make a sketch of this graph and therefore
obtain an estimate for the three roots of the equation.
Your graph should look something like this
You can see that a reasonable estimate for the three roots is
= -2.7, = 0.5, Æ = 2.1
Now to obtain a more accurate value for these roots we can attempt to apply an iterative
process. To achieve this we rearrange x 3 - 6x + 3 = 0 as x = g (x). This can be done in
many ways.
Check that the following are all possible rearrangements.
x3 + 3
(1) x =
6
3
(2) x =
6 - x2
(3) x = x3 - 5x + 3
2x3 - 3
(4) x =
3x2 - 6
You can probably think of some more yourself.
Using one of these rearrangements we can now set up an iterative process to try to
obtain the roots of the original equation to a greater degree of accuracy.
Rearrangement (1) will give us the iterative scheme
x3n + 3
xn + 1 = 6
We can now check to see how effective this is for finding the roots.
Let x0 = 0.5 then we obtain the following results (correct to 6 decimal places).
x0 0.5
x1 0.520833
x2 0.523548
x3 0.523918
x4 0.523968
x5 0.523975
x6 0.523976
x7 0.523976

xn 0.523976
Therefore it seems that this iterative process has converged to the root near 0.5 and,
correct to 6 decimal places, this root is 0.523976
(A graphic calculator can do these calculations very quickly. For example, we can
achieve the result using the following steps
• key in x0 = 0.5 and press enter.
• now key in (ANS^3+3)/6
• Pressing enter repeatedly will give x1 , x2 , x3 , ...)
However, look what happens if we try to find the root near -2.7
x0 - 2.7
x1 - 2.7805
x2 - 3.082758
x3 - 4.382778
x4 - 13.531274
x5 - 412.418921
x6 - 11691345.03

xn
For x0 = -2.7 the sequence diverges and we are unable to find a more accurate value
for the root near -2.7
Q26: Using the same iteration formula as above find out what happens when you take
x0 = 2.1
Having considered different iterative schemes with different values for x 0 it should now
be obvious to you that not every iterative scheme will converge to the required solution.
Convergence will depend on the iterative scheme and on the value of x 0 used.
Q27: There is an interactive version of this question on the course web site.
Copy and complete the following table to show how each iteration formula behaves near
the given values of x0
x0 = -2.7 x0 = 0.5 x0 = 2.1
xn + 1 = x3n + 3 Diverges Converges to

6
0.523976
3
xn + 1 = 6 - x2n
xn + 1 = x3 - 5xn + 3
2x3n - 3
xn + 1 = 3x2n - 6
Q28: Which of the previous iterative schemes gave all three solutions?
Exercise 5
25 min
Q29: The equation xe x = 1 has one solution near x = 0.5
Find an iterative scheme that will find this solution, and write down your answer to 6
decimal places.
Q30: The equation sin x + x - 2 = 0 has one solution near x = 1

Find an iterative scheme that will find this solution and write down your answer to 6
decimal places.
Q31: The equation x 3 + x - 1 /2 has one solution near x = 0
x = 1 /2 - x3 and x =
1 - x13 are both rearrangements of the above equation.
2
What results do you obtain when you apply iterative schemes to these rearrangements?
If possible write down the value of the solution to 6 decimal places.
Q32: The equation e x - x = 3 has two solutions, one near -3 and the other near 1.5
Show that x = ex - 3 and x = ln (x + 3) are both rearrangements of this equation.
What results do you obtain when you apply iterative schemes to these rearrangements?
Give any solutions to 6 decimal places.
3.2.4 Locating an initial approximation
Learning Objective

ÆUse a graphical technique to locate the approximate solution x 0

In practice we may often obtain an initial approximation to roots of the equation f (x) = 0
by using a graphic calculator to graph y = f (x) and then reading the x-coordinates of the
point(s) where the curve cuts the x-axis.
It is useful to have another method which will allow us to approximate x 0 without relying
on a graphic calculator. The following example shows how this is possible.
Example Find an approximate solution to the equation x 3 - 2x + 3 = 0
Answer
We can rewrite x3 - 2x + 3 = 0 as x3 = 2x - 3
Now if we graph y = x3 and y = 2x - 3 on the same diagram then the intersection of these
two lines will give an approximate solution for x 3 - 2x + 3 = 0
(Since y = x3 and y = 2x - 3 are familiar graphs, this is probably a better method than
trying to graph y = x3 - 2x + 3).
You should try to draw these graphs as accurately as you can so that you can achieve a
good estimate for x0 (2mm graph paper is recommended)
A table of values will help and then the graph can be plotted more accurately.
x -2 - 1.5 -1 - 0.5 0 0.5 1 1.5 2

y = x3 -8 - 3.38 -1 - 0.13 0 0.13 1 3.38 8
(2 d.p.)
x -2 - 1.5 -1 - 0.5 0 0.5 1 1.5 2

y = 2x - 3 -7 -6 -5 -4 -3 -2 -1 0 1
Notice that the scales on the axes are different. We need to choose a large scale for the
x -axis so that we can achieve a good estimate for x 0
From the graph it would seem that there is only one solution for the equation and a first
approximation to this solution is x0 = -1.9
To find the solution to a greater degree of accuracy we then use a suitable iterative
scheme with x0 = -1.9
You can check that x = (2x - 3)1/3 is a rearrangement of x3 - 2x + 3 = 0
The iterative scheme xn + 1 = (2x - 3)1/3 with x0 = -1.9 then gives us the root more
accurately. To 6 decimal places the root is -1.893289
Exercise 6
20 min
Q33: a) Use a graphical method to find an approximate solution for the equation
ex - x - 4 = 0 between 1 and 2.
b) The equation e x - x - 4 = 0 can be rewritten as x = ln (x + 4). Use your value for x0 and
the iterative scheme xn + 1 = ln (xn + 4) to find the solution of the equation to 6 decimal
places.
Q34: a) The equation x 3 - x + 4 = 0 has one real solution. Use a graphical method to
obtain an approximation to this solution.
b) Find an iterative scheme that will give this solution more accurately and write down
the value of the solution to 6 decimal places.
Q35:
a) Show graphically that the equation ln x + x - 3 = 0 has only one real solution and find
an approximate value for this solution.
b) Find an iterative scheme that will give this solution more accurately and write down
the value of the solution to 6 decimal places.
Q36:
a) Show graphically that the equation x 4 - 5x + 1 = 0 has two real solutions for
0 x 2 and write down an approximate value for these solutions.
b) Find iterative schemes that will give you these solutions to 6 decimal places.
3.2.5 Conditions for convergence
Learning Objective

ÆDetermine when an iterative process will converge to a solution

From the work that we have done up to now it should be obvious that the success of an
iterative scheme in the form xn + 1 = g (xn ) depends on:
i the starting value used
ii and the particular iterative formula used.
When we were dealing with a linear first order recurrence relation we were able to
determine the condition that -1 a 1 for the recurrence relation u n + 1 = aun + b
to converge to a limit.
It would be useful now if we were able to determine for the iterative process x n + 1 = g (xn )
whether or not a particular formula and starting value will converge to the appropriate
solution. In a similar way for recurrence relations we can represent x n + 1 = g (xn )
geometrically in either a staircase or cobweb diagram by the lines y = x and
y = g (x)
Study the following diagrams and try to arrive at some conclusions. The solution of the
equation x = g (x) is where the two lines intersect at the fixed point . Notice that the
gradient of the curve y = g (xn ) at the fixed point is indicated by mT
In diagrams 1 and 3 the iterative process converges towards the fixed point . In
diagrams 2 and 4 the iterative process diverges away from the fixed point
Hopefully you will have spotted that the iterative process converges towards the fixed
point when -1 m T 1 ie. mT 1 near the fixed point.
Therefore the iterative process xn + 1 = g (xn ) converges towards the fixed point when
g (x) 1 in an interval close to
Example x = 2x3 + 2
5x - 2 1
3 4x3 - 2
5 , x= 2 , and x = 6x2 - 5
are all rearrangements of the equation 2x 3 - 5x + 2 = 0

These rearrangements give the iterative process x n + 1 = g (xn )
The equation has three solutions. One of these solutions is near 0.5
By considering the value of g (x) at 0.5 for the iterative processes determine which
of them will converge to the solution near 0.5
Answer
1. For the iterative scheme

2x3n + 2
xn + 1 = 5
we have
2x3 + 2
g (x) = 5
g (x) = /5 x2 and
6 g (0.5) = 0.3
Since g (0.5) 1 we can expect this iteration to converge to the solution near
0.5

5x - 2 1
3
xn + 1 = n
2
we have
g (x) =
5x - 2 1
3
-
2
2.1
2
g’ (x) = 5 5x - 2
6 2
3
and g ’ (0.5)
Since g (0.5) 1 we should not expect this iterative scheme to converge to the
solution near 0.5

4x3n - 2
xn + 1 = 6x2n - 5
we have
4x3 - 2
g (x) = 6x2 - 5
12x2 (6x2 - 5) - 12x(4x3 - 2)
g’ (x) =
(6x2 - 5)2
and g (0.5) = -0.1 approx
Since g (0.5) 1 we can expect this iteration to converge to the solution near
0.5
Using either of the iterative methods in 1 or 3 we are able to calculate the root as
0.432320 (to 6 decimal places).
You should also note that unlike a linear first order recurrence relation it will make a
difference to the iterative scheme xn + 1 = g (xn ) which value we choose for x0 . Since
g (x) represents the equation of a curve its gradient will vary, so given that the iterative
process converges and g () 1, x0 should be chosen as close to the root as
possible to ensure that g (x0 ) 1 also.
Now try the questions in the following exercise.
Exercise 7
20 min
Q37:
The equation 4x + 5 - x 3 = 0 has one solution near x 0 = 2
a) Show that the formula x n + 1 = (4xn + 5)1/3 with x0 = 2 should give a convergent process
for this solution.
b) Does the formula xn + 1 = 1
x3 - 5 with x = 2 converge?
4 n 0
Q38:
The equation x 3 - 4x - 1 = 0 has three solutions, near -1.9, -0.3 and 2.1
1
xn + 1 = x2n - 4
is an iterative scheme for this equation in the form x n + 1 = g (xn )
a) Calculate g (x0 ) for the initial values x0 = -1.9, -0.3 and 2.1 and therefore determine
for which of these values you would expect the iterative formula to converge to the
required solution.
b) Calculate the value of the solution or solutions that this iterative scheme gives you (to
6 decimal places).
Q39:
a) The following are all rearrangements of the equation 2x 3 - 5x + 2 = 0
(A) x = 2x3 + 2
5x - 2 1
3 4x3 - 2
5 , (B) x = 2 , and (C) x = 6x2 - 5
These rearrangements give the iterative process x n + 1 = g (xn )
The equation has three solutions. One of these solutions is near -1.8
By considering the value of g (x) at -1.8 for the above iterative processes determine
which of them will converge to the solution near -1.8
b) Calculate the value of this solution to 6 decimal places.
3.2.6 Order of convergence
Learning Objective

ÆDetermine the order of convergence for an iterative process

We now have a way for deciding whether a particular iterative process will converge or
not to a particular solution. You may also have noticed that some iterative processes
converge more quickly than others. This depends on the order of convergence. The
following example should explain this.

Example The equation x 2 - 7 = 0 has two solutions, one of which is near 2.6 and the
other near -2.6. Notice that solving this equation will give the exact roots = 7
Check that the following are both rearrangements of this equation
x= x+7
and x = 1
x + 7
x+1 2 x
xn + 7 x+7
For xn + 1 = , g (x) =
xn + 1 x+1
-6
and g ’(x) =
(x + 1)2
g ’ (2.6) = - 0.463 approx
1
7

1

7
For xn + 1 = xn + , g (x) = x+
2 xn 2
1
x
7

and g ’ (x) = 1- 2
2 x
g ’ (2.6) - 0.018 approx
Since g (2.6) 1 for both of these iterative processes they converge to the solution
near 2.6 as follows.
xn + 7

1 7

The iteration xn + 1 = The iteration xn + 1 = xn +

xn + 1 2 xn
gives the following iterates gives the following iterates
x0 2.6 x0 2.6
x1 2.666667 x1 2.646154
x2 2.636364 x2 2.645751
x3 2.650000 x0 2.645751
x4 2.643836
x5 2.646617
x6 2.645361
x7 2.645928
x8 2.645672
x9 2.645787
x10 2.645735
x11 2.645759
x12 2.645748
x13 2.645753
x14 2.645751
x15 2.645752
x16 2.645751
x17 2.645751
Both iterations give the solution 2.645751 (to 6 decimal places). However, it is very
obvious that the second iterative process converges far more quickly.
The value of g (x) gives some indication of the speed of convergence for an iterative
process. Notice that for the second iteration g (2.6) was much closer to zero. This
indicates that it should converge more quickly.
In this case we can actually calculate g () since =
7
-6
For g ’ (x) =
(x + 1)2
g ’ () = -6
= - 0.451 approx
( 7 + 1)2
1
7

For g ’ (x) = 1- 2
2
x

g ’
1 7
2
1- 2
0
7
The previous two iterative processes are examples of first order and second order
convergence.
xn + 7
xn + 1 = is an iteration with first order convergence

xn + 1
1 7

xn + 1 = 2 xn + xn is an iteration with second order convergence.
order of convergence
For an iterative process in the form xn + 1 = g (xn )
The order of convergence is first order when g (x) 1 but g () 0
The order of convergence is second order when g (x) 1 and g () = 0 but
g () 0
Similar statements can also be made about higher orders of convergence.
Now try the following questions.
Exercise 8
15 min
Q40: Show that the following iterative processes have first order convergence i.e. show
that g (x) 1 and that g () 0
2
a)xn + 1 = xn + 4 for the solution of x2 + 4x - 2 = 0 near x = 0
b) xn + 1 = 1 - sin xn for the solution of sin x + x - 1 = 0 between 0 and 1 radians

Q41: The iterative process xn + 1 = 12 xn 3 - 5x2n can be used to determine 1

5
. Show
1
that this process is second order and apply it to obtain
5
to 6 decimal places. (You can
check this on your calculator.)
3.3 Summary
Learning Objective

ÆRecall the main learning points from this topic

1. A power series is an expression of the form

an xn = a0 + a1 x + a2 x2 + + an xn +
n=0
where a0 , a1 , a2 , , an , are constants and x is a variable.
2. The Maclaurin series generated by the function f (x) is

xr x x2 xn
f (r) (0) = f (0) + f (1) (0) + f (2) (0) + + f (n) (0) +
r! 1! 2! n!
r=0
3. For the linear recurrence relation u n + 1 = aun + b, un tends to

a limit, L, when -1 a 1
b
L is given by the formula L = 1-a
4. The iterative process xn + 1 = g (xn ) ) will converge to the fixed

point when g (x) 1 for x close to
5. The order of convergence for x n + 1 = g (xn ) is first order when
g (x) 1 but g () 0
The order of convergence for x n + 1 = g (xn ) is second order
when g (x) 1 and g () = 0 but g () 0

There are links on-line to a variety of web sites related to this topic.
Colin Maclaurin (1698 - 1746)
Colin Maclaurin was born in Kilmodan, Scotland. He was the youngest of three brothers
but never knew his father who died when he was only six weeks old. His mother also
died, when Colin was nine years old, and so he was brought up by his uncle who was a
minister at Kifinnnan on Loch Fyne.
At the age of 19, in August 1717, Maclaurin was appointed professor of mathematics at
Marischal College, Aberdeen University. Maclaurin at this time was a great supporter
of Sir Isaac Newton and is reported to have travelled to London to meet him. He was
elected a Fellow of the Royal Society during one of these visits.
On 3 November 1725 Maclaurin was appointed to Edinburgh University where he spent
the rest of his career. He married Anne Stewart the daughter of the Solicitor General for
Scotland and had seven children. His teaching at Edinburgh came in for considerable
praise and he is said to have been keen to aid the understanding of his students. If they
had difficulty with a concept then he was likely to try another method of explanation in
order to give them a clearer understanding.

Maclaurin is best remembered for his publication the Treatise of fluxions where he
demonstrates the special case of the Taylor series which is now named after him.
The Taylor series generated by the function f (x) at x = a is
2 n
f (a) + f(1) (a)(x - a) + f(2) (a) (x -2!a) + + f(n) (a) (x -n!a) +
Notice that Maclaurin series are Taylor series with a = 0. The Maclaurin series was
not an idea discovered independently of the Taylor series and indeed Maclaurin makes
acknowledgement of Taylor’s influence.
Maclaurin’s other interests include the annual eclipse of the sun, the structure of bees’
honeycombs and actuarial studies.
’He laid sound actuarial foundations for the insurance society that has ever since helped
the widows and children of Scottish ministers and professors.’
Maclaurin also became involved in the defence of Edinburgh during the Jacobite
rebellion of 1745. However, when the city fell to the Jacobites he fled to England but
returned when the Jacobites marched further south. Much weakened by a fall from
his horse in combination with his exertions defending Edinburgh and a difficult journey
through winter weather to return to his home city, he became very ill in December 1745.
He died the next year and was buried in Greyfriars Churchyard, where his grave can still
be seen.
Calculating
The series for tan-1 x was first discovered by James Gregory in 1671.
x3 x5 x7
tan - 1 x = x - 3 + 5 - 7 +
When x = 1 in the above then we obtain Leibniz’s formula
1 1 1 1 1 ( - 1)n - 1
=1- + - + - + +
4
3 5 7 9 11 2n - 1
1 1 1 1 1 ( - 1)n - 1
=4 1- + - + - + +
3 5 7 9 11 2n - 1
This series converges very slowly and so is not used to approximate to many decimal
places. The series for tan-1 x converges more quickly when x is near zero. Therefore
if you want to use the series for tan-1 x to calculate then you could consider various
trigonometric identities.
For example we could use the following trigonometrical identity with
= tan - 1 12 and = tan - 1 31
tan + tan
tan ( + ) =
1 - tantan
(12) + (13)
1 - (16)
=
= 1

= tan
4
Therefore
1 1
= + = tan - 1 + tan - 1
4 1
2
1

3
= 4 tan - 1 + tan - 1
2 3
Now we can use the expansion for tan -1 x with x = 1 /2 and x = 1 /3
Since these values for x are nearer to zero the above method will give accurate results
for more quickly than Leibniz’s formula.
3.5 Review exercise

Review exercise in further sequences and series
An on-line assessment is available at this point, which you might find useful.
15 min
Q42: Find the first three terms of the Maclaurin series for f (x) = exp (3x)
Q43: The equation x 3 - 3x + 2 = 0 can be rewritten as x = (3x - 3)13 . This equation has
a solution which lies in the interval -3 x -2
By using the simple iterative scheme xn + 1 = (3xn - 3)13 with x0 = -2, find an
approximation to this solution. Give your answer correct to three decimal places.

Advanced review exercise in further sequences and series
30 min

Q44: Use Maclaurin’s theorem to write down the expansion, as far as the term in x 4 , of
1 + 2x
x 12
Q45: Use Maclaurin’s theorem to write down the expansion, as far as the term in x 3 , of
(3 - x)-2 , x 1
Q46: The equation xe x = 2 can be rewritten as x = 2e-x and has one solution near x = 1
a) Show that the iterative scheme x n + 1 = 2e - xn will converge to give the solution near
x = 1 and calculate the value of this solution to 3 decimal places.
b) Determine the order of convergence for this iterative scheme.
Q47: The solution of the equation x 3 - 6 = 0 will give the cube root of 6
2x3 + 6
a) Show that x = 3x2
is a rearrangement of this equation.
Use the iterative scheme, which comes from this rearrangement with x 0 = 2, to write
down the values you get for x 0 , x1 , x2 , x3 , x4 , ... Find the cube root of 6 to 3 decimal
places.
b) Determine the order of convergence for this iterative scheme.

Set review exercise in further sequences and series
An on-line assessment is available at this point, which you will need to attempt to have
these questions marked. These questions on the web site are not randomised and 10 min
may be posed in a different manner, but you should have the required answers in your
working.
Q48: Find the first three terms of the Maclaurin series for f (x) = sin 3x
Q49: The equation x 3 + 2x - 4 = 0 can be rewritten as x = (4 - 2x)1/3 . This equation has

a solution which lies in the interval 1 x 2
By using the simple iterative scheme xn + 1 = (4 - 2xn )13 with x0 = 1, find an
approximation to this solution. Give your answer correct to three decimal places.
125
Topic 4
Further Ordinary Differential

Equations
Contents
4.1 First order linear differential equations . . . . . . . . . . . . . . . . . . . . . . . 126
4.1.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126
4.1.2 Solving first order linear equations . . . . . . . . . . . . . . . . . . . . . 127
4.1.3 Applications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 132
4.2 Second order, linear, differential equations with constant coefficients . . . . . . 137
4.2.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137
4.3 Homogeneous, second order, linear, differential equations . . . . . . . . . . . . 138
4.3.1 Real and Distinct Roots . . . . . . . . . . . . . . . . . . . . . . . . . . . 139
4.3.2 Equal Roots . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 140
4.3.3 Complex Roots . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 141
4.3.4 Initial value problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . 143
4.4 Non-homogeneous, second order, linear, differential equations . . . . . . . . . 146
4.4.1 Finding particular integrals . . . . . . . . . . . . . . . . . . . . . . . . . 148
4.5 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 152
4.6 Proofs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 154
4.8 Review exercise . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 160
Learning Objectives
Solve further ordinary differential equations.
• Solve first order linear differential equations using the integrating factor method.
126 TOPIC 4. FURTHER ORDINARY DIFFERENTIAL EQUATIONS
Prerequisites
Before attempting this unit you should be familiar with:
• The rules for differentiating and integrating functions and in particular the product
rule and chain rule.
• Properties of the exponential and logarithmic functions and in particular a ln x = ln
xa and eln x = x
• Solving quadratic equations and using the quadratic formula.
• The complex number i and its use when evaluating the square root of a negative
number.
Try the following exercise to test your skills.

Some revision may be necessary if you find this difficult.
Revision exercise
These questions are designed to practice skills that you should already have. If you
15 min have difficulty you could consult your tutor or a classmate.
Q1: For f (x) = x2 sin x calculate f (x)
Q2: For y = e4x calculate dy /dx
Q3: Integrate
3 dx
x
Q4: Simplify e3 ln x
Q5: Rewrite
(-49) in terms of i
Q6: Find the complex roots of the equation x 2 + 4x + 13 = 0
4.1 First order linear differential equations

4.1.1 Introduction
Learning Objective

ÆRecognise a first order linear differential equation

first order linear differential equation
A first order linear differential equation is an equation that can be expressed in the
standard form
dy / + P (x)y = f (x)
dx
In general, such equations contain only simple terms in y such as y and dy / and not
dx
more complicated nonlinear terms such as y 2 , ydy /dx , ey , sin y etc
The examples that follow are first order linear differential equations.
4.1. FIRST ORDER LINEAR DIFFERENTIAL EQUATIONS 127
Examples
1. dy / + 3xy = x
dx
( P (x) = 3x, f (x) = x)
2. dy / - y = ex
dx
(P (x) = -1, f (x) = ex )
3. dy / + y sin t = t2
dx
(P (t) = sin t, f (t) = t2 )
Some algebraic manipulation may be necessary in order to express an equation in this

form.
Examples
1. xdy /dx - 3x2 = y can be expressed as dy /dx - y /x = 3x
2. dy / = x2 + xy - 3 can be expressed as dy /dx - xy = x2 - 3

dx
4.1.2 Solving first order linear equations
Learning Objective

ÆSolve first order linear differential equations using the integrating factor method

The first order linear differential equation dy /dx + P (x)y = f (x) can be solved by multiplying
both sides of the equation by a suitable function called the integrating factor, I (x)
integrating factor
For the first order linear
Ê differential equation dy /dx + P (x)y = f (x), the integrating factor
P(x) dx
is given by I (x) = e
Note that
Ê Ê
d
dx I (x) = dxd e P (x) dx = P (x) e P (x)dx
= P (x) I (x)
Here we have used
• the chain rule

d
eu (x) = u’(x)eu (x) where u (x) = P(x)dx
dx
• and the fact that

d
P(x)dx = P(x)
dx
Ê Ê
Using the the product rule and then d
dx I (x) = dxd e P (x) dx = P (x) e P (x)dx
= P (x) I (x)
we obtain
d
dx
I (x)y = d (Idx(x)) y + I (x) dx
dy
dy
= P (x) I (x)y + I (x)
dx
We can now see why multiplying both sides of dy / + P (x)y = f (x) by I (x) enables us to
dx
solve the equation.
dy
If + P(x)y = f (x)
dx
dy
then I (x) + I (x) P(x)y = I (x) f (x)
dx
and so
d
dx
I (x)y = I (x) f(x)
Integrating both sides we obtain
I (x)y =
I (x) f (x)dx
which can now be solved for y
Hence a first order differential equation can be solved using the following strategy.
Strategy for solving first order linear differential equations
1. Write the equation in standard linear form

dy
dx + P(x)y = f(x)
and thus identify P (x) and f (x)
Ê
P (x)dx
2. Calculate the integrating factor I (x) = e
3. Mutiply the equation in standard linear form by I (x) and obtain the equation
d
dx I (x)y = I (x) f (x)
I (x) f (x)dx
4. Integrate both sides to give
I (x)y =
5. Rearrange to solve for y.
general solution
The general solution of a differential equation contains one or more arbitrary constants
and gives infinitely many solutions that all satisfy the differential equation.
Example Find the general solution of

dy / + 2y = e3x
dx
Answer
1. Note that the equation is given in standard linear form so we can easily identify
P (x) = 2 and f (x) = e3x
2. We can now calculate the integrating factor
Ê Ê
P (x)dx 2dx
I (x) = e =e = e2x
Note that we do not use a constant of integration in finding the integrating factor.
3. Now use the fact that the equation can be written as
d
dx I (x)y = I (x) f (x)
With I (x) = e2x and f (x) = e3x the equation becomes
d
e2x y = e2x e3x
dx
d
e2x y = e5x
dx
d
4. Integrating both sides of this last equation gives
e2x y dx = e5x dx
dx
e2x y = 1 5x
5e +C
5. Rearranging to solve for y we obtain the general solution

y = 1 /5 e3x + Ce-2x
Note that in order to apply this method it is first necessary to ensure that the equation is
expressed in standard linear form.
dy / + P (x)y = f (x)
dx
See the following example.
Example Find the general solution of

xdy /dx + y = sin x
Answer
We can rewrite the equation in standard linear form as dy / = y /x = sin x /x
dx
Thus P (x) = 1 /x
Ê
1xdx
Hence the integrating factor is I (x) = e = elnx = x
Multiplying both sides of dy /dx + y /x = sin x /x by I (x) we obtain
d
dx
I (x)y = I (x)
sinx
x
i.e.
d
dx
xy = x = sinx
x sinx
Integrating we obtain
xy =
sinxdx = - cosx + C
Rearranging we obtain the general solution

y = - cos x /x + C /x
We saw in Unit 2, Further Integration (7.4) that if we are given an initial condition for a
first order differential equation then it should be possible to find a particular solution of
the differential equation satisfying the initial condition.
initial condition
For a differential equation an initial condition is an additional condition which must be
satisfied by the solution. This could be a coordinate on a curve, a velocity at t = 0, the
amount of money in a bank account on 1st January 2000, etc.
particular solution
The particular solution of a differential equation is a solution which is often obtained
from the general solution when values are assigned to the arbitrary constants.
The following example illustrates this point.
Example Solve the initial value problem

dy
x + 3y = 5x2
y dx
(1) = 0
Answer
In standard linear form the equation is
dy / + 3y /x = 5x
dx
Thus P (x) = 3 /x
Ê
3xdx 3
Therefore the integrating factor is I (x) = e = e3 lnx = elnx = x3
Multiplying both sides of dy /dx + 3y /x = 5x by I (x) we obtain
d
I (x)y = I (x)5x
dx
d
x3 y = x3 .5x = 5x4
dx
Integrating we obtain x 3 y =
5x4dx = x5 + C
Now we are given that y (1) = 0 and so we can find C
Substituting x = 1 and y = 0 into x3 y = x5 + C gives 0 = 1 + C so C = -1
Therefore x3 y = x5 + C becomes x3 y = x5 - 1 and rearranging to solve for y gives
y = x2 - x-3
Exercise 1
40 min
Use the strategy given previously to find a general solution for the following differential
equations.
Q7: dy / - 2y = e3x
dx
Q8: dy / + y /x = 4x2
dx
Q9: 2dy /dx + y = ex/2
Q10: xdy /dx + 2y = cos x /x

dy xy + 4
Q11: dx = x2
Q12: x dy
dx + y = xsinx
Solve the following initial value problems.
dy 2y
Q13:
+ =6
ydx(1) =x1
Q14: (Note that this linear equation is also a separable equation and so there is a choice

of methods for solving it.)

dy
+ 5y = 2

dx
y
1
5
=
3
5
dy
Q15:
- cos(x)y = 2x esin(x)
y(dx) = 0
dy
Q16:
x2 + y = x2 e1x
y (1)
dx
=0
dy
Q17:
+ 3x2y = 6x2
ydx(0) = 1
Extra Help: Differential Equations - linear 1

An online exercise is provided to help you if you require additional assistance with this
material, or would like to revise this subject.
Extra Help: Differential Equations - linear 2

4.1.3 Applications
Learning Objective

ÆSolve differential equations in context

As we have already seen in Unit 2, Further Integration, differential equations have
important applications in engineering and science. In this section we look at some
further examples that involve first order linear equations.
Examples
1. Bacterial Growth
Bacteria grow in a certain culture at a rate proportional to the amount of bacteria present.
Given that there are 100 bacteria present initially, find an equation for bacterial growth
and by solving an appropriate differential equation find the number of bacteria present
at any subsequent time.
Answer
We have already seen how to deal with this type of problem when we used the method
of separating variables to solve first order differential equations. However, we can use
an integrating factor as an alternative method.
Since bacteria grow at a rate proportional to the amount of bacteria present, the
differential equation describing bacterial growth is
dB / = kB
dt
where B (t) is the number of bacteria present at time t and k is the constant of
proportionality.
In standard linear form the equation is dB / - kB = 0 and therefore P (t) = -k and f (t) = 0
dt
Ê
- kdt
The integrating factor is I (t) = e = e - kt
The differential equation becomes
d
dt
I (t)B = I (t) f(t)
d
i.e. (e - kt B) = 0
dt
Integrating both sides gives
e-kt B = C
Where C is a constant.
Rearranging to solve for B gives B = Ce kt
(You should recognise this formula from Further Integration, Unit 2 (7.4.3).)
Now we are given that there are initially 100 bacteria present in the culture. i.e.B = 100
when t = 0
Therefore 100 = Ce 0 and so C = 100
Hence, the number of bacteria present at time t is given by
B = 100ekt
2. A mixing problem
A tank contains 50 kg of salt dissolved in 200m 3 of water. Starting at time t = 0, water
which contains 2 kg of salt per m3 , enters the tank at a rate of 4 m 3 /minute and the
well-stirred solution leaves the tank at the same rate. Calculate how much salt there is
in the tank after 30 minutes.
Answer
Let M (t) represent the total mass of salt (in kg) in the tank at time t
Clearly the rate of change of mass of salt is equal to dM /
dt
However, we can also calculate the rate of change of mass by calculating the rate at
which the total amount of salt in the tank changes and so obtain a differential equation
for M
The rate at which salt enters the tank is 2 4 = 8kg/min
The rate at which salt leaves the tank = mass of salt in the tank proportion of solution
which leaves the tank per minute
= M 4 /200 = 0.02 per minute
Thus the rate of change of mass = 8 - 0.02 M kg/minute
Hence M (t) must satisfy the differential equation
dM / = 8 - 0.02M
dt
In standard linear form the equation is

dM / + 0.02M = 8
dt
and therefore we can identify P (t) = 0.02 and f (t) = 8

Ê
0.02dt
The integrating factor is I (t) = e = e0.02t
Therefore the differential equation becomes
d
I (t)M = I (t) f (t)
d dt

e0.02t M = 8e0.02t
dt
Integrating both sides with respect to t gives
8 0.02t
e0.02t M = e +C
0.02
= 400e0.02t + C
M= 400 + Ce - 0.02t
Since there is initially 50 kg of salt in the tank, M (0) = 50 and so 50 = 400 + C i.e.
C = -350
Hence our equation becomes M = 400 - 350e -0.02t
When t = 30 then
M = 400 - 350e -0.06 = 207.9 kg
After 30 minutes there is 207.9 kg of salt in the tank.
3. An inductance-resistance circuit
This diagram represents an electrical
circuit which contains a constant DC
voltage source of 12 volts, a switch, a
resistor of size 6 ohms, and an inductor of
size L henrys. The switch is initially open
but is closed at time t = 0 and the current
begins to flow at that time. Let i (t) denote
the current t seconds after the switch is

closed. Then it is known that i (t) satisfies
di
L + Ri = V
dt
i (0) = 0
Find a formula for i (t) in terms of L,R and V.
(Note that in this question L,R and V are constants.)
Answer
Rewrite the equation in standard linear form
di Ri V
dt + L = L
and so we can easily see that P (t) = R /L and f (t) = V /L
Calculate the integrating factor
Ê R R
I (t) = e L dt = eLt
Therefore the differential equation becomes
d
I (t)i = I (t) f (t)
dt
d
V
eRtL i = eRtL
dt L

Integrating both sides we obtain
d
V RtL
eRtL i dt = e dt
dt L
V L RtL
eRtL i = e +C
LR
V
= eRtL + C
R
We now substitute in the initial values and solve for y
When t = 0 then i = 0, and so we have
V
eo .0 = e0 + C
R
V
0= +C
R
V
i.e.,C = -
R
Therefore
V RtL V
eRtL i = e -
R R
V V - RtL
and so i = - e
R R
V
= 1 - e - RtL
R
Exercise 2
50 min
Q18: In the first few weeks after birth a baby gains weight at a rate proportional to its
weight. A baby weighing 3.5 kg at birth weighs 4.2 kg after two weeks. How much did it
weigh after 5 days (give your answer to two decimal places)?
Q19: Sugar dissolves in water at a rate proportional to the amount still undissolved.
There was 80 kg of sugar initially and after 3 hours there is still 50 kg undissolved.
How long will it take for half of the sugar to dissolve (give your answer to the nearest 5
minutes)?
Q20: A tank contains 1000 litres of salt water. 20 litres of fresh water flow into the tank
and 20 litres of salt water flow out of the tank each minute.
Let M (t) represent the total amount of salt in the water at time t then the rate of change
of salt in the tank can be described by the equation
dM 20
= - M
dt 1000
dM
i.e. = - 0.02M
dt
If the concentration of salt in the water was initially 30g/litre calculate the concentration
of the salt in the water after 40 minutes. (Give your answer to 1 decimal place.)
Q21: A tank contains 100 kg of salt dissolved in 500m 3 of water. Starting at time t = 0,
water which contains 3 kg of salt per m 3 enters the tank at a rate of 5 m 3 /minute and the
well-stirred solution leaves the tank at the same rate.
Let M (t) represent the total amount of salt in the water at time t then the rate of change
of salt in the tank can be described by the equation
dM 5
= 15 - M
dt 500
dM
i.e. = 15 - 0.01M
dt
a) Calculate how much salt there is in the tank after 1 hour.
b) How long does it take until there is 1000 kg of salt in the water?
c) What happens to the level of salt in the water as t ?
Q22: An object of mass m is falling towards the earth. As it gets near the surface it is
slowed down by air resistance proportional to its velocity so that, according to Newton’s
second law of motion
mdv /dt = mg - kv
where v = v (t) is the velocity of the object at time t and g is the acceleration due to
gravity near the surface of the earth. Assuming that the object falls from rest at time
t = 0 i.e. v (0) = 0 then find an expression for v (t) for t 0
Q23: A patient is fed glucose intravenously at a constant rate. The change in the
concentration of glucose in the blood with respect to time can be described by the
differential equation
dc G
dt = 100V - kc
G, V and k are positive constants.
G is the rate at which glucose is administered, in milligrams per minute and V is the
volume of blood in the body, in litres. The concentration c (t) is measured in milligrams
per centilitre. The term -kx is included because it is assumed that the glucose is
continually changing into other molecules at a rate propotional to its concentration.
a) Solve the equation for c (t), given that c (0) = 100
b) Find the limit of the concentration of glucose in the blood as t
Save the fish

A rare species of fish is to be stored in a large tank containing 5000 litres of salt water.
15 min Initially the concentration of salt in the water is 20g/litre but the fish will survive best in a
concentration of 12g/litre. Fresh water is poured into the tank and salt water is drained
out at the same rate to reduce the concentration of salt in the water. What flow of water
is needed for the concentration of salt in the water to reach the correct level after
minutes when the fish is first put in the water? Give your answer to the nearest whole
number of litres.
There is a simulation on the course web site which allows you to check your answers for
various values of t.
4.2. SECOND ORDER, LINEAR, DIFFERENTIAL EQUATIONS WITH CONSTANT
COEFFICIENTS 137
4.2 Second order, linear, differential equations with constant

coefficients
4.2.1 Introduction
Learning Objective

ÆRecognise second order, linear, differential equations
second order, linear, differential equation

A second order, linear, differential equation is an equation which can be expressed
in the standard form
2
a ddxy2 + b dy
dx + cy = f (x)
where a, b and c are constants and a 0.

This equation is second order because it contains a second derivative (but no derivatives
of higher order).
non-homogeneous
A second order, linear, differential equation is non-homogeneous when it can be
expressed in the standard form
2
a ddxy2 + b dy
dx + cy = f (x)
where a, b and c are constants and a 0 and f (x) 0

homogeneous
A second order, linear, differential equation is homogeneous when it can be expressed
in the standard form
2
a ddxy2 + b dy
dx + cy = 0
where a, b and c are constants and a 0

If an equation is second order we would expect it to be solved by carrying out two
integrations and so we should also expect it to have a general solution containing two
arbitrary constants.
In this topic we shall see how to find such solutions for second-order, linear,
homogeneous and non-homogeneous differential equations with constant coefficients.
4.3 Homogeneous, second order, linear, differential

equations
Learning Objective

Æ
Solve homogeneous, second order, linear, differential equations with constant
coefficients
We would like to find the general solution of

2
a ddxy2 + b dy
dx + cy = 0
i.e. a solution containing two arbitrary constants.

Suppose that we could find two independent solutions y 1 and y2 (i.e. y1 is not a multiple
of y2 )
Then it is easy to see that
1. y1 + y2 is also a solution
2. If A is any constant then Ay1 is also a solution
2
It follows that if y1 and y2 are two independent solutions of a ddxy2 + b dy
dx + cy = 0, then Ay1
+ By2 is also a solution where A and B are arbitrary constants,
2
i.e. Ay1 + By2 is the general solution of a ddxy2 + b dy
dx + cy = 0.
Thus we can find the general solution provided we can find two independent solutions
y1 and y2 for the equation.
We now see how to find these two distinct solutions.
The method used is to try and find appropriate values of m such that y = e mx is a solution.
If y = emx , then
dy d2 y
dx = memx and dx2
= m2 emx
Substituting the above into the second order linear differential equation we obtain
d2 y
dy
a 2
+ cy = 0
+b
dx dx
am2 emx + bmemx + cemx = 0

emx am2 + bm + c = 0
auxiliary equation
The second order, linear, homogeneous differential equation
2
a ddxy2 + b dy
dx + cy = 0
has auxiliary equation

am2 + bm + c = 0
Hence y = emx is a solution of the differential equation if and only if m satisfies the
quadratic equation
4.3. HOMOGENEOUS, SECOND ORDER, LINEAR, DIFFERENTIAL EQUATIONS 139
am2 + bm + c = 0
This quadratic equation, which is termed the auxiliary equation (or characteristic
2
equation) of a ddxy2 + b dy
dx + cy = 0, has roots given by

- b b2 - 4ac
m 2a
and these provide two independent solutions which enable us to find the general
solution.
Three possible cases arise in solving the quadratic:
1. b2 - 4ac 0 when the quadratic has real distinct roots.
2. b2 - 4ac = 0 when the quadratic has a repeated real root.
3. b2 - 4ac 0 when the quadratic has roots which are complex numbers.
We now see how the general solution can be found in each case.
4.3.1 Real and Distinct Roots

Learning Objective

Æ
Learn the form of the general solution when the roots of the auxiliary equation are
real and distinct
If the roots are m1 and m2 , then y1 = em1 x and y2 = em2 x are independent solutions of
d2 y dy
a dx2 + b dx + cy = 0 and the general solution is
y(x) = Aem1 x + Bem2 x

You can see how this works in the following example.
Example Find the general solution for the differential equation

d2 y
dx2
- 4 dy
dx + 3y = 0
Answer
The auxiliary equation for this equation is
m2 - 4m + 3 = 0
We can factorise this to give
(m - 1)(m - 3) = 0
m = 1 or m = 3
Therefore the roots of the auxiliary equation are m 1 = 1 and m2 = 3
Hence the general solution of the differential equation is
y = A ex + B e3x
4.3.2 Equal Roots
Learning Objective

Æ
equal
If b2 - 4ac = 0, the auxiliary equation has a single repeated root given by

m = -b /2a
2
Hence a ddxy2 + b dy
dx + cy = 0 has the solution y1 = e
mx
A straightforward calculation shows that in this case y2 = xemx is also a solution.

(See Proof 1 for a complete explanation of this.)
Hence, in this situation, the equation has independent solutions e mx and xemx and so
has general solution
y = A emx + Bx emx
Example Find the general solution of the differential equation

d2 y
dx2
- 4 dy
dx + 4y = 0
Answer
The auxiliary equation is
m2 - 4m + 4 = 0
i.e.
(m - 2)(m - 2) = 0
m = 2 (repeated root)
Hence the equation has independent solutions y 1 = e2x and y2 = xe2x
and therefore has general solution
y = A e2x + bx e2x
4.3.3 Complex Roots
Learning Objective

Æ
complex
If b2 - 4ac 0, the auxiliary equation has complex roots given by

b
b2 - 4ac - b b2 - 4ac
m1 2a and m2 2a
which can be written as the complex conjugate pair

m1 = p + iq amd m2 = p - iq
The functions em1 x and em2 x now involve complex numbers and are not appropriate
solutions of our differential equation. However, it is possible to deduce ( see Proof 2)
2
that in this case a ddxy2 + b dy
dx + cy = 0 has independent real solutions.
y1 = epx cos qx and y2 = epx sin qx
Example Find the general solution for the differential equation

d2 y
dx2
+ 2 dy
dx + 5y = 0
Answer
The auxiliary equation for the above equation is
m2 + 2m + 5 = 0
This equation has roots given by the following

2 2
m
2
4 - 20
2
- 16 -2
2
4i
=-1 2i
Therefore p = -1 and q = 2 and the equation has independent solutions

y1 = e-x cos 2x and y2 = e-x sin 2x
Hence the general solution is
y = e-x (A cos 2x + B sin 2x)
We can summarise all of the above information as follows.
Strategy for solving

2
a ddxy2 + b dy
dx + cy = 0
1. Write down the auxiliary equation

am2 + bm + c = 0
2. Solve the auxiliary equation by factorising or using the quadratic formula.
3. Write down the general solution, depending on the nature of the roots of the
auxiliary equation, as follows:
a) if roots m1 and m2 are real and distinct then the general solution is
y = Aem1 x + Bem2 x
b) if there is a repeated root m then the general solution is
y = A emx + Bx emx
c) if there are complex roots p iq then the general solution is
y = epx (A cos qx + B sin qx)
You can now try the questions in Exercise 3.
Exercise 3
50 min
Find the general solutions of the following second order differential equations.
d2 y dy
Q24: dx2
- dx - 6y = 0
d2 y
Q25: dx2
+ 6 dy
dx + 5y = 0
d2 y
Q26: dx2
+ 6 dy
dx + 9y = 0
d2 y
Q27: dx2
+ 4 dy
dx + 13y = 0
d2 y
Q28: dx2
= 10 dy
dx - 25y
d2 y
Q29: dx2
+ 4 dy
dx + 8y = 0
2
Q30: 2 ddxy2 + 5 dy
dx = 3y
2
Q31: 4 ddxy2 - 4 dy
dx + y = 0
2
Q32: 9 d y
+ 12 dy
dx + 4y = 0
dx2
2
Q33: 2 ddxy2 + 2 dy
dx + 5y = 0
d2 y
Q34: dx2
+ 4 dy
dx + 6y = 0
Q35:
In this question find the solutions for the auxiliary equation to two decimal places.
d2 y
dx2
+ 4 dy
dx + 2y = 0
4.3.4 Initial value problems
Learning Objective

Æ
Given initial conditions find the particular solution for a second order differential
equation
If we are given a second order differential equation and initial conditions which must be
satisfied by the solution we can find a particular solution of the equation satisfying the
initial conditions.
As the general solution contains two arbitrary constants it is appropriate to specify two
initial conditions.
Examples
2
1. Solve the initial value problem

d y - 6 dy + 9y = 0

2
ydx(0) = 2 and y’ (0) = 9
dx
Answer
The auxiliary equation for the differential equation is
m2 - 6m + 9 = 0
(m - 3)2 = 0
m = 3 (twice)
The roots of the above auxiliary equation are equal and so the general solution of the
differential equation takes the form
y (x) = (A + Bx)emx = (A + Bx)e3x
We can now find A and B to ensure that the initial conditions are satisfied.
Since y (0) = 2 we require
(A + B 0)e0 = 2
A = 2
Hence we must have y (x) = (2 + Bx)e3x
We will now find B to ensure that y (0) = 9. First we need to calculate y (x)
y (x) = (2 + Bx)e3x
y (x) = Be3x + 3(2 + Bx)e3x by the product rule
Thus we require
9 = y (0) = Be0 + 3(2 + B 0)e0 = B + 6
B=3
Hence the solution for the initial value problem is

y (x) = (2 + 3x)e3x
2. Simple harmonic motion
A mass is attached to a vertical spring and hangs in
equilibrium. The mass is then pulled a distance of 10
cm down from its equilibrium position and released
from rest so that it oscillates at the end of the spring.
If x (t) denotes the distance of the mass below its
equilibrium position at time t, it can be shown that
x (t) satisfies the differential equation
d2 x
dt2
+ 9x = 0
2
(The equation ddt2x + w2 x = 0 is the equation of simple harmonic motion. There is
further discussion of this equation in the Extended information chapter.)
Find x (t) and hence determine the position of the mass when t = 2
Answer
The differential equation
d2 x
dt2
+ 9x = 0
has auxiliary equation m2 + 9 = 0

which has solutions m = (-9) = 3i
Obviously, we have complex roots of the form p iq where p = 0 and q = 3
Hence the equation has independent solutions y 1 = e0t cos 3t = cos 3t and
y2 = e0t sin 3t = sin 3t and so has general solution x = A cos 3t + B sin 3t
We are given that at t = 0
• the spring is 10cm from the equilibrium position i.e. x = 10 at t = 0
• the mass is at rest i.e. dx / = 0 at t = 0
dt
We now choose A and B to obtain a particular solution which satisfies these initial
conditions.
Since x = 10 when t = 0 we require
10 = A cos 0 + B sin 0 A = 10
Hence x (t) = 10 cos 3t + B sin 3t and so x (t) = -30 sin 3t + 3B cos 3t
Since dx /dt = 0 when t = 0 we require
0 = -30 sin0 + 3B cos 0 3B = 0 B = 0
Thus the position of the mass is given by
x = 10 cos 3t
Finally when t = 2 the position of the mass is x = 10 cos 6 9.6 cm below the equilibrium
position.
Exercise 4
60 min
Find the particular solution for each of the following differential equations with the given
initial conditions.

Q36:

d2y + dy - 6y = 0

2
ydx(0) = 0 andy’ (0) = 10
dx
Q37:
d2 y
dx2
- 4 dy
dx + 4y = 0 with y = 0 and
dy /
dx = 3 when x = 0
Q38:
d2 y
dx2
+ 2 dy
dy
dx = 3 when x = 0
2
Q39:

d y - 6 dy + 5y = 0
dx2
y (1) = 7 and y’ (1) = 15
dx
2
Q40:

d y - 8 dy + 16y = 0
dx2
y(1)
dx
= 7 and y’ (1) = 30
Q41:
d2 y
dx2
- 6 dy
dy
dx = 11 when x
Q42: A mass is attached to a vertical spring and hangs in equilibrium. At time t = 0 the
mass is lowered 40 cm from its equilibrium position and given an initial velocity of 12
cm/second in the upward direction (i.e. dx /dt = -12).
If x (t) denotes the distance of the mass below its equilibrium position at time t, it can be
shown that x (t) satisfies the differential equation
d2 x
dt2
+ 16x = 0
Find x (t) and hence determine the position of the mass when t = 5
Q43: A mass is attached to a vertical spring which is immersed in a sticky fluid. At time
t = 0 the mass is lowered 0.25 metres from its equilibrium position and given an initial
velocity of 1m/second in the upward direction. (i.e. dx /dt = -1)
If x (t) denotes the distance of the mass below its equilibrium position at time t, it can be
shown that x (t) satisfies the differential equation
d2 x
dt2
+ 2 dx
dt + x = 0
Find x (t)
Q44: The motion of a certain mass on the end of a spring is governed by the differential
equation
d2 x
dt2
+ 0.1 dx
dt + 0.125x = 0
where x (t) denotes the distance in metres of the mass below its equilibrium position at
time t. When t = 0 the mass is lowered 0.7 metres from its equilibrium position and is
released from rest. Find x (t)
4.4 Non-homogeneous, second order, linear, differential

equations
Learning Objective

Æ
Solve non-homogeneous, second order, linear, differential equations with constant
coefficients
In this section we see how to find the general solution of the non-homogeneous (or
inhomogeneous) equation
2
a ddxy2 + b dy
dx + cy = f (x)
where f (x) is a simple function such as a polynomial, exponential or sine function.

As the equation is of second order the general solution should contain two arbitrary
constants. We can find the general solution in the following way.
1. Find a general solution, y c of the corresponding homogeneous equation

2
a ddxy2 + b dy
dx + cy = 0
complementary function
The non-homogeneous equation
2
a ddxy2 + b dy
dx + cy = f (x)
has a corresponding homogeneous equation
2
a ddxy2 + b dy
dx + cy = 0
The homogenous equation has a general solution, y c , which contains two
constants.
yc is the complementary function for the non-homogeneous equation.
2. Find (by means of the method of undetermined coefficients, which you will learn
about later) an ’obvious’ special solution y p of
2
a ddxy2 + b dy
dx + cy = f (x)
particular integral yp is a particular integral of the differential equation
2
a ddxy2 + b dy
dx + cy = f (x)
4.4. NON-HOMOGENEOUS, SECOND ORDER, LINEAR, DIFFERENTIAL
EQUATIONS 147
when yp is a solution to the equation.

2
3. Write down the general solution of a ddxy2 + b dy
dx + cy = f (x) as
y = yp + yc
(This is because y contains two arbitrary constants and
d2 y dy d2 y dy d2 y dy
a +b + cy = a 2P + b P + cyP + a 2C + b C + cyC
dx2 dx dx dx dx dx
= f (x) + 0 = f (x)
so that y is a solution of the equation.)
Example a) Show that y = e 2x is a particular integral of the differential equation

d2 y dy
dx2
+ dx - 2y = 4e2x
b) Hence find the general solution for this equation.
Answer
a) If y = e2x then
dy d2 y
dx = 2e2x and dx2
= 4e2x
Therefore
d2 y dy
dx2
+ dx - 2y = 4e2x + 2e2x - 2e2x = 4e2x
and so yp = e2x is a particular integral of the equation.

b) The corresponding homogeneous equation is
d2 y dy
dx2
+ dx - 2y = 0
This has auxiliary equation
m2 + m - 2 = 0
(m + 2)(m - 1) = 0
m = -2 or m = 1
Thus the complementary function is
yc = A e-2x + B ex
for A and B arbitrary constants.
Therefore the equation has general solution
y = yp + yc
= e2x + Ae-2x + Bex
Thus we can use the following strategy to find the general solution of a non-
homogeneous, second order, linear, differential equation.
Strategy for solving

2
a ddxy2 + b dy
dx + cy = f (x)
1. Find the complementary function y c

i.e. find the general solution of the corresponding homogeneous equation
2
a ddxy2 + b dy
dx + cy = 0
2. Find a particular integral yp
3. Write down the required general solution

y = yp + yc
The difficult step in the method is finding the particular solution y p . We shall see how this
can be done in the next section but you may find it useful to first obtain some practice
with the overall method in Exercise 5.
Exercise 5
15 min
Q45:
a) Show that y = e4x is a particular integral for the equation
d2 y
dx2
- 6 dy
dx + 9y = e
4x
Q46:
a) Show that y = e3x is a particular integral for the equation
d2 y
dx2
+ 2 dy
dx + 5y = 20e
3x
4.4.1 Finding particular integrals
Learning Objective

Æ Find particular integrals using the method of undetermined coefficients

We now see how to find particular integrals of the differential equation
2
a ddxy2 + b dy
dx + cy = f (x)
by using the method of undetermined coefficients. This relies on the idea that for simple
functions it is easy to guess the general form of a particular integral and then unknown
coefficients in the general form can be found by substituting into the differential equation.
We will examine how this can be done in the cases where f (x) is a polynomial, where
f (x) is an exponential and where f (x) is a sine or cosine function.
EQUATIONS 149
Polynomial functions
If f (x) is a polynomial of degree n, it is always possible to find a particular solution of the
equation
(c
2
a ddxy2 + b dy
dx + cy = f (x) 0)
of the form yP = an xn + an - 1 xn - 1 + + a1 x + a0
Example Find a particular integral for

d2 y dy
dx2
- dx + 2y = 2x
Answer
In this example the particular integral will have the form y p = ax + b where a and b are
the undetermined coefficients.
We find values for a and b to ensure that y p is a particular integral.
dyP d2 yP
When yP = ax + b then dx = a and dx2
=0
Hence yp is a solution, provided
d2 yPdyP
2
+ 2yP = 2x
-
dx dx
i.e. 0 - a + 2(ax + b) = 2x
2ax + (2b - a) = 2x
Equating coefficients we obtain
2a = 2 a = 1
and
2b - a = 0 b = a /2 b = 1/2
Thus we now have the particular integral
yp = x + 1 /2
Note
• If f (x) is a quadratic e.g. f (x) = 3x2 - 4x + 1 or f (x) = x2 - 5 we would seek a

particular solution of the form yp = ax2 + bx + c
• If f(x) is a constant e.g. f (x) = 3 we would seek a particular integral of the form
f (x) = a
Exponential functions
It is usually possible to find a particular integral for the equation
2
a ddxy2 + b dy
dx + cy = e
rx
of the form yp = kerx where k is the undetermined coefficient.
Example Find a particular integral for

d2 y
dx2
+ 2 dy
dx - 3y = e
2x
Answer
We make a guess and try yp = k e2x Then
dyP d2 y P
dx = 2ke2x and dx2
= 4ke2x
Hence yp is a particular integral provided

4ke2x + 2 2ke2x - 3ke2x = e2x
i.e.5ke2x = e2x
5k = 1
1
k= 5
Thus the function yp = 1 /5 e2x is a particular solution for the differential equation.
Note
The above method for finding a particular solution of
2
a ddxy2 + b dy
dx + cy = e
rx
fails when r coincides with one of the roots of the auxiliary equation am 2 + bm + c = 0
In this case we can still find a particular integral of the form:
• yp = kx erx if r is a single root of the auxiliary equation;
• yp = kx2 erx if r is a repeated root of the auxiliary equation.
Sine and Cosine functions

It is always possible to find a particular integral of the equation
2
a ddxy2 + b dy
dx + cy = sin nx, (b 0)
or of
2
a ddxy2 + b dy
dx + cy = cos nx,(b 0)
of the form
yp = p sin nx + q cos nx
where p and q are the undetermined coefficients.
It would have been much more convenient if the particular integrals could have been
respectively yp = p sin nx and yp = q cos nx. Unfortunately, this simpler and more
obvious approach does not work and the general form y p = p sin nx + q cos nx should
always be used.
Example Find a particular solution for the differential equation

d2 y
dx2
+ 2 dy
dx + 2y = 5 sinx
EQUATIONS 151
Answer
We try a particular solution of the form yp = p sin x + q cos x
Hence
dyP d2 y P
dx = p cos x - q sin x and dx2
= - p sin x - q cos x
d2 y
Hence yp is a particular solution of dx2
+ 2 dy
dx + 2y = 5 sinx provided
- p sinx - q cosx + 2(p cosx - q sinx) + 2(p sinx + q cosx) = 5 sinx

i.e. (p - 2q)sinx + (2p + q)cosx = 5 sinx
i.e. p - 2q = 5 and 2p + q = 0
i.e. p = 1 and q = - 2
The particular solution is therefore
yp = sin x - 2 cos x
Note
For the differential equation
2
a ddxy2 + b dy
dx + cy = f (x)
if f (x) is a sum of terms i.e. f (x) = p (x) + q (x) and

dy d2 y
y = u (x) is a particular integral of a + cy = p (x) and
2
+b
dx dx
d2 y dy
y = v (x) is a particular integral of a 2 + b + cy = q (x)
dx dx
then it can be shown that y = u (x) + v (x) is a particular integral of
2
a ddxy2 + b dy
dx + cy = p (x) + q (x)
Exercise 6
50 min
In the following questions find
a) the complementary function, yc
b) a particular solution, yp
c) the general solution
of the non-homogeneous equation.
d2 y
Q47: dx2
- 6 dy
dx + 9y = e
4x
d2 y
Q48: dx2
- 5 dy 2
dx + 4y = 8x + 3
d2 y
Q49: dx2
+ 4 dy
dx + 4y = 25 sinx
d2 y
Q50: dx2
+ 2 dy
dx + 10y = 8 - 10x
d2 y
Q51: dx2
- 6 dy
dx + 9y = 18
d2 y
Q52: dx2
+ 2 dy
dx + 17y = 10e
x
d2 y
Q53: dx2
+ 3 dy
dx - 10y = 21e
2x
d2 y
Q54: dx2
- 5 dy
dx + 4y + 50 cos3x = 0
d2 y dy
Q55: dx2
+ dx - 2y = 3 sinx + 19 cosx
Q56: Find the general solution of the differential equation

d2 y dy
dx2
- dx - 2y = f (x)
in each of the cases
a) f (x) = 12x
b) f (x) = 8 e3x
c) Hence find the general solution of
d2 y dy
dx2
- dx - 2y = 12x + 8e3x
4.5 Summary
Learning Objective

Æ Recall the main learning points from this topic

A. First order linear differential equations can be solved in the following
way
1. Write the equation in standard linear form

dy
dx + P(x) y = f (x)
2. Calculate the integrating factor

Ê
p (x) dx
I (x) = e
3. Multiply the equation in standard linear form by I (x) and obtain the
equation
d
dx I (x) y I (x ) f (x)
I (x) f (x) dx
4. Integrate both sides of this last equation to give
I (x) y =
5. Rearranged to solve for y
4.5. SUMMARY 153
B. When an initial condition is given for a first order, linear, differential

equation we should then be able to find a particular solution.
C. The homogeneous, second order, linear differential equation
2
a ddxy2 + b dy
dx + cy = 0
can be solved in the following way
1. Write down the auxiliary equation

am2 + bm + c = 0
2. Solve the auxiliary equation by factorising or using the quadratic

formula.
3. Write down the general solution depending on the nature of the

roots of the auxiliary equation as follows
a) if roots m1 and m2 are real and distinct then the general

solution is
y = Aem1 x + Bem2 x
b) if there is a repeated root m then the general solution is
y = A emx + Bx emx
c) if there are complex roots p iq then the general solution is
D. When we are given two initial conditions we can find a

particular solution of a second order linear differential equation
satisfying these initial conditions.
E. The non-homogeneous, second order, linear differential equation
2
ad y
+ b dy
dx + cy = f (x)
dx2
can be solved in the following way
a) Find the complementary function y c

i.e. find the general solution of the corresponding homogeneous
equation
2
a ddxy2 + b dy
dx + cy = 0
b) Find a particular integral yp
c) Write down the required general solution

y = yp + yc
F. In certain cases it is possible to find a particular integral for the

equation
2
a ddxy2 + b dy
dx + cy = f (x) c 0
1. If f (x) is a polynomial of degree n then the particular integral is of

the form yP = an xn + an - 1 xn - 1 + + a1 x + a0
2. If f (x) = erx then the particular integral is usually of the form

yp = k erx
3. If f (x) = sin nx or f (x) = cos nx then the particular integral is usually

of the form yp = p sin nx + q cos nx
4.6 Proofs
Proof 1
For a homogeneous, second order, linear, differential equation with equal roots show
that when y1 = emx is a solution then y2 = xemx is also a solution.
Proof
2
a ddxy2 + b dy
dx + cy = 0

am2 + bm + c = 0
The auxiliary equation has equal roots when b 2 - 4ac = 0 and from the quadratic formula

b b2 - 4ac
m 2a
the repeated root is given by m = - b /2a

We saw in section 14.3 that y1 = emx is one solution of the homogeneous equation. In
the case where m is a repeated root we try y 2 = xemx as another solution.
dy2 d2 y2
Then = mx emx + emx and = m2 x emx + 2m emx
dx dx2
Hence
d2 y2 dy2
a +b + cy2 = am2 xemx + 2amemx + bmxemx + bemx + cxemx
dx 2 dx

= am2 + bm + c xemx + (2am + b)emx
b

2
= 0 since am + bm + c = 0 and m = -
2a
Therefore, if y is a repeated root of the auxiliary equation, the differential equation has
solutions y1 = emx and y2 = xemx and we can write the general solution as
y = A emx + Bx emx
Proof 2
Show that the homogeneous, second order, linear, differential equation with complex
roots has independent real solutions, y 1 = epx cos qx and y2 = epx sin qx.
Proof
2
a ddxy2 + b dy
dx + cy = 0

am2 + bm + c = 0
If b2 - 4ac 0 then the auxiliary equation has complex roots which can be written as
m = p iq
Now if m = p iq is a solution to the auxiliary equation then y = e mx = e(p + iq)x is a
solution of the differential equation. Unfortunately y involves complex numbers and we
would like to find solutions of the equation invoving only real numbers.
But
y = ep + iqx = epx eiqx
q
= epx eix
= epx cos x + i sin xq (see Further sequences and series)
= epx cosqx + i sin qx (by De Moivre’s theorem)
It can be checked that both the real and imaginary parts of y give rise to solutions of the
differential equation and hence we obtain the independent solutions y 1 = epx cos qx and
y2 = epx sin qx and we can write the general solution as

There are links on the course web site to a variety of web sites related to this topic.
These include references to The Tacoma Bridge Disaster of 1940, an example of where
harmonic motion went wrong. The Millenium Bridge in London has also had oscillation
problems and we include a site giving interesting background information. There is also
a site which can take you on to more in-depth study of differential equations, if you wish,
and there are two sites which demonstrate harmonic motion.
Leonhard Euler (1707 - 1783)
Leonhard Euler was born on the 15th April 1707 in Basel, Switzerland.
While living with his grandmother, he went to school in Basel but learned no mathematics
at all there. His interest in mathematics was first sparked by his father, Paul, and indeed
Euler read mathematics texts on his own and even took some private lessons.
He entered the University of Basel in 1720 at the age of 14 and in 1723 he completed his
Master’s degree having compared and contrasted the philosophical ideas of Descartes
and Newton. He then went on to study mathematics and completed his studies in
1726. In 1727 he submitted an entry for the Grand Prize of the Paris Academy on
the best arrangement of masts on a ship. His essay won him second place which was
an excellent achievement for a young graduate. In later life he did go on to win this prize
on two occasions, in 1738 and 1740.
He served as a medical lieutenant in the Russian navy from 1727 to 1730. However,
when he became professor of physics at the St Petersburg Academy he was able to give
up his Russian navy post.
On 7th January 1734 he married Katharina Gsell who was from a Swiss family and the
daughter of a painter. They had 13 children altogether, although only five survived their
infancy. It is said that Euler claimed to have made some of his greatest mathematical
discoveries while holding a baby in his arms while others played around his feet.
Euler is best known for his analytical treatment of mathematics and his discussion of
calculus concepts, but he is also credited for work in acoustics, mechanics, astronomy,
and optics. Indeed his work in mathematics is so vast that he is considered the most
prolific writer of mathematics of all time.
He discovered the procedure for solving linear differential equations with constant
coefficiants in1739. This arose out of his work on various problems in dynamics that
led to differential equations of this type.
We owe to Euler the notation f (x) for a function, i for
(-1), for pi, for summation and
many others.
In 1735 Euler had a severe fever and almost lost his life. In1738 he started to have
problems with his eyesight and by 1740 he had lost the use of an eye. By 1766 he
became almost entirely blind after an illness. Incredibly after the age of 59, when he
was now totally blind, he produced almost half of his total works. Upon losing the sight
in his right eye he is quoted as saying ’Now I will have less distraction.’
Euler died in 1783. His last day has been described as follows.
On 18 September 1783 Euler spent the first half of the day as usual. He gave a
mathematics lesson to one of his grandchildren, did some calculations with chalk on
two boards on the motion of balloons; then discussed with Lexell and Fuss the recently
discovered planet Uranus. About five o’clock in the afternoon he suffered a brain
haemorrhage and uttered only ’I am dying’ before he lost consciousness. He died at
about eleven o’clock in the evening.
After his death the St Petersburg Academy continued to publish Euler’s unpublished
work for nearly 50 more years.
Robert Hooke
Robert Hooke was born on the 18th July 1707 in Freshwater, Isle of Wight, and died on
3rd March 1703 in London, England.
He went to school in Westminster where he learned Latin and Greek and in 1653 he
went to Christ College, Oxford.
In 1660 he discovered the law of elasticity, known as Hooke’s law. He worked on optics,
simple harmonic motion and stress in stretched springs. He applied these studies in his
designs for the balance of springs in watches.
In 1662 he was appointed curator of experiments to the Royal Society of London and
was elected a fellow the following year. In 1665 he became professor of geometry
at Gresham College, London where he remained for 30 years. In addition to this he
also held the post of City Surveyor and was a very competent architect. He was chief
assistant to Sir Christopher Wren in his project to rebuild London after the great fire of
1666.
In 1665 he first achieved world wide fame after the publication of his book Micrographia
("Small Drawings"). This book contained beautiful pictures of objects Hooke had
studied through a microscope that he made himself, including the crystal structure
of snowflakes. His studies of microscopic fossils led him to become one of the first
proposers of a theory of evolution.
There is no portrait of Hooke that is known to exist. Perhaps because he has been
described as a ’lean, bent and ugly man’ and so was maybe reluctant to sit and have his
portrait painted.
Simple harmonic motion
One of the most useful applications of second order differential equations is in the study
of vibrations or oscillations. This type of motion occurs in many situations but the classic
example is the motion of a spring.
Consider a weight of mass m, suspended from a spring of natural length L. The weight
will stretch the spring to a new length, L + s
Hooke’s Law tells us that the tension in the spring acts back up towards the equilibrium
position and is proportional to the amount the spring has stretched. Thus the upward
force exerted by the spring is ks and since the mass is in equilibrium this is balanced by
the force of gravity mg acting downwards on the mass. Thus we have
ks = mg
Suppose that the spring is pulled downwards by an additional amount x (positive
direction downwards) from this equilibrium position. Then the spring has been stretched
by a total of s + x and so the spring now exerts a force of k(s + x) upwards. Thus the net
downward force acting on the spring is
mg - k(s + x) = mg - ks - kx = -kx
But by Newton’s Second Law
mass acceleration = force
and so we obtain the differential equation
2
m ddt2x = - kx
Dividing this equation by m and rearranging, we obtain the equation
d2 x
dt2
+ w2 x = 0, where w2 = k
m
equation of simple harmonic motionThe equation

d2 x
dt2
+ w2 x = 0
is the equation of simple harmonic motion.
It has general solution
x = A cos wt + B sin wt
This second order differential equation has auxiliary equation m 2 + w2 = 0 which has
solutions

m = (-w2 ) = wi
Obviously, this has complex roots of the form p qi where p = 0 and q = w. Hence the
general solution is
x = e0t A coswt + B sinwt
= A coswt + B sinwt
This can be rewritten in the alternative form
y = R cos(wt - )
B
with R = A2 + B2 and tan - 1 A
This equation represents simple harmonic motion of amplitude R and period 2 /w The
angle is the phase angle.
The graph for this solution is shown here
This graph represents the graph of simple harmonic motion of a frictionless spring. It is
referred to as undamped vibration.
Damped harmonic motion
If the motion of the spring is retarded by a friction force, which is proportional to the
velocity, then the equation of motion becomes
2
m ddt2x = - kx - c dx
dt
cdx /dt is the friction force with c the constant of proportionality.

This equation can then be rewritten as
d2 x
dt2
+ 2b dx 2
dt + w x = 0,where 2b =
c
m and w2 = k
m
The auxiliary equation is m 2 + 2bm + w2 = 0 which has roots

m = -b
(b2 - w2)

If b w then the differential equation has complex roots - b qi, q= w2 - b2 , and
the general solution takes the form
x = e - bt A cosqt + B sinqt
= e - bt R(cosqt - )
As you can see, this is similar to the equation for undamped harmonic motion. However,
due to the factor e-bt the amplitude diminishes exponentially as t because
e-bt 0 as t
The graph for x = e-bt R (cos wt - ) follows. This solution represents damped vibratory
motion as the result of friction. Notice that the vibrations tend to die out as time
progresses.
4.8 Review exercise

Review exercise in further ordinary differential equations
An on-line assessment is available at this point, which you might find useful.
15 min
Q57: Obtain the general solution of the first-order linear differential equation
dy / = 5y = e-x
dx
dy 2
dx + 2xy = e - x cos x

Advanced review exercise in further ordinary differential equations
20 min
Q59: Find the particular solution of the initial value problem
x dy 4
dx + 2y = 6x , x = 1 when y = 5
Q60:
a) Show that the differential equation
dy
dx+ (tanx) y = cos2 x
1
has integrating factor I (x) = cosx
b) Hence find a solution of the differential equation for which y = 3 /2 when x = /4
Q61: Find the general solution of the differential equation
d2 x
+ 2 dx
dt + 5x = 0
dt2
Q62: Find the particular solution of the differential equation with the given initial
conditions
d2 y
dx2
- 10 dy
dx + 25y = 0, y (0) = 3, y’ (0) = 20
Q63: Find the solution of the differential equation

d2 y
dx2
- 4 dy
dx + 3y = f (x)
in each of the cases
a) f (x) = 40 cos x
b) f (x) = 40 sin x
c) f (x) = 40 cos x + 40 sin x

Set review exercise in further ordinary differential equations
An on-line assessment is available at this point, which you will need to attempt to have
these answers marked. These questions are not randomised on the web site. The 15 min
questions on the web may be posed in a different manner but you should have the
required answers in your working.
dy / = y /x = 4x2
dx
dy 3
dx + 3x2 y = (2x + 1)e - x
163
Topic 5
Further number theory and further

methods of proof
Contents
5.1 Revision . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 164
5.2 Implications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 164
5.3 Further proof by contradiction . . . . . . . . . . . . . . . . . . . . . . . . . . . . 171
5.4 Further proof by induction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 174
5.5 Proof using the contrapositive . . . . . . . . . . . . . . . . . . . . . . . . . . . . 177
5.6 Direct methods of proof . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 178
5.7 The division algorithm and number bases . . . . . . . . . . . . . . . . . . . . . 181
5.7.1 The division algorithm . . . . . . . . . . . . . . . . . . . . . . . . . . . . 181
5.7.2 Number bases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 184
5.8 The Euclidean algorithm and its applications . . . . . . . . . . . . . . . . . . . 187
5.8.1 The Euclidean algorithm . . . . . . . . . . . . . . . . . . . . . . . . . . . 187
5.8.2 The gcd as a linear combination . . . . . . . . . . . . . . . . . . . . . . 190
5.9 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 193
5.11 Divisibility rules and integer forms . . . . . . . . . . . . . . . . . . . . . . . . . 195
5.12 Review exercise . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 196
Learning Objectives
• Use further number theory and direct methods of proof.
• Use proof by mathematical induction.
• Use Euclid’s algorithm to find the greatest common divisor of two positive integers.
164 TOPIC 5. FURTHER NUMBER THEORY AND FURTHER METHODS OF PROOF
Prerequisites
A sound knowledge of the following subjects is required for this unit:
• Basic number theory concepts.
• Standard results involving summations.
5.1 Revision
revise the techniques before starting the topic. Ask a tutor for advice.
Q1: Explain the relationship between the terms ’theorem’, ’conjecture’ and ’proof’.
Q2: Find S4 for

r + 3

r
r1
4
Q3: If x2 + 3y 4x + 5
y 5x4 5x x y
1
2 , show that xy = - 5
4
Q4: State the meaning of the symbols:

1.
2.
3.
5.2 Implications
Learning Objective

ÆRecognise and use implications on propositions.

Mathematical propositions by their nature are either true or false. They must have one
truth value (either 1 for true or 0 for false) and cannot have both.
Example Consider the proposition 2 + 3 = 5; this is a true proposition.

On the other hand 2 + 3 = 3 is a false proposition.
In number theory, the truth of a proposition has a bearing on the symbolism used.
Many of the symbols and terms that are important in mathematics come from the
fascinating topic called mathematical logic, which is not part of this course.
Necessary and sufficient conditions
5.2. IMPLICATIONS 165
In topic 10 the following definitions of ’implies’ and ’implied by’ were given.
’Implies means that the first statement can be used to logically deduce the next
statement. It is denoted by the symbol ’.
’Implied by means that the first statement is a logical consequence of the second
statement. It is denoted by the symbol ’.
The expression P Q reads ’P implies Q’.
P is called the antecedent, the premise or the hypothesis.
Q is the conclusion or consequent.
There are however many other precise ways of saying exactly the same thing. The
following statements are all ways of saying that P Q:
• P implies Q
• if P then Q
• Q follows from P
• Q if P
• Q is necessary for P
• P is sufficient for Q
Also P Q, meaning P is implied by Q, has a similar list:
• Q implies P
• if Q then P
• P follows from Q
• P if Q
• P is necessary for Q
• Q is sufficient for P
These two lists demonstrate clearly that the symbol gives the reverse implication of
the symbol
The remainder of this section will concentrate on the sign but clearly the same
explanations can be found for the implication (which is much less commonly used).
The exercise which follows is similar, but not identical, to that given in topic 10. It is still
worthwhile taking a few extra minutes to complete it at this stage.
Symbol exercise
Learning Objective

Æ
5 min
Use the correct symbol
There is an alternative version of this exercise on the web for you to try if you wish.
Q5: Insert the correct symbol between proposition A and proposition B in the following:
1. A: The liquid is colourless and clear.

B: The liquid is drinking water.
2. A: x2 = -2
B: x is a complex number.
3. A: R is a square matrix.
B: The inverse matrix R-1 exists.
4. A: n is an odd prime.
B: n is an odd number.
Q6: Reword the following implications using either P Q or P Q
1. If x + 8 = 4 then x is a negative number.

2. x 0 is sufficient for x + 1 to be positive.
3. sin = 0 is necessary for =
The two phrases ’Q is necessary for P’ and ’P is sufficient for Q’, both of which mean P
Q, give a different emphasis to the propositions.
Look at a truth table for P Q. The table is constructed by labelling the columns with
the propositions P and Q and the implication P Q. The rows are then completed by
considering the combinations of true (T = 1) and false (F = 0) which can occur. (T = 1
and F = 0 are known as Boolean constants.)
P Q P Q
1 1 1
1 0 0
0 0 1
0 1 1
The last two lines may seem puzzling, but can be explained by considering the meaning
of P Q as ’if P then Q’. Hence if P is true, Q follows by some logical deduction: but if
P is not true, Q might still be true for some other reason.
Examples
1. If P is the statement ’it is Saturday’ and Q is the statement ’it is the weekend’ then P
Q. But statement Q is true on Sundays too, when P is false.
2. If P is the statement ’x 3’ and Q is the statement ’x2 9’ then P Q. But if, for
example, x = -4, then P is false, but Q is true.
So when P Q either P is true - and then Q must be true - or P is false and no

conclusion can be made about Q.
Hence P Q is logically equivalent to the statement ’either P is false, or Q is true, or
both’ and when P is false, the implication P Q is true.
This may seem at odds with the usual sense of the word ’implication’ but it is correct
within the framework of mathematical reasoning.
3. If P is the statement ’x is an odd number and 1 x 9’ and Q is the statement ’x is

a prime number’ then P Q.
If P is false, then Q can be either true - for example when x = 13 - or false as when
x = 4 or x = 14.
However as a single mathematical proposition P Q is true.
To say ’P is sufficient for Q’ indicates that if P is false the implication is true regardless
of P but that if P is true then Q has to be true for the implication to hold.
P is termed the sufficient condition.
Sufficient condition
If P Q then P is a sufficient condition for Q.
The phrase ’Q is necessary for P’ however, indicates that Q has to be true if P is true for
the implication to hold but Q can be true independently of P and give a true implication.
Q is termed the necessary condition.
Necessary condition
If P Q then Q is a necessary condition for P.
The same concepts apply to the lists of related phrases for P Q which were shown
earlier.
Bi-implication or equivalence The implication P Q can be read as:
• P if and only if Q
• P is equivalent to Q
• P implies Q and Q implies P
The definition as given in topic 10 follows.

Equivalent to
Is equivalent to means that the first statement implies and is implied by the second
statement. (The first statement is true if and only if the second statement is true.)
The truth table for is
P Q P Q
1 1 1
1 0 0
0 0 1
0 1 0
Example Let P be the statement:

A triangle has side lengths a, b and c (with c the largest) and contains a right angle.
Let Q be the statement:
A triangle has side lengths a, b, c (with c the largest) and a 2 + b2 = c2
Then P Q
Equivalence exercise
There is a short exercise similar to this on the web for you to try if you prefer it.
5 min
Q7: State, with explanation, which of the following statements is correct with the
equivalence symbol of replacing the question mark.
1. In the Smith family, Joe is Amy’s father ? Amy is Joe’s daughter.

2. x is divisible by 6 ? x is an even number.
3. I study Maths ? I am good at Maths.
4. x is an odd number ? x is a prime 2
Negation
The symbol is used to denote negation in this topic. There are alternative symbols
such as - and ~ . The definition of negation is as follows.
Negation
• If the proposition P is true then its negation P is false.
• If the proposition P is false then its negation P is true.
Both propositions P and P cannot be true at the same time.
Q8: Complete the truth table for negation.

P P
1
0
Example State the negation of the proposition P : x 2

Answer:
The proposition is P and

the negation is P : x 2
This can be checked as follows:
Let x = 4 then P is true and P is false.
Let x = 1 then P is false and P is true.
Negation exercise
There is another small exercise on the web for you to try if you wish.
5 min
Q9: For each of the propositions P, state its negation P.

1. P: x is rational.
2. P: I like Maths.
3. P: x is a negative real.
4. P: x 0, x .
Inverse and converse

Returning to the original implication P Q, the inverse implication applies to the
negation of both propositions P and Q.
Inverse
The inverse of the statement P Q is P Q
The converse is also worth mentioning at this point.
Converse
The converse of the statement P Q is Q P
Look at the truth tables of both.
Inverse
P P Q Q P Q
1 0 1 0 1
1 0 0 1 1
0 1 0 1 1
0 1 1 0 0
Converse
Q P Q P
1 1 1
0 1 1
0 0 1
1 0 0
Notice that the results of the two truth tables are the same.
That is, P Q is logically the same as Q P:
the inverse is the logical equivalent of the converse.
Example State the converse of the implication P Q where P is the proposition

1/ and
x
Q is the proposition x 0
Answer:
The implication as it stands reads ’if 1 /x then x 0’.
The converse is Q P and reads ’ If x 0 then 1 /x ’.
Note that the converse implication does not need to hold.
Converse exercise
There is a similar web exercise for you to try if you wish.
5 min
Q10: Consider the implication S: ’if it is Thursday then I will be at home’ and state the
converse.
Q11: State the converse of the following implications:
1. I am over fifty I am over forty.

2. If it rains then it is cloudy.
3. I work only if I am at Heriot Watt University.
4. Joyce cycles to school if it is dry.
Contrapositive
The final implication to consider is the contrapositive.
It is the inverse of the converse.
Since the converse and the inverse are logically equivalent, from this alone it can be
deduced that the contrapositive and the original implication are logically equivalent. (It
is similar to taking the inverse of an inverse in matrix algebra or reflection twice in an
axis in functions).
Contrapositive
The contrapositive of the implication P Q is Q P
Activity
Complete, compare and confirm that the truth tables for the implications P Q and
Q P are the same.
’When a philosopher says something which is true then it is trivial. When he says
something that is not trivial then it is false.’
Carl Gauss 1777 - 1855
5.3. FURTHER PROOF BY CONTRADICTION 171
5.3 Further proof by contradiction

Learning Objective

ÆApply the method of proof by contradiction

The following definition may be familiar.
Proof by contradiction
Proof by contradiction assumes the conjecture to be false and shows that this leads to
the deduction of a false conclusion.
Apart from assuming that the original conjecture is false, the strategy in proof by
contradiction depends, as do other types of proof, on exactly what the conjecture states.
Many proofs by contradiction can use the odd and even integer forms.
For example any even integer m takes the form m = 2s for some other integer s (since
all even integers are divisible by 2).
An odd integer k takes the form k = 2t + 1 for some other integer t
Further details on this and on other integer forms will be given in the section on the
division algorithm and more of these integer forms are shown in a separate section
called Divisibility rules and integer forms.
Here is an example of a proof by contradiction using the odd and even forms of integers.
Example If m2 is even then m is even. Prove this conjecture using the proof by
contradiction method.
Answer:
Assume that the conjecture is false. So assume that m 2 is even but that m is odd.
Let m = 2t + 1 where t is an integer.
Then m2 = (2t + 1)2 = 4t2 + 4t + 1 = 2 (2t2 + 2t) + 1 = 2s + 1 where s = 2t2 + 2t
This is in the form of an odd integer and contradicts the assumption that m 2 is even.
Thus the conjecture is true.
Sometimes straightforward algebraic manipulation is needed, as in the next example.
Example Prove, by contradiction, that

3 + 5 17
Answer:
Assume that the conjecture is false
Make a new assumption that
3 + 5 17
Square both sides to give
5)2 = 3 + 215 + 5 = 8 + 215
( 3+

( 17)2 = 17

8 + 2 15 17
215 9
Square both sides again to give
60 81 which is a contradiction.
Thus the assumption that
3 + 5 17 is untrue and the original conjecture is true.
That is,
3 + 5 17
In an example such as the last one it is important to realise that ’prove’ means provide
a reasoned logical argument. Calculation to show that the result holds is not a proof
(though the proof is a calculation).
Divisibility rules are also useful. For example, an integer which is not a multiple of three
has a remainder of 1 or 2 if divided by 3. This is the basis of the next example.
Example Prove, by contradiction, that for any integer n, if n 2 is a multiple of 3 then n

itself is a multiple of 3
Answer:
Assume that the conjecture is untrue and make a new assumption that n 2 is a multiple
of 3 but n is not a multiple of 3. There are therefore two possibilities to consider
1. n is of the form 3k + 1 for some integer k
2. n is of the form 3k + 2 for some integer k
Look at each case in turn.
1. Suppose n = 3k + 1
Then n2 = (3k + 1)2 = 9k2 + 6k + 1 = 3(3k2 + 2k) + 1
which is of the form 3t + 1 where t = 3k2 + 2k
But, by the assumption, n2 is a multiple of 3 and this contradicts the fact that
n2 = 3t + 1
The assumption is false and so n is not of the form 3k + 1
2. Suppose n = 3k + 2
Then n2 = (3k + 2)2 = 9k2 + 12k + 4 = 3(3k2 + 4k + 1) + 1
which is of the form 3t + 1 for another integer t
But, by the assumption, n2 is a multiple of 3 and this contradicts the fact that
n2 = 3t + 1
The assumption is false and so n is not of the form 3k + 2
The conjecture is therefore true and for any integer n, if n 2 is a multiple of 3 then n itself
is a multiple of 3
5.3. FURTHER PROOF BY CONTRADICTION 173

Irrationality is another common area for proof by contradiction and leads to a famous
theorem which states that 2 is irrational. This is the subject of one of the questions in
the exercise which is to come but the next example is very similar.

Example ’ 6 is not a rational number.’ Prove this conjecture.
Answer:
Assume that
6 is rational.

Therefore 6 = pq where p and q are positive integers with no common factors.
(Otherwise the fraction could be reduced.) (p and q are relatively prime or coprime.)
So 6q2 = p2 p2 is even.
Now if p is odd, then p 2 is odd.
But here p2 is even: so p is even
and p = 2m for some integer m
Thus 6q2 = 4m2 3q2 = 2m2
Using the same argument again
3q2 = 2m2 q2 is even and so q is even.

Thus both p and q are even but both were assumed to have no common factors. This is
a contradiction and the assumption that 6 is rational is false.
The original conjecture that
6 is not a rational number is therefore true.
There are many other techniques used in proofs, including the use of upper and lower
bounds, largest or smallest integer and so on. With practice, recognising how to proceed
does become easier.
Contradiction exercise
There is a similar exercise on the web.
Q12:

Prove that 2 + 3 10 by using proof by contradiction.
20 min
Q13: If m2 = 14 then m is not a rational number. Prove this conjecture using proof by
contradiction.
Q14: If x, y such that x + y is irrational then at least one of x, y is irrational. Prove
this using proof by contradiction.
Q15: Prove, by contradiction, that for any integer n, if n 2 is a multiple of 5 then n itself is
a multiple of 5
Q16: ’There is no largest positive integer.’ Prove this conjecture by contradiction.
Q17: Prove, using proof by contradiction, the Archimedean Property which states that
given any two positive real numbers x and y, there is a positive integer n, such that
nx y
Q18: Prove the conjecture that

2 is not a rational number using proof by contradiction.
Q19: ’There are infinitely many primes’. Prove this using proof by contradiction.
5.4 Further proof by induction

Learning Objective

ÆApply the method of proof by induction

Axioms are statements that are accepted without proof and the Peano axioms are a
famous set which were constructed for the natural numbers in 1899.
The full set of axioms are shown in the section on divisibility rules and integer forms.
One axiom of particular interest in this section is the axiom of induction that is the basis
for the method of proof by induction. This states:
’ If S is a subset of the natural numbers such that
• if 1 S
• if x S then x + 1 S
then S = ’.
’Proof by induction (or proof by the principle of mathematical induction) is an advanced
technique for showing that certain mathematical statements about the positive integers
are true.
The proof has two parts: the basis and the inductive step.
The basis is an explicit check of the result for some positive integer.
The inductive step is as follows:
• Assume the statement to be true for some unspecified value of n, say n = k

• Use this to construct an argument that the statement is true for n = k + 1
• Once this is done the conclusion is that the statement is true for all positive integers
the value of the basis step.’
The proofs of the Binomial Theorem and De Moivre’s Theorem were given in topic 10
and in the relevant earlier topics 1 and 8. They will not be given again but it is important
that both proofs are fully understood.
The technique is shown in the variety of examples which follow. Some may be familiar.
Again it is possible to categorise the proofs and the first example shows one of the many
summation results that can be proved using induction.
Example If n is a positive integer, then

n r = 1
2n (n + 1)
r=1
Prove the conjecture by using the principle of mathematical induction.
5.4. FURTHER PROOF BY INDUCTION 175
Answer:
Check the first value for which the conjecture is stated: when n = 1
then 1 = 1
2 1 (1 + 1) = 1
Suppose the result is true for n = k then 1 + 2 + 3 + ... + k = 12 k (k + 1)
Consider n = k + 1 then
1
1 + 2 + 3 + + k + k + 1 = k (k + 1) + k + 1
2
(k + 1) (k + 2)
=
2
The result is true for n = k + 1 if it is true for n = k. But it is true for n = 1 and so by the
principle of mathematical induction, the conjecture is true.
It is always important to check for a starting value which relates to the conjecture and
not assume that this will include n = 1. In fact many conjectures do not hold for smaller
values. The next example shows an induction proof which has a basis step of n = 3 (and
a conjecture which is untrue where n = 1 or 2).
Example n2 2n + 1 when n is a positive integer 3. Prove this conjecture by the

principle of mathematical induction.
Answer:
Check the first value of n = 3
Then 9 7 and the result is true.
Assume that the result is true for n = k
Then k2 2k + 1
Consider n = k + 1
Then (k + 1)2 = k2 + 2k + 1 2k + 1 + 2k + 1 2k + 1 + 1 + 1 = 2 (k + 1) + 1
So (k + 1)2 2 (k + 1) + 1
The statement is true for n = k + 1 if it is true for n = k but it is true for n = 3. So by the
principle of mathematical induction the conjecture is true for all n greater than or equal
to 3.
The starting value for the conjecture is sometimes an essential link in the argument
under the step for n = k + 1. This can be seen clearly in the next example where the fact
that the conjecture is only being proved for n 3 is used to reach a conclusion that
2 + 2n 3
Example 2n 1 + 2n for all n 3. Prove this conjecture by using induction.

Check the first value for which the conjecture has to hold i.e. n = 3
Then 23 1 + 6 which is true.

Assume that the conjecture is true for n = k
Then 2k 1 + 2k
Consider n = k + 1 (This has the effect of multiplying by 2 on LHS so multiply RHS by 2
also.)
So 2k + 1 2(1 + 2k) = 2 + 2k + 2k
(The next line is very important.)
But 2 + 2k + 2k 3 + 2k since 2 + 2n 3 for all n 3
and 3 + 2k = 1 + 2(k + 1)
So 2k + 1 1 + 2(k + 1)
The result holds for n = k + 1 when it is true for n = k. But it is also true for n = 3 so by
the principle of mathematical induction the conjecture is true.
Divisibility rules are also used.
Example Prove by induction that 3 2n + 7 is divisible by 8 for n

Answer:
Check the conjecture for n = 1
Then 32 + 7 = 16 which is divisible by 8
Then 32k + 7 = 8a for some integer a
Consider n = k + 1
Then 32(k + 1) + 7 = 32 (32k )+ 7 = 32 (8a - 7) + 7
= 72a - 63 + 7 = 8(9a - 7)
The result holds for n = k + 1 when it is true for n = k. But it is also true for n = 1 so by
the principle of mathematical induction the conjecture is true.
Induction exercise
There is a web exercise for you to try if you wish.
20 min
Q20: By induction, if n , prove

n r2 = n(n + 1)(2n + 1)
6
r=1
Q21: Prove by induction that n 2 - 2n - 1 0 for n 3

Q22: Find the smallest integer m for which n! 3 n
Hence prove by induction that n! 3 n for all n m
Q23: Prove, by induction, that n 2 - 1 is divisible by 8 for all positive odd integers.
Q24: Prove by induction that 10 n n10 for all n 10

5.5. PROOF USING THE CONTRAPOSITIVE 177
The induction principle is similar to the falling domino chain. If the dominoes are
arranged correctly, knocking the first one over produces a chain reaction causing the
rest to follow. Proof by induction is a very strong argument which works by the same
principle. Care must be taken, however, to ensure that the correct starting point is
chosen.
Activity
n
A famous number theorist, Fermat, conjectured that numbers of the form 2 2 + 1 are
prime for all .
Inspect the first four numbers of this form and comment on the results. Is the fifth number
prime? Use a graphics calculator to help by checking division by 641
This is a case where the initial results would seem to support his claim but even induction
will fail to prove the result.
5.5 Proof using the contrapositive

Learning Objective

ÆUse proof by assumimg the contrapositive appropriately

There are two types of indirect proof. The first is proof by contradiction which has been
fully explored in this topic and in topic 10. The second is the subject of this section and
is proof using the contrapositive.
Recall that earlier in the topic, the contrapositive implication of P Q was defined as
Q P. Although the two implications are logically equivalent, there are cases where
it is easier to construct a proof starting with the contrapositive implication and facts about
P and Q.
Examples

1. The Pythagorean conjecture ’ 2 is irrational’ is sometimes worded as ’if x2 = 2, then
x is not rational.’ It is difficult to find properties and techniques for the two propositions
’x2 = 2’ and ’x is not rational’. The contrapositive implication however reads ’if x is rational
then x2 2’.
2. Prove that if 7 is not a factor of n2 then 7 is not a factor of n

Answer:
The conjecture looks cumbersome and is certainly a case for using the contrapositive.
This states ’ if 7 is a factor of n then 7 is a factor of n2 ’
Thus n = 7m for some m n 2 = 49m2 = 7(7m2 )
That is, 7 is a factor of n2 if 7 is a factor of n. The conjecture is proved.
Proof using the contrapositive should only be used where it is easier to find relationships
and properties for the negations than for the original propositions.
Contrapositive exercise
20 min
Q25: Prove that if n3 is not even, then n is not even using the contrapositive.
Q26: Prove by using the contrapositive that the integer n cannot be expressed as a sum
of two square integers if n has remainder 3 on division by 4
Q27: Prove that if n 2 has the form 4m + 1 then n is odd.
Extra Help: Proofs - using the contrapositive

5.6 Direct methods of proof

Learning Objective

ÆApply direct methods of proof where possible

Direct methods of proof are the final type of proof to be examined.
There are no set rules for such a proof. It is constructed by using the properties and
known results which apply to the propositions and the implication as its stands.
Saying this does not mean that the proof should be any less rigorous than the other
methods of proof already used. In all cases, the deductions from one line of the proof to
the next must be logical and correct mathematically.
The following example shows how a proof can fail by not applying logical and
mathematically correct deductions.
This is an incorrect proof that 1 = 2
Let x = 1 and x = y
Then xy = y2
So xy - x2 = y2 - x2
x(y - x) = (y + x)(y - x)
Thus x = y + x and 1 = 2
This proof fails because the second last step divides through by y - x which of course is
zero. As everyone should know, zero cannot be used as a divisor.
Now some correct proofs will be demonstrated. The first is very straightforward and
uses the property that an integer greater than one can be expressed as m + 1 for some
positive integer m.
5.6. DIRECT METHODS OF PROOF 179
Example Prove that if x 1 then x 2 1

Answer:
Let x = m + 1 where m 1 then x2 = (m + 1)2 = m2 + 2m + 1 and this is greater than 1
since m 1
Therefore if x 1 then x2 1
As in other methods of proof, the divisibility rules and basic properties of numbers are
used in this next example.
Example Prove that a number under 100 is divisible by 3 if the sum of its digits is
divisible by 3.
Answer:
Let n be the number and n = 10a + b where a and b are natural numbers 9
But a + b = 3k by the conjecture that the sum of the digits is divisible by 3
So n = 10a + b = 9a + a + b = 9a + 3k = 3(3a + k)
The conjecture is proved.
The next example is straightforward algebraic manipulation.
Example Prove that a b if a 2 + b2 2ab for a, b

Answer:
a2 + b2 2ab
a - 2ab + b2 0
(a - b)2 0
a - b 0 or a - b 0
a b or a b
Thus a b as required.
In the next case the form of an odd integer is encountered again.
Example Prove that the square of an odd integer is odd.

Answer:
Let n = 2m + 1 for some integer m
Then n2 = (2m + 1)2 = 4m2 + 4m + 1 = 2(2m2 + 2m) + 1 which is in the form of an odd
integer as required.
Summation results can sometimes be proved by this method and do not need induction.
Example Prove that for the arithmetic series with first term a and common difference d,
S = n
2 2a + (n - 1)d
Answer:
The solution uses a trick of expressing Sn in two different ways

n
S = a + n - 1d
r1
= a + a + d + a + 2d + + a + n - 2d + a + n - 1d
but
S = (a + (n - 1)d) + (a + (n - 2)d) + + (a + 2d) + (a + d) + a
Add the two expressions for Sn to give
2Sn = (2a + (n - 1)d) + (2a + (n - 1)d) + ... + (2a + (n - 1)d) + (2a + (n - 1)d)
That is 2Sn = n(2a + (n - 1)d)
Hence S = n
2 2a + (n - 1)d
There are also many summation results that can be proved using the common
summations and the combination rules as in the next example.
Example Prove that

n r(r + 1) = 1 n(n + 1)(n + 2) by using the summation results for
3
n r and n r2 r=1
r=1 r=1
Answer:
By the combination rules and then using the summation results

n
n
n
r(r + 1) = r2 + r
r=1 r=1 r=1
1
+ 1)(2n + 1) + 12 n(n + 1)
6 n(n
n(n + 1)(2n + 1) + 3n(n + 1)
=
6
n(n + 1)(2n + 4)
6
n(n + 1)(n + 2)
3
Direct proof exercise

20 min
Q28: Prove that for all , cos2 + sin2 = 1
5.7. THE DIVISION ALGORITHM AND NUMBER BASES 181
Q29: Prove that

n (2r - 1) = n2
r=1
Q30: Prove that if a 0 and b c then ab ac
Q31: Prove directly using the results of the summations to n terms of r, r2 and r3 that
n r(r + 1)(r + 2) = n(n + 1)(n + 2)(n + 3)
4
r=1
Extra Help: Direct proof

5.7 The division algorithm and number bases

5.7.1 The division algorithm
Learning Objective

ÆUse the division algorithm and produce the proof

At this point it is convenient to introduce an extra piece of symbolism. Consider the
equation 6 = 3 2. A less precise way of stating this could be 2 divides 6 or 3 divides
6. For ease this is commonly written as 2 6 or 3 6.
This symbolism not only is a neat shorthand way of expressing division but also has
another advantage over the words: when written in this form, a b, then a is assumed
0 provided that b 0. If, in fact, a does not divide b this can be shown as
Obviously in the integers, 3 7 but to cope with this most people are taught at an early
age to write this sum with a remainder such that 7 = 2 3 + 1
In general terms if there are two integers a and b (a b) where b a then this can written
as a = qb + r where q and r are also integers. It is this type of division which forms the
basis of the division algorithm.
There are two forms of the algorithm which differ only in the condition placed on the
remainder r.
The original form of the division algorithm has the condition 0 r b where b is the
divisor.
The general form of the division algorithm replaces this condition with 0 r b and
is given in the definition which follows.
Division algorithm
If a and b are integers, b 0, then there are unique integers q and r such that
a = qb + r where 0 r b
As in any division, q is the quotient and r is the remainder.

Note that the proof of the division algorithm is based on the ’the well ordering principle’
which states that every nonempty set S of nonnegative integers contains a least
element. The proof is shown for the original form with 0 r b
Proof of the division algorithm
Let S be the set {a - xb such that x and a - xb 0}
The steps in the proof are:
• show that the set S is nonempty.

• show that S contains a least element.
• show that r b
• show the uniqueness of q and r
• S is nonempty: by the condition on b in the definition of the division algorithm

b 1 so a b a
Thus a - (- a )b = a + a b a + a 0
and so if x = - a , a - xb S
• S contains a least element: by the well ordering principle, since S is nonempty, it
contains a least element say r such that r = a - qb for some integer q and r 0
• r b: assume r b then
a - (q + 1)b = a - qb - b = r - b 0
but r - b r which contradicts the fact that r is chosen as the least element in S.
The original assumption is incorrect and r b
• q and r are unique: suppose a = bq 1 + r1 and a = bq2 + r2 with the conditions
0 r1 b and 0 r2 b
So bq1 + r1 = bq2 + r2
b ( q1 - q2 ) = r2 - r1
But then the left hand side is a multiple of b and the condition on r 1 and r2 dictates
that
-b r2 - r1 b
But the only multiple of b between -b and b is zero.
Hence r2 - r1 = 0 r2 = r1 and b ( q1 - q2 ) = 0 q1 = q2
Division algorithm example

There is an example on the web of this algorithm.
5 min Here are two examples of the division algorithm.
Examples
1. Find q and r by the division algorithm, such that a = qb + r

where a = -29 and b = 6
Answer:
-29 = -5 . 6 + 1 where a = -29, q = -5, b = 6 and r = 1
2. Find q and r by the division algorithm, such that a = qb + r

where a = 33 and b = 4
Answer:
33 = 8 . 4 + 1 where a = 33, q = 8, b = 4 and r = 1
Note the convention of expressing the calculation using a dot. This is an important and
useful approach to adopt with both the division and the Euclidean algorithm.
At first glance, the subtleties of the division algorithm may go unnoticed.
In example one note that, in keeping with the division algorithm r is positive.
So -29 = -4 . 6 - 5 is not a solution.
In example two, 33 = 7 . 4 + 5 is not a solution in keeping with the division algorithm
since r b (5 4).
Division algorithm exercise

5 min
Q32: Using the division algorithm find q and r such that a = qb + r where
1. a = -42 and b = 5
2. a = 25 and b = -7
3. a = 14 and b = -9
Q33: Using the division algorithm find q and r such that a = qb + r where
1. a = -1 and b = 3
2. a = 4 and b = 9
3. a = -12 and b = 3
It may not have been apparent at the time but the division algorithm has been used
in some form in many of the proofs in the previous sections. The application of the
algorithm produces the integer forms such as 2k for an even integer and 2k + 1 for an
odd integer.
It is the condition 0 r b that is used to great effect in determining the form that a
particular number can take.
Example Suppose that b = 4 then in the equation a = qb + r, the division algorithm

shows that a will have one of the following forms
• 4k where r = 0 with q = k
• 4k + 1 where r = 1
• 4k + 2 where r = 2
• 4k + 3 where r = 3
In all cases r satisifes the condition 0 r b
Q34: Determine all the even forms of a number on division by 8
These types of integer forms are the basis of many proofs.
5.7.2 Number bases

Learning Objective

Æ
Convert decimal integers into other number bases such as binary, octal or
hexadecimal
The division algorithm is a simple method to use to convert integers to different number
bases. The standard number base used in mathematics is 10, that is the place values
of a number are powers of 10. The decimal system, as it is known, has the structure as
shown.
A number can be converted and expressed in a different number system using any
integer as a number base, but three of the most useful are the bases 2, 8 and 16.
These are called binary, octal and hexadecimal respectively and are used extensively in
subjects such as computing and electronics.
Binary
In the decimal number system there are ten digits, 0 to 9. In binary, which has a base
of two, there are just two digits, 0 and 1. The place values in a binary system are all
powers of 2 and another diagram showing these place values can be constructed.
The diagram shows how easy it is to find the decimal form of a binary number.
11010 is 1 x 24 + 1 x 23 + 1 x 21 = 26
The division algorithm is used to find the binary form from the decimal form and is best
described by an example.
Example Convert 30 into a binary number.

Answer:
Use the algorithm repeatedly for a = qb + r; in the first case a = 30 and b = 2 (for binary);
then a = 15 and b = 2 and so on until the algorithm can be taken no further.
30 = 15 . 2 + 0
15 = 7 . 2 + 1
7=3.2+1
3=1.2+1
1=0.2+1
The binary number is taken directly from the remainders in this calculation by reading
them off upwards and is 11110
Octal
This is just as straightforward but uses the powers of 8 as place values. The octal system
only uses the eight integers 0 to 7
Example Convert the number 53467 into octal form.

Answer:
In this case the equation a = qb + r has a = 53467 and b = 8
53467 = 6683 . 8 + 3
6683 = 835 . 8 + 3
835 = 104 . 8 + 3
104 = 13 . 8 + 0
13 = 1 . 8 + 5
1=0.8+1
Again the octal number is taken straight from the remainders in the calculation by
reading upwards and is 150333
This can be checked in reverse as 3 x 1 + 3 x 8 + 3 x 8 2 + 5 x 84 +1 x 85 = 53467
Hexadecimal
This system uses base 16 and has sixteen ’digits’. Since each digit can only occupy one
place value a slight modification is needed. In this case the system uses integers 0 to 9
and then A for 10, B for 11, C for 12, D for 13, E for 14 and F for 15
Example Find the hexadecimal form of the number 298047

Answer:
Repeated application of the division algorithm with b = 16 gives
298047 = 18627 . 16 + 15 note: 15 F
18627 = 1164 . 16 + 3
1164 = 72 . 16 + 12 note: 12 C
72 = 4 . 16 + 8
4 = 0 . 16 + 4
The hexadecimal number read upwards is 48C3F
Number bases exercise

There is a web exercise for you to try if you prefer.
10 min
Q35: Convert the following numbers into binary numbers:
a) 332
b) 501
c) 43
d) 63
Q36: Find the octal forms of the following numbers:
a) 347
b) 2924
c) 3012
d) 534
Q37: Find the hexadecimal equivalent numbers to
a) 4014
b) 364
c) 2179
d) 5034
’Mathematics is the queen of sciences and number theory is the queen of mathematics.’
Carl Gauss 1777 - 1855
5.8. THE EUCLIDEAN ALGORITHM AND ITS APPLICATIONS 187
5.8 The Euclidean algorithm and its applications

In the last section the definition of the division algorithm gives expressions of the form
a = qb + r
Consider what happens when r = 0
In this case, the division algorithm leads to a definition. ’An integer b is said to be
divisible by a, that is ab if b = ac for some integer c. So a is a divisor of b.’
Suppose that there is an integer d such that d a and d b for two integers a and b,
one non-zero. In this case d is a common divisor of a and b. An extra condition gives a
definition for the greatest common divisor.
Greatest common divisor
The greatest common divisor of two integers a and b, where at least one is nonzero,
denoted gcd (a, b) is the positive integer d such that:
• d a and d b
• if c a and c b then c d
The greatest common divisor is also well known as the highest common factor (hcf).
Example Find gcd ( 42, 68)

Answer:
The positive divisors of 42 are 1, 2, 3, 6, 7, 14, 21, 42 and for 78 they are 1, 2, 3, 6, 13,
26, 39, 78. The common divisors are 1, 2, 3, 6 of which 6 is the greatest.
gcd (42, 68) = 6
A particular case occurs when, for two integers a and b, the gcd (a, b) = 1
Relatively prime
The integers a and b, where at least one is non-zero, are relatively prime (coprime)
when gcd (a, b) = 1
Q38: Show that 2210 and 399 are relatively prime.
It would be tedious to take large integers and search for their factors, as the last question
demonstrates. In fact the division algorithm again plays its part in producing another
famous algorithm to make this task much easier.
5.8.1 The Euclidean algorithm
Learning Objective

ÆUse the Euclidean algorithm to find the greatest common divisor of two integers

This algorithm is very similar to the calculation required for the change of number basis
except that at each step the value of b also changes. By repeated use of the division
algorithm the calculation is subsequently reduced to give the greatest common divisor
in the final step. An example will help to demonstrate this.
Example Find the gcd of 140 and 252

Answer:
Using the expression a1 = qb1 + r1 let a be the larger number
a1 = 252 and then b 1 = 140
Use the division algorithm to obtain
252 = 1 . 140 + 112
At this stage a2 becomes 140 and b 2 becomes 112 and the next step gives
140 = 1 . 112 + 28
The process continues
112 = 4 . 28 + 0
Once r = 0 after i steps the algorithm stops and the last value of b i gives the greatest
common divisor of the two numbers.
Here is diagram indicating the steps in the algorithm for the last example.
To find the gcd of 252 and 140 with the Euclidean algorithm
Euclidean algorithm
On the web there is a demonstration of the algorithm which allows you to input two
10 min integers and watch the algorithm in action.
It is common to write the gcd of two numbers as, for example, gcd (252, 140) = 28 as in
the last diagram.
Example Find the gcd of 1980 and 3696

Answer:
Let the larger number be a so a = 3696 and let b = 1980. Now use the division algorithm
to give
3696 = 1 . 1980 + 1716

1980 = 1 . 1716 + 264
1716 = 6 . 264 + 132
264 = 2 . 132 + 0
The greatest common divisor (or highest common factor) of 3696 and 1980 is 132
An interesting use of this algorithm is to reduce fractions.
19712
Example Simplify the fraction 55040
Answer:
Use the Euclidean algorithm to find the gcd.
55040 = 2 . 19712 + 15616
19712 = 1 . 15616 + 4096
15616 = 3 . 4096 + 3328
4096 = 1 . 3328 + 768
3328 = 4 . 768 + 256
768 = 3 . 256 + 0
Thus the gcd is 256 and now divide both numerator and denominator by 256 to give the
77
fully simplified fraction as 215
Euclidean algorithm exercise

There is a web version of this exercise for you to try if you prefer it.
15 min
Q39: Use the Euclidean algorithm to find the gcd of 299 and 1365
Q40: What is the highest common factor of 5187 and 760
Q41: Use the Euclidean algorithm to show that 10465 and 5643 are coprime.
Q42: Use the Euclidean algorithm to find the gcd of 3024 and 5184
372
Q43: Reduce the fraction 4340 to its simplest form.
Activity
It is possible, in a few simple steps, to program a graphics calculator to do this algorithm
and give the gcd. Try this if you have time.
Fibonacci numbers
The Fibonacci sequence, mentioned in topic 9 on sequences and series, produces some
fascinating results in this topic.
Recall that the sequence is defined by u n + 2 = un + 1 + un where u1 = 1 and u2 = 1

The sequence is: 1, 1, 2, 3, 5, 8, 13, 21, 34, 55, 89, 144, 233, ...
It can be shown that gcd (u n , un + 1 ) = 1 for all n 1. So successive Fibonacci numbers
are relatively prime.
Activity
Try to prove this result, viz, gcd (un , un + 1 ) = 1 for all n 1
Now consider non-successive Fibonacci numbers by first answering the following
question.
Q44: Find gcd (610, 8) and gcd (610, 55)
The last question used the Fibonacci numbers 8, 55 and 610. The answers were also
the Fibonacci numbers 2 and 5 and in general it can be proved that the greatest common
divisor of two Fibonacci numbers is in fact another Fibonacci number. This result can be
taken further.
The question first asked for gcd (610, 8). These are the Fibonacci numbers u 15 and u6
and the answer of a greatest common divisor of 2 is the Fibonacci number u 3
Thus the gcd (u15 , u6 ) = u3 and this leads to a very useful theorem.
The greatest common divisor of two Fibonacci numbers is also a Fibonacci number such
that
gcd ( un , um ) = ud where gcd (n , m) = d
Example Find the gcd (987, 144)

Answer: 987 = u16 and 144 = u 12
But gcd (16, 12) = 4 and so gcd (987, 144) = u 4 = 3
Activity
Use the Euclidean algorithm and check that the gcd (987, 144) = 4
Activity
Investigate the proposition that every positive integer can be expressed as a sum of
Fibonacci numbers with none used more than once. Try to prove it as a challenge.
5.8.2 The gcd as a linear combination
Learning Objective

Æ
Use the Euclidean algorithm to express the greatest common divisor as a linear
combination of the two integers a and b
Having used the Euclidean algorithm to find the gcd of two integers it is possible to take
the algorithm and retrace the steps taken to arrive at a linear combination of the two
integers which equals the gcd.
Example Use the Euclidean algorithm to obtain integers x and y which satisfy the
equation
gcd (2695, 1260) = 2695x + 1260y
Answer:
First use the Euclidean algorithm to find the gcd
2695 = 2 . 1260 + 175
1260 = 7 . 175 + 35
175 = 5 . 35 + 0
The gcd (2695, 1260) = 35
Now taking the algorithm it is possible using algebraic manipulation to work backwards.
Start at the second last step of the algorithm and work back each step eliminating the
remainders.
working down working up
35 = -7 . 2695 + 15 . 1260
2695 = 2 . 1260 + 175 35 = 1260 - 7(2695 - 2 . 1260)
1260 = 7 . 175 + 35 35 = 1260 - 7 . 175
175 = 5 . 35 + 0
So 35 = -7 . 2695 + 15 . 1260 and x = -7 and y = 15
Thus the gcd is found in the last entry in the first column and the linear combination,
having worked upwards is given in the first entry of the second column.
Note that this is not the only way to express the gcd as a linear combination but it will
always give a solution. The set of all solutions can, however, be found from this solution.
A simpler example will be easier to follow.
Example Find all solutions of the equation gcd (48, 20) = 48x + 20y
Answer:
Using the Euclidean algorithm gives:
48 = 2 . 20 + 8
20 = 2 . 8 + 4
8=2.4+0
Working backwards to find one linear combination gives
4 = 20 - 2 . 8
4 = 20 - 2(48 - 2 . 20) = 5 . 20 - 2 . 48
Thus one solution is 4 = 48x + 20y where x = -2 and y = 5
So 48x + 20y = -2 . 48 + 5 . 20
Rearranging this gives
48(x + 2) = 20(5 - y)Now divide through by the gcd which is 4 to give

12(x + 2) = 5(5 - y)
Since 12 and 5 have no common factors x + 2 must be a multiple of 5 and 5 - y is a
multiple of 12
Thus x + 2 = 5m x = 5m - 2
5 - y = 12m y = 5 - 12m
The general solution is 4 = 48(5m - 2) + 20(5 - 12m)
There is a shortcut to the general solution.

Suppose that ax + by = d where d is the gcd (a, b)
If, by the Euclidean algorithm, the particular solutions for x and y are x 0 and y0 then the
general solutions xn and yn are found as follows
xn = x0 +
b m
d
and
yn = y0 -
a m
d
In the last example the particular solutions were x 0 = -2 and y0 = 5

The coefficient of x = b = 48 and the coefficient of y = a = 20
The gcd is 4
Thus the general solution is 4 = 48x n + 20yn
4 = 48(-2 + (20 /4 )m) + 20(5 - (48 /4 )m)
That is, 4 = 48(5m - 2) + 20(5 - 12m)
A linear combination of two integers equal to the gcd is a special example of the simplest
type of Diophantine equation.
Diophantine equation
A Diophantine equation is an equation with integer coefficients in one or more
unknowns which is to be solved in the integers.
In its simplest form in two unknowns the linear Diophantine equation is ax + by = c
If the gcd (a, b) = d is found then such an equation has a solution if and only if d c
The full method of then solving the equation is very similar to that for finding the general
solution of the equation ax + by = c; with one extra step. Suppose that c = dn then after
finding the particular solution ax + by = d the whole equation is multiplied up to give
nax + nby = dn = c and the general solution is found in the same way as before.
Example Find the solution to the linear Diophantine equation 18 = 90x + 12y if it exists.
Answer:
First find the gcd ( 90, 12)
5.9. SUMMARY 193
90 = 7 . 12 + 6
12 = 2 . 6 + 0
The gcd (90, 12) = 6
Since 6 18 there is a solution to the Diophantine equation.
Working backwards 6 = 90 . 1 + 12 . (-7)
Multiply by 3 to give 18 = 90 . 3 + 12 . (-21) which is one particular solution to the
Diophantine equation.
Using the shortcut the general solution is
18 = 90(3 + (12 /6 )m) + 12(-21 - (90 /6 )m) = 90(3 + 2m) + 12(-21 - 15m)
Note: Fermat’s conjecture or, as more commonly known, Fermat’s last theorem states
that there is no solution to the Diophantine equation x n + yn = zn in the integers for
n 2. This was one of the most famous ’unproved’ results until Andrew Wiles presented
a proof in the mid 1990s.
Linear combinations exercise

There is another exercise on the web to find linear combinations of the gcd of two
integers and to solve linear Diophantine equations. 20 min
Q45: Use the Euclidean algorithm to find integers x and y such that
gcd (693, 84) = 693x + 84y
Q46: Use the Euclidean algorithm to find integers x and y

such that gcd (10080, 3705) = 10080x + 3705y
Q47: Use the Euclidean algorithm to find a solution to the equation

gcd (336, 180) = 336x + 180y
Q48: Use the Euclidean algorithm to solve for x and y when 14 = 1078x + 420y
Q49: Find the general solution of the equation gcd (585, 104) = 585x + 104y using the
Euclidean algorithm.
Q50: Solve the linear Diophantine equation 20 = 204x + 56y
5.9 Summary
At this stage you should be familiar with the following techniques and the methods:
• Use of the correct implication.
• Proof by contradiction.
• Proof by induction.
• Proof using the contrapositive.
• Direct methods of proof.
• The division algorithm.
• The Euclidean algorithm.
• Uses of the Euclidean algorithm.

Learning Objective

ÆDisplay a knowledge of the additional information available on this subject

There are links on the web which give a selection of interesting sites to visit. These sites
can lead to an advanced study of the topic but there are many areas which will be of
passing interest.
Aristotle
His logical arguments are the basis of much of the mathematical logic of the present
day.
Euclid
The Euclidean algorithm is given in Euclid’s seventh book of the Elements but there is
evidence to suggest that it was known before this time.
de Morgan
An 18th century mathematician who formalised the method of proof by induction. The
method however was known long before.
Boole
This English mathematician showed in 1847 that algebra could be used to express
logical relationships. His work referred to as Boolean algebra is widely used in
electronics.
Peano
He was a relatively modern Italian mathematician and logician who lived in the late
19th/early 20th century. His axioms form a sound basis for the study of integers. He
also introduced the membership sign in his work on set theory.
Lame
The French mathematician Gabriel Lame produced a theorem which states that the
number of steps required in the Euclidean algorithm is never greater than 5 times the
number of digits in the lesser of the two numbers.
5.11. DIVISIBILITY RULES AND INTEGER FORMS 195
Godel
A 20th century mathematician famous for his incompleteness theorem. Look this one
up on the web.
5.11 Divisibility rules and integer forms

The Peano axioms are stated below in their original form. Later Peano modified his
axioms to include zero.
• 1 is a natural number.
• Every natural number x has a successor x’ = (x + 1)

• There is no number whose successor is one. (no x such that x + 1 = 1)
• The function ’successor’ is a one to one function. (if x’ = y’ then x = y and no two
numbers have the same immediate successor)
• Induction: If S is a subset of the natural numbers such that
– 1 belongs to S
– if x belongs to S then x’ belongs to S
then S =
Here the some of the divisibility rules for integers:

An integer N is divisible by:
2 if the last digit is even

3 if the sum of its digits is divisible by 3
4 if the number formed by the last two digits is divisible by 4
5 if the last digit is 5 or 0
6 if it is divisible by 2 and 3
7 if the number formed by removing the last digit and subsequently
subtracting twice the last digit is divisible by 7
8 if the number formed by the last three digits is divisible by 8
9 if the sum of its digits is divisible by 9
10 if it ends in 0
The following integer forms are useful to know:
• An odd integer has the form 2m + 1 for m an integer.

• An even integer has the form 2m for m an integer.
• A rational integer has the form p /q for p and q integers.
• An integer is of the form 2k or 2k + 1
• An integer is of the form 3k, 3k + 1 or 3k + 2
• An integer is of the form 4k, 4k + 1, 4k + 2 or 4k + 3 and so on.
• Every odd integer is of the form 4k + 1 or 4k + 3
• The square of any integer is of the form 3k or 3k + 1
• The cube of any integer is of the form 9k, 9k + 1 or 9k + 8
• The product of two or more integers of the form 3k + 1 is of the same form.
• The product of two or more integers of the form 4k + 1 is of the same form.

Review exercise
There is an alternative exercise available on the web for you to try if you prefer.
15 min
Q51: Use proof by induction to prove that

n r3 = n (n + 1)
2 2
4
r=1
Q52: Use the Euclidean algorithm to find the greatest common divisor of the integers
943 and 779
Q53: Prove, by induction, that the sum of the cubes of any three successive integers is
divisible by 9
Q54: Use the Euclidean algorithm to find the greatest common divisor of 5141 and 3763

Advanced review exercise
There is an alternative exercise available on the web for you to try if you prefer.
15 min
247x + 139y = 1
252x + 160y = 4
gcd(297, 180) = 297x + 180y

Set review exercise
differently but will require the same answers. One question on proof and one on the
Euclidean algorithm should be completed in 20 minutes. However, if time permits, try
them all.
Q58: Use proof by induction to show that

n r(r - 1) = n(n + 1)(n - 1)
3
r=2
Q60: Use proof by induction to prove

n 2r(r - 1)(r + 1) = n(n - 1)(n + 1)(n + 2)
2
r=2
199
Topic 6
End of unit three tests
Contents
200 TOPIC 6. END OF UNIT THREE TESTS
Online tests are provided on the web. The first two tests are set to provide questions at
level ’C’ competency and cover work from the five topics for unit 3.
The third test gives questions at level ’A’ or ’B’.
GLOSSARY 201
Glossary
addition of two matrices
If A and B are two matrices of the same order, the matrix A + B is formed by adding
the corresponding elements of each matrix.
auxiliary equation
The second order, linear, homogeneous differential equation
2
a ddxy2 + b dy
dx + cy = 0
am2 + bm + c = 0
cartesian equation of a plane
The Cartesian equation of a plane containing the point P (a 1 , a2 , a3 ) and
perpendicular to the vector n = n 1 i + n2 j + n3 k is xn1 + yn2 + zn3 = d where d
= a1 n1 + a2 n2 + a3 n3
column matrix
A column matrix has only one column but any number of rows.
complementary function
The non-homogeneous equation
2
ad y
+ b dy
dx + cy = f (x)
dx2
has a corresponding homogeneous equation
2
a ddxy2 + b dy
dx + cy = 0
The homogenous equation has a general solution, y c , which contains two
constants.
yc is the complementary function for the non-homogeneous equation.
contrapositive
The contrapositive of the implication P Q is Q P
convergent series
A convergent series is one for which the limit of partial sums exists. This limit is
called the sum and is denoted by S
converse
The converse of the statement P Q is Q P
determinant of a matrix
The determinant of a matrix is a value representing sums and products of a square
matrix.
a11 a12 a1n

If the matrix A =

an1 an2 ann

then the determinant of A is det A =
a11 a12 a1n

an1 an2 ann
202 GLOSSARY
determinant value of a 2 x 2 matrix

If A =
a b
c d
then the value of det A = a b
c d
= ad - bc

determinant value of a 3 x 3 matrix
a b c
If A =
d e f
then the value of det A =

g h i
a e f
h i
- b d f
g i
+ c d e
g h
=
a(ei - hf) - b(di - gf) + c(dh - ge)

dilatation or scaling matrix
0

x

The matrix acting by multiplication on the column matrix

0 y
represents the effect on the point (x, y) thus scaling the x-coordinate by and
the y-coordinate by
diophantine equation
A Diophantine equation is an equation with integer coefficients in one or more
unknowns which is to be solved in the integers.
divergent series
A divergent series is one which is not convergent. For example 1 + 2 + 3 + 4 + 5
... is a divergent series. The sum of this series will continue increasing the more
terms we add. It will not tend towards a limit.
division algorithm
If a and b are integers, b 0, then there are unique integers q and r such that
a = qb + r where 0 r b
elementary row operations

The three ways in which a matrix can be manipulated to solve a system of
equations are called elementary row operations. They are:
• Interchange two rows.

• Multiply one row by a non-zero constant.
• Change one row by adding a multiple of another row.
equation of simple harmonic motion

The equation
d2 x
dt2
+ w2 x = 0
is the equation of simple harmonic motion.
It has general solution
x = A cos wt + B sin wt
GLOSSARY 203
equivalent to
Is equivalent to means that the first statement implies and is implied by the
second statement. (The first statement is true if and only if the second statement
is true.)
first order linear differential equation

A first order linear differential equation is an equation that can be expressed in
the standard form
dy / + P (x)y = f (x)
dx
first order recurrence relation

A first order recurrence relation is a recurrence relation where u n + 1 depends
only on un and not on values of u r where r n
fixed point
The recurrence relation u n + 1 = aun + b has fixed point given by b //(1 - a) . The
recurrence relation will only converge to this fixed point if (-1 a 1)
general solution
The general solution of a differential equation contains one or more arbitrary
constants and gives infinitely many solutions that all satisfy the differential
equation.
greatest common divisor

The greatest common divisor of two integers a and b, where at least one is
nonzero, denoted gcd (a, b) is the positive integer d such that:
• d a and d b
• if c a and c b then c d
homogeneous
A second order, linear, differential equation is homogeneous when it can be
2
a ddxy2 + b dy
dx + cy = 0
where a, b and c are constants and a 0
identity matrix
An n x n matrix which has only ones on the leading diagonal and zeros elsewhere
1 0 0

is called an identity matrix and takes the form

0 1 0 n

0 0 1 It is denoted by I.
n
204 GLOSSARY
initial condition
For a differential equation an initial condition is an additional condition which
must be satisfied by the solution. This could be a coordinate on a curve, a velocity
at t = 0, the amount of money in a bank account on 1st January 2000, etc.
integrating factor
For the first order linear differentialÊ equation dy / + P (x)y = f (x), the
dx
integrating factor is given by I (x) = e P(x) dx
inverse
The inverse of the statement P Q is P Q
inverse of a 2 x 2 matrix
a b

Let A be the square non-singular matrix

c d
a b

d -b

1
The inverse of A is denoted A -1 = = ad - bc
c d -c a
iteration
Iteration is the successive repetition of a mathematical process using the result of
one stage as the input for the next.
length of a vector
a
Let p be the vector then the length of p is defined as p = a2 + b2 + c2
b
c
linear first order recurrence relation

A linear first order recurrence relation is a recurrence relation of the form
un + 1 = aun + b, a 0
maclaurin’s theorem
Maclaurin’s theorem states that

xr
f (x) = f (r) (0)
r!
r=0
x x2 x3 xn
= f (0) + f (1) (0) + f (2) (0) + f (3) (0) + + f (n) (0) +
1! 2! 3! n!
matrix
A matrix is a rectangular array of numbers.
multiplication by a scalar
To multiply an m x n matrix A by a scalar , take each element and multiply it by
to give the matrix A.
multiplication of matrices
The product of the two matrices A mn and Bnp = Cmp
GLOSSARY 205
a11 a12 a1n

b11 b12 b1p

and B =
a21 a22 a2n b b22 b2p
If A = 21
am1 am2 amn bn1 bn2 bnp
then the matrix AB =


am1 b11 + + amn bn1 am1 b12 + + amn bn2 am1 b1p + + amn bnp
necessary condition
If P Q then Q is a necessary condition for P.
negation
• If the proposition P is true then its negation P is false.

• If the proposition P is false then its negation P is true.
non-homogeneous
A second order, linear, differential equation is non-homogeneous when it can be
2
a ddxy2 + b dy
dx + cy = f (x)
where a, b and c are constants and a 0 and f (x) 0
numerical analysis
Numerical analysis is a branch of mathematics. It can be described as the
analysis and solution of problems which require calculation.
order of convergence
For an iterative process in the form xn + 1 = g (xn )
The order of convergence is first order when g (x) 1 but g () 0
The order of convergence is second order when g (x) 1 and g () = 0 but
g () 0
orthogonal matrix
A square matrix A is orthogonal A T = A-1
AT = A-1 det A = 1
parametric equation of a plane
The parametric equation of the plane is
r = a + b + c where a is a position vector of a point in the plane, b and c are two
non-parallel vectors parallel to the plane and and are real numbers.
parametric form of the equation of a line

The parametric equations of a line through the point P = (a 1 , a2 , a3 ) with direction
d = d1 i + d2 j + d3 k are
x = a1 + d1 , y = a2 + d2 , z = a3 + d3 where is a real number.
206 GLOSSARY
particular integral
yp is a particular integral of the differential equation
2
a ddxy2 + b dy
dx + cy = f (x)
when yp is a solution to the equation.
particular solution
The particular solution of a differential equation is a solution which is often
obtained from the general solution when values are assigned to the arbitrary
constants.
position vector
A position vector is a vector which starts at the origin.
power series
A power series is an expression of the form
a xn

a0 + a1 x + a2 x2 + a3 x3 + + an xn +
n
n=0
where a0 , a1 , a2 , a3 , ..., an , ... are constants and x is a variable. It is called a power
series as it is made up of a sequence of powers of x with coefficients a 0 , a1 , a2 ,
a3 , ..., an , ...
proof by contradiction
Proof by contradiction assumes the conjecture to be false and shows that this
leads to the deduction of a false conclusion.
recurrence relation
A recurrence relation describes a sequence of terms u0 , u1 , u2 , u3 , ..., un , un + 1 ,
... where each term is a function of previous terms.
reflection matrix
cos 2 sin 2

The matrix acting, by multiplication on the column matrix

x
sin 2 - cos 2
, represents the effect of reflection on the point (x, y) in the line through the
y
origin which makes an angle of Æ , -90Æ 90Æ
relatively prime
The integers a and b, where at least one is non-zero, are relatively prime (coprime)
when gcd (a, b) = 1
rotation matrix
cos - sin

The matrix R = acting, by multiplication on the column matrix

x
sin cos
A= , represents the effect of rotation through Æ anticlockwise about the
y
origin on the point (x, y)
row matrix
A row matrix has only one row but any number of columns.
GLOSSARY 207

scalar product in component form
a d
If p = and q = e
b
c f
then the scalar product is the number p q = ad + be + cf
scalar product in geometric form
The scalar product of two vectors a and b is defined as
ab= a b cos where is the angle between a and b, 0 180 Æ
scalar triple product

The scalar triple product a (b x c) is the number given by the scalar product of
the two vectors a and (b x c)
second order, linear, differential equation
A second order, linear, differential equation is an equation which can be
2
a ddxy2 + b dy
dx + cy = f (x)
where a, b and c are constants and a 0.
second order recurrence relation
A second order recurrence relation is a recurrence relation of the form un + 2 =
aun + 1 + bun + c, with a and b 0. Note that u n + 2 depends on the previous two
terms in the sequence.
singular matrix
A matrix whose determinant is zero is called a singular matrix.
square matrix
A square matrix has the same number of rows and columns.
subtraction of two matrices
If A and B are two matrices of the same order, the matrix A - B is formed by
subtracting the corresponding elements of matrix B from those in matrix A.
sufficient condition
If P Q then P is a sufficient condition for Q.
symmetric form of the equation of a line
If x = a1 + d1 , y = a2 + d2 , z = a3 + d3 are parametric equations of a line, the
symmetric equation of this line is
x - a1 y - a2 z - a3
d1 = d2 = d3
transpose of a matrix
The transpose of a matrix A of order m n is found by reflecting the matrix in its
main diagonal.
It is denoted AT or A .
The effect is to interchange the rows of the matrix with the columns and A T has
order n m.
208 GLOSSARY
vector equation of a plane

The vector equation of a plane containing the point P and perpendicular to the
vector n is (r - a) n = 0 where a = OP
vector form of the equation of a line

The vector equation of a line through the point P with direction d is

r = a + d where a = OP, d is a vector parallel to the required line and is a real
number.
vector product in component form

If p = ai + bj + ck and q = di + ej + fk
then the vector product is defined as the vector
p x q = (bf - ec)i - (af - dc)j + (ae - db)k
vector product in geometric form

The vector product of a and b is defined with
• magnitude of a x b = a b sin , where is the angle between a and

b, 0 180Æ
• direction perpendicular to both vectors a and b as determined by the right
hand rule.
vector quantity
A vector quantity is a quantity which has both direction and magnitude.
HINTS 209
Hints for activities

Topic 4: Further Ordinary Differential Equations
Save the fish
Hint 1: Let S = total mass of salt in the tank in and let w = the flow of water in
litres/minute. Then we can write down the differential equation
dS w
dt =- 5000 S
In standard linear form this becomes
dS w
dt + 5000 S =0
w
with P(t) = 5000 and f (t) = 0
Hint 2:
e - wt5000S = 0
We can now obtain the differential equation
d
dt
which we can integrate and solve for S to give
S = Ce-wt/5000 where C is an arbitrary constant.
20 5000
Hint 3: There is initially S(0) = 1000 = 100kg of salt in the tank.
Therefore C = 100.
12 5000
We need there to be S = 1000 = 60kg when the fish is put in.
We can therefore write down the equation
60 = 100e-wt/5000
210 ANSWERS: TOPIC 1
Answers to questions and activities

1 Vectors
Revision exercise (page 3)
Q1:
Q2: The equation will take the form y = mx + c

y2 - y1 15
The gradient m = x2 - x1 5 3
Thus y = 3x + c
By substituting the values for one of the points into this equation the value of c is obtained
and is -7
The equation of the line is y = 3x - 7
Q3: (ab + 2c)(ac - b)(c + b) = (a2 bc + 2ac2 - ab2 - 2bc)(c + b)

= a2 bc2 + 2ac3 - ab2 c - 2bc2 + a2 b2 c + 2abc2 - ab3 - 2b2 c
If a2 = 2, b2 = 1 and c2 = 1 then this simplifies to a(b + c)
Q4: x = 1, y = -1 and z = 3
Vector length exercise (page 7)
Q5: The vector to the point (-1, -2, 4) has length 4.58 to 2 decimal places.
The vector to the point (2, 5, 4) has length 6.71 to 2 decimal places.
The vector to the point (4, 2, -5) has length 6.71 to 2 decimal places.
The vector to the point (3, -5, -2) has length 6.16 to 2 decimal places.

Q6: The length is - 2)2 + 22 + 12 = 3
ANSWERS: TOPIC 1 211
Vector notation exercise (page 7)
Q7: -3i + 3j - 4k
1
Q8: 2
1
Direction ratios and direction cosines exercise (page 9)
Q9:
a) -3 : 2 : -1
b) 1 : 1 : 1
c) 2 : -1 : -1
Q10:
a) The modulus of the vector is 7. The direction cosines are 2 /7 , 3 /7 and 6 /7

b) The modulus of the vector is
3. The direction cosines are
1

- 31 and - 31
3
c) The modulus of the vector is 6. The direction cosines are 2 / , -2 / and 1 /
3 3 3
Basic skills exercise (page 12)
Q11: 3i - 3k
0
Q12: The vector addition gives 3
1
Q13: -5i + 2j - 5k
7

Q14: The vector is 5
-3
Q15: -2 (2i - j + 3k) = -4i + 2j - 6k

These are scalar multiples of each other. The vectors are therefore parallel.
Q16: Let the vector be ai + bj + k

Then ai + bj + k = (2i - 4j + k)
Thus a = 2, b = -4 and c =
But c = 3 and so = 3
The vector parallel to 2i - 4j + k with z = 3 is 3(2i - 4j + k) = 6i - 12j + 3k
Algebraic scalar product exercise (page 14)
Q17: a b = -6 + 6 - 12 = -12
Q18: 2 - 9 + 2 = -5
Q19: The scalar product is -16
Answers from page 14.
Q20: b + c = -4i + j + 2k
a (b + c) = -8 - 2 = -10
a b = -6 - 3 = -9 and a c = -2 + 1 = -1
so a b + a c = -10 as required.
b
a1
Q21: a b = b12 = a1b1 + a2 b2 + a3b3 and
a2
a3
b3
b1
a1

b a = b2 a2 = b1 a1 b2 a2 b3 a3
b3 a3
But a1 b1 + a2 b2 + a3 b3 = b1 a1 + b2 a2 + b3 a3 by the laws of algebra.
Thus a b = b a
Geometric scalar product exercise (page 15)
Q22: 22.627
Q23: -10.825
Q24: -6 = 4 3 cos
So cos = -1 /2 which gives = 120Æ
But the acute angle is always stated, so the angle is 60 Æ
Q25: let a = 3i - 4j and b = 12i + 5j

a b = 16
a has length = 5
b has length = 13
So cos = 16 / (5 13) = 16 /65
Thus = 75.75Æ

Q26: CA = i - 2j + 2k

CB = 3i - 4k
ab -5 -1
cos ACB = a b = 35 = 3 so angle ACB = 109.47 Æ
Perpendicular vectors exercise (page 16)
Q27: c = -3
Q28: c = 9
Q29: a b = (-4 2) + (-3 -2) + (2 1 ) = 0

Therefore the vectors are perpendicular.
Algebraic vector product exercise (page 17)
Q30: The answers are:

a) -6i + 3j - 3k
b) -7i - 7j - 7k
c) 0
Q31: b + c = 3i - j - 4k
a x (b + c) = 9i + 15j + 3k
a x b = 3i + 4j - k
a x c = 6i + 11j + 4k
Thus (a x b) + (a x c) = 9i + 15j + 3k as required.
Vector product properties exercise (page 19)
Q32: a x a = (1 - 1)i - (3 - 3)j + (3 - 3)k = 0

Q33: p x q = (15 - 4)i - (10 - 16)j + (2 - 12)k = 11i + 6j - 10k
q x p = (4 - 15)i - (16 - 10)j + (12 - 2)k = -11i - 6j + 10k
Thus p x q = -q x p
1
Q34: a + b = 0

-1
2
So (a + b)x c = 0

2
-3 5
But a x c = 6 and b x c = - 6

9 -7
2
So (a x c) + (b x c) = 0 as required.
2
Geometric vector product exercise (page 21)
Q35: Find the vector product. This will be perpendicular to both vectors.
It is -6i + 3j + 5k
Q36: a x b = 4 5 sin 45 Æ = 14.14 units
Q37: 12.728 square units
Q38: The area of the triangle is 1

axb
2

a x b = 2i + 4j - 5k and so the area is 1 /2 45 = 3.354 square units.

Q39: The area of the triangle is 3 2 or 4.243 square units.
Scalar triple product exercise (page 23)
Q40: -4
Q41: 9
Q42: 20 cubic units
Q43: Let a = i + 2j + 3k
b = 4i - j + 2k
c = 2i - 5j - 4k
a (b x c) = 0 and so the three points are coplanar; they lie in the same plane.
Q44: b x c = -i + 2j + k
c x a = -i - k
a x b = i - 2j + k
a ( b x c) = -1 - 1 = -2
b (c x a) = -2
c (a x b) = 1 - 2 - 1 = -2
They are all equal.
Equation of a line exercise (page 28)
Q45: r = 3i + 4j + 5k + (-2i - j + 3k)
Q46: r = 3i - 2j + 4k + (4i - 3j - k)
Q47: The parametric equations are:
a) x = -1 - 2, y = 3 - and z = 4
b) x = 2 - , y = -1 and z = 4
The symmetric equations are:

x+1 y-3 z-0
a) For line one -2 = -1 = 4
x-2 y+1 z-0
b) For line two -1 = 0 = 4
Note the form of the middle term:

The zero is used in cases such as this since y = -1 and is not involved.
Q48:
x-0 y+1 z-2
a) For line one -1 = 1 = -1
x-2 y+2 z+2
b) For line two -1 = 3 = -1
Q49: The symmetric equations are x = 2 + 3, y = 1 - , z = -3 -

If y = -1 then -1 = 1 - = 2 and the point is ( 8, -1, -5)
Q50: a = (3, 4, 5) and d = i + j + k so this gives

x = 3 + , y = 4 + , z = 5 +
y+2
Q51: Solving each for gives x-0
-1 = 1 = z-3
2
Equation from two points exercise (page 29)
Q52: Using the general equations

x - a1 y - a2 z - a3
r = p + (q - p) and b1 - a1 = b2 - a2 = b3 - a3
where p = (a1 , a2 , a3 ) = (3, -1, 6) and q = (b1 , b2 , b3 ) = (0, -3, -1) gives
parametric equations as x = 3 - 3, y = -1 - 2 and z = 6 - 7
x-3 y+1 z-6
symmetric equations as -3 = -2 = -7

Q53: AB = b - a = -2i - 2j + 2k
This gives a direction vector of i + j - k by taking a multiple of it.
a = i + 2j - k
Thus the parametric equations are
x = 1 + , y = 2 + , z = -1 - and the symmetric equations are
x-1 y-2 z+1
1 = 1 = -1
Intersection points and angles exercise (page 34)
Q54: Let be the constant for line L 1 so

y-2
x-2
1 = 3 = z-3
1 = so x = 2 + , y = 2 + 3 and z = 3 +
Let be the constant for line L 2 so
y-3
x-2
1 = 4 = z-4
2 = so x = 2 + , y = 3 + 4 and z = 4 + 2
If they intersect then

2+=2+
2 + 3 = 3 + 4
3 + = 4 + 2
These solve to give = = -1 and satisfy all equations.
The lines do meet at (1, -1, 2)
Q55: L1 has parametric equations x = -3 + , y = 2 + , z = 1 - 3

L2 has parametric equations x = -4 - 2, y = 1 + , z =
Trying to solve the set of equations
-3 + = -4 - 2
2+=1+
1 - 3 = gives no consistent solution for and
Thus the equations do not intersect.
Q56: L2 has parametric equations x = -3 - , y = 4 + , z = -

Thus solving the equations
-2 + 2 = -3 -
1 - 3 = 4 +
-1 + = - gives = -2 and = 3
Substituting in either set of equations gives the point of intersection as (-6, 7, -3)
The direction vectors are d1 = 2i - 3j + k and d2 = -i + j - k
d1 d2
cos = d½

d2 =
- 6
14 3
The angle of intersection is 157.79 Æ but the acute angle is given and this is 22.21 Æ
Q57: (1, 3, 2) and the angle is 72.95 Æ
Q58: (9, 3, 0) and the angle is 77.47 Æ
Q59: They intersect at the point (5, 9, 4) at an angle of 49.11 Æ
Equation of a plane exercise (page 38)
Q60: (r - a) n = 0 or r n = a n
Here n = 6i - j + 1 /3 k
and a = i + 2j + 3k
The equation is 6x - y + 1 /3 z = 6 - 2 + 1 = 5
That is, 18x - 3y + z = 15

Q61: AB and AC are two vectors in the plane. AB x AC will be perpendicular to both
vectors and therefore to the plane. This vector product will give a normal to the plane.

AB = i - j + 3k and AC = -i - 3j + 2k

so AB x AC = 7i - 5j - 4k
Thus a normal to the plane is n = (7, -5, -4)
A with position vector a = (1, 1, -1) is on the plane and the equation of the plane is
r n = a n which gives 7x - 5y - 4z = 6

Q62: AB = -2i - 2j + 2k

AC = -i + 3j - 2k

AB x AC = -2i - 6j - 8k
The equation is x + 3y + 4z = 1 + 3 2 + 4 1 since the plane goes through A
That is, x + 3y + 4z = 11
Q63: Let y = and z = then x = - 3 + 2 + 6

The parametric equation is a + b + c where a = -3i, b = 2i + j and c = 6i + k
Intersection and angles exercise (page 41)
Q64: Since it is parallel it takes the form x - 2y + z = d

A normal to the plane is n = i - 2j + k it passes through (1, -3, 2) so a = i - 3j + 2k
Since r n = a n then x - 2y + z = 1 + 6 + 2 = 9
The equation is x - 2y + z = 9
Q65: Let P1 3x - 6y +9 z = -12 and P2 = -x + 2y - 3z = 2

Let P1 have normal n and P 2 have normal m.
Then -3m = n so P 1 and P2 are either parallel or coincident.
The set of equations 3x - 6y + 9 z = -12 and -x + 2y - 3z = 2 have no solutions so P 1 and
P2 do not intersect.
Q66: A normal to the plane is n = 3i + j + k

a = i - j + 3k
Since r n = a n then 3x + y + z = 5
This is Cartesian form.
y+2
Q67: x 1- 0 = -1 = z-1
2 is the symmetric equations of the line of intersection of the
planes.
Q68: The normals are not multiples of each other. Therefore the planes do intersect.
- 13 + 5 - 14 + 7
The equation of the line in parametric form is x = , y = -3 ,z= -3
Q69: The normals are n = i + j - 4k and m = 2i - 3j + 4k

n m = 2 - 3 - 16 = -17

n = 18 and m = 29
cos =
- 17

18 29
so = 138.1Æ
Taking the acute angle gives the angle between the two planes as 41.9 Æ
Q70: The normals are n = 2i - j and m = i + j + k

n m = 1, n = 5, m = 3
and so cos =
1
which gives = 75.04Æ
15
Intersection of three planes exercise (page 43)
Q71: Using Gaussian elimination gives the point (2, -2, 3)
Q72: Use Gaussian elimination.

They intersect at the point (2, 1, 3)
Q73: Using Gaussian elimination the system reduces to a zero line on the bottom. Thus
the intersection is a line where z = -1 and x + y = 2. In symmetric form this equation of
the line is x +1 0 = y--12 = z +0 1
Q74: Using Gaussian elimination, the bottom row reduces to 0 = 42, which is
impossible.
The three planes do not intersect.
Q75: The normals are n = i + j - 4k and m = 2i - 3j + 4k

n m = 2 - 3 - 16 = -17

n = 18 and m = 29
cos =
- 17
so = 138.1Æ
18 29
Taking the acute angle gives the angle between the two planes as 41.9 Æ
Q76: The normals are n = 2i - j and m = i + j + k

n m = 1, n = 5, m = 3
and so cos =
1
15
which gives = 75.04Æ
Lines, points and planes exercise (page 46)
Q77: The general point has coordinates (1 + 2, -1 - , 3)

Substituting in the plane gives 3 + 6 - 2 - 2 - 3 = 5
and so = 4
The coordinates of the point of intersection are (9, -5, 12)
Q78: d = 2i + 3j + 6k
n = 10i + 2j - 11k
cos = nd
n d = - 40
105 and so = 112.4 Æ
Thus the acute angle is 67.6 Æ
The angle between the line and plane is 90 - 67.6 = 22.4 Æ
Q79: n = i + 2j - 2k and m = 5i - 2j - k
The vector product of these is along the direction of the line of intersection of the planes
since it is perpendicular to both normals.
n x m = -6i - 9j - 12k
d = 2i + 3j + 4k
But n x m is a multiple of d and so it is parallel to it.
Q80: n = 2i + j - k and m = i + 2j + k
n x m = 3i - 3j + 3k thus a normal to the plane is i - j + k
2i + j - k is also on this plane
The equation of the plane is r n = a n
That is, x - y + z = 0
Review exercise in vectors (page 49)
Q81: a x b = -i - j - 3k
b x c = 6i - j + 4k
a b x c = -3 - 4 = -7
Q82: The vector is -i + 2j + 6k

In parametric form the equations are
x = 2 - , y = -3 + 2 and z = 1 + 6
Q83: The normal is 2i + j + 3k

So the equation has the form 2x + y + 3z = c
The point (1, 2, 3) is on the plane thus 2 + 2 + 9 = c
The equation is 2x + y + 3z = 13
1+ 12 - 13
Q84: x = , y = - 5 and z = 5
Q85: Using Gaussian elimination gives the point as (3, -1, 2)
Advanced review exercise in vectors (page 50)
Q86: a) u v = 9 and u x v = -6i -3j + 6k

b) i) The angle is 45 Æ
ii) Q is (1, -1, 2)
x-1 y+1 z-2
iii) L is the line with equations -6 = -3 = 6
It meets the xy-plane where z = 0. The point R = (3, 0, 0)
c) a x b = c so a is perpendicular to c and b is perpendicular to c
b x c = a so a is perpendicular to b
Since the vectors are mutually perpendicular
a x b = c gives a b = c and b x c = a gives b c = a

so a
c b = b1
=
Therefore b = 1 That is, b is a unit vector
Q87: a) p q = -9 and p x q = i + j + k
b) i) The point is (1, -1, 0)
1 1
x-0 y+ 4 z- 4
b)ii) The equation of the line is 1 = 3 = 1
- 4 - 4
b)iii) It meets the yz-plane at (0, -1 /4 , 1 /4 )
Q88: a) The equation is x - 3y - 2z = -5

b) The angle QPR is 11.74 Æ
c) The angle is 138.11 Æ i.e. the acute angle is 41.89
The angle between the line and the plane is 90 - 41.89 Æ = 48.11Æ
Set review exercise (page 51)
Q89: The answers are only available on the web.
Q90: The answers are only available on the web.
Q91: The answer is available on the web.
2 Matrix algebra
Q1: The element a 23 is 1
Q2: x = 2, y = -1 and z = 3
Q3:
1. (2 -3)
2. (-3, 2)
Q4: (-a2 b3 - a3 b2 )i - (-a1 b3 + a3 b1 )j + (-a1 b2 - b1 a2 )k
Adding matrices exercise (page 57)

-2 -5

Q5:
4 2
3
Q6: 0
1
-4 11 -3

Q7: 2 -1 1
-1 0 0
1 0

Q8:
0 1
Subtracting matrices exercise (page 58)

-3 -3

Q9:
-5 -2
-1

Q10: 0
1
2 4 -3

Q11: 2 1 -4
1 0 4
-5 6

Q12:
-2 -3
Multiplication of a matrix by a scalar exercise (page 59)
Q13:
4 2 -6

Q14: = -1
-4 0 4

Q15: A = -8 -4 8
- 12 8 - 16
Multiplying matrices exercise (page 60)
Q16: The answer in this case is the 1 x 1 matrix (10)

Q17: The answer is the 3 x 4 matrix
-4 -2 -1 -3

6 0 0 -2
8 -4 3 -3
Q18: The following products can be formed: AB, BA, AC, AD, CB and CD.
The product matrices are
0
8
-3 1 7 -3
AB =
7
2
7
, BA = -1
AC =
-4 3
-4 4
7
9
9
1
-6 3
11
9 -5 4 4

CB = 4 0 , CD =
5 1 -7
AD = 3 3 1 -5
9 -1 4 -2
-7 - 10 - 11 3 -5 0
-4 -1 -5 5

6 0 9 -5
Q19: 2 0 3 1
-5 4 2 3
Q20:
1. These are not equal. There is no value for y which satisfies the equations.
2. Yes, these are equal with x = 1 and y = 6
3. These are not equal as the matrices are not of the same order.
4. Yes, these are equal with x = 3 and y = 0
2 x 2 determinant exercise (page 62)
Q21:
a) 5
b) 8
c) -1
d) 4
e) 0
3 x 3 determinant exercise (page 64)
Q22:
a) 2
b) 4
c) 1
d) 20
2 x 2 inverse exercise (page 65)
Q23:
1 -2

1
- 2

1 5 5
a) 5 = 3 1
2

3 -1 5 - 5
1 1
1 2 2 2
b) 4 = 1 3
1 1 3
0

4 4
-1 0

c) -1 =
1 3
-2
-1

1
-3 1
- 1

1 4 2
d) 4 = 1 1
3

1 2 4 2
1 1
1 -2 4 - 6
e) 12 = 1 1
3 2 4 6

1 0

Q24: The 2 x 2 identity matrix is

0 1
3 x 3 inverse exercise (page 67)
Q25:
1 4 6

1. 0 2 3
-1 1 2
-1 0 0

2. -2 -1 0
2
-3 -4 1
1 4
-
3. 9
1
9
9
5
9
2
9
9
4 2 1

- 9 - 9 9
6 8 9
-
4. -
11
2
11
11
1
11
11
3
11

1 5 7
- 11 - 11 11
-5 2 2
5. -
3
11
6
3
5
6
3
1
3

4 1 1
3 - 3 - 3
Properties exercise (page 75)
- 2 + ( - 2) 3+1
-4 4

Q26: A + B = 0+2 -2+1 = 2 -1
1+1 -1+2 2 1
- 2 + ( - 2) 1+3
-4 4

and B + A = 2+0 1 + ( - 2) 2 - 1 as required.
1 + 1 2 + ( - 1) 2 1
6 -5

-1 -1

Q27: AB = and BA =
2 -2 3 5
0 1 0 -4

Q28: AB = and
-9 2 2 13

-4
-1

(AB)C =
0 1 0
-9 2 2 13
1
-2

=
-3
20
1
1

1
BC = 3 and A(BC) = 2 0
-1 2
-1
3
= 3
-3
20
as required.
-5 5
0 0

Q29: B + C = and
1 -3
2 -1

-1 3

A(B + C) = -3 0 01 0
-3
= 0 0
1 1 1 -3
-5
0 4 3

AB = 6 and AC =
3 -6 -3
-1 -3 2 0
-5 0
4 3
-1 3

so AB + AC = 6 3 + -6 -3 = 0 as required.
-1 -3 2 0 1 -3

-1 3

Q30: AT =
-1 2
3 1 2
4
and transposing this matrix gives 2 1
4 2
So (AT )T = A
1 3 -2

Q31: A + B = 0 1 2 and
-5 2 3
1 0

-5
(A + B)T = 3 1 2
-2 2 3
2 -2 -3
-1 2 -2

But AT = 1 1 -1 T
and B = 2 0 3
0 1 2 -2 1 1
2 -2 -3
-1 2 -2
1 0 -5

so AT + BT = 1 1 -1 + 2 0 3 = 3 1 2
0 1 2 -2 1 1 -2 2 3
as required.
8 -2

8 1

Q32: AB = so ABT =
1 -4 -2 -4
3 1

3 1

But BT = and AT =
0 2 -1 -2
3 1

3 1

8 1

which gives B T AT = = as required.

0 2 -1 -2 -2 -4
2 -3

Q33: AB = and det AB = -24

- 10 3
3 3

1
Thus (AB)-1 = 24 10 2
3 0

-1 3

1 1
det A = 3 and det B = -8 so A -1 = 3 and B-1 = 8
2 1 2 2
-1 3

3 0

3 3

1 1 1
Thus B-1 A-1 = 8 times 3 = 24 as required.
2 2 2 1 10 2
Q34: The point under a rotation of -90 Æ will go to the point (2, -5)
a b

a b

5
2

Let the matrix R = then =

c d c d 2 -5
This gives two equations to solve:
5a + 2b = 2

5c + 2d = -5 . The simplest solutions are a = 0, b = 1, c = -1 and d = 0
0 1

The matrix R such that RA = B is

-1 0

5

(
0 1
-1 0
2
=
2
-5
)

cos sin

Q35: R-1 =
- sin cos
Rotation exercise (page 78)

6 428
Q36:
4 866
cos - sin

2
1 537
Q37: =
sin cos 1 1 624
This gives the two equations
2 cos Æ - sin Æ = 1.537 and
2 sin Æ + cos Æ = 1.624
Rearrange the first to give sin Æ = 2 cos Æ - 1.537
Substitute this into the second to give cos = 0.9396
Thus = 20Æ
Q38: RA = B so A = R-1 RA = R-1 B

cos sin

As shown earlier R-1 = and the point is (-1, -2)

- sin cos
Reflection exercise (page 80)

0.6

Q39:
- 3.1
cos 2 sin 2

-1

-2

Q40: =
sin 2 - cos 2 -2 -1
This gives the two equations
-cos 2 Æ - 2sin Æ = -2 and
-sin 2 Æ + 2cos Æ = -1
Rearrange the first to give cos 2 Æ = 2 - 2 sin 2 Æ
Substitute this into the second to give -sin 2 + 4 - 4 sin 2 = -1
Thus solving for gives = 45Æ
Q41: The line y = -4x makes an angle of -76 Æ with the x-axis.
If S is the rotation matrix and A is the original point with image B

Then SA = B so A = S -1 SA = S-1 B
- cos ( - 152) - sin ( - 152)

cos ( - 152) sin ( - 152)

S-1 = - 1 =
- sin ( - 152) cos ( - 152) sin ( - 152) - cos ( - 152)
and the point is (0.9, 0.5)
Scaling exercise (page 82)

-3 0

Q42: The matrix is and the image of the point is (-9, 1)

0 1
a 0

Q43: Let the matrix be

a

0 b

Then
0
0
b
-1
-2
=
-2
-1
This gives two simple equations which solve to give a = 2 and b = 0.5
0
The matrix is
0 0.5
a 0


0 b
Then 3a = -1 and 4b = 0

Thus a = -1 /3 and b = 0
- 1
0

The matrix is 3
0 0
General transformation exercise (page 83)
Q45: Answer:
b
Let the matrix be
a

6
c d
Thus
b
c d
-3
2
=
7
gives the equations
1) -3a + 2b = 6 and
a
2) -3c + 2d = 7

5

Also
b
c d
4
=
-8
- 13
gives the equations
3) 5a + 4b = -8 and
4) 5c + 4d = -13
Solving the set of equations, 1) and 3) gives a = -20 / 3
11 and b = /11
Similarly the set of equations 2) and 4) gives c

- 20 3
= -27 /11 and d =-2 /11
The matrix is 11 11
27 2
- -
11
a 11

c d
This gives equations 2a - 3b = -8 and -a - 4b = -7
These solve to give a = -1 and b = 2
Similarly the equations 2c - 3d = 16 and - c - 4d = 3 solve to give c = 5 and d = -2

-1 2
The matrix is
5 -2
Q47: The point is (18, -16)

-1 0

Q48: The matrix is

2 3
a b


c d
Take the two points on the first line and find two sets of equations to solve. The two sets
here are
1) 2a + 2b = 0 and 3a + 4b = -2
2
2) 2c + 2d = 8 and 3c + 4d = 13
-2

The matrix is
3 1
Review exercise (page 88)
Q50:
4 17

a)

- 14 7
-3 2
b)
11
-4 3
-9
c)
15 9
d) 36
Q51:
-8 2

a)
- 14 -4
- 1 2

b) 3 3
1 1
1
6 6
-3

c)
-9 -7
d) -31
Advanced review exercise (page 88)
Q52: 2A2 = A + I rearranges to give A (2A - 1) = I

Thus 2A - 1 is the inverse matrix and so A is invertible. A -1 = 2A - 1
Q53: A2 - 4I = 0
2 0

-2 0

So (A - 2I)(A + 2I) = 0 thus A = 2I or A = -2I i.e. A = or

0 2 0 -2
cos ( - 60) - sin ( - 60)

cos - 60
Q54: The rotation matrix to use is followed by the scaling
-2 0

sin ( - 60)
matrix . The reflection is straightforward from there and the image of the
0 3
point has coordinates (3.196,9.696) to one decimal place.
Q55: Since A is orthogonal A T = A-1 and AAT = I

Thus (AAT )-1 + A(3AT - A-1 ) = I-1 + A(2A-1 ) = I + 2I = 3I
Q56: The answer is only available on the web.
3 Further sequences and series

Revision Exercise (page 92)
Q1: 32
Q2: 1+ ( - 1)
1! x + ( - 1)( - 2) 2
2! x + ( - 1)( - 2)( - 3) 3
3! x +
Q3: f (x) = 2e2x

f (x) = 4e2x
f (x) = 8e2x
Q4:
u0 5
u1 12.5
u2 17.75
u3 21.425
u4 23.9975
As n , un 30, L = 30
Q5: S10 = 2.718282 e
Exercise 1 (page 95)
Q6:
x2 x4 x6 x8 x10
a) 1 - 2! + 4! - 6! + 8! - 10! +
b) 1 + nx + n(n2!- 1) x2 + n(n - 1)(n
3!
- 2) 3
x + n(n - 1)(n4!- 2)(n - 3) x4 +
c) x - 12 x2 + 13 x3 - 14 x4 + 15 x5 - 16 x6 +
Q7: f(4) (0) = (4 3 2 1)a4 = 4!a4
Q8: We obtain the following results

f (x) = lnx f (0) = ln0
1 1
f (1) (x) = f (1) (0) =
x 0
-2 -2
f (2) (x) = 2 f (2) (0) =
x 0
f (x) = ln x and its derivatives are undefined at x = 0 therefore we cannot find a Maclaurin
series expansion for ln x

f (x) = x = x12 f (0) = 0
1 1
f (1) (x) = f (1) (0) =
2x12 0
1 1
f (2) (x) = - 32 f (2) (0) = -
4x
0

The derivatives of f (x) = x are undefined at x = 0 therefore we cannot find a Maclaurin
series expansion for x

1 1
f (x) = cotx = f (0) =
tanx 0
1 1
f (1) (x) = - cosec2 x = - 2 f (1) (0) = -
sin x 0
(2) cosx 2 2
f (x) = 2cotxcosec x = 2 3 f (2) (0) =
sin x 0
f (x) = cot x and its derivatives are undefined at x = 0 therefore we cannot find a Maclaurin
series expansion for cot x
(5x)2 (5x)3 (5x)4 (5x)5

Q11: a) e5x = 1 + 5x
1! + 2! + 3! + 4! + 5! +
3 5 7 9
(2x) (2x) (2x) (2x)
b) sin2x = 2x
1! - 3! + 5! - 7! + 8! -
(3x)2 (3x)4 (3x)6
c) cos3x = 1 - 2! + 4! - 6! +
x2 x4 x6 x8 x10
d) cos( - x) = 1 - 2! + 4! - 6! + 8! - 10! +
e)
Q12:
a)
ix (ix)2 (ix)3 (ix)4 (ix)5 (ix)6
eix = 1 + + + + + + +
1! 2! 3! 4! 5! 6!
x2 x3 x4 x5 x6
= 1 + ix - -i + +i - +
2! 3! 4! 5! 6!
b) eix = cos x + i sin x
Q13:
a) 1 - 12 x - 18 x2 - 1 3
16 x - 5 4
128 x
b) 1 - 5x + 15x2 - 35x3 + 70x4
c) 1 + 32 x + 38 x2 - 1 3
16 x + 3 4
128 x
Q14:
a)16 + 32x + 24x2 + 8x3 + x4
b) 1 - 6x + 12x2 - 8x3
Q15:
1
a) 3 - 29 x + 4 2 8 3
27 x - 81 x
1 3 3 2 5 3
b) - 8 - 16 x - 16 x - 32 x
Q16: a), b) and i) all diverge.

c) converges to 20
d) converges to 8
e) converges to 5
f) converges to 4
g) converges to 31 /3
h) converges to 2.5
Q17: -1 a 1
Q18:
a) converges to 12.5
b) converges to 0.75
c) converges to -0.25
d) converges to -5
Q19: Changing b will not change whether a recurrence relation converges or not. It will
change the value of the limit and may also alter the rate at which the recurrence relation
converges to this limit.
Q20: In each case the recurrence relation converges to 7.5
Q21: It does not seem to matter what value we choose for u 0 . The recurrence relation
always converges to the same limit.
Q22:
a) -4 /3 b) 20 c) 31 /3
Q23: a) un + 1 = -0.3un + 7
b) un + 1 = 2un + 5
c) un + 1 = 0.4un + 6
d) un + 1 = 0.8un + 4
e) un + 1 = -0.2un + 3.6
f) un + 1 = 1.5un + 15
Q24: a), c), d) and e)
Staircase and Cobweb diagrams (page 109)
Q25: a)
b)
c)
d)
Testing Convergence (page 109)
Q26: This time the sequence converges to 0.523976 This is the root that we found
already. We have not found a more accurate value for the root near 2.1
Q27:
x0 = -2.7 x0 = 0.5 x0 = 2.1
xn + 1 = x3n +3 Diverges Converges to Converges to

6
0.523976 0.523976
3 Converges to Converges to Converges to
xn + 1 = 6 - x2n
0.523976 0.523976 0.523976
xn + 1 = x3n - 5xn + 3 Diverges Diverges Diverges
2x3n - 3 Converges to Converges to Converges to
xn + 1 = 3x2n - 6 -2.669079 0.523976 2.145103
2x3n - 3
Q28: Of the iterative schemes given, only x n + 1 = 3x2n - 6
gave values for all three roots.
Q29: xn + 1 = e - xn gives the solution 0.567143 (to 6 decimal places)

Q30: xn + 1 = 2 - sin xn gives the solution 1.106060 (to 6 decimal places)
Q31:
x = 1 /2 - x3 gives the solution 0.423854

x = 12 - x
1
3
does not give the solution. It oscillates between the values -0.845263 and
1.103915
Q32:
x = ex - 3 has only one solution, -2.947531 (to 6 decimal places)
x = ln (x + 3) has only one solution, 1.505241 (to 6 decimal places)
Q33: a) x0 = 1.7
b) 1.749031
Q34: a) x0 = -1.8
b) The iterative scheme xn + 1 = (x - 4)1/3 will give the solution -1796322 (to 6 decimal
places)
Q35:
a) x0 = 2.2
b) The iterative scheme xn + 1 = 3 - ln xn will give the solution 2.207040 (to 6 decimal
places)
Q36:
a) x0 = 0.2 or x0 = 1.6
b) xn + 1 = (5x - 1)1/4 will give the solution

(to
decimal places)1.637306 (to 6
decimal places) and xn + 1 = 1 /5 (x4 + 1) will give the solution 0.200322 (to 6 decimal
places)
Q37:
a)
g (x) = (4x + 5)13
4
g’ (x) = (4x + 5) - 23
3
g’ (2) = 0.24 approx
g
(2) 1 therefore the iterative process will converge to the solution.
b)
1 3
g (x) = x -5
4
3
g ’ (x) = x2
4
g ’ (2) = 3
g
(2) 1 therefore the iterative process will diverge.
Q38:
a) g (-1.9) = 24.9, g (-0.3) = 0.039 and g (2.1) = -24.9 Therefore we should only
expect the given iterative process to converge to the solution near -0.3
b) -0.254102 (to 6 decimal places)
Q39:
a)
For (A) g ’ ( - 1.8) = - 3.88 approx
For (B) g ’ ( - 1.8) = 0.27 approx
For (C) g ’ ( - 1.8) = 0.07 approx
(A) will not converge to the required root but (B) and (C) will.
b) -1.752332 (to 6 decimal places).
Q40:
a)
2
g (x) = = 2(x + 4) - 1
x+4
g ’ (x) = - 2(x + 4) - 2
-2
=
(x + 4)2
For close to zero g ’ () = -2
( + 4)2
0
Therefore this iterative process is first order.
b)
g (x) = 1 - sinx
g ’ (x) = - cosx
We are given that 0 < < 1 radians therefore g ’ () 0
3 3
(cosx = 0 when x = or radians and , > 1)
2 2 2 2
Q41:
1
g (x) = x 3 - 5x2
2
3 5
= x - x3
2 2
3 15 2
g ’ (x) = - x
2 2
At = , g ’
1 1 3 15 1
= - =0
5 5 2 2 5
Therefore this iterative process has second order convergence.
Review exercise in further sequences and series (page 122)
9x2
Q42: 1 + 3x + 2
Q43: -2.104
Advanced review exercise in further sequences and series (page 122)
x2 x3
Q44: 1 + x - 2 + 2 - 58 x4
1 2 1 2 4 3
Q45: 32
+ 33
x + 33
x + 35
x
Q46:
a)
g’ (x) = - 2e - x
g’ (1 0.736 1
Therefore the iterative process will converge to the solution near x = 1
= 0.853 (to 3 decimal places).
b) g’ -2
e «
0
Therfore the convergence is first order.
Q47:
a)
x0 2
x1 1.83333
x2 1.81726
x3 1.81712
x4 1.81712
To three decimal places = 1.817
b)
2 4
g’ (x) = -
3 x3
2 4
g’ (613 ) = - = 0
3 6
Therefore the convergence is second order.
Set review exercise in further sequences and series (page 123)
Q48: This answer is only available on the course web site.
4 Further Ordinary Differential Equations

Q1: f (x) = 2x sin x + x2 cos x
Q2: dy / = 4e4x
dx
Q3: 3 ln (x) + C
Q4: x3
Q5: 7i
Q6: -2 3i
Q7: y = e3x + C e2x
Q8: y = x3 + C /x
Q9: y = 12 ex2 + Ce - x2

sinx C
Q10: y = x2
+ x2
2
Q11: y = - x + Cx
sinx C
Q12: y = - cosx + x + x
Q13:

2
Q14: y = + 15 e(1 - 5x)

5
Q15: y = esin(x) x2 - 2
Q16: y = e1/x (x - 1)
3
Q17: y = 2 - e - x
Q18: 3.74 kg
Q19: 4 hours 25 minutes
Q20: 13.5 g/litre
Q21:
a) 731.7 kg
b) Approximately 103 minutes.
c) As t then M = 1500 - 1400e -0.01t 1500 since e-0.01t 0

Q22: v (t) = mg
k 1 - e - ktm

Q23: a) c (t) = G
100Vk 1 - e - kt + 100e - kt
b) As t , e-kt 0. Therefore the limit of concentration in the blood is 100Vk

G
Save the fish (page 136)
The formula for the correct answer is

5000 ln0.6
w=- t
Q24: y = Ae-2x + B e3x
Q25: y = A e-x + B e-5x
Q26: y = (A + Bx) e-3x
Q27: y = e-2x (A cos 3x + B sin 3x)
Q28: y = (A + Bx)e5x
Q29: y = e-2x (A cos 2x + B sin 2x)
Q30: y = A e-3x + Bex/2
Q31: y = (A + Bx)ex/2
Q32: y = (A + Bx)e - 2x3

Q33: y = e - x2 A cos( 32 x) + B sin( 32 x)

Q34: y = e - 2x A cos 2x B sin 2x
Q35: y = Ae - 3.41x + Be - 0.59x
Q36: y = -2e-3x + 2e2x
Q37: y = 3xe2x
Q38: y = e-x (2 cos x + 5 sin x)
Q39:

Q40: y = (5 + 2x)e4x - 4
Q41: y = e3x - cos2x + 4 sin2x

Q42: x (t) = 40 cos 4t - 3 sin 4t

When t = 5, x = 13.6 cm
Q43: x = 0.25e-t - 0.75te-t
Q44: x = e-0.05t (0.7 cos 0.35t + 0.1 sin 0.35t)
Q45: a) When y = e4x then

2
dy
dx = 4e4x and ddxy2 = 16e4x
d2 y
Therefore dx2
- 6 dy 4x 4x
dx + 9y = 16e - 24e + 9e
4x
= e4x
Hence e4x is a particular integral for the equation and we can write y p = e4x
b) y = e4x + A e3x + Bx e3x
Q46: a) When y = e3x then
dy d2 y
dx = 3e3x and dx2
= 9e3x
d2 y
Therefore dx2
+ 2 dy
dx + 5y = 9e
4x
+ 6e4x + 5e4x = 20e4x
Hence e3x is a particular integral for the equation and we can write y p = e3x
b) y = e3x + e-x ( A cos 2x + B sin 2x)
Q47:
a) yc = A e3x + Bx e3x
b) yp = e4x
c) y = e4x + A e3x + bx e3x
Q48:
a) yc = A ex + B e4x
b) yp = 2x2 + 5x + 6
c) y = 2x2 + 5x + 6 + A ex + B e4x
Q49:
a) yc = A e-2x + Bx e-2x
b) yp = 3 sin x - 4 cos x
c)y = 3 sin x - 4 cos x + A e2x + Bx e2x
Q50:
a) yc = e-x (A cos 3x + B sin 3x)
b) yp = 1 - x
c)y = 1 - x + e-x (A cos 3x + B sin 3x)
Q51:
a) yc + A e3x + Bx e3x
b) yp = 2
c)y = 2 + A e3x + Bx e3x
Q52:
a) yc = e-x ( A cos 4x + B sin 4x)
b) yp 1 /2 ex
c)y = 1 /2 ex + e-x ( A cos 4x + B sin 4x)
Q53:
a) yc = Ae-5x + Be2x
b) yp = 3x e2x
c)y = 3x e2x + A e-5x + B e2x
Q54:
a) y = A ex + B e4x
b) yp = cos 3x + 3 sin 3x
c) y = cos 3x + 3 sin 3x + A ex + B e4x
Q55:
a) yc = A e-2x + B ex
b) yp = sin x - 6 cos x
c)y = sin x - 6 cos x + A e-2x + B ex
Q56:
a) y = 3 - 6x + A e-x + B e2x
b) y = 2 e3x + A e-x + B e2x
c) y = 3 - 6x + 2e3x + A e-x + B e2x
Review exercise in further ordinary differential equations (page 160)
Q57: y = 1 /4 e-x + Ce-5x
Q58: y = e - x sin x + C
2
Advanced review exercise in further ordinary differential equations (page 160)
4
Q59: y = x4 + x2
Q60:
Ê

tanx dx
a) I (x) = e so we need to find tanx dx first.
sinx
tanx dx = dx
cosx
du
Let u cosx then = - sinx
sinx
dx
sinx dx

We now have dx = du
cosx u du
1

= - du
u
= - ln u
= - ln(cosx)
= ln(cosx) - 1
Ê
tanx dx -1
Hence I (x) = e = eln(cosx) = (cosx) - 1 = 1
b) y = sin x cos x +
2 cos x cosx
Q61: x = e-t (A cos 2t + B sin 2t)
Q62: y = (3 + 5x)e5x
Q63:
a) y = 4 cos x - 8 sin x + A ex + Be3x

b) y = 8 cos x + 4 sin x + A ex + Be3x
c) y = 12 cos x - 4 sin x + A ex + Be3x
Set review exercise in further ordinary differential equations (page 161)
5 Further number theory and further methods of proof

Q1: In the first topic on number theory the following diagram and definitions were given
to explain the connection.
A conjecture is a precise, unambiguous statement for which a convincing argument is

needed.
A proof is a logically rigorous and complete argument that a mathematical statement is
true.
A theorem is a mathematical conjecture which has been proved.
Q2: 16.25
S4 is the notation for the sum of the first four terms.
This is the answer to (1 + 3) + (2 + 1.5) + (3 + 1) + (4 + 0.75)
The sigma sign has a starting value (the lower limit) and an ending value (the higher
limit) to indicate which values of r should be used.
Q3: Using simple algebraic manipulation
5x2 5x2
4x3 + 12xy + + 15 = 4x3 +
y y
12xy + 15 = 0
xy = - 45
Q4: The terms mean:
1. Implies (from left to right).
2. Implied by (from right to left).
3. If and only if (implication in both directions).
Symbol exercise (page 166)
Q5:
1. The correct symbol is . There are many clear, colourless liquids which are
certainly not drinking water.
2. The correct symbol is . x = 3 - 4i is a complex number which does not imply that
x2 = -2
3. The correct symbol is . R may be a square singular matrix.
4. The correct symbol is . 15 is an odd number which is not prime.
Q6:
1. If P is ’x + 8 = 4’ and Q is ’x is a negative number’ then P Q

2. If P is ’ x 0’ and Q is ’x + 1 is positive’ then P Q
3. Careful: if Q is sin = 0 and P is = then P Q
Equivalence exercise (page 168)
Q7: The answers are:
1. Equivalence is correct here with the assumption that step parents are not being
considered.
P = Joe is Amy’s father Q = Amy is Joe’s P Q
daughter
1 1 1
1 0 0
0 0 1
0 1 0
2. Equivalence is incorrect here. The correct sign is If x is an even number, this
does not imply that it is divisible by 6: for example, 4 is even.
3. Equivalence is incorrect here. None of the implication signs necessarily apply here.
I study Maths does not imply that I am good at Maths. Similarly I am good at Maths
does not imply that I study Maths.
4. Equivalence is incorrect. In this last case the correct sign is
Take x = 9; x here is odd and does not imply that x is a prime since 9 is not a prime.
However any prime greater than 2 is odd and so the implication works from Q to
P where Q is the proposition that x is a prime 2 and P is the proposition that x is
odd.
Q8: The truth table is simply

P P
1 0
0 1
Negation exercise (page 169)
Q9:
1. P:x is irrational.
2. P: I dislike Maths.
3. P: x is not a negative real. Note that x could be zero and this rules out putting the
word positive into the answer.
4. P: x 0, x .
Converse exercise (page 170)
Q10: ’If I am at home then it is Thursday.’ The proposition P is ’it is Thursday’ and the
proposition Q is ’I will be at home’. So S : P Q and the converse is Q P
Q11:
1. P: I am over fifty. Q: I am over forty.

So Q P states ’I am over forty I am over fifty’!
2. P: It rains. Q: It is cloudy.
So Q P, meaning if Q then P reads ’If it is cloudy then it rains’.
3. P: I work. Q: I am at Heriot Watt University.
So Q P meaning Q only if P reads ’I am at Heriot Watt University only if I am
working’.
4. P: It is dry. Q: Joyce cycles to school.
So Q P, meaning P if Q reads ’It is dry if Joyce cycles to school’. Remember in
this case that the phrase Q if P means P Q and this is why P and Q are defined
as shown.
Contradiction exercise (page 173)
Q12: Assume that the conjecture is false and so

2 + 3 10

2 + 2 6 + 3 10 2 6 5

24 25 which is a contradiction.
2 + 3 10 is untrue and the conjecture is true. That is,
2 + 3 10
Thus the assumption that
Q13: Assume the conjecture is false and make a new assumption that m 2 = 14 but m is
rational.
p
Therefore m = q where p and q are relatively prime.
So 14q2 = p2 p2 is even since 14q2 = 2 (7q2)
By the recently-proved conjecture (see the example) that ’if m 2 is even then m is even’
then p is even and p = 2k for some integer k

Thus 14q2 = 4k2 7q2 = 2k2
7q2 = 2m2 q2 is even and so q is even.
Thus both p and q are even but both were assumed to be relatively prime. This is a
contradiction and the assumption that m is rational is false.
The original conjecture that if m 2 = 14 then m is not a rational number is therefore true.
Q14: Assume that the conjecture is untrue and make the assumption that x + y is
irrational but both x and y are rational.
a c
Thus x = b and y = d for some integers a, b, c and d
a c ad + bc
x+y= b + d = bd
This is rational and contradicts the assumption that if x + y is irrational then both x and y
are rational.
The conjecture that if x + y is irrational then at least one of x, y is irrational is true.
Q15: Assume that the conjecture is untrue and make a new assumption that n 2 is a
multiple of 5 but n is not a multiple of 5. There are therefore four possibilities to consider.
n takes one of these forms:
a) 5k + 1
b) 5k + 2
c) 5k + 3
d) 5k + 4
Taking each in turn:
a) Suppose n = 5k + 1 then
n2 = (5k + 1)2 = 25k2 + 10k + 1 = 5(5k2 + 2k) + 1
Which is of the form 5t + 1 for another integer t
Thus n2 is not a multiple of 5 which is false and the new assumption is contradicted.
n is not of the form 5k + 1
b) Suppose n = 5k + 2
then n2 = (5k + 2)2 = 25k2 + 20k + 4 = 5(5k2 + 4k) + 4
which is of the form 5t + 4 for another integer t
Thus n2 is not a multiple of 5 which is false and the new assumption is contradicted.
n is not of the form 5k + 2
c) Suppose n = 5k + 3
Then n2 = (5k + 3)2 = 25k2 + 30k + 9 = 5(5k2 + 6k + 1) + 4 which is of the form 5t
+ 4 for another integer t. Thus n 2 is not a multiple of 5 which is false and the new
assumption is contradicted. n is not of the form 5k + 3
d) Suppose n = 5k + 4
Then n2 = (5k + 4)2 = 25k2 + 40k + 16 = 5(5k2 + 8k + 3) + 1 which is of the form 5t
+ 1 for another integer t. Thus n 2 is not a multiple of 5 which is false and the new
assumption is contradicted. n is not of the form 5k + 4
The assumption that n is not a multiple of five is contradicted. The conjecture is therefore
true and for any integer n, if n 2 is a multiple of 5 then n itself is a multiple of 5
Q16: Assume that the statement is false: that is, assume that there is a largest integer
say, P
Since 2 is a positive integer, P 2
But P2 is a positive integer, thus P P2
Divide by P to give 1 P
But P 2 and 1 P cannot both be true. The statements contradict.
Therefore the assumption that P is the largest integer is false.
The original conjecture that there is no largest positive integer must then be true.
Q17: Assume that the conjecture is false. Assume that nx y for every choice of
positive integer n
Thus n y /x
all positive integers are bounded above by y/x
a largest integer exists.
By the previous question ’There is no largest positive integer.’ The assumption is
contradicted and the original conjecture is true.
Q18: Assume that

2 is rational.

Therefore 2 = pq where p and q are positive integers with no common factors.
(Otherwise the fraction could be reduced.)
So 2q2 = p2 p2 is even.
Now if p is odd, then p 2 is odd.
But here p2 is even: so p is even
and p = 2m for some integer m
Thus 2q2 = 4m2 q2 = 2m2
q2 = 2m2 q2 is even and so q is even.

Thus both p and q are even but both were assumed to have no common factors. This is
a contradiction and the assumption that 2 is rational is false.

The original conjecture that 2 is not a rational number is therefore true.
Q19: Suppose that the conjecture is untrue and assume that there are a finite number
of primes.
So there is a collection of all primes P1 , P2 , P3 , ..., Pk
Let N = (P1 P2 ... Pk) + 1
Then N is obviously larger than all P k
Since P1 , ..., Pk is the collection of all primes then N is composite.
If N is composite then N has a prime divisor Q
But since all the primes P1 , P2 , P3 , ..., Pk leave remainder 1 when dividing N, none of
these is equal to Q
This is impossible.
The assumption that there is a finite number of primes is false.
The conjecture that there is an infinite number of primes is true.
Induction exercise (page 176)
Q20: Check when r = 1

1(1 + 1)(2 + 1)
Then 1 = 6 = 1. The result is true for r = 1
Assume true for r = k
k(k + 1)(2k + 1)
Then 1 + 4 + 9 + ... + (k - 1)2 + k2 = 6
Consider r = k + 1
Then 1 + 4 + 9 + ... + (k - 1)2 + k2 + (k + 1)2
k(k + 1)(2k + 1)
= + (k + 1)2
6
(k + 1)(2k2 + 7k + 6)
=
6
(k + 1)(k + 2)(2k + 3)
=
6
(k + 1)(k + 2)(2(k + 1) + 1)
=
6
So the result is true for r = k + 1 if it is true for r = k. But it is true for r = 1 and so by the
principal of mathematical induction the conjecture is true for all r = n
Q21: Check for n = 3

Then 32 - 2 3 - 1 = 2 and 2 0
The result holds for n = 3
Then k2 - 2k - 1 0
Consider n = k + 1
Then (k + 1)2 - 2(k + 1) - 1 = (k2 - 2k - 1) + 2k - 1
But (k2 - 2k - 1) + 2k - 1 k2 - 2k - 1 0 since n 3
Thus the result is true for n = k + 1 if it is true for n = k
But it is also true for n = 3 and so by the principle of mathematical induction the
conjecture is true for all n 3
Q22: A bit of number crunching will find that when m = 7 the result holds.
Then 7! 37
Assume the result is true for n = k then k! 3k
Consider n = k + 1
Then (k + 1)! (k + 1) 3k 3k + 1 since k 7 from the basis step.
Thus the result holds for n = k + 1 if it is true for n = k but it is true for n = 7.
So by the principle of mathematical induction the conjecture is true for all n 7
Q23: The first step here is to find a conjecture for all positive integers which will be
easier to prove.
Since n is a positive odd integer, let n = 2m + 1
Then n2 - 1 becomes (2m + 1)2 - 1 = 4m2 + 4m and the conjecture can then be restated
as 4m2 + 4m is divisible by 8 for all positive integers m
Check m = 1 which gives 8 and 8 is divisible by 8. It is true for m = 1
Assume that the conjecture is true for m = k then 4k2 + 4k = 8a for some integer a
Consider m = k + 1
Thus 4(k + 1)2 + 4(k + 1) = 4k2 + 8k + 4 + 4k + 4 = 4k2 + 12k + 8
So the effect in letting m = k + 1 from m = k is to add 8k + 8
Thus 4(k + 1)2 + 4(k + 1) = 8a + 8k + 8 = 8(a + k + 1)
That is 4(k + 1)2 + 4(k + 1) is divisible by 8
Thus the result holds for m = k + 1 if it is true for m = k
But it is true for m = 1 and so by the principle of mathematical induction the new
conjecture is true for all m positive integers.
But n = 2m + 1 and so the original conjecture that n 2 - 1 is divisible by 8 for all positive
odd integers is now true.
Note that the new conjecture of 4m 2 + 4m is divisible by 8 for all positive integers m
could have been simplified and stated as m 2 + m is divisible by 2 for all positive integers
m.
Q24: Check for n = 10

Then 10 10 = 1010
The result is true for n = 10
Then 10k k10
Consider n = k + 1
Then 10k + 1 10 k10 (k + 1)10 (check using the binomial expansion).
So the result is true for n = k + 1 if it is true for n = k
But it is true for n = 1 and so by the principle of mathematical induction the conjecture is
true for all n
Contrapositive exercise (page 178)
Q25: The contrapositive states that if n is odd then n 3 is odd.

So n = 2k + 1 for some k
Thus n3 = (2k + 1)3 = 8k3 + 12k2 + 6k + 1 = 2(4k2 + 6 k2 + 3k) + 1
That is, if n is odd then n3 is odd and the original conjecture is proved.
Q26: The contrapositive implication is if n can be expressed as a sum of two square

integers then n does not have remainder 3 on division by 4.
If n does not have remainder 3 on division by 4 then n takes the form 4k, 4k + 1
or 4k + 2. This proof now involves four cases (proof by exhaustion).
Let n = x2 + y2 where x and y are integers. Then x and y are either even or odd.
a) Consider when x is even and y is even

then x = 2m and y = 2l for some integers m and l
Then n = x2 + y2 = 4m2 + 4l2 = 4(m2 + l2 ) and n has the form 4k
b) When x is even and y is odd
then x = 2m and y = 2l + 1 for some integers m and l
Then n = 4m 2 + (2l + 1)2
= 4m2 + 4l2 + 4l + 1
= 4(m2 + l2 + l) + 1 and n has the form 4k + 1
c) When x is odd and y is even the same result as for part 2 is found.
d) When x is odd and y is odd
then x = 2m + 1 and y = 2l + 1 for some integers m and l
Then n = (2m + 1)2 + (2l + 1)2
= 4m2 + 4m + 1 + 4l2 + 4l + 1
= 4(m2 + m + l2 + l) + 2 and n has the form 4k + 2
The contrapositive conjecture and so the original conjecture is proved.
Q27: Using the contrapositive which states

if n is even then n2 is not of the form 4m + 1, there are two cases to consider.
• When n = 4m then n 2 = 16m2 = 4(4m2 ) and is not of the form 4m + 1

• When n = 4m + 2 then n 2 = 16m2 + 16m + 4 = 4(4m2 + 4m + 1) which is not of the
form 4m + 1
The contrapositive holds and the original conjecture is true.
Direct proof exercise (page 180)
Q28: Sketch the triangle
cos = b
c and sin = a
c
b2 a2
so cos2 + sin2 = c2
+ c2
But by Pythagoras, a2 + b2 = c2
b2 a2 c2
so c2
+ c2
= c2
=1
Q29: Using the combination rules and known results

n
n
(2r - 1) = 2 r -n
r=1 r=1
2n(n + 1)
= -n
2
= n2 + n - n
= n2
Q30: b c b - c c - c = 0
a(b - c) 0
ab - ac 0
ab ac
Q31: r(r + 1)(r + 2) = r3 + 3r2 + 2r so

n
n
3

n
2

n
r(r + 1)(r + 2) r +3 r +2 r
r=1 r=1 r=1 r=1
2
n2 (n
4
+ 1)
n(n + 1)(2n
2
+ 1)
2n(n2+ 1)
n(n + 1)(n2 + n + 4n + 2 + 4)
4
n(n + 1)(n + 2)(n + 3)
4
Division algorithm exercise (page 183)
Q32: The answers are
1. q = -9 and r = 3: -42 = -9 . 5 + 3
2. q = -3 and r = 4: 25 = -3 . -7 + 4
3. q = -1 and r = 5: 14 = -1 . -9 + 5
Q33: The answers are
1. q = -1 and r = 2: -1 = -1 . 3 + 2
2. q = 0 and r = 4: 4 = 0 . 9 + 4
3. q = - 4 and r = 0: -12 = -4 . 3 + 0
Q34: All the possible forms of a number divided by 8 are

8k, 8k + 1, 8k + 2, 8k + 3, 8k + 4, 8k + 5, 8k + 6 and 8k + 7
Of these the even forms are 8k, 8k + 2, 8k + 4, 8k + 6 as each of these is of the form 2m
for some integer m
Number bases exercise (page 186)
Q35: The binary equivalents are :
a) 101001100
332 = 166 . 2 + 0
166 = 83 . 2 + 0
83 = 41 . 2 + 1
41 = 20 . 2 + 1
20 = 10 . 2 + 0
10 = 5 . 2 + 0
5=2.2+1
2=1.2+0
1=0.2+1
b) 111110101
c) 101011
d) 111111
Q36: The octal forms are:
a) 533
347 = 43 . 8 + 3
43 = 5 . 8 + 3
5=0.8+5
b) 5554
c) 5704
d) 1026
Q37: The hexadecimal answers are :
a) FAE
4014 = 250 . 16 + 14 (E)
250 = 15 . 16 + 10 (A)
15 = 0 . 16 + 15 (F)
b) 16C
c) 883
d) 13AA
Q38: The factors of 2210 are 1, 2, 5, 13, 17, 2210 and the factors of 399 are 1, 3, 7, 19,
399
The only common factor is 1, thus gcd (2210, 399) = 1 and by definition the integers are
relatively prime.
Euclidean algorithm exercise (page 189)
Q39: gcd (299, 1365) is 13

1365 = 4 . 299 + 169
299 = 1 . 169 + 130
169 = 1 . 130 + 39
130 = 3 . 39 + 13
39 = 3 . 13 + 0
Q40: The highest common factor is the same as the gcd.

The gcd (5187, 760) = 19
Q41: For coprime integers the gcd is 1

10465 = 1 . 5643 + 4822
5643 = 1 . 4822 + 821
4822 = 5 . 821 + 717
821 = 1 . 717 + 104
717 = 6 . 104 + 93
104 = 1 . 93 + 11
93 = 8 . 11 + 5
11 = 2 . 5 + 1
5=5. 1+0
This gcd is 1 and the numbers have no common factor other than 1. They are coprime.
Q42:
5184 = 1 . 3024 + 2160
3024 = 1 . 2160 + 864
2160 = 2 . 864 + 432
864 = 2 . 432 + 0 and the gcd (3024, 5184) = 432
Q43: By the Euclidean algorithm the numerator and denominator have a gcd of 124.
3
Dividing top and bottom by this gives a simplified fraction of 35
Q44:
610 = 76 . 8 + 2
8=4.2+0
The gcd (610, 8) is 2
610 = 11 . 55 + 5
55 = 11 . 5 + 0
The gcd (610, 55) is 5
Linear combinations exercise (page 193)
Q45:
693 = 8 . 84 + 21
84 = 4 . 21
The gcd is 21
working backwards 21 = 693 . 1 - 84 . 8 and x = 1, y = -8
Q46:
10080 = 2 . 3705 + 2670
3705 = 1 . 2670 + 1035 (line a)
2670 = 2 . 1035 + 600 (line b)
1035 = 1 . 600 + 435 (c)
600 = 1 . 435 + 165 (d)
435 = 2. 165 + 105 (e)
165 = 1 . 105 + 60 (f)
105 = 1 . 60 + 45 (g)
60 = 1 . 45 + 15 (h)
45 = 3 . 15 + 0 (i)
The gcd is 15
Working backwards gives 15 = 60 - 1 . 45 using line (h)
15 = 60 - 1(105 - 1. 60) = - 1 . 105 + 2 . 60 using line (g)
15 = -1 . 105 + 2(165 - 1 . 105) = 2 . 165 - 3 . 105 using line (f)
15 = 2 . 165 - 3(435 - 2 . 165) = -3 . 435 + 8 . 165 using line (e)
15 = -3 . 435 + 8(600 - 1 . 435) = 8 . 600 - 11 . 435 using line (d)
15 = 8 . 600 - 11(1035 - 1 . 600) = -11 . 1035 + 19 . 600 using line (c)
15 = -11 . 1035 + 19(2670 - 2 . 1035) = 19 . 2670 - 49 . 1035 using line (b)
15 = 19 . 2670 - 49(3705 - 1 . 2670) = -49 . 3705 + 68 . 2670 using line (a)
15 = -49 . 3705 + 68( 10080 - 2 . 3670) = 68 . 10080 - 185 . 3705
x = 68 and y = -185
Q47: 12 = 336 . 7 - 180 . 13 so x = 7 and y = -13
Q48: x = -7 and y = 18
Q49:
585 = 5 . 104 + 65
104 = 1 . 65 + 39
65 = 1 . 39 + 26
39 = 1 . 26 + 13
26 = 2 . 13 + 0
The gcd is 13
Working backwards eliminating the remainders to solve for x and y gives
13 = 39 - 1 . 26
13 = 39 - 1(65 - 1 . 39) = -1 . 65 + 2 . 39
13 = -1 . 65 + 2(104 - 1 . 65) = 2 . 104 - 3 . 65
13 = 2 . 104 - 3(585 - 5 . 104) = -3 . 585 + 17 . 104
Thus x = -3 and y = 17
For the general solution x = -3 + (104/13)m = -3 + 8m
and y = 17 - (585/13)m = 17 - 45m
The general solution is 13 = 585(-3 + 8m) + 104(17 - 45m)
Q50:
204 = 3 . 56 + 36
56 = 1 . 36 + 20
36 = 1 . 20 + 16
20 = 1 . 16 + 4
16 = 4 . 4 + 0
The gcd is 4 and 4 20 so there is a solution.
Thus working backwards
4 = 20 - 1. 16
4 = 20 - 1(36 - 1 . 20) = -1 . 36 + 2 . 20
4 = -1 . 36 + 2(56 - 1 . 36) = 2 . 56 - 3 . 36
4 = 2 . 56 - 3(204 - 3 . 56) = -3 . 204 + 11 . 56
Multiply by 5 since 5 4 = 20
20 = -15 . 204 + 55 . 56
The general solution is given by x = -15 + ( 56 /4 )m and y = 55 - (204 /4 )m
That is, x = -15 + 14m and y = 55 - 51m
The general solution is 20 = 204(-15 + 14m) + 56(55 - 51m)
Review exercise (page 196)
1(2)2
Q51: Check when r = 1 then 1 = 4 = 1 as required.
Assume that it is true for n = k
k2 k + 1)2
Then 1 + 23 + 33 + ... + k3 = 4
Consider n = k + 1 then
1 + 23 + 33 + ... + k3 + (k + 1)3
k2 k + 1)2
+ (k + 1)3
4
k2 k + 1)2 + 4 (k + 1)(k + 1)2
4
k + 1) k + 4k + 4)
2 2
4
k + 1 k + 2)2
4
(k + 1)2 (k + 1) + 1)2
4
The result is true for n = k + 1 if it is true for n = k. But it is also true for n = 1 and so by
the principle of mathematical induction it is true for all n
Q52:
943 = 1 . 779 + 164
779 = 4 . 164 + 123
164 = 1 . 123 + 41
123 = 3 . 41 + 0
The greatest common divisor is 41
Q53: Let the integers be n, n +1 and n + 2

The conjecture states that n3 + (n + 1)3 + (n + 2)3 = 9a for some integer a
Check when n = 1
13 + 23 + 33 = 36 and 36 = 9 4
The result is true for n = 1
Then k3 + (k + 1)3 + (k + 2)3 = 9b for some integer b
Consider n = k + 1 then the effect is the difference between k 3 + (k + 1)3 + (k + 2)3 and
(k + 1)3 + (k + 2)3 + (k + 3)3 which is 9k2 + 27k + 27
Thus when n = k + 1
(k + 1)3 + (k + 2)3 + (k + 3)3 = 9b + 9k2 + 27k + 27 = 9(b + k2 + 3k + 3). That is, it is
divisible by 9
The result is true for n = k + 1 if it is true for n = k
But it is also true for n = 1 and so by the principle of mathematical induction the
conjecture is true for all n
Q54:
5141 = 1 . 3763 + 1378
3763 = 2 . 1378 + 1007
1378 = 1 . 1007 + 371
1007 = 2 . 371 + 265
371 = 1 . 265 + 106
265 = 2 . 106 + 53
106 = 2 . 53 + 0
The greatest common divisor is 53
Advanced review exercise (page 196)
Q55:
247 = 1 . 139 + 108
139 = 1 . 108 + 31
108 = 3 . 31 + 15
31 = 2 . 15 + 1
15 = 15 . 1 + 0
So working backwards
1 = 31 - 2 . 15
= 31 - 2(108 - 3 . 31) = - 2 . 108 + 7 . 31
= -2 . 108 + 7(139 - 1 . 108) = 7 . 139 - 9 . 108
= 7 . 139 - 9(247 - 1 . 139) = -9 . 247 + 16 . 139
so 1 = 247x + 139y where x = -9 and y = 16
Q56:
252 = 1 . 160 + 92
160 = 1 . 92 + 68
92 = 1 . 68 + 24
68 = 2 . 24 + 20
24 = 1 . 20 + 4
20 = 5 . 4 + 0
so working backwards gives
4 = 24 - 1 . 20
4 = 24 - 1(68 - 2 . 24) = -1 . 68 + 3 . 24
4 = -1 . 68 + 3(92 - 1 . 68) = 3 . 92 - 4 . 68
4 = 3 . 92 - 4(160 - 1 . 92) = -4 . 160 + 7. 92
4 = -4 . 162 + 7(252 - 1 . 160) = 7 . 252 - 11 . 160
So 41 = 252x + 160y where x = 7 and y = -11
Q57:
297 = 1 . 180 + 117
180 = 1 . 117 + 63
117 = 1 . 63 + 54
63 = 1 . 54 + 9
54 = 6 . 9 + 0
So gcd (297, 180) = 9
Working backwards gives
9 = 63 - 1 . 54
9 = 63 - 1(117 - 1 . 63) = -1 . 117 + 2 . 63
9 = -1 . 117 + 2(180 - 1 . 117) = 2 . 180 - 3 . 117
9 = 2 . 180 - 3(297 - 1 . 180) = -3 . 297 + 5 . 180
so 9 = 297x + 180y where x = -3 and y = 5
Well done! This is the end of the final exercise in the last topic in this Advanced Higher
in Maths. We hope that you have found the course worthwhile and interesting.

Scholar Advanced Higher Maths Unit 3

Uploaded by

Copyright:

Available Formats

Scholar Advanced Higher Maths Unit 3

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

Scholar Advanced Higher Maths Unit 3

Uploaded by

Copyright:

Available Formats

SCHOLAR Study Guide

SQA Advanced Higher Mathematics

Distributed by Heriot-Watt University.

3 Further sequences and series 91

3.7 Set review exercise . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123

4 Further Ordinary Differential Equations 125

5 Further number theory and further methods of proof 163

6 End of unit three tests 199

Hints for activities 209

Answers to questions and activities 210

• Use vectors in three dimensions.

Minimum Performance Criteria:

• Calculate a vector product.

• Find the equation of a line in parametric form.

• Plotting points in three dimensions.

1.1 Revision exercise

Q1: Draw and label a three-dimensional graph with axes x, y and z

Example If Q is 45 km North West of P then the distance is 45 km and the direction is

There is a particular type of vector called a position vector.

2. These are all position vectors in 2 dimensions.

These are position vectors in 3 dimensions.

Vector length exercise

Vector notation exercise

Q7: Express the vector 3 using the standard basis vectors.

Q8: Express the vector i + 2j + k as an ordered triple.

1.3 Direction ratios and cosines

The ratios of a : b : c are called the direction ratios of the vector p

1. Find the direction ratios of the vector -2i - 3j + 4k

2. Find the direction cosines of the vector p = 2i - 2j - k

Direction ratios and direction cosines exercise

1.4 Basic arithmetic and scalar product of vectors

Vectors in 2d addition and subtraction

If p and q are written in terms of the standard bases then

Example Add the vectors -2i + 3j - 4k and -3i - 6j - 3k

2. Evaluate (i + 4j + 2k) - (2i -2j + k)

Vectors can also be multiplied by a scalar.

Using the standard bases, the multiplication is just as straightforward.

Example If a = 3i - j + 2k find -4a

Basic skills exercise

Q13: Subtract 3i - 2j + 5k from -2i

Q15: Show that 2i - j + 3k and -4i + 2j - 6k are parallel vectors.

Q16: Find a vector parallel to 2i - 4j + k with z component equal to 3

1.4.2 Scalar product of vectors in three dimensions

Algebraic scalar product exercise

Q18: Find the scalar product of the vectors 2i + 3j - 2k and i - 3j - k

Algebraic rules of scalar products

The second way of expressing the scalar product is geometrically.

The proof of this is shown at Proof 1 in the section headed Proofs.

Example Find the angle between i + 4j and -8i + 2j

Example Show that the two vectors a = -2i + 2j + k and b = 3i + 2j + 2k are

Geometric scalar product exercise

Perpendicular vectors exercise

Q28: Find c so that 3i + cj + 4k is perpendicular to 2i -2j + 3k

Q29: Show that the vectors -4i - 3j + 2k and 2i - 2j + k are perpendicular.

1.5 Vector product - algebraic form

2. Find p x q where p = -2i + 3j - k and q = 4i + 2j - 2k

Algebraic vector product exercise

Q31: Show that the distributive property a x (b + c) = (a x b) + (a x c) holds for the

From the result a x b = -b x a, then (a x b) + (b x a) = 0

They are not the same.

Example Show that a (a x b) = 0 and that b (a x b) = 0

Since a (a x b) = 0 and b (a x b) = 0 the vector a x b is perpendicular to both a and

• magnitude of a x b = a b sin , where is the angle between a and b,

Example Find a (b x c) when a = i + 2j - 3k, b = 4i - j + 2k and c = 3i - 2j + 5k

Q41: Find a (b x c) if a = i + 2j + 3k, b = 2i + j + 3k and c = 4i + k

Then solve for x, y and z in terms of

The equations are x = -3 + 2, y = 1 + , z = 4 - 3

This has symmetric equations x = 9 - 4, y = 8 - 6, z = 4 - 2

Example Suppose that L 1 has parametric equations x = 1 + , y = -1 + , z = 2 +

If is the angle between L 1 and L2 then

The vector equation is r = a + d = + 23