©

2012 Colin W. Clark

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, electronic, mechanical, photocopying, recording, or otherwise, without the written
permission of the author.

First published by Dog Ear Publishing

4010 W. 86th Street, Ste H
Indianapolis, IN 46268
www.dogearpublishing.net

ISBN: 978-1-4575-1481-4

This paper is acid free and meets all ANSI standards for archival quality paper. Printed in the United States
of America
Contents 1 Addition and Multiplication
1.1 The decimal system for whole numbers
1.2 Addition of whole numbers
1.3 Multiplication of whole numbers
1.4 Binary arithmetic
1.5 Decimal-point numbers
1.6 Scientific notation
1.7 The rules of arithmetic
1.8 The number line
1.9 How to learn mathematics
1.10 Review problems

2 Subtraction and Division

2.1 Subtraction
2.2 Multiplication
2.3 Division and fractions
2.4 The real number system
2.5 Review problems

3 Using Elementary Mathematics

3.1 Calculator math
3.2 Per cent
3.3 Areas
3.4 Scale effects
3.5 Proportionality
3.6 Speed and distance
3.7 Units of measurement and their conversion
3.8 Review problems
4 Algebra
4.1 Exponents
4.2 Exponential growth and compound interest
4.3 Operations with fractions
4.4 Polynomial algebra
4.5 Linear and quadratic equations
4.6 Inequalities
4.7 The binomial theorem
4.8 Fractional exponents
4.9 More about polynomials
4.10 Review problems

5 Euclidean Plane Geometry

5.1 Lines and angles
5.2 Triangles
5.3 Similarity and congruence
5.4 Solving a right triangle using trigonometry
5.5 Isosceles triangles
5.6 Circles and arcs
5.7 The Scientific calculator
5.8 Area of a circle
5.9 Review problems

6 Analytic Geometry
6.1 The rectangular coordinate system
6.2 Straight lines
6.3 Circles
6.4 Transformations
6.5 Conic sections
6.6 Review problems

7 Functions and Graphs

7.1 Sets
 7.2 Functions
7.3 The graph of a function
7.4 Inverse functions
7.5 Graphing sums and products
7.6 Polynomial calculus
7.7 Numerical solution of equations
7.8 Interpolation
7.9 Review problems Solutions to Review Problems

Diagnostic Test
Common Errors

Greek Alphabet

Mathematical Symbols

Index
Acknowledgements
Special thanks are owing to my wife Janet, who set the book in type, using the
mathematical typesetting program TeX. Getting all the equations, graphs and
problem solutions in the right position was a mammoth undertaking.
My daughter Jennifer drew the cover picture. The rear-cover pboto was taken
by Deborah Kieselbach.
Thanks also to colleagues who encouraged me to write and publish Math
Overboard! including Marc Mangel, Michael Mesterton-Gibbons, Peter Rastall,
Jim Zidek and Leah Keshet.

Colin Clark
July, 2012
How to use this book Math Overboard
will help you to raise your level of
understanding of basic mathematics,
from Arithmetic to Algebra, Geometry,
Trigonometry and other topics. You are
assumed to have studied these topics (or
most of them) in school, but like many
former math students you now realize
that your understanding of elementary
mathematics is not satisfactory. You need
help, and this is it!
Math Overboard can be used in various ways. For example, you can look up
a specific topic, or review an entire subject, or even review all of elementary
math. If you’re not sure where to begin, try the following:
1. Do the Diagnostic Test on page 405. This will help to identify the areas you
need to work on.
2. Read about Common Errors in Mathematics, page 411. Understanding
mathematics well enough to avoid these errors is essential for success in
later math and science courses.
3. To decide whether you need to study a specific chapter, start by trying to
solve the Review Problems at the end of the chapter. If you have difficulty
with these problems, you clearly need to work through the chapter.
 Math Overboard is based on the principle that to learn mathematics you
must take the trouble to understand it. Many students get into the habit of trying
to learn math by memorization alone. This is a sure recipe for disaster. Working
with Math Overboard will help you to supplement memorization with
understanding. The whole subject will begin to “make sense.” You will develop a
new mental outlook, which will help you immensely in your future studies.
Let me mention one example, adding fractions:

My question is, how can you check whether this is correct? Try to answer this
before reading on.

Here is the answer:

Are you already in the habit of using this check every time you combine
fractions? It is easy to do, and will prevent many errors. Knowing how to do this
implies that you understand cancellation, and also the logic behind adding
fractions. It’s a far cry from merely memorizing Eq. 1 above, with no clue as to
why it is true. (See Chapter 2 for detailed discussion.) Another example: a math
teacher recently asked me whether there was an easy explanation of the “invert
and multiply” rule for dividing fractions. She said that she has to teach this
method every year, but has no idea why it is true. None of the textbooks in the
school library gave any explanation. Isn’t that a sad story? Especially since the
explanation is very simple, and only requires that you understand that division is
the reverse operation to multiplication. Sure, students can be made to memorize
the rule, but is that an education?
(This topic led to a big controversy in math education some years ago.
Someone decided that students should understand that division means “goes
into.” Kids should be taught to divide by for example, by answering the
question “How many times does go into ” This nonsense even got into
textbooks. Believe it or not.) Here are the main reasons why understanding
mathematics is important:
1. Knowing why a given rule (or formula, or method) is valid is a big help in
remembering that rule.
2. Knowing why things are true means that your mathematical knowledge
 “hangs together” and makes sense as a whole. As you study new topics, you
see how they relate to what you already know.
3. Your confidence grows as you learn more mathematics, because you
yourself understand the subject in detail. You do not have to rely on what
your teacher tells you.
So how do you go about understanding math as you learn it? This is where Math
Overboard comes to the rescue. Some of its main features are:
1. Frequent problems (with answers over the next page) allow you to test your
understanding as your proceed. Be sure to work out these problems, using
pen and paper. Do not look up the solution before you have tried hard to
solve the problem.
2. Complete, easy-to-follow explanations are provided for every topic. Read
each explanation carefully, and make sure that you understand each step.
Then put the book down and repeat the explanation for yourself.
3. Advice on how to learn mathematics occurs throughout the book. If you
want to read this advice by itself, look up “mathematics—
learning/understanding” in the index.
4. The index can be one of your main aids to learning. Use it frequently to
chase down any term or topic that you are not absolutely clear about. Each
index entry leads you directly to the main presentation of that topic.
Let me offer one bit of advice right now. Before starting to use Math
Overboard, buy yourself a dedicated math notebook. Whenever you read a
section of the book, start making notes. List the section heading and page
number. Next, write down each major point as you encounter it. Write out
(carefully!) the solution of each Problem. If you had trouble, make a note of that,
and if possible explain where you went wrong.
The notebook will become a permanent record of your endeavors. Would it
not be a good idea to write it using a pen? I myself never do math with a pencil!
Pencil writing encourages sloppiness, and shows a lack of confidence. If you
make a mistake, put a line through it – you might discover later that it wasn’t a
mistake after all.
Math Overboard includes many applications of elementary mathematics to
actual situations in Science, Business and Economics, and many other areas.
Study these applications not only for their own interest, but also to assist your
learning process. Knowing how a certain math topic is used in practice, is part of
understanding that topic. For example, learning about musical scales can help
you to understand trigonometric functions (Part 2).
You will soon realize that Math Overboard is not a typical (boring) math
text. It is designed for self-study, not classroom use. Also, the book is much more
comprehensive than any textbook.
A final word of warning. Math Overboard is not designed as an introductory
textbook for any specific topic. For one thing, there are not enough exercises for
an introductory text. If you need additional exercises, two approaches are, first to
search the web, where there are thousands of math exercises, and second to make
up your own exercises. I’m not kidding – making up exercises as you read the
book is a creative and productive approach to understanding math. Try it!
Chapter 1
Addition and Multiplication

1.1 The decimal system for whole numbers

Numbers are the very core of mathematics, science, and economics. We are so
accustomed to using numbers in our everyday lives that we seldom stop to think
about what an amazing intellectual feat our current method of writing numbers
really is. It wasn’t until the seventeenth century that the present “decimal” (i.e.,
base ten) system came into use. The Romans, in spite of their vast empire and the
great organization that must have been required to conquer and maintain it, had
developed only the unwieldy method of Roman numerals. How much is
MCMLXXVIII? (In fact, it’s 1978.) How merchants in Roman times kept their
accounts is hard to imagine (it seems they used counting devices, similar to the
abacus still used today in China). Other early civilizations also had their own
equally unwieldy systems for writing numbers.
The feature that makes the modern method of writing numbers so successful
is positional significance of the digits. The expression 532 means five hundreds,
three tens (thirty) and two ones. Once this simple way of writing numbers
became known, it spread over the entire world, and is now used almost
everywhere. The invention of positional notation ranks right up there with other
great inventions, like the wheel and the transistor. Simple, but profound.
We call our method of positional notation the decimal system, “decimal”
meaning that it is based on tens. Each position, as we saw with the example 532,
represents 10 times the next position:

Because the decimal system is based on ten, we need ten digits (0, 1, 2, 3, 4, 5,
6, 7, 8, 9). The word “digit” also means finger (or toe), of which we have ten.
Well, you know the rest of the story. If humans actually had six fingers per hand,
no doubt our number system would use twelve as its base, and it would need 12
“digits.”
There is one important modern device that uses a base different from ten: the
computer (including the hand calculator). Because they are electronic devices,
computers and calculators use base two. This is called the binary system. Luckily
for us, we don’t have to know anything about binary numbers to use a calculator.
We just punch in a decimal number and the calculator itself automatically
translates it into binary notation, which the user never sees at all. Still, binary
arithmetic is kind of interesting (it’s actually a bit easier than decimal
arithmetic). If you want, you can read about it later.
By the way, you may be wondering about the term “decimal system” – what
became of the decimal point? Well, decimal points will be discussed later on in
this chapter.
Numbers versus numerals
Is 7 a number? Is it the same thing as “seven”? Or If we were to be very
precise, we would say that there are numbers, names of numbers, and numerals.
“Seven” is the English name of the number seven, and 7 is the symbol, or
numeral, for this number. Other names for seven include “sept” (French),
“sieben” (German), and so on. Other numerals include VII (Roman).
But what exactly is the number seven itself? Well, seven is the number of little
circles on the next line.

Or, seven is the number of Snow White’s dwarfs – Happy, Doc, Sleepy, Dopey,
Grumpy, Sneezy, and Bashful. “Oh, daddy, everyone knows what seven is.”
Right, but there is a difference between the number seven, its name “seven,” and
its numeral 7. Just as there is a difference between a chair, the name “chair,” and
the symbol .
If you ask somebody “What’s this?”

they will probably answer “a chair.” They know perfectly well it’s not a chair –
it’s only a crude drawing of one. But we all happily accept the slight inaccuracy,
which is preferable to circumlocution: “That’s a crude sketch that looks a bit like
a chair.” So, in this book we will not hesitate to say “the number 7,” and not “the
number whose numeral is 7.” This should never cause any confusion.
Powers of ten
The phrase “the powers of 10” refers to the numbers

and so on. An easy, organized way to write these numbers is to use exponents.

and so on. (The expression 107 is read as “ten to the seventh.” But for reasons you
can perhaps guess, 102 is usually read as “ten squared,” rather than “ten to the
second.” Also 103 is read as “ten cubed.” Do you see why? Think of a large square
with 10 units on a side. How many little one-unit squares are there? One
hundred, or 102.)
In other words, for any n, the expression 10n (“ten to the n’th”) equals 10 × 10
× ··· × 10 (n times). So how would you write 1 million using exponents? One
billion (i.e., a thousand million)? Answer: 1 million (1,000,000) is 106. One
billion (1,000,000,000) is 109.
When written out in full, 10n becomes 1 followed by n zeros. For example,
102 = 100 (two zeros), 107 = 10,000,000 (seven zeros).
A comment on how to write large numbers, such as 1,000. This book uses
commas between every third digit, starting from the right. Alternative methods
used by other authors are:

Problem 1.1 Actually, we can take the n’th power of any number. For example 32
= 3 × 3 = 9, or 26 = 2 × 2 × 2 × 2 × 2 × 2 = 64. Your problem is, write out the
powers of 2 (i.e. 2n) for n = 1, 2, 3, up to n = 10.

Note: to see the solution to any problem, look at the bottom of the next even-
numbered page.
You can probably see what the powers of 10 have to do with decimal
notation. For example,

(In case anyone is confused, × means “times” and + means “plus.”) In other
words, each digit, according to its position, tells us how many multiples of each
power of 10 are contained in the number. This is the positional system, based on
powers of 10 – in other words, the decimal number system.

Problem 1.2 Think about this number: Write it in a different way. What
would it be in ordinary decimal notation? [This problem could really blow your
mind!]
The number zero
Our decimal system depends critically on the use of the zero symbol, 0. For
example, 204 is two hundreds plus no tens plus four units. All this seems
elementary to us today, but historically speaking, people had great difficulty
accepting the idea of zero. For example, although the Babylonians in 1600 BC
used positional notation (with base 60, not 10), they simply left a blank where we
would put a zero. Greek astronomers by 100 AD did use zeros, but in Western
Europe the decimal system with zeros was not adopted until the 17th century.
But is zero really a number? We have a name, zero, and a symbol, 0, but does
that mean that zero has to be a number? This rather philosophical difficulty
apparently helped to delay the acceptance of decimal notation for many
centuries. The modern view is this: if we want to say that zero is a number, fine.
There’s no law against it. What must be done is to develop a logically consistent
and useful numbering system. When we consider subtraction in Chapter 2, we
will find that zero is indispensable as a number, not just a symbol.
The number 0 can be considered as a counting number – it counts no, or 0
objects. Thus we would have 9 + 0 = 9, and so on. More on this in the next
section.
The numbers 0, 1, 2, 3, … comprise the system of whole numbers. This
system excluding 0, in other words, the system 1, 2, 3, … is called the system of
natural numbers. Other number systems will be discussed later in this chapter,
and in Chapter 2.

Problem 1.3 The main feature of the natural number system is that any given
natural number n has an immediate successor n + 1. Your problem is to specify
exactly the rule for adding 1 to a natural number given in decimal notation, and
to illustrate with examples.

1.2 Addition of whole numbers

Numbers are used for two main purposes – counting and measuring. In this
section we discuss the addition of natural, or counting numbers. Later we will
study the measuring numbers.
I suggest you take a moment here to express in words what addition of
numbers means, exactly. Start with a simple example – when we assert that 5 + 3
= 8 (five plus three equals eight), what do we mean? What about 756 + 39 = 795?
Children begin counting about as soon as they start to talk. In the first grade
they learn the decimal (base ten) system for writing whole numbers. Next they
master the addition algorithm.
Algorithm – what’s that? An algorithm is a method, or routine, for
calculating something. Most algorithms are repetitive: you keep repeating the
same steps until the calculation is finished. Because of this repetitive
characteristic, algorithms are easy to learn – you don’t even have to really
understand what you’re doing, you just do it. For example, let’s look at the
addition algorithm.

Problem 1.4 Add 627 and 415 (writing one above the other). Note the repetitive
steps. Specify the addition algorithm in detail.

There are several things worth noting about the addition algorithm. (1) The
algorithm includes starting and stopping instructions. (2) It is repetitive. (3) You
can probably perform the algorithm without even understanding what you’re
doing. (Of course, to be able to use the addition algorithm, you have to know
your “plus table:” 7 plus 5 equals 12, and so on. But this was presumably
programmed into your brain in first grade.)
The advantage of any algorithm is that it is repetitive, and “mindless.”
Computers can be (and are) programmed to perform algorithms, and a
computer certainly doesn’t understand what it’s doing. The mindless nature of
algorithms can also be a major educational disadvantage, if mastering algorithms
is stressed over understanding mathematics. Learning an algorithm is easy.
Gaining understanding is more difficult. But both are important – you need to
master the algorithms, and also understand the mathematics.
I once had a remarkable experience of this in an engineering math course I
was teaching. One problem I put on the Christmas exam had two parts; part (a)
used an algorithm from the course, and part (b) asked the students to explain in
plain words what the answer to part (a) meant. Of fifty-two students, 51 got (a)
completely right, but only one had even the vaguest idea what it meant. This
certainly convinced me that students can easily master an algorithm without
having the faintest idea about what they’re doing. (I thought that I had stressed
the meaning of the topic when I taught it; the underlying concept was pretty
basic to the whole course. It was shocking and embarrassing to find that most of
the students had missed the whole point.)
But why does it matter? If a student memorizes the algorithms (techniques)
in a course, he or she will probably pass the course. Isn’t that good enough?
Understanding everything is a lot more work, so who needs it? The answer to
these questions is subtle, in part because the very meaning of “understanding” is
a complicated matter. I will return to this question at various times in this book,
but let’s agree for now that understanding a topic in mathematics includes:
1. Knowing why the particular topic (algorithm, formula, etc.) is true.
2. Knowing how the topic is related to other topics.
3. Being familiar with the uses of the topic, both elsewhere in mathematics,
and in practical applications.
The disadvantages of trying to rely on memorization, rather than
understanding, are:
1. Anything that is memorized without understanding is quickly forgotten, or
incorrectly recalled at a later time. Once fully understood, however, a
mathematical technique or concept is usually remembered for life. The
mental effort exerted in understanding some topic appears to establish
permanent synapses in the brain, much more so than rote memorization.
2. Failure to fully understand the mathematics taught at a particular stage will
often return to haunt the student in later courses. Advanced mathematical
topics often depend strongly on understanding elementary material. For
example, learning calculus requires a secure understanding of basic algebra
and geometry, as well as a mastery of techniques in these subjects.

Problem 1.5 Explain in some detail why the addition algorithm works.
Addition as counting
Both addition and multiplication are closely related to counting. We adults
don’t actually count anything when doing arithmetic, because we memorized our
plus and times tables as children, and learned the algorithms. Nowadays we have
electronic calculators, so perhaps less emphasis needs to be given to teaching
arithmetic skills to children. Does this mean that teachers can stop teaching
arithmetic altogether? I don’t think so.
Every citizen needs to be “number-conscious” (a better word is numerate), in
order to deal with numbers in everyday life, and to have some “feel” for the
meaning of numbers. Arithmetic is as important as ever. But endless drill
exercises aren’t. Better to teach children the practical uses of arithmetic. Most
schools do seem to be going in the right direction these days.
Politicians and salesmen are always out to take advantage of people’s
innumeracy. Why does every price end with 9? A bottle of milk is $1.99. Why
not call it $2? Most ads for new cars these days only list the monthly lease charge,
not the price of the car. I can’t believe anyone would ever lease a new car. It’s
much more expensive than buying the car outright, even if you have to borrow
from the bank. But many people don’t, or can’t do the arithmetic to compare
these options. They couldn’t even do it using a calculator, perhaps.
Not paying off your monthly credit card promptly and completely is another
way to throw money away. Wouldn’t a decent school system use such examples
to illustrate the uses of math?
But let’s get back to pure math for a while. We need to discuss the “rules of
arithmetic.” One of these rules is

Rule 1 is true because of the basic relation between addition and counting.
What is this relation? If you ask a seven-year-old to add 3 and 5, she (or he) may
count on her fingers 1,2,3,1,2,3,4,5 and then recount up to finger 8. Now ask her
to add 5 and 3. Why does she think she gets the same answer?
What I’m leading up to is this. We need to formulate a basic definition of a +
b, the sum of a and b. (I’ll talk later about the role of definitions in mathematics.
Few students graduating from high school seem to have any idea how utterly
important definitions are. This seems to be part of the “memorize, don’t try to
understand” syndrome.)

In other words, to add a and b, you first count up a objects followed by b

more objects. Children are right!
Three things about this basic definition. First, it’s completely general: a and b
can be any whole numbers. Second, it implies that a + b = b + a (see below). And
third, it’s virtually useless in practice! Would you add 83, 746 + 19, 772 by
counting? That’s why we have an addition algorithm.
Another point: the basic definition applies if either of a or b equals 0. It
implies that a + 0 = a, for example.
So why is a + b = b + a? By definition, a + b is obtained by counting all the
objects in the a-set followed by those in the b-set. Example:

But the total set of objects is the same for both cases. Therefore you get the same
result. This is an instance of a fundamental principle.

That’s all there is to Rule 1. Thus a + b means first count a objects, then b
more. And b + a means count the b objects first, then a more. You always will get
the same total (unless you make a mistake). No doubt your grade one teacher
explained this very well, and you’ve never thought about it since. Okay, what
about the next rule?

What does this mean, and why is it true? First, what about the brackets, ()? In
mathematics, brackets specify the order of performing multiple operations, with
quantities inside brackets being calculated first. Thus, 4 + (5 + 6) means 4 + 11,
which becomes 15. (The brackets can be removed when the calculation inside
them is finished.) This is a different calculation than (4 + 5) + 6, which means 9
+ 6, which turns out to be 15 again.

Now we can see why Rule 2 is valid. We have three sets of objects, one with a
objects, one with b, and one with c. Expression a + (b + c) says first count the b
and c objects, then continue on with the a objects. Expression (a + b) + c does it a
different way. But the counting principle tells us we have to get the same answer
both ways.
Rules 1 and 2 are second nature to most people. But think how convenient
they are. More generally, given any list of whole numbers, we could group them
in any way and add the groups in any order, and always get the same sum.
(Before calculators – or even with them – one way to check a long sum was to do
it twice, in different order. If you get the same answer both times, it’s probably
correct. If not, it’s certainly incorrect!)
Rules 1 and 2 may seem pretty obvious. Students rarely make mistakes in
using these two rules. But they do make mistakes with some of the other rules of
Arithmetic (as will be discussed later). Mistakes are easy to make when one relies
entirely on memorization of rules and formulas – because we all are subject to
faulty memories.
In today’s world, mathematical mistakes can have drastic consequences.
When a nurse makes a mistake in dispensing a dose of medicine, a patient may
die. When an engineer makes a mistake in calculating a safety allowance, a
building may collapse. Mathematical education must emphasize methods for
eliminating mistakes, and this is a leading principle of Math Overboard!
Here, then, are some fundamental ways of avoiding mistakes.

Other methods of eliminating mistakes will be discussed throughout this

book.
Returning to Rules 1 and 2, you need to start now by being sure you
understand why these rules are valid. Put down the book and mentally review
both rules and their rationales.

Problem 1.6 Rule 2 says that brackets aren’t needed in addition. We can simply
write 2 + 15 + 8, without any brackets. To see that brackets are sometimes
needed in other situations, calculate 2 × (15 + 8) and compare it with (2 × 15) +
8.

What would 2 × 15 + 8 mean? Do we do the multiplication or the addition

first? Problem 1.6 shows that the results are different. To avoid ambiguity (a
disaster in mathematics!), we must use brackets to specify the order of the
operations. (Actually, there is a convention in Algebra that, unless otherwise
specified by brackets, multiplications are done before additions. I will use this
convention later in the book, but to keep things simple I won’t use it yet. I’ll
always put in brackets.)
A brief final comment about brackets. The pedantically correct name for () is
parentheses; brackets are { }. However, many people use the term brackets for
either. Sometimes () are referred to as “round brackets,” { } as “curly brackets,”
and [ ] as “square brackets.” The rule of brackets applies to any and all of these
types. Different types of brackets are often used to improve readability in a
complicated expression. For example, calculate [(2 + 5) × 7] + 8. Answer: 57.

1.3 Multiplication of whole numbers.

We start this section with:

Here a × b is read as “a times b.” For example, 3 × 5 = 15, as you can check by
actually counting 3 groups of 5 objects:

There are several ways of expressing the product of two numbers a and b:

In other words, we can use a cross × , a dot ·, or nothing other than

“juxtaposition”(writing the two symbols a and b one beside the other). However,
there could be ambiguity in the case of juxta position. Does 35 means 3 times 5,
or thirty-five? The convention is that it always means thirty-five – this is our
accepted positional notation for writing numbers. In order for juxtaposition to
mean multiplication, at least one of the factors must be a symbol. Thus 3m
means 3 times m and ax means a times x. This convention is very commonly
used in mathematics. If you want to write three times five, you write 3 × 5.
 As far as learning basic math goes, one has to pay as much attention to such
conventions as to “facts.” A student who doesn’t understand that 3m means 3
times m is clearly going to have major difficulties in later courses.
Some basic terminology: the expression 3 × 5 is called the product of 3 and 5.
Also, the numbers 3 and 5 are called the factors of the product 3 × 5. A product
can have many factors (3 × 5 × 9); some or all of the factors may be letters (3ab),
in which case it is understood that each letter represents a number, perhaps
unspecified. For example, the value of 3ab when a = 5,b = 9 is 3 × 5 × 9, or 135.
To go back a bit, what about 3 + 5? This is called the sum of 3 and 5. The
numbers 3 and 5 are called the summands of this sum.
Returning to the definition of multiplication given at the beginning of this
section, and the counting principle of Section 1.2, we can now deduce two more
basic rules of arithmetic.

Problem 1.7 See if you can think up a good explanation for Rule 3. If not, please
read the explanation given in Solution 1.7.
 Next, how would we deduce the Associative Law of Multiplication? Here we
have three numbers a, b, and c, so we need to think of a stack of objects (for
example, children’s square blocks), as shown. Then a × (b × c) equals the total
number of blocks in a layers, each containing b × c blocks (left diagram). On the
other hand, (a × b) × c equals the number of blocks in c “slices,” each containing
a × b blocks. No matter how you arrange to count these blocks you get the same
answer. This proves Rule 4.
Rules 3 and 4 tell us that to multiply three numbers, for example on a hand
calculator, we can group them and multiply them in any order. This is similar to
Rules 1 and 2, which tell us we can add numbers in any order. Because of these
facts, we don’t need to use parentheses when writing sums or products. Thus 21
+ 105 + 273 is unambiguous, and so is 21 × 105 × 273.
Our next rule, which combines addition and multiplication, is also explained
by the counting principle.

Note that, according to the convention that a × b can also be written as ab,
Rule 2 could be stated as
 Let’s start with an example: 3 × (4 + 9) = 3 × 13 = 39, whereas (3 × 4) + (3 ×
9) = 12 + 27 = 39. Try another example for yourself.
Before explaining the reasoning behind Rule 5, let us consider the notation
involved. The left side, a × (b + c), is straightforward; to calculate this, first you
add b and c, then you multiply the result by a; see the example. (Recall the rule of
brackets: operations inside brackets are performed first.) The expression on the
right side, (a × b) + (a × c), also uses brackets. First you do the multiplications,
then the addition.
Now for the explanation of Rule 5, which once again is based on the counting
principle. First, consider a × (b + c). By the definition of multiplication, this is
the total number of objects in a groups of (b + c) objects each. The diagram
shows the example 3 × (2 + 4).

One way to count up these dots is in groups: first 3 × 2, then 3 × 4, and add.
This is what the symbols (3 × 2) + (3 × 4) mean. A second way to count the dots
is to first count one line (2 + 4), then combine the 3 lines, getting 3 × (2 + 4). By
the counting principle, these are equal:

We don’t even have to calculate the numerical values to know that the two
results will be the same.
I hope you see that the above argument could be used to prove that Eq. 1.6 is
true for every possible case. The special example used here, 3 × (2 + 4), is strictly
representative of the general case, a × (b + c). We can therefore conclude that in
general

I wish to emphasize that, although the above discussion was based on a

certain numerical example, that example was accompanied by an argument (here
using a diagram) that could have been applicable to any example.
“Proof by example” is often claimed to be unacceptable in mathematics. But
there are two kinds of proof by example, one acceptable and the other not
acceptable. You can’t prove a general rule (such as a + b = b + a, to recall an
earlier case) by exhibiting one, or several numerical examples (such as 3 + 5 = 8
and 5 + 3 = 8). But you can prove it with a general argument illustrated by a
typical, or generic, example. I hope you recall what this general argument was,
for the case of the Commutative Law of Addition.
Let us agree to call the two types of examples isolated examples, and generic
examples, respectively. Math Overboard! often uses isolated examples as
illustrations or exercises, but never as proofs of general laws. We often use
generic examples to prove general laws. I hope you will always take the trouble to
understand these generic examples and the arguments showing that they are in
fact generic.
To summarize, the five Laws of Arithmetic are valid for all whole numbers.
All five laws follow directly from the counting principle. In other words,
arithmetic is just organized counting. We will see that this holds also for the rest
of arithmetic, including subtraction and division. Furthermore, all of Algebra is
based on arithmetic, so Algebra is also based on counting.

Multiplying by 10
How do you multiply a given number by 10? How much is 763 × 10? Answer:
7,630 – you just put a zero on the end of the number. Can you explain why this
works? Take a moment to think about this before reading on.
It has to do with positional notation and powers of 10. Thus
and therefore

(Did you notice the use of the Distributive Law in this argument? If not, look
at it again.) Similarly, to multiply a number by 100 (i.e. 102), you put 2 zeros on
the end of the number, and so on.
Nested brackets
The above calculation used nested brackets, that is, brackets inside other
brackets. How are these handled? Remember the rule of brackets: operations
inside brackets are carried out before operations outside these brackets. This rule
can be applied repeatedly. If one pair of brackets contains an expression that
itself involves further brackets, then the operation inside those brackets must be
done first. For example:

Here, the inner brackets are dealt with, and removed, first. Try this example

Answer: 11 × 13 = 143. This example used two kinds of brackets, round brackets
(), and square brackets [ ]. This makes the whole expression easier to read, but is
not required; the above example could be written as (8 + 3) × ((2 × 6) + 1).
Again, the innermost brackets are dealt with first.
Using brackets correctly is important. Weak students often make errors in
using brackets, and sometimes get the wrong answer because of such an error.
For example, what would this mean:

Did the writer intend (6 + 5) × 3, or perhaps (6 + 5 × 3)? These are different,

being equal to 33 and 21, respectively (see below). Or how about 6 + 3 × 2) × (7 +
8? This is totally meaningless. If you make such errors on a computer, you will
get an error message, often quite rude.
Precedence rule
Calculate 6 + 5 × 2. The answer is 16, not 22. How come? This is a consequence
of the following

This means that, in any expression involving both multiplication (×) and
addition (+), the multiplications are carried out before the additions, unless
indicated otherwise by brackets. Thus
However

Here brackets are needed to specify that the addition is done before the
multiplication.
The precedence rule allows us to use fewer brackets than would otherwise be
needed. This in turn makes mathematical expressions easier to read – provided
you are familiar with the rule. However, whenever you are in doubt, the brackets
can be left in. For example, 3 × 6 + 2 looks confusing to me, and I would usually
write (3 × 6) + 2. But I don’t find xy + 2 confusing, and I would never
misinterpret this as x(y + 2).

Problem 1.8 Calculate (a) 3 × 4 + 2 × 9; (b) 3 × (4 + 2) × 9; (c) 3 × (4 + (2 × 9)).

Another example: where are the “hidden” brackets in the expression 3xy +
2z. Evaluate this expression for the case x = 2, y = 5 and z = 4. (Answer: 38.)
Would this be the same as 3x(y + 2z)? (Answer: no, the latter evaluates to 78 in
this case.)

Problem 1.9 Remove all brackets: (a) (3x)(y + 4); (b) (7p)(4 + 2q).
The multiplication algorithm
The multiplication algorithm that most people use is illustrated by the following
example, for the calculation of 67 × 3:

You probably do the carryover in your head, and normally do such a simple
multiplication on one line: 67 × 3 = 201. This calculation is again based on the
powers-of-ten, positional (decimal) notation:

As in doing addition, you start with the units column (7 × 3 = 21) and keep
track of the carryover. Do another example or two for yourself.
A more complicated example is done in the same way:

If you’re like me, you hardly ever do long multiplications any more, what
with hand calculators being readily available. But I think it’s worthwhile knowing
how to, and also understanding why the old school algorithm is valid. It’s all part
of acquiring a firm foundation for all your mathematical knowledge. I’m not
embarrassed to be rusty at multiplication, but I would be embarrassed if I had no
idea how to do it by hand, or why the hand method is valid.
1.4 Binary arithmetic
You don’t have to read this section if you don’t want to; it’s for curiosity’s sake.
However, understanding the binary number system can strengthen your
understanding of the decimal system. Besides, it’s fun. The good news is that
binary arithmetic is much easier than decimal arithmetic. The bad news is that
the binary representation of a given number is longer to write down than the
decimal representation. (Because this section is for really keen readers, I’ve
condensed the writing style quite a bit. You’ll have to figure out some of the
details on your own.)
The binary number system is much like the decimal system, except that it
uses base 2 instead of base 10. Consequently it uses only two symbols, 0 and 1.
Every whole number can be written as a string of 0s and 1s. For example, the
binary number 10110 is the number twenty-two, i.e. 22 [in base 10], written in
binary form. The symbols 0 and 1 in a binary number are called “bits.”
How do we know what the number 10110 is? Well, reading from the right,
the bits are the units bit, the two’s bit, the four’s bit, the eight’s bit, etc. This
means that

Compare this with a decimal number:

We have special English names for some of the powers of ten, so we can read off
any base-ten number. Thus 9,536 becomes “nine thousand, five hundred and
thirty-six.” We don’t have names for the powers of 2, however, because nobody
uses base-two numbers in everyday life. Thus 10110 is just “one oh one one oh
(base 2).”

Problem 1.10 Write the numbers from one to ten in base two.

Problem 1.11 (Optional) Write 54 (base 10) as a base-two number.

Binary Addition Algorithm
One example should suffice to illustrate the binary addition algorithm.

Easy, eh? This is basically how computers do it.

Multiplying by two. How do you multiply by two in binary notation? Figure this
out for yourself, and compare with multiplying by ten in decimal notation.

Binary Multiplication Algorithm. Basically, binary multiplication reduces

immediately to addition:

If binary arithmetic is so simple, why do you think the schools haven’t

switched entirely to binary arithmetic? I think the answer is that the binary
representation of any moderate-size number is just too long – we can’t quite
grasp that 1101110010110 is the number we know as 7,062. Binary numbers are
for computers, not people. But now when aliens land with 4 digits on each hand,
you’ll be able to discuss octal arithmetic (base 8) with them right away.

1.5 Decimal-point numbers

How much is 8.372? This is an example of a “decimal point” representation of a
number. It is not a whole number, but consists of a whole number (8) plus a
decimal part (.372). The decimal part is a certain number between 0 and 1. Thus
8.372 is a number somewhere between 8 and 9. (We read 8.372 as “eight point
three seven two,” or “eight decimal three seven two.”)
Everyone is familiar with decimal-point numbers in the case of currency.
Thus $3.49 is three dollars and 49 cents. In other applications, decimal-point
numbers may have any number of digits after the decimal. Decimal-point
numbers are used especially in science and technology, where a high degree of
precision is often required.
Inexpensive hand calculators typically handle up to 7 decimal places. For
example, try expressing 1/7 (one-seventh) on your calculator: punch 1÷7= . My
kitchen calculator gives 0.1428571. Multiplying this by 7 should give 1 – try it. (I
get 0.9999997 – why? More about this in Chapter 3.) From now on I assume you
have a calculator. An inexpensive non-scientific calculator will be OK for
Chapters 1 to 3, but a scientific calculator is needed from Chapter 5 on. See
Section 3.1 for information on using your calculator.
You remember that the numbers (1,2,3,4,…) are counting numbers – they’re
used for counting objects. The decimal-point numbers are measuring numbers –
we use them for all kinds of measurements, such as lengths, weights, and so on.
Decimal-point numbers allow us to specify measurements with any required
degree of precision.
We must now explain carefully what decimal-point numbers are, and how to
add and multiply them. (Later, in Chapter 2, we discuss subtraction and
division.)
Let us start with the case of a single digit after the decimal, for example, 0.3.
We will use an imaginary ruler to locate numbers like 0.3.

This ruler has unit length, and we label the ends 0 and 1. Now we mark off the
ruler into ten equal lengths. Each little segment has length (one-tenth) of a
unit. We can label the points and so on.
 Thus, the point labeled is three-tenths of the distance from 0 to 1. To be

consistent, we could label 0 as and 1 as Keep in mind that

However, another way to label the ruler is to use decimals. We write

and and so on. Now we have our ruler neatly labeled, using decimals.

Remember, etc. To improve readability, we usually write .3 as 0.3 (“zero

point three”).

Next, we imagine a long ruler, also labeled in tenths.

Although this picture doesn’t show all the decimal labels, you can easily locate
any number such as 2.7 on this ruler. You also realize that 2.7 means two and
seven-tenths:
Notice that a whole number can also be considered to be a decimal point
number. For example, 3 = 3.0.

Addition. For simplicity, let’s stay with single-digit decimals a bit longer; the
principle is the same for all decimal-point numbers. How do we add two such
numbers, for example 1.2 + 2.5? Nothing could be simpler. First add the decimal
parts, then the whole-number parts, so 1.2 + 2.5 = 3.7. In terms of tenths

To repeat, you first add the tenths, then the units.

In some examples, you also have to carryover. Consider 3.6 + 2.8; writing this
in tenths,

But what is It’s Thus Therefore 3.6 + 2.8 = 6.4.

Fortunately, you don’t use this method of adding tenths in practice, because
the addition algorithm for decimals is just an extension of the algorithm for
whole numbers.

The new algorithm is this: line up the decimal points, and use the usual
addition algorithm.

Problem. Add 6.7 + 18.5. (Answer 25.2.)

Addition and the Ruler
Addition has an interesting and important interpretation in terms of our number
ruler. To add two numbers using the ruler, start at the first number, and then
measure to the right a distance given by the second number. For example,
consider 1.4 + 2.8 = 4.2:

Keep this in mind: addition a + b of two decimal-point numbers corresponds,

on the ruler, to starting at location a and then measuring distance b to the right of
a. This assumes that the numbers get bigger as you move right along the ruler,
and this is the usual assumption in textbooks. But a ruler could be placed
vertically, or in any direction. The point is that a + b is at a distance b out, or
forwards, along the ruler from point a.

General decimal-point numbers

What about numbers with two or more decimal digits, such as 2.85, or 2.853? We
now imagine our ruler with each of the tenths segments themselves divided into
10 equal segments of length of the original unit lengths. We can locate any
number with two decimal digits, such as 3.67, as one of the marks on this
hundredth-ruler. For three digits, as in 3.672, we need

to imagine the hundredth segments again divided into 10 segments, each of

length one-thousandth, and so on. Thus

The same addition algorithm still implies: line up the decimal points and add
as always. For example

This just keeps track of thousandths, hundredths, tenths, units, tens, and so
on.
Problem Find 261.09 + 77.62. Check using your calculator.
Powers of Ten
Perhaps you recall our discussion of the powers of 10, and how this relates to the
positional decimal notation for whole numbers. For example,

The same powers of ten scheme applies also to decimal numbers, once we use
minus exponents:

(10−1 is read as “ten to the minus one,” etc.) For example,

To make this even more orderly, let’s define 100 = 1, so we have

Then

which clearly shows the positional significance of all the digits, both before and
after the decimal point. This is our modern decimal system as used throughout
the world today.

Multiplying or dividing by 10
Earlier we explained that, to multiply a whole number (written in decimal
notation) by 10, we simply tack a zero on the end of the number. For example,
672 × 10 = 6,720. Each digit of the number gets “upgraded” when we multiply by
10. The 1s digit becomes the 10s digit, while the 10s digit becomes the 100s digit,
and so on. The new units digit is 0. Read the discussion of Multiplying by 10
again if you don’t remember this.
A similar thing happens when we multiply a decimal point number by 10.
Each digit gets “upgraded” by 1. Example:

(Do you see why 10−2 × 10 = 10−1 on the second line? This says that 10
hundredths equal one tenth. That’s exactly what hundredths are: ten hundredths
make one tenth.)
The rule for multiplying a decimal point number by 10 is: shift the decimal
point one digit to the right. Thus 4.18 × 10 = 41.8. This amounts to upgrading
each digit by one. Actually, the same rule works for whole numbers, if we
remember that a whole number is also a decimal number. For example

Problem 1.12 What is the rule, in terms of shifting the decimal point, for
multiplying by 100? by 1000? Give examples.
 If multiplying a number by 10 amounts to upgrading its digits by 1, which
means shifting the decimal point one position to the right, what does dividing by
10 do? You’re right – it “downgrades” each digit by 1, which means that the
decimal point is shifted one place to the left. For example, 37.4 ÷ 10 = 3.74. We
discuss division in Chapter 2, but this particular case is easy to describe:

Problem 1.13 Find (a) 9.03 × 1, 000. (b) 9.03 ÷ 100.

Multiplication of decimal numbers
The multiplication algorithm for decimal numbers is exactly the same as for
whole numbers, except for one detail – where to put the decimal point. Consider
the example 102.7 × 3.6. First, you do the long multiplication, ignoring the
decimal points: 1027 × 36 – you’ll get 36972. Next, count up the total number of
digits to the right of the decimal in both numbers – it’s 2. Put the decimal point 2
places from the right.

Maybe you remember this algorithm from school.

A quick approximate calculation can and should be used to check that the
decimal point is correctly located. 102.7 is close to 100, and 100 × 3.6 is 360, quite
close to 369.72. If the decimal point was anywhere else, the result would be
completely incorrect.
Why is the multiplication algorithm valid? For once I’m going to apparently
break my rule that you must understand everything. Who ever does hand
multiplication of decimal numbers nowadays? The multiplication algorithm isn’t
important for any later topics in math, so let’s not spend any more time on it,
even though it is fairly easy to show why it is correct, arguing in terms of powers
of ten. Whenever we need to do a decimal-point multiplication, we will use a
calculator.

Quick, approximate calculations

A very useful skill for everyday use in many situations is quick, approximate
calculation. I’ll feature this idea often in this and the next chapter, but let’s look
at the example 314 × 87. Very roughly, this should be approximately 300 × 90,
agree? Thus 314 × 86 is approximately 300 × 90 = 27, 000. This approximate
answer isn’t that far off the actual answer 27,318. At least it’s in the right ball
park. I hope you see how I did the mental arithmetic for 300 × 90:
 You can also do quick, approximate addition. This can be useful for checking
your bill at a restaurant, or at the supermarket. My technique at the supermarket
is to round off each item to the nearest dollar, and keep a running total as I put
items into the shopping cart. I usually come out within a few dollars of the cash
register total. If not, I question the checkout clerk – sometimes the mistake is
mine, but not always.
Try my method on the following list: $4.85, $.97, $6.28, $2.75, $3.80, $1.25. (I
got $20. The actual total is $19.90. See, it works!) Try this on your next shopping
trip. It’s fun and worthwhile.
I always do quick approximate checks of restaurant bills. Recently I got a bill
that was quite a bit larger than my estimate. It turned out I had been charged for
an extra main course! I won’t be going back there again soon.

1.6 Scientific notation

It’s a shame that so few people understand scientific notation. Perhaps the word
“scientific” scares people off. Maybe it should be called something else, like
“compact notation,” or “easy to understand numerical notation.”
Whatever it’s called, scientific notation is very useful, for example in
approximate calculations involving very large numbers. Here’s an example. The
total U.S. federal debt on June 7, 2010 was about $13,055,000,000,000, that is, 13
trillion, 55 billion dollars. The population of the U.S. in 2010 was approximately
308 million. How much was the per-capita federal public debt in that year? Your
inexpensive hand calculator can’t solve this problem, because it won’t accept the
number 13,055,000,000,000.
Another example: you may read in the paper that the nearest star, Alpha
Centauri, is about 250,000,000,000,000 (250 trillion) miles distant from earth.
This number is almost incomprehensible (although it is miniscule compared to
other astronomical distances, such as the distance to the nearest galaxy,
Andromeda, at 1,200,000,000,000,000,000 miles distance). There is really no way
to grasp the meaning of numbers like this, at least in terms we are familiar with.
However, these huge numbers can at least be written much more simply using
scientific notation, as I will now explain. (Astronomers usually quote distance in
light-years; Alpha Centauri being 4.3 and Andromeda about 2 million light years
away. That might help – if you’re an astronomer.)
 Scientific notation can be explained by an example:

First, check that this is correct: to multiply 3.5 by 108 we shift the decimal point 8
places to the right, which puts in 7 zeros after the 5, as above. “Scientific
notation” means that a number is written in the form

Two more examples:

When written as ordinary decimal numbers, these become

This can be done by shifting decimals, but the following is probably easier:
 Similarly

(You may have never before seen commas after the decimal. This is because
scientific notation is usually used for such numbers.) In cases like this, scientific
notation is much easier to read than normal decimal notation.

Problem 1.14 Express the U.S. debt for June 7, 2010 in scientific notation. Also
the distance to Andromeda.

If you have a scientific calculator, it can use scientific notation, using the E-
symbol. For example 2.6E5 means the same as 2.6 × 105. But you can also do the
calculation on your non-scientific calculator. For example, let’s calculate the per
capita federal debt in the U.S. The population is 308,000,000, or 3.08 × 108.
Therefore

You can use your calculator, for the number parts, and keep track of the powers
of ten yourself:

In June 2010 every U.S. citizen was in debt about $42,400, on behalf of the federal
government. If you’re married with 2 children, your family’s share of the debt is
$169,600. No wonder taxes are so high!
The interesting thing about this calculation is that it is virtually impossible to
do at all if you don’t understand scientific notation, but quite easy if you do. A
rough calculation might be just as informative: the debt was about 1.2 × 1013 and
the population about 3 × 108. Therefore the per-capita debt was about $.4 × 105,
or $40,000. Close enough. Again, you can’t even do this rough calculation if you
don’t know scientific notation. (Division is discussed fully in Chapter 2.)

Problem 1.15 Try to “guesstimate” the total value of all residential property in
the U.S. (or in your own country).

I hope you are convinced that scientific notation is useful, and not just for
scientists. I think it should be taught in school. Then newspapers could start
using scientific notation, both in the financial columns and in articles about
science (fat chance?). Large numbers in the paper are usually stated in words -
million, billions, trillions, and so on. For your information, here is what these
words mean, in base-ten and scientific notation.

There are more such names, for example, a quadrillion is 1015, but these are
seldom used. Billions, or trillions are large enough to describe such things as the
world’s population, and national debts. (By the way, to make matters even more
confusing, the British use the word billion to mean 1012, not 109. Universal use of
scientific notation would remove this confusion.)
Numerical precision
Mathematically speaking, the decimal number system is capable of arbitrary
precision. In real life we can never obtain (and never require) unlimited
precision. For example, no one even knows what the exact population of the U.S.
is at any given moment. The census can’t track every last American. Some people
are traveling, others are dying, others are being born. Anyway, the census is only
taken once every 10 years, so there’s additional uncertainty between censuses.
When a number is written in scientific notation, the number of digits is
called the “number of significant digits.” For example, the U.S. population figure
3.08 times 108 has 3 significant digits. If you read somewhere that the latest figure
for the U.S. population is 308,409,618, be suspicious. This figure is bogus – the
population can’t be known to the last person.
Listing the U.S. population as 308,409,618 gives a false sense of precision.
Better to simply say that the population in December, 2009 was approximately
308 million.
Similar comments apply to almost every instance of numerical data,
including financial data, scientific values, and so on. The precision with which
such values are listed should give some indication of the accuracy with which the
values are known.
Next, what about addition and multiplication, in terms of significant digits?
How much is

The answer may surprise you: 4.21 × 106. Purely mathematically, you would
get 4.21033 × 106, but this is over precise. To leave the answer in this form would
imply a precision of 6 significant digits, whereas the original data are only precise
to 3 digits.
Multiplication is treated in the same way:

Your calculator will give 4.21 × 3.3 = 13.893, but this is once again over precise,
and should be rounded off to 13.9. (“Round off to 3 figures” means to replace
13.893 by the closest 3-digit number, 13.9. Scientific calculators can be set to
display a given accuracy, and will then automatically round off all results to this
accuracy.)
Remember, avoid over precision when working with actual data, which are
always of limited precision. And be suspicious of data or statistics presented with
high precision. In many cases they will have been calculated without due
attention to the limits of precision involved. Calculators and computers can give
answers with many digits, but some of these digits will not be meaningful in
terms of the known precision of the input numbers.
Orders of Magnitude
Scientists and other people sometimes use the term “orders of magnitude.” To
say that A is about 3 orders of magnitude larger than B just means that A is
approximately 103 × B. For example, let’s compare the population of China with
that of New York City. In rough figures, China has about a billion (109) people,
and New York about ten million (107). Therefore the population of China is
about two orders of magnitude greater than that of New York, or about 100
times as large.

Another example: compare the speed of light to the speed of sound. At sea
level, sound travels about 300 meters/sec; the speed of light is 300,000 km/sec.
Thus we are comparing (in meters/sec)3 × 108 with 3 × 102, so that the speed of
light is about six orders of magnitude greater than the speed of sound at sea level.

An order-of-magnitude statement gives a very rough idea of how large two

quantities are in relative terms.

Problem 1.16 (a) Using a calculator, find 1.76 × 35.3, rounding off appropriately.
(b) A chickadee weighs about 10 grams, and a swan about 10 kilograms. By how
many orders of magnitude does the swan’s weight exceed the chickadee’s? (One
kilogram equals 1,000 grams.)

1.7 The rules of arithmetic

The rules of arithmetic – that is, Rules 1 to 5 discussed earlier – are very
important throughout mathematics. Using these rules incorrectly in Algebra is
one of the leading causes of failure among college math students. Having the
logical explanation of these rules in mind, even if way back in your mind,
provides confidence in understanding what you’re doing later, and helps avoid
errors that can result from insecurity. Well trained students virtually never make
errors in using these rules.

Fortunately, the five Rules of Section 1.1 remain valid for all decimal
numbers. As a reminder, here they are again:

Here the five rules are written in the simplest possible way. For example, ab
means a × b. Also the expression in Rule 5, ab + ac means (a × b) + (a × c). This
is in accord with the Precedence Rule: × precedes +.

Since these rules are important, you should again understand why they’re
true. You remember that, for whole numbers (which are the counting numbers),
each rule is a consequence of the counting principle. But decimal numbers are
also a kind of counting number. They don’t count units, but tenths, hundredths,
and so on.

For example, consider the number 6.4. This equals which is the same as
that is, 64 tenths. (Why? Well, ten tenths equals 1, so 60 tenths equals 6.

Therefore Or, you can just remember that

by shifting the decimal.) This means that 6.4 can be thought
of as counting up 64 tenths. What does 2.576 count? Thousandths, namely 2,576
thousandths. In other words, any decimal number counts a kind of mathematical
objects. What the objects are depends on how many digits there are after the
decimal point.
How can we now conclude that, for example, 6.4 + 2.75 equals 2.75 + 6.4,
based on the counting principle? We have seen that 6.4 counts tenths, while 2.75
counts hundredths. Do you see how to consider that both 6.4 and 2.75 count the
same kind of mathematical objects?
The answer is that we can write In other words, both numbers can
be thought of as counting hundredths. Therefore the counting principle
guarantees that 6.4 + 2.75 = 2.75 + 6.4.
Surely the above argument works for any two decimal numbers a and b. The
same argument also works for each of the Rules 2-5, which are all consequences
of the counting principle. Without further ado, we can conclude that the five
rules of arithmetic are valid for all decimal numbers.

1.8 The number line

The decimal point number system discussed in this Chapter consists of all
decimal numbers, such as 306.2174 for example. Any such number is expressed
as a string of digits, with a single decimal point (which may be omitted if the only
digits after the decimal are zero).
These numbers can all be located on the number line, which extends
indefinitely far to the right, and includes indefinitely fine subdivisions. To give
the number line a starting point, we also include the number 0 (zero) at the left
end, as shown.

What other kinds of numbers are there? This will be discussed fully in
Chapter 2; the other types include
 negative numbers
infinite decimals, either repeating or non-repeating.
All these numbers correspond to points on an extended number line, as will
be explained in the next chapter.
Larger and smaller

The symbol < is called the inequality sign, but we read it as “is less than” or
“is smaller than.” For whole numbers a and b, the statement a < b means that a
occurs before b, in terms of counting. Thus 17 < 19. Also, because decimal
numbers can be thought of as counting numbers, we can readily compare any
two decimal numbers. For example, 2.05 < 2.1 because while

Deciding which of two decimal point numbers is the smaller is immediately

possible by inspection; this is yet another advantage of the decimal system. In
words, given two numbers a and b, we first look at the whole number parts (i.e.,
the digits before the decimal point). If these are different, then a and b have the
same relative order as their whole number parts. For example

If the whole number parts of a and b are equal, we look at the tenths digit:

If the tenths digits are the same, we compare the 100ths digits, and so on. (Yet
another algorithm!) Example: 2.1409 < 2.141.
In terms of the number ruler (pointing to the right) a < b means that a lies to
the left of b.
 Sometimes it is convenient to use the symbol >. Thus a > b is read as a is
greater than b, or also a is larger than b. To be specific,

Example: 56 > 55.7.

The use of inequality signs may seem rather pointless at this stage. However,
they are useful in later Chapters, so it’s a good idea to become familiar with them
now. Here are the basic Laws of Inequalities:

All numbers considered so far are positive. However, negative numbers will
be studied in Chapter 2, and Rule 2 is not valid if c is negative.
These rules are almost self evident for the case of whole numbers a, b, and c.
For example, Rule 3 says that if a comes before b, and b before c, then a comes
before c. Rules 1 and 2 are equally obvious. To show that the rules are also valid
for decimal point numbers, we remember that a decimal point number can be
thought of as a type of counting number. Therefore the rules also apply to such
decimal point numbers.

Problem 1.17 List all whole numbers x for which 9 < x < 15. (This double
inequality means that 9 < x and x < 15.) How many decimal point numbers
satisfy this double inequality?

Problem 1.18 Using Rules 1-3, show that: if a < b then a2 < b2. (Hint: Use Rule 2
twice.) This might stump you for a while! Assume that a and b are positive.
1.9 How to learn mathematics
Most people find mathematics confusing, at times. This book tries to eliminate
confusion as much as possible. Nevertheless, you the reader may still get
confused occasionally. Everyone who studies mathematics experiences mental
blocks once in a while, something that just doesn’t seem to make sense. This is
not a sign of stupidity! The question is, what should you do to overcome a mental
block? Here are three possible approaches:
1. Forget it, it’s probably not important anyway.
2. Forget about understanding the point, just memorize the result.
3. Take the time to identify the difficulty, and then try to resolve it.
One of the reasons, I think, that many people “drop out” of math is that they
start adopting strategies 1 or 2. Either of these strategies is a sure-fire recipe for
eventual failure. In learning mathematics, any point of confusion must be
eliminated as soon as it occurs. If not, everything that follows on from the point
of confusion will also probably be confusing. Pretty soon the whole subject
becomes incomprehensible – and then hateful.
Suppose that you have encountered a difficulty, and that you wish to adopt
strategy 3. How should you proceed? First of all, you may not even know that
you didn’t fully understand some particular point. The Problems that occur
throughout this book are designed to help you to quickly recognize any points of
confusion. Be sure to pause and solve these Problems. Don’t be too hasty in
looking up the solution, especially if the Problem seems confusing. Try hard to
work out the solution on your own. Then check, and make sure you understand
the given solution. Can you make up another similar problem, and solve it?
Still, you may sometimes encounter a confusing point not related to any
given Problem. While reading this book, you should always be asking yourself
“Do I understand this point?” Hopefully, the answer will usually be a firm “Yes.”
If not, here are my suggestions.
1. Pause immediately. (Well, you might just glance at the next sentence to see
if it explains the problem.)
2. Try to pinpoint exactly what the difficulty is. If it’s an unfamiliar word, try
to find out where in the book the word was first used. Look in the index.
3. If it’s a logical, or “mathematical” question, try to express your difficulty in
simple words. Or, if it’s a certain sentence that doesn’t make sense, find out
 why not. Re-read the whole paragraph. Rewrite the sentence another way.
4. Perhaps the problem is one of ambiguity - something that could be
interpreted in two or more different ways. Write out both, or all, reasonable
interpretations. Are they really reasonable? Which, if any, is likely the
intended meaning? (We authors are sometimes guilty of ambiguity, no
matter how hard we try to avoid it. No one’s perfect!)
5. If the difficulty is that the mathematics has become too complicated for you
at this point, see if you can invent a simpler problem based on the same
idea. Try to solve the simpler problem – maybe this will give you the clue to
the more complicated one.
6. If nothing works, put the book aside until tomorrow. Then try rereading the
whole section. (Of course, you can also think about the matter in the
meantime, for example just before going to sleep. I often used this approach
successfully when I was in graduate school.)
Everyone has difficulties with mathematics at some time, even professional
mathematicians. Overcoming these difficulties is a challenge. At first your self-
confidence plunges. You may get angry at yourself. But persevere. Eventually you
will crack it, often with a sudden flash of insight. Then the elation will be
sublime. Or maybe you will kick yourself for being so stupid. In any case,
overcoming a mathematical difficulty is an encouraging experience, one that will
improve your confidence in the future. Don’t ever say to yourself that you can’t
“do” mathematics. Worst of all, don’t ever abandon understanding in favor of
memorization.
The next time you seem to be confused, try the 6-step procedure described
above. Remember, no one is immune to the occasional “mental block” while
studying mathematics. I bet not one student in a hundred ever makes a
conscious effort to overcome mental blocks. No math teacher of mine ever
warned me about blocks, or explained how to fix them. There are countlessly
many ways to misunderstand a given mathematical topic, far too many for a
teacher (or book author) to anticipate them all. Everyone’s blocks are different.
You have to learn to recognize them, and then overcome them. Any time you
seem confused, stop and try to discover what’s confusing you. The six-step
approach works for me, and I recommend you try it.

1.10 Review problems

1. Write out the following powers of 10 in full decimal notation, and express each
number in words: 104, 108, 1010.

2. Multiply each number by 1,000: 846, 0.0372.

3. What are the five Rules of Arithmetic discussed in this chapter? (Write them
in symbols, not by name.)

4. Show how the Distributive Law a(b + c) = ab + ac follows from the counting
principle. (Use a diagram.)

5. Explain why 105 × 103 = 108. What is the general formula?

6. (a) The distance from earth to the sun is about 93 million miles. Express this
distance in kilometers, using scientific notation. (One mile is about 1.6 km.)

(b) If I told you that 1 mile is 1.60934 km, what would your answer be?

7. Review the basic definitions of addition and multiplication for whole numbers,
based on the counting of objects.

8. (a) Write in ordinary decimal notation: (6 × 102) + (9 × 101) + (3 × 100) + (5 ×

10−1).

(b) Write out in terms of powers of 10 (as in part a): 3.1416.

9. Calculate (either by calculator, or by hand):

(a) 2.7 × 3 + 4.8 × 0.9;

(b) 3 × (15 + 8 × 8);

(c) 6.3 × 103 × 1.8 × 106.

10. [Optional] Computer programmers sometimes use the octal (base 8)) system.
Write out the first 20 whole numbers in octal notation.
11. List in increasing order: 30.7, 84.1, 0.9, 16.5, 0.99.

12. Use the Distributive Law a(b + c)= ab + ac to prove that a(b + c + d)= ab + ac
+ ad.

Note: Solutions to the Review Problems are given at the end of the book.
Solution 1.1 The first ten powers of 2 are: 2, 4, 8, 16, 32, 64, 128, 256, 512, and
1024.

Solution 1.2 First, we have 1010 = 10,000,000,000, or ten billion. Therefore

= 1010,000,000,000. Now suppose you were to try to write this out: it would be 1
followed by 10 billion zeros. How big a number is that? Well, one page of this
book holds about 2,000 letters or digits. So you’d fill 10,000,000,000 ÷2, 000 or
about 5,000,000 pages – about ten thousand books! Just to write the number
down! Isn’t mathematics fantastic? [The number is sometimes called the
“googol.”]
Solution 1.3 (a) If the final digit is not 9, change the given number by increasing
the final digit by 1. Example: 34 + 1 = 35. (b) If the final digit is 9, replace the
final digit by 0 and inspect the previous digit. (c) If this digit is not 9, change the
given number by increasing this digit by 1. Example: 239 + 1 = 240. (d) If this
digit is 9, repeat the instructions from (b). Example: 399 + 1 = 400. (e) If you run
out of digits (which can only happen if all the digits were 9s), replace the last 9 by
10. Example 99 + 1 = 100.
Solution 1.4

The algorithm is:

1. Write the numbers, one above the other, keeping the positions of the digits
(units, tens, hundreds, etc.) aligned. Draw a horizontal line.
2. Start by adding the units digits, writing the units digit of the sum below the
line. Carry over any tens digit to the tens position.
3. Repeat step 2 for the tens position (including the carryover, if any), carrying
over now the hundreds position.
 4. Repeat these steps until finished.
(Note that the same algorithm applies to adding more than two numbers.)
Solution 1.5 Often the best way to explain something is to look at a simple
example that illustrates the general principle. Let’s consider the sum 17 + 46:

This shows how the 10 in 13 (obtained from 7 + 6) gets “carried over” to the 50,
making it 60. This is how carryovers work in general. You start by adding the
units position because any carryover from there may affect the sum in the tens
position, including the carryover, and so forth.
This argument applies to any addition, and explains the usual way of doing it, as
in Problem 1.4. Try a more difficult example, first using the usual carryover
method, and then writing out the units 10s, 100s, and so on. Use a calculator to
check the answer.

Solution 1.6 2 × (15 + 8) = 2 × 23 = 46, but (2 × 15) + 8 = 30 + 8 = 38. They’re

different.

Solution 1.7 By definition a × b is the number obtained by counting a groups of

b objects each. Similarly b × a is the number obtained by counting b groups of a
objects each. Consider the example 3 × 5 (the argument applies equally to any
product a × b). Arrange objects (dots) in 3 rows of 5:

If we count these objects row by row, we are counting 3 groups of 5 objects, i.e. 3
× 5. And if we count them column by column we are counting 5 groups of 3
objects, i.e. 5 × 3. By the counting principle, these must be the same so 3 × 5= 5 ×
3.
(At this stage, some readers may be asking “What’s the point? I know that 3 ×
5 = 15 and 5 × 3 = 15, so they’re the same. Why do I have to draw all those dots?”
What would you say to help such a reader?)

Solution 1.8 (a) 12 + 18 = 30; (b) 3 × 6 × 9 = 162; (c) 3 × (4 + 18) = 3 × 22 = 66.

Solution 1.9 (a) 3xy + 12x; (b) 28p + 14pq.

Solution 1.10 1, 10, 11, 100, 101, 110, 111, 1000, 1001, 1010.

Solution 1.11 The way to do this is to find the biggest power of 2 in the given
number, subtract it, and repeat the process:

Therefore 54 equals 110110 (base two).

Solution 1.12 To multiply a number by 100, shift the decimal point two places to
the right, appending zeros if necessary. Examples: 4.18 × 100 = 418; 6.5 × 100(=
6.50 × 100) = 650.
Solution 1.13 (a) 9.03 × 1, 000 = 9, 030. (b) 9.03 ÷ 100 = .0903. (If you didn’t get
these answers, try writing the calculation out in terms of powers of 10, as in the
text. Then do the calculation by shifting the decimal point.)

Solution 1.14 The debt was $1.3055 × 1013. The distance to Andromeda is 1.2 ×
1018 miles.
Solution 1.15 First, I guess that, on average about 3 people live in each
residential unit. Therefore the number of residential units (houses and
apartments) in the U.S. is approximately 3 × 108 ÷ 3= 108. Next, I suppose the
average unit to be worth about $200, 000 = $2 × 105. Therefore, the total value of
U.S. residential property is something like $108 × 2 × 105 = $2 × 1013 (in words,
20 trillion dollars). You can see that this is about times the federal debt of $13
× 1012, which puts the debt in some perspective. It also suggests that the debt
can’t be allowed to grow much larger, or else Americans will be paying about as
much to service government debt charges as they pay on their mortgages.
Solution 1.16 (a) 62.1. (b) 3 orders of magnitude.
Solution 1.17 The whole numbers are x = 10, 11, 12, 13, and 14. There are
infinitely many decimal point numbers between 9 and 15.

Solution 1.18 If a < b then a × a < b × a (Rule 2), or a2 < ab. Also, since a < b we
have a × b < b × b,or ab < b2. Therefore, by Rule 3, a2 < b2.
Chapter 2
Subtraction and Division

2.1 Subtraction
In popular language, to subtract means to take something away. This is also the
mathematical meaning of the term. For example,

says that “12 take away 5 leaves 7,” although we more commonly say “12 minus 5
equals 7.” If we start with 12 items and then remove 5, there will be 7 items left.
This is the basic definition of subtraction, in terms of counting:

Now, if 5 objects are removed from 12 (leaving 7) then replacing the 5 objects
restores the original number, 12. In other words

We could just as well use this idea as our basic definition of subtraction:
Please check that the statement in Box 2.2 exactly fits the preceding example.
Let me reassure the reader about the use of symbols m, n and p here. Making
sense out of a mathematical statement containing symbols always requires a
special effort. You first need to understand that the symbols stand for
unspecified whole numbers. To comprehend what the statement means, it helps
to pause and make up a few numerical examples, such as

which fits the pattern

So now you realize that subtraction is intimately related to addition, and Box
(2.2) specifies how.
Many people find it easier to remember Eq. 2.2 if it is written as

This at least keeps the letters in the same order. Also, in Chapter 4 (Algebra) we
will learn about “transposition” in equations, and Eq. 2.2a involves the
transposition of the symbol n across the equals sign.

Problem 2.1 Use Eq. 2.2a to check that (a) 9 − 8 = 1; (b) 92 − 45 = 47; (c)28 − 13
≠ 14.

Problem 2.2 What goes awry with the definition of subtraction if m < n?
Consider some examples.

Equation 2.2 or 2.2a can be understood as saying that subtraction is the

reverse operation to addition. The following figure explains this idea graphically.
Adding 5 to a number moves you up a distance 5 on the number line.
Subtracting 5 from a number moves you down by 5. (Question: is this numerical
example generic?)
One question I’m often asked, after being introduced to somebody as a math
prof, is “Oh, Dr.Clark, could you explain to me why two minuses make a plus? I
always found this confusing at school.” Fortunately, this is easy to explain (in
fact, there are four explanations in this chapter!). Consider an example:

Remember, first you must do the calculation inside the brackets. Therefore 12 −
(7 − 4) = 12 − 3 = 9. But this is the same as writing

(both are equal to 9). The two minus signs in front of 4 have turned into + 4! In
other words, in 12 − (7 − 4) you first take 4 away from 7 before taking anything
away from 12. From 12 you take away 4 less than 7. It’s exactly the same as
adding 4 to 12 − 7.
At this stage I usually get a bemused “Oh, thanks,” as the new acquaintance
ambles off. No, seriously, this is the explanation. I’m sure it makes sense to you!
Please pause to make sure that you do understand the way the number line
works here. Adding corresponds to moving up the number line, while
subtraction moves down. We will use this idea often in this chapter.
 Showing that two minuses make a plus: 12 − (7 − 4) = 12 − 7 + 4

Next we consider the subtraction algorithm. Start with an example:

I hope you remember (at least vaguely) how to do this. Also, remember that
you can immediately check the answer by addition: 233 + 82 = 315.

Problem 2.3 Write out the subtraction algorithm (for whole numbers) in detail.

Negative numbers
The “take-away” concept of subtraction doesn’t help if the number to be
subtracted is larger than the number it’s subtracted from. For example, how
much is 12 − 23? The counting idea in Box 2.1 doesn’t apply here. But there are
lots of practical situations where we want to be able to subtract a larger number
from a smaller one. One example is temperature. Financial gains and losses is
another. (See Chapter 3 for other examples)
Negative numbers provide the answer. They are best understood in terms of
the extended number line:

Here, our previous number line including 0, 1, 2, 3, etc., is now extended in

the opposite direction (here, to the left), to include the negative numbers −1, −2,
−3, etc. These negative numbers are new to our system. There is also the number
0, which is neither positive nor negative. The system of all these numbers is
called the system of integers.
Now we can immediately go ahead and also include negative decimal
numbers in our system. Any negative number, for example −1.7, is located to the
left of 0, at a distance equal to its positive counterpart (1.7). In other words, any
positive number (e.g., 1.7) and its negative counterpart (−1.7) are located
symmetrically relative to 0 on the number line. The point 0 is called the origin of
the number line.

Problem 2.4 Draw (a freehand sketch is OK, or you could use a ruler) a vertical
line, pointing upwards. Locate points 0 and 1 arbitrarily on the line. Now locate
(approximately) the numbers 0.3, −0.8, −2.25 on this number line.

We will call the system of all positive and negative decimal point numbers
(including 0) the full decimal-point number system. This number system
includes the following kinds of numbers:
whole numbers: 0,1,2,3, …
integers: …, −3, −2, −1, 0, 1, 2, 3,…
decimal-point numbers (positive, negative, and zero): ±n.d1d2 …dk
A positive number, for example 1.6, can be written with or without a + sign
(1.6= + 1.6), but we normally omit the + sign. Of course, a negative number
always has the − sign. (Many school textbooks insist that positive and negative
numbers are indicated by small + or − signs written above and before the digits:
+
5.1, −2, etc. These would be called “positive 5.1” and “negative 2.” This notation
is unnecessary, and will not be used in this book. It is almost never encountered
in later math courses, or in science or economics. Some people do say “negative
2” instead of “minus 2.” Use which ever you prefer.)
Using negative numbers, we can now subtract without constraint. For
example, 7 − 10 = −3. You just figure out the deficit. How much is 12.3 − 18.6?
How do you calculate it? What’s the general method? First, calculate 18.6 − 12.3=
6.3, then attach a minus sign. This can be written as

The general method is the same. To calculate a − b (where a and b are positive
numbers):

Examples: 7.1 − 6.8= ? 6.1 − 7.8 = ? (Answers: 0.3 and −1.7.)

Arithmetic for the full system

To repeat, using negative numbers allows us to calculate a − b even if b > a. We
now need to develop the complete arithmetic of this system, including addition,
multiplication, subtraction, and division. This is fairly large undertaking, which
will occupy us for the rest of the chapter.
Order
The number line indicates that the decimal-point number system is “linearly
ordered,” as we say. This means that, given any two different numbers a and b
(positive, negative, or zero) we have either a < b or b <a. (a < b is read as “a is
less than b,” or “a is smaller than b.”)

In any instance, a and b are given in decimal notation. We can tell by

inspection which is the larger. I hope this is clear to you without further
discussion. Pay careful attention to the next problem.

Problem 2.5 Arrange the following numbers in increasing order: 5.6, −2.1, −8.3,
0, −8.8.

Sometimes it is convenient to use the symbol > (greater than)

For example, 6 > 2, and −5.1 > −5.11 (right?).

We often use the symbols a > 0 rather than the phrase “a is positive.”
Similarly, a < 0 means the same as “a is negative.”

Magnitude
The magnitude of a number is its absolute size, regardless of sign. The symbol |a|
is read as “the magnitude of a” (or else “the absolute value of a”), and |a| equals
a itself if a > 0 (or a = 0), and |a| equals the positive counterpart of a if a < 0. For
example,

In terms of the number line, |a| measures the distance from the point a to 0,
regardless of direction.

Addition
Addition is easy to understand in terms of the number line.

In other words, adding a positive number moves up the number line, whereas
adding a negative number moves down the number line. For example

How about (−8) + 3 or (−8) + (−3)? I hope you get the answers −5 and −11,
respectively.

Problem 2.6 Add: (a) (−5) + (−8); (b) (−5) + 8; (c) 5 + (−8); (d) 8 + (−8). Also
add these in commuted order: (−8) + (−5), etc.

As you can see, it is always possible to add two numbers by referring to the
number line. But we need an algorithm for addition, which will give the correct
sum without needing to bring in the number line. Here is the complete addition
algorithm for adding a + b, for decimal point numbers:

Cases 1 and 2 are not new. Cases 3 to 5, which involve adding a negative number,
are new. Here are some generic examples of these cases:
Case 3: (−9) + (−6) = −(9 + 6) = −15.
Case 4: 4 + (−13) = 4 − 13 = −9.
Case 5: (−20) + 15 = 15 − 20 = −5.
Please pause to check (i) that the answers agree with the up-or-down the number
line prescription given in Box 2.4, and (ii) that the Addition algorithm of Box 2.5
has been correctly used.

Problem 2.7 Calculate by using the Addition algorithm. Check using a

calculator. (a) (−3.9) + (−7.7); (b) (−8.5) + 12.2; (c) 15.8 + (−22.0).

Let’s emphasize the fact that

 I expect that you consider this to be obvious without any discussion. For
example, 5 + (−5) = 0 – but why, exactly? Does this follow logically from the
basic definition, Eq. 2.4, or from the Addition algorithm, Eq. 2.5? Yes, both
actually. But I am going to be Mr. Nice guy, and omit any more discussion of the
point. If you’re curious, probably you can complete the argument for yourself.
Next, you might be wondering, what’s all the fuss? Isn’t a + (−a) = a − a
which is zero for sure? Yes, OK – except that we haven’t yet defined subtraction
for the full number system. We come to that topic soon.
Let’s talk briefly about concepts and algorithms. The concept of adding two
numbers a + b is defined in terms of shifting positions up or down the number
line. The above algorithm (Cases 1–5) is entirely “mechanical,” and, when used
to calculate, doesn’t depend on the number line at all. However, the validity of
the algorithm follows logically from the number line concept expressed in the
definition of addition.
Similar situations have often been encountered in this book. For example, the
concept of multiplying two whole numbers is related to counting up groups of
objects, whereas the multiplication algorithm is mechanical, and by-passes
counting entirely. Understanding math requires that you understand concepts,
as well as learning algorithms. This book tries to emphasize concepts and
algorithms equally, in line with the philosophy that learning mathematics
demands both understanding and mastery of technique. However, I expect you
to use your calculator freely, at least in later chapters. You also need to master
that technique.

The commutative and associative laws

Two of the basic laws of arithmetic are:

We discussed these laws for positive numbers in Chapter 1. They both

remain valid (as do the other laws of arithmetic) for the full system of decimal
point numbers. For example, 3 + (−7) = −4 and (−7) + 3 = −4. You could try a
couple of other examples, if you’re interested. But why, exactly, do these basic
laws remain valid for all numbers? What’s the logic?
Establishing that these two laws are valid, case by case as specified in the
addition algorithm, is possible, but tedious. For example, let’s check the
commutative law for case 4 (a > 0, b < 0). The addition algorithm says that a + b
= a − |b| in this case. What about b + a? This fits into case 5 (b < 0, a > 0), but
with the roles of a and b reversed (check this point). According to case 5,
therefore b + a = a − |b|. This shows that indeed a + b = b + a, for this case.
Similar, careful arguments can be used to check the four other cases, and also to
laboriously check the associative law. If you are interested, you can try one more
case. Without further ado, we will now accept the validity of these two laws of
arithmetic for the full number system. Try one example:

You probably figured this out as (8 − 5) − 6= 3 − 6= −3. You could have also
done the calculation as 8 + ((−5) + (−6)) = 8 + (−11) = −3. This illustrates the
associative law.

Problem 2.8 Calculate (with care): (a) (−17) + 6 + (−9); (b) 82 + (−101) + (−5);
(c) (−33) + (−34) + (−35).

Subtraction
Subtraction of general numbers is defined as the inverse of addition.

This definition of subtraction is very important, and should be memorized. It

means that subtraction is the inverse of addition, but states exactly what is meant
by that language. Note that equation 2.7 is the same as Eq. 2.2a.
 Some generic examples (check these carefully):
1. 6 − 4 = 2 because 6 = 4 + 2
2. 6 − (−4) = 10 because 6 = (−4) + 10
3. (−6) − 4= −10 because −6= 4 + (−10)
4. (−6) − (−4) = −2 because −6= (−4) + (−2)
Note especially that:

Check that this is correct for the above examples.

Now that we understand the basic definition of subtraction, we need an
algorithm. (You will recall that we have followed the same procedure –
definition, then algorithm – on many occasions previously.) Nothing could be
simpler:

This says, to subtract b you simply add −b. We already have an Addition
algorithm, Box 2.5, so Eq. 2.9 is all we need for our subtraction algorithm. Again,
check the above generic examples in this light, especially examples (2) and (4).
For example (2):

Note especially that

– again, two minuses make a plus! In general,

 How do we know this, you ask? Well, we can write −(−b) = 0 − (−b), and

by Eq.2.7, and (−b) + b = 0 by Eq. 2.6.

An aside here: you may feel a bit confused about the minus sign, −. Does this
mean minus (as in −4), or does it mean subtract (as in 6 − 4)? The answer is,
both. The number −4 is four units down from 0 on the number line, and the
number 0 − 4 is also four units down from zero. Moreover, Eq. 2.9 tells us that 0
− 4 is the same as 0 + (−4), and the definition of addition, Box 2.4, Case 3, tells us
that this number is also located four units down from zero. Everything fits like a
glove.
So how should you read 6 − (−4)? Just as “six minus minus four.” If you
absolutely insist, you could read it as “six subtract negative four,” but I don’t
think that terminology helps at all.

Problem 2.9 Calculate (a) mentally: 12 + (−15); −25 − (−8); 61 − (+ 20); −33 −
(−33); (b) by calculation: −14.73 + (−8.07); 521.9 − 7, 650.0; 24.75 − (−24.65);
−1.87 + 7.63 − 5.09.

In Chapter 4 you will learn about “transposing” terms across the = sign.
When a term is transposed, it changes sign. Thus, from a + (−a) = 0 we get a = −
(−a), by transposing the term (−a). This is probably how most students think of
“two minuses equal a plus.” There are many explanations for this fact. Yet
another explanation is based on the number line.

Addition, subtraction, and the number line

To subtract a positive number you move down (left) that amount. This is the
reverse of adding the given number.

To subtract a negative number, you go up (right) by its magnitude. This is

the reverse of adding the original negative number.

Study these two diagrams carefully, as they fully explain the relation of
addition and subtraction in the full decimal-point number system.

Problem 2.10 Calculate (carefully!) (a) −1.8 − (−7.4 − 7.6). (b) −(4.3 − 6.1) + (2.7
− 7.0). (c) 3.5 − (2.6 − (4.2 − 5.8)).

A general pattern worth taking note of is:

Note how “two minuses make a plus” comes in here. Alternatively, you can think
of the “take away” example given at the very beginning of this Section.

2.2 Multiplication
Our next task is to discuss multiplication for the full number system. What
would we want 6 × (−9) to be? It should be 6 times as negative as −9, do you
agree? Therefore
You just multiply 6 × 9 and tack on a minus sign.
This takes care of multiplying a positive times a negative number. Next, how
about (−5) × (−7)? Well, first (5) × (−7) = −35. We would therefore expect that
(−5) × (−7) = 35. Each minus sign reverses direction on the number line. Two
minus signs, when multiplied, reverse direction twice, which amounts to no
change in direction.
You might be wondering, does this prove that two minuses (when multiplied)
make a plus, or is it just an assumption? The answer is that it is a definition,
motivated by the need for consistency, as in the example.
Our complete definition of multiplication is:

Note that the definition of multiplication is itself the Algorithm for

multiplication. Roughly speaking, the algorithm is that a × b is calculated by first
multiplying the magnitudes |a| × |b|, and then attaching a plus or minus sign
using a simple rule.
Here are the remaining three Laws of Arithmetic. In the equations, a, b and c
stand for arbitrary decimal point numbers (positive, negative, or zero).
Problem 2.11 State explicitly all the conventions that are used in writing these
formulas.

Let’s get some practice using these Laws. First, with numbers:

Problem 2.12 Calculate: (a) −2 × (13 − 6 × 6); (b) (−8 + 3) × (3 − 3) + 7; (c) (−5 −
2) × (−8 + 2).

Next, with symbols:

Be sure you follow these examples. Chapter 4 goes into Algebra in much more
detail.

Problem 2.13 Remove brackets: (a) 2c − 3(c − d); (b) x(y − 2z); (c) (w + 2)(w −
3).
 How do we know that the above laws of arithmetic are valid in the full
number system? A professional mathematician (like me) would be able to give a
completely logical proof, but no self-respecting student (like you) would sit still
for it. It’s sufficient that you remember how these rules were justified for whole
numbers, and accept as reasonable the statement that the Laws of Arithmetic are
universally valid. Of course, you might like to check out a few examples chosen
at random. For example, let’s check that 2 × (3 − 5) = (2 × 3) + (2 × (−5)):

(Remember, you could do a thousand examples such as this, but that still
wouldn’t prove that the five Laws of Arithmetic are valid.)

Here are two special cases to keep in mind:

I expect that you consider these equations to be obvious, but from strictly logical
standpoint we should prove them by using the basic definition of multiplication,
Eq. 2.11. Let’s start with the first, 1 × a = a. This is certainly true if a is positive,
or zero (remember that 1 × a is what you get by counting up one group of a
objects). If a is negative, Box 2.11 says to first multiply 1 × |a|, which equals |a|,
and then attach a minus sign:

But when a is negative we have − |a| = a, right? (Try a generic example if you
don’t see this.) This completes the proof of Eq. 2.12.
 Equation 2.13 is just a restatement of the first Convention listed above.
To end this section, let’s make sure that you fully understand the important
special result

(This is another of those stumbling points for some math students.) What you
need to remember from now on is that the minus sign (−) can always be
interpreted as (−1) × , and also that two minuses when multiplied always give a
plus. The fact that −(b − c)= −b + c is then apparent, by applying the Distributive
Law: −(b − c)= (−1)(b + (−1)c)= −b + c.

Problem 2.14 Calculate (a) 7 − (12 − 15); (b) (7 − 12) − 15; (c) 7 − 12 − 15; (d)
−7 − 12 − 15.

2.3 Division and fractions

When children first learn about subtraction, the sequence of steps they go
through is:
1. subtraction of whole numbers as “take away.”
2. subtraction as the inverse of addition.
3. subtraction with “deficits.”
4. negative numbers (and zero), i.e. integers.
5. addition and subtraction for the system of positive and negative integers.
6. multiplication of positive and negative integers.
This is the same sequence of steps that we have gone through in the first part of
this Chapter (with the additional twist of including decimal point numbers).
A similar progression of steps is encountered in learning about division:
1. division of whole numbers as “divide into.”
2. division as the inverse of multiplication.
3. division with remainders.
4. fractions, or rational numbers.
5. multiplication and division for the system of rational numbers.
 6. addition of rational numbers.
We will follow the same steps.
Before starting, however, let me throw out some comments. The number
system is certainly pretty complicated. It took mankind many centuries – no,
many millennia – to perfect the number system we use today. It takes many years
of schooling to absorb and master all the ideas and techniques. No wonder some
students come “unstuck” somewhere along the way. What can easily happen is
that some important detail is not fully understood. Whether this was the
student’s fault, or nobody’s fault doesn’t matter. The point is that failure to
understand one essential detail eventually snowballs into widespread confusion,
effectively ending that student’s academic career as far as math goes.
Recovering the lost thread, and then re-learning the subject from that point
on, is not easy. It’s what this book is about. It’s why the book continually stresses
understanding of details. I do hope it’s working (and will continue to work) for
you.

Division of whole numbers

In everyday parlance, to divide a group of m objects means to partition it into a
number of subgroups, all having the same size. For example, 15 candies could be
divided among 5 children, with each child getting 3 candies. This is written as

or sometimes, 15/5 = 3. Both are read as “15 divided by 5 equals 3.” Also, 15/5 =
3 can be read as “15 over 5 equals 3.”
Now, if the 5 groups of 3 candies are counted up together, there will of course
be 5 × 3 = 15 candies altogether. Thus

In other words, division is the inverse of multiplication. In general,

 Thus 30 ÷ 6 = 5 (because 6 × 5 = 30), and 77 ÷ 11 = 7. Equation 2.14 can also
be written in the form

Problem 2.15 (a) How could you check that 27, 024 ÷ 563 = 48 without using
long division or a calculator: (b) List all numbers n that divide (i.e., evenly) into
30, and find q = 30 ÷ n for each such number. Same for 31.

This is the first sense in which children learn about division. They say that “5
divides into 15 three times.” However “7 doesn’t divide into 15,” or “you can’t
divide 15 by 7.”
Later on, the children are told that you can divide 7 into 15 after all, if you
use “fractions.” Or, as an alternative, you can use “remainders.” Then (“15
over 7”) is an example of a fraction, whereas 15 ÷ 7 = 2 with remainder 1 is an
example of division with remainder.
Are these two types of division related, or perhaps the same? In practical
terms, if there are 15 candies and 7 children, then each child can get 2 candies
and one will be left over. This is quite different from each child getting
candies, which would mean that each child actually got candies. The one
leftover candy is sliced into 7 equal pieces and distributed. However, the two
types of division are closely related mathematically, as we shall see.
First, we consider division with remainder.
 Here the condition 0 ≤ r < n means that the remainder r is greater than or
equal to zero and also less than n. Both q and r are assumed to be whole
numbers, or zero. Also, in Box 2.15, n is called the divisor (it does the dividing)
and m is the dividend.
For the candies example we have

In other words, 7 goes into 15 two times with remainder 1. Check that this agrees
with Box 2.15, with quotient q = 2 and remainder r = 1. Another example: divide
53 by 8, with remainder. Answer: 53 = 6 × 8 + 5. To do this division mentally,
you think of the candies example: 8 goes into 53 six times, leaving remainder 5.

Problem 2.16 Divide (mentally) with remainder; specify the values of m, n, q,

and r in Eq. 2.15 for each example. (a) 3 into 23. (b) 5 into 49. (c) 12 into 96.

Problem 2.17 In terms of Eq. 2.15 what is the condition that n divides evenly
into m? Give an example.

Given two whole numbers m and n, it is always possible to calculate the

quotient q and r, as given by Eq. 2.15. Here is how:
1. Find the largest whole number q for which nq ≤ m. This is the quotient.
2. Put r = m − nq. This is the remainder.
For example, let m = 33 and n = 9. Then 9 × 3 = 27, but 9 × 4 = 36. Therefore
q = 3 and r = 33 − 9 × 3 = 6. Result 33 = 9 × 3 + 6.
I expect this is how you solved problem 2.16. If you had trouble with that
problem, try it again now.
 The above 2-step algorithm for calculating q and r is fine, but how do you
figure out q, especially when m and n are fairly large numbers? For example,
divide 502 by 14, using the remainder approach. I’ll show you two methods.

Method 1 (Calculator Method). Enter 502 ÷ 14 on a calculator. You get 35.86

approximately. Use the integer part of this: q = 35. Now calculate nq = 14 × 35 =
490, and put r = m − nq = 502 − 490 = 12. That’s the answer: 502 = 14 × 35 + 12.
(You could check this with the calculator, to make sure. Notice that r is less than
n (12 < 14), which is one of the requirements for the remainder.)
Do you understand why this method works? The calculation 502 ÷ 14 =
35.86 tells us that 14 goes into 502 thirty-five times plus something, but not 36
times. Therefore q must be 35 and r follows as r = 502 − 14 × 35 = 12.

Problem 2.18 Using a calculator, divide 67 into 9,300 with remainder.

Method 2 (Long Division Algorithm) The example 1956÷23 will be used to

explain the long division algorithm. Remember, 1956 is called the dividend, and
23 is the divisor. Here is the algorithm.
1. Set up the format divisor

2. Select the first few (left-most) digits of the dividend, until they themselves
make a number ≥ the divisor. Call this number the “first segment” of the
dividend. Here the first segment is 195, because 23 > 19, while 23 ≤ 195.
3. Determine, by trial and error, the largest multiple of the divisor that is ≤ the
first segment. Write this number (always a digit between 1 and 9) on the top
line, above the last digit of the first segment. This is the first digit of the
quotient.

4. Multiply this quotient digit by the divisor, and write the result below the
first segment. Draw a line, and subtract. The number you get must be
greater than or equal to zero, and less than the divisor. If not, you chose the
wrong digit on the top line. (Here 0 ≤ 11 < 23, so it’s OK.)
 5. “Bring down” the next digit of the dividend, if there is one. This produces
the next “segment.”

6. Repeat steps 3–5 for this segment, until you run out of digits to bring down.
Each cycle gives another digit (between 0 and 9) of the quotient.

7. The division is now finished. The quotient is on the top line, and the
remainder is on the last line. Thus 1956 = 23 × 85 + 1. Let’s check that the
result is correct.

Therefore, divisor × quotient + remainder = dividend, as in Eq. 2.15. If you

carefully compare the last two calculations, you’ll see that the long division
algorithm basically reverses the multiplication algorithm.
Problem. For practice, make up a couple of examples of long division. Check by
multiplication.

Problem 2.19 A pirate ship has 54 pirates, including the captain. They capture a
ship carrying a treasure chest with 896 gold doubloons. The rule is that the
treasure is divided evenly among everyone, but the captain also gets any
remainder. Who gets what?

Factors and primes

Any number that divides evenly into a whole number n is called a factor of n.
Thus in Problem 2.15 you showed that the factors of 30 are 1, 2, 3, 5, 6, 10, 15,
and 30.
What are the factors of 7? By trial and error, you will find that the only
factors of 7 are 1 and 7. These are not very interesting – any whole number n has
factors 1 and n. Any factor of n other than 1 and n is called a proper factor of n.
A whole number that has no proper factors is called a prime number, or a
prime, for short.
Prime numbers have interested mathematicians for thousands of years. The
great Euclid studied prime numbers, and gave a proof that there are infinitely
many primes. We present this proof later, in Section 2.6. Meanwhile, let’s list the
first few primes:

These are all the primes less than 30. You should check mentally that all these
numbers have no proper factors, and that all the omitted numbers do have
proper factors. For example, 8 = 2 × 4, and 21 = 3 × 7, etc. (Non-prime numbers
are sometimes called composite numbers.) Do you see any pattern in the above
list of primes, which would allow you to continue the list without actually
checking each new number? If so, you are probably mistaken – no one has ever
found such a pattern.
Here is one reason why prime numbers are considered interesting:
 The product so obtained is called the prime factorization of n. For example:

Complete this list. Extend it further, if you want.

The algorithm for finding the prime factorization of any given number n is
simply to try every prime less than n in order. Whenever you find one, you can
continue by working with its quotient. For example,

Problem 2.20 Find the prime factorization of 220.

Some factors can be determined by inspection. For example, 2 is a factor of n

if and only if the final digit of n is even. Also, 5 is a factor of n if and only if the
final digit of n is either 0 or 5. It happens that 3 is a factor of n if and only if the
sum of the digits of n is divisible by 3. This is called the rule of three. For
example, 783 is necessarily divisible by 3, but 784 is not, as you can check.
Here is a slick little proof of the rule of three for 3-digit numbers ABC (these
are digits). We have, because of decimal notation

Now, A × 99 + B × 9 is divisible by 3, so ABC will be divisible by three if and only

if A + B + C is so. The same proof works for any number of digits. (It also shows
that ABC is divisible by 9 if and only if A + B + C is so. This is called the rule of
nine. For example, 5607 is divisible by 9 – try it!)
Let us apply the rule of nine to test whether the number 764,296 is divisible
by 9. The sum of the digits is 34, which is not divisible by 9. Therefore the given
number is not divisible by 9 either. (It’s also not divisible by 3, since 34 isn’t.)
The rule of nine can be simplified. Start by adding the digits, beginning at the
left, say. As soon as the sum becomes ≥ 9, subtract 9, and continue. Example:
34716 : 3 + 4 + 7 = 14 → 5 + 1 + 6 = 12 → 3, not divisible by 9. Here an arrow
indicates that 9 has been subtracted. Try another example for yourself.
This algorithm is called “casting out nines.” Why does it work? It works
because a given number, s is divisible by 9 if and only if s − 9 is also.
Question: why doesn’t the method work for checking divisibility by 7, for
example? Think of an example, such as 16.

Fractions
As we pointed out above, children first learn that, for example, “you can’t divide
7 by 3.” But later they learn about division with remainders: 7 divided by 3 equals
2 with remainder 1. Finally, they learn that 7 divided by 3 is a perfectly good
number, denoted by (“seven over three,” or “seven thirds”). This number is an
example of a fraction, that is, a number of the form ,where m and n are natural
numbers. We need to explain how such fractions are interpreted and then to
develop the arithmetic of fractions. This takes some care, so please follow the
discussion carefully. Errors in working with fractions are common among
students who are experiencing difficulty with their college math courses.
Consider a line segment of length 1 unit, that is, the part of the number line
between 0 and 1. If n is a natural number, we imagine dividing this unit interval
into n equal parts. (The figure shows the case n = 3.)

The number (pronounced “one over n”) represents the length of one of
these pieces. We can mark off the numbers etc., as shown. Thus is
the number obtained by moving to the right from 0 by m times the length . In
this way, the number can be thought of as a kind of counting number, which
counts up the total length of m equal “pieces,” each of length . The expression
is called a fraction. Note that m and n are whole numbers (0, 1, 2,…), and n ≠
0. The meaning of is given in:
 The number m is called the numerator and n the denominator, of this
fraction. Thus the denominator specifies the size of the “pieces” , and the
numerator specifies how many pieces are included in the fraction . For
example, the fraction is the total length of 17 [numerator 17] pieces, each of
size [denominator 5].
The fact that fractions are, in this sense, counting numbers, means that the
arithmetic of fractions is quite similar to the arithmetic of the counting numbers
(i.e., natural numbers). We will now develop this arithmetic methodically.
First, we have

because n pieces of length make up the entire distance from 0 to 1.

Next question: where would the number be located on the number line?
To answer this, we note that

(because of Eq. 2.17). We usually write

which is read as “one and five sixths.”

We see, therefore, that is located to the right of the number 1: This
method can be used to locate any fraction .
Problem 2.21 Explain where the following numbers are located on the number
line:
A fraction with m > n is called an improper fraction. To figure out where
such an improper fraction is located, we need to write it as

where q is a whole number, and r < n. (To follow this general argument, I suggest
that you keep track of a numerical example of your own choice.) This last
equation means that

Surprise! We are back to division with remainder! To repeat: to locate the

fraction on the number line, first divide m by n with remainder:

and write this as

Try some more examples:

Problem 2.22 Locate the following fractions:

In the case that m < n, the fraction is called a proper fraction. Therefore, to
locate an improper fraction, we use division with remainder to rewrite it as a
whole number plus a proper fraction. (Note that a proper fraction is a number
between 0 and 1.) Next, can you see how to reverse this procedure? How would
you write as an improper fraction? See if you can figure this out before
reading on. The answer is that
Problem 2.23 Express as improper fractions:

We have now expanded our number system to include fractions where m

and n are positive whole numbers. (We could also introduce negative fractions
now, but to keep things simple we will delay that discussion until later on.) Our
next step is to explain how to add, multiply or divide fractions. But first we need
to discuss the important procedure of “cancellation.”

Cancellation
Question: locate the fraction on the number line. Can you see a simple way to
express this number? Answer: from a sketch you will see that is exactly half-
way between 0 and 1:
This means that in fact

Similar simplifications hold for many other fractions, because of the following:

Here the letters m, a and b all represent positive whole numbers. The above
example fits this framework as follows:
Here the common factor 3 can be “cancelled out,” and disappears from the
fraction.
Try to do a similar calculation with First, you factor the numerator and
denominator:

by cancellation of the common factor 4. The fractions and are located at the
same spot on the number line.

Problem 2.24 Simplify the following fractions using cancellation:

Simplifying a fraction by cancellation is called reducing the fraction. If no

further cancellations are possible, the fraction is said to reduced to its lowest
terms. Example:

It so happens that errors in cancellation are among the most common

mistakes made by beginning math students. To avoid making these errors, you
need to clearly understand why the Cancellation Law is true. Let us use the
example

to explain the Cancellation Law. Consider the number line between 0 and 1,
marked into 15 equal segments.
 Next, we group these segments into groups of 5, as shown. Three such groups
make up the whole length, which means that each group has total length
This tells us that

(and also that but we already know this). In other words, the fraction
has 2 groups of 5 little segments in the numerator, and 3 groups of 5
little segments in the denominator, so this fraction equal
The same argument applies in general. Namely, represents a group of m
little segments out of b groups of m little segments, which is the same as

This explains the Cancellation Law.

Errors in cancellation
Cancellation is frequently used in Algebra and other topics. Unfortunately,
errors are often made in cancellation. Students are sometimes taught to “strike
out” common factors, as in

I strongly advise you not to use strike-outs, for the following reasons:
1. Strike-outs are easily used incorrectly.
2. Strike-outs make your calculations look messy and hard to check.
3. Strike-outs are completely unnecessary.
For example,
is perfectly understandable. Although strike-out might be OK for this example,
what about

What’s the final result? (I have often encountered this kind of situation while
grading math exams. Often the student doesn’t seem to know what the final
result is!) The answer is that

(because 4 = 4 × 1, right?).
Another common mistake is to cancel numbers that aren’t factors, as in

Of course, you might make this mistake without striking out the 2s, but for some
reason the habit of using strike-outs seems to encourage this kind of error. A
student who has understood the logic behind the Cancellation law is not likely to
make this error.
Let me be explicit:

Recall the basic terminology: Whenever we have a number which is the

product of two other numbers (“product” means that the two other numbers are
multiplied together, as in 7 × 5), each of the numbers is called a factor of the
given number. For example,
7 × 5 has factors 7 and 5
 ab has factors a and b
But
7 + 5 does not have any factors
Things can get a bit more complicated, however:
3 × (7 + 5) has factors 3 and (7 + 5)

Problem 2.25 Identify the factors in: (a) 36 × 8; (b) 36 + 8; (c) 36x + 8y; (d) 36(x
+ y); (e) 36xy.

Here is a quick review of cancellation: Simplify

Answers: (c) No cancellation possible.

Slash notation
Henceforth we will often write the fraction as m/n. This saves space in
printing. Both are read as “m over n.” (Another possibility is m ÷ n, which we
sometimes use.)

Order among fractions

Consider the two fractions

Both are located between 0 and 1 on the number axis, but which is larger? The
answer is not obvious just from inspection. However we can answer the question
by using the Cancellation Law in an unexpected way:
The fractions on the right have the same denominators, so it is now evident that
the second fraction is the larger. The same method works for any pair of
fractions. Try it on 13/6 and 20/9 (answer: 20/9 is larger).
Any two fractions a/b and c/d can be rewritten to have the same
denominators:

Be sure that you understand this, because we will use it again later, when adding
and subtracting fractions.

Problem 2.26 Which is larger, 8/53 or 7/52?

You may have thought of a second way to compare fractions – use a

calculator! Any fraction can be written as a decimal number, and your calculator
will do just that, within its accuracy limitations. Try it now on 13/6 and 20/9.
You should get

Thus you see that 20/9 is the larger number. But notice also that both these
fractions appear to involve repeating decimals. We next explain all this.

The decimal representation of a fraction

We have shown how to locate any fraction m/n on the real number line, in terms
of combining line segments of length 1/n. But any real number must also have a
decimal representation. To calculate the decimal expression of a fraction we use
extended long division, which is just an extension of the previous algorithm for
long division with remainder.
We start with the same setup as before, divisor except that we
now include a decimal point and trailing zeros in the dividend. Here’s the
calculation for the fraction
Notice how the 0s from the dividend are brought down one by one.
In this example we eventually got a 0 on the bottom line. This terminates the
calculation, because to continue would give nothing but zeros in the quotient.
The result,

can immediately be checked by multiplying, 8 × 1.625 = 13 exactly.

Problem 2.27 Find the decimal representation of (One method: first

find then multiply by 2,3,etc.)

Next, we look at the example 7/30:

Note that the pattern now repeats indefinitely, giving the result
 To check this, note that

So what happened – we are supposed to get 7. Well, first, if we include the

repeated 3s we get the answer 6.999… (repeated 9s). But in fact, 6.999 … = 7.
Some people find this confusing, but it is a fact that

I won’t spend a lot of time on this. It is not an approximation, but an exact result.
One way to explain it is to write

which is equal to 0. Checking repeating decimals by multiplication will always

involve this situation.
The two foregoing examples suggest that, in general:

The repetition need not be a single digit, but can be a group of digits, as in

If you will work out this example for yourself, you may understand why the
decimal expansion of any fraction must always repeat, if it doesn’t terminate.
So how can it be proved that the decimal expansion of a fraction m/n must
terminate or repeat? Upon doing the long division, once the same non-zero
remainder occurs a second time, the pattern repeats from then on. Since there
are only n − 1 possible non-zero remainders in m/n,you in fact must either
eventually get remainder 0, or repeat an earlier remainder. For example, to see
how this works, find the decimal expansion of 1/7. How long is the repeating
group of digits?

Problem 2.28 Find the decimal expansion of (a) 5/6; (b) 2/9 by long division.

You may wonder whether every repeating decimal is necessarily equal to a

fraction. The answer is yes. How to find the fraction is explained in Section 2.4.
Fractions are also called rational numbers. You may also wonder whether
nonrational numbers exist. Yes, they do exist. Here’s an example:

where the 1s occur in longer and longer groups, separated by single zeros. Since
this is a nonrepeating decimal, it is not a rational number. Such numbers are
called irrational. Another example is (the square root of 2) – we will show in
Section 2.4 that is irrational. So are and so on. To repeat,
rational numbers have terminating or repeating decimals. Any other decimal
number is irrational.
In passing, note that our notion of decimal-point numbers has suddenly
expanded, to include the possibility of infinitely many digits to the right of the
decimal point. These digits may occur in a repeating pattern (in the case of a
rational number), or a nonrepeating pattern (irrational number). The resulting
“real number system” is discussed further in Section 2.4.

Addition of fractions
We next study the arithmetic of fractions – addition, subtraction, multiplication,
and division. Of course, since any fraction is equal to a decimal-point number,
we might dispense with fractions altogether. But because fractional expansions
are widely used in math and science, you need to master the arithmetic of
fractions.
The basic definition for adding fractions is
 (I switch to capital letters because this seems to make the formulas easier to
look at.) Here the letters A, B and C represent whole numbers. An example:

Recall that is the total length of 3 little segments of length Hence is the
total length of 3 + 7 or 10 such little segments, which says that
The same argument applies to any example, as in Box 2.21. Another example:

Eq. 2.21 tells us how to add two fractions if they have the same denominators.
But how would you add

If you said go to the bottom of the class! That’s incorrect. Can you think of
some way to use Eq. 2.21 for this case?
Well, we could first rewrite the two fractions so that they have the same
denominator, as explained in the section on Order of fractions:

Therefore
(which could be reduced to This is called the method of common
denominators. It applies to any addition of fractions.

Problem 2.29 Add, and reduce if possible:

The general formula is:

The proof of this formula uses the same method as in the numerical examples.

Problem 2.30 Prove Eq. 2.22.

Now, should you memorize Eq. 2.22? Of course not! Just remember how to
do the calculation.
Also, note that you can easily – and should – check any addition of fractions:

(check this by mental cancellation before continuing).

The method for adding two fractions with different denominators is
therefore, first to rewrite the fractions using a common denominator, and then
to add the numerators:

Problem 2.31 Combine fractions and check:

Subtraction and the rational numbers
We can subtract fractions in the same way as adding them:

(You can check this in the same way as before.) This might result in a negative
fraction, however. No problem! The combined system of positive and negative
fractions (and zero) is called the system of rational numbers.

Problem 2.32 Calculate:

Problem 2.32 indicates a short-cut that can often be used in adding or

subtracting fractions. Namely, the common denominator can sometimes be
smaller than BD. For example

Try this method on the rest of Problem 2.32. For example, in (d) you can use the
denominator 24; this is called the least common denominator. In simple cases you
can find the least common denominator by trial and error. But you can always
use BD as your denominator.

Multiplication of fractions
Now, where have we got in developing the arithmetic of fractions? We have
positive and negative fractions including 0, and we can add or subtract such
fractions, always obtaining another fraction.
Next comes multiplication. Here’s the rule:
This formula is easy to remember (“to multiply fractions, multiply their
numerators and multiply their denominators”). It is also very easy to use.
For example

Problem 2.33 Calculate Simplify

where possible.

So the foregoing formula for multiplying is easy to use and easy to remember.
But why is it true? How come

for example? The formula, Eq. 2.24, tells us this, but what’s the reasoning behind
this formula?
I now have to admit a little secret. Nowhere in this book have we yet defined
multiplication for anything but whole numbers. If you look back to Section 1.5,
you will find the multiplication algorithm for decimal-point numbers. Did you
understand why this rule was correct? Why is 0.7 ×0.9= 0.63, for example? How
can you base your mastery of math on a complete understanding if your author
doesn’t explain these things? The explanation is now forthcoming. It’s a bit
finicky, and I suggest that you don’t spend too much time on it unless you’re
really interested. But do try Problem 2.24, and continue reading from there on.
The first step is to define what is meant by multiplication of fractions. We
need a definition that makes good sense. Our definition will apply to a special
case:
 The graph shows the case n = 5.

The line segment from 0 to a is divided into 5 equal segments, each having
length
For example, suppose that a = m, a whole number. Consider the example m =
8. It happens that

This might seem obvious to you, but keep in mind that these two expressions
actually have different definitions (what is the definition of ?). To prove that is
the same as according to our definition, we need to show that is one-fifth
the distance from 0 to 8, or in other words that 5 segments of length will have
length 8. But the total length is just which equals 8 by cancellation.
Clearly, this example is entirely generic, so we can conclude that

whenever n and p are whole numbers (n ≠ 0). This is a special case of Eq. 2.24
(write to see this).
Did you understand that argument? I admit it’s pretty confusing! It just
seems to be proving the obvious in a non-obvious way. Perhaps you have known
that for many years, so you can’t see the point of “proving” a
particular case from scratch. But mathematics always has to proceed from
carefully stated definitions of basic concepts to usable results, or algorithms. The
multiplication formula, Eq. 2.24, is just that – the multiplication algorithm for
fractions. Like most other algorithms, it is not obviously true. Its truth has to be
deduced from the basic definition.
The last argument applies to any number a, so let us take We then
conclude that

because indeed the total length of n segments each of length is equal to

and this establishes the truth of the above formula.

Finally, must be the same as which equals Therefore we
have proved that

for all whole numbers m, n, p, and q. This is exactly the same as Eq. 2.24 (but
with different symbols, which were used to emphasize that they are whole
numbers).
Dear reader, I offer my apologies for this extremely abstruse discussion. No
elementary math text that I have seen makes any attempt to explain the logic
behind the multiplication formula, Eq. 2.24. So is it essential that you take the
effort to understand this completely? Maybe not, for the simple reason that the
multiplication formula is easy to remember and use correctly. Try the next
problem.

Problem 2.34 Calculate:

As suggested by Problem 2.34, we can now perform addition, subtraction,

and multiplication of rational numbers. All the Laws of Arithmetic that were
developed in Chapter 1 remain valid for the system of rational numbers. Why?
Because rational numbers are related to counting numbers (i.e., natural
numbers), so that the entire logical content of Chapter 1 remains in force,
without further ado.
This is very reassuring. However, an important warning: as I said before,
errors in calculating with rational numbers are unfortunately common. I wish I
had a dollar for every time I’ve had to take marks off college exam papers for
blunders like

Perhaps you would like to explain why this is incorrect. What is 1/x − 1/2,
actually? Answer, (2 − x)/2x, which certainly doesn’t look anything like 1/(x − 2).
Indeed, if you try some value of x, such as x = 1, you immediately see how silly
the above equation is. The value x = 1 is said to be a counterexample to the
above equation.

Problem 2.35 Use counterexamples to show that the following “calculations” are
incorrect. What is the correct version?

Properties of rational numbers

Some further properties of rational numbers are:
These equations are fairly obvious. For example, in Eq. 2.26, m/1 represents the
total length of m line segments each of length 1, which is clearly m. Likewise, in
Eq. 2.27, 0/n represents the total length of zero line segments, which is zero.
By the way, what does m/0 mean? See if you can figure this out; re-read the
beginning of the present subsection on Fractions. First try to figure out what
1/0means.
Any luck? It says there that 1/n is the length of the line segment obtained by
cutting a unit segment into n equal parts. Therefore 1/0 is the length of the line
segment obtained by cutting a unit segment into 0 equal parts. Come again? Zero
parts – what does that mean? It doesn’t mean anything! It’s completely
meaningless:

Therefore, m/0 is also meaningless.

You have probably heard that “you can’t divide by zero.” This is absolutely,
inescapably true in mathematics. Division by zero is meaningless. We’ll return to
this point again later on. (Try division by zero on your calculator, in the
meantime.)
Next, we have

These equations are easy to remember and use, but they take a bit of explaining.
Equation 2.28 is in fact a definition. The expression has not been defined
previously in our discussion. This number would be located at −m times the
length 1/n, on the negative number line, and this is what −(m/n) has been
defined as. As a check, try (−7) ÷ 3 on your calculator (punch in “7 ÷3 =
“), and observe that this equals −(7 ÷ 3). Note: means “Change Sign.”
Your calculator may have a different symbol. The display for a negative number
will have a − sign somewhere.
Next, Eq. 2.30 can be deduced from the Cancellation Law (Eq. 2.20), if we
assume that this law must also hold for negative numbers (a consistency
assumption).

and this is Eq. 2.30. Once again, you can use your calculator to check this
equation.
Finally, Eq. 2.29 can be deduced by a similar argument:

Pause to review
Dear reader, you have almost completed your review journey through
Arithmetic. Let’s pause here to briefly review the voyage so far, from Chapters 1
and 2. The experience has been one of progressive enrichment.
1. Counting, or natural numbers
– addition and multiplication as counting
– the decimal representation of natural numbers (positional significance)
– algorithms for addition and multiplication
– the Counting Principle implies the Laws of Arithmetic
2. Positive decimal-point numbers
– extension of the idea of positional significance of digits
– algorithms for addition and multiplication
– the number line
 – Laws of Arithmetic remain valid
3. Integers
– subtraction as the inverse of addition
– integers: 0, ±1, ±2, …
– addition and subtraction; algorithms
– Laws of Arithmetic remain valid
4. Fractions
– division with remainder
– fractions m/n
– location on number line
– addition and multiplication
– Laws of Arithmetic remain valid
5. Rational numbers
– subtraction of fractions
– rational numbers ±m/n
– positive and negative number line
– addition, subtraction, and multiplication
– Laws of Arithmetic remain valid

The one step that is yet to be discussed is division of rational numbers. When
this is completed, we will have a number system (the rational numbers) in which
all four operations, + , −, ×, ÷ can be carried out. (However, there are yet further
numbers on the number line, called irrational numbers. We discuss these
numbers briefly later.)

Division
We will use this definition of division of rational numbers:
 Here, is read as “a over b,” or “a divided by b.” This can also be written as
a/b, or a ÷ b. Thus Eq. 2.31 could thus be expressed as

This definition of division is of course familiar in the case that a and b are
whole numbers and b divides evenly into a. For example 8/4= 2 means that 8 = 4
× 2. Now we can extend the definition to allow for a and b to be rational
numbers.
Equation 2.31 or 2.32 just says that “division is the inverse of multiplication.”
For example, consider We have

Can you see what c must be? What can you multiply by, to get Think about
it for a moment.
Answer: because

Any example could be done in this way, which is like guessing the answer.
However, the following formula avoids guessing:
 Here, A, B, C, D are assumed to be whole numbers (with B, C, D ≠ 0).
This is the famous invert and multiply rule: To divide by a fraction p/q invert
this fraction and then multiply. (Invert just means to switch the numerator and
denominator.) For example,

This agrees with the answer that we obtained earlier.

Let’s look at one more example, and then show why Eq. 2.33 is true in
general. What is

Write this as

What has to go into the space (?) to make this work? It has to cancel out the
factor and replace it by Do you get it? In fact

So the answer is that

– invert and multiply. Try a few more examples.

Problem 2.36 Prove Eq. 2.33 by emulating the above argument.

This gives me another opportunity to expostulate about memorization versus

comprehension in mathematics. Many people – students, acquaintances, even a
math teacher, have told me that they never understood why the “invert and
multiply” rule was true, although they had no difficulty remembering or using
the rule. This suggests to me that these people are uncomfortable about such an
understanding gap. So they should be. After reading this section, you will fully
understand that the invert and multiply rule reflects the basic fact that division is
the inverse of multiplication.
The fraction q/p is often referred to as the reciprocal of p/q. For example, 3/4
is the reciprocal of 4/3, and 4/3 is the reciprocal of 3/4. Also, the reciprocal of 6 is
1/6. The rule for dividing fractions can therefore be stated as: to divide by a
fraction, multiply by its reciprocal.

Problem 2.37 Calculate (a) (16/9) ÷ (64/81); (b) (12/13) ÷ (3/13); (c) (11/50) ÷
(−7/150); (d) (12/7) ÷ 6.

Question: calculate 3/2/5 – or is this not possible? Answer: the given

expression is ambiguous, because of a lack of brackets. It could be either of

In any combination of two or more divisions, brackets are needed to remove

ambiguity. The same holds if the fraction if written vertically:

Sometimes fraction lines of different length are used:

The longer line serves to group things above and below it separately. However,
using brackets is usually preferable.

Problem 2.38 Calculate (3/4 − 1/8)/(5/3 + 1/6).

On the same topic of ambiguity, or lack of it, consider the following

examples:
2/3 + 7 = ?
5 − 1/8 = ?
4 − 6 − 1 = ?
On the face of it, these might all seem ambiguous. For example, is 2/3 + 7 to
be interpreted as (2/3) + 7, or 2/(3 + 7)? The other examples could also be
interpreted in different ways. However, in actuality none of these expressions is
ambiguous, because of the following precedence convention.

Applying these conventions, we obtain

These conventions do not cover cases involving multiple divisions and

multiplications, such as 2/3 × 5. Here, brackets must be used to prevent
ambiguity.
You can always use brackets whenever you want to clarify the meaning of an
expression, even if the brackets are not actually required. For example, 2 + 3/11 −
8 is equal to 2 + (3/11) − 8, because of the above precedence convention. But the
second form is easier to grasp, and should be preferred for the sake of clarity.

Division by zero
We saw earlier that m/0 is an undefined, or “illegal” expression, if m is a whole
number. More generally, division by 0 is always undefined: a/0 is meaningless, or
illegal, if a is any number. I’m often asked “Professor, why is division by zero
impossible?” My whimsical answer is “Try it on your calculator.” But this is no
explanation. The proper answer is “Well, what is the basic definition of
division?” Remember,

Try this with 1/0 = c. Then 1 = 0 × c. If c exists, then 0 × c = 0, and therefore 1 =

0. In other words, if division by zero were possible, we would have 1 = 0.
Therefore, it is not possible, i.e. division by zero is undefined.
Do you remember what I wrote earlier about the importance of definitions? I
said that very few math students seem to be aware of how important definitions
are. Both of the mysterious questions, division by zero, and invert-and-multiply,
can be resolved by referring to our basic definition of division, Eq. 2.31. A person
who asks “why can’t you divide by zero?” does not know the meaning of the
word “divide.” Definitions are crucial.

Long-division of decimal-point numbers

The algorithm discussed earlier for calculating the decimal-point representation
of a fraction m/n can be adapted to calculate a/b for any decimal-point numbers
(b ≠ 0). We begin by setting up the usual format but ignoring the decimal
point in the divisor a. Consider the example 12.7/5.8:
and so on (the calculation would eventually repeat).
This calculation says that

Why? Think this out for yourself.

The general algorithm for long division of decimal point numbers is:
1. Set up the usual format for long division. Retain the decimal point in the
dividend, but ignore the decimal point in the divisor.
2. Append zeros to the dividend if needed.
3. Divide as usual, placing the decimal point in the same location as in the
dividend. The result at this stage equals the dividend ÷ the divisor with the
decimal point ignored.
4. To obtain the final answer, shift the decimal point left a number of places
equal to the number of digits beyond the decimal point in the original
divisor.
Why this algorithm is correct should now be obvious from the above example.

Problem 2.39 Calculate by long division: 18.73÷2.7. What is the repeating

group? (Check by calculator.)

I have included the long-division algorithm for the sake of completeness. I’m
not convinced that every one needs to remember it.

The laws of arithmetic

In Chapter 1, we discussed the five Laws of Arithmetic, pertaining to
addition and multiplication. In this chapter, these laws have been supplemented
to cover subtraction and division. Here is the full list, which applies to all rational
(and decimal) numbers a, b, c.
 The Zero Product Rule (which is very important in Algebra) is new to our
list. Most students seem to think that the rule is obviously true – but why? Well,
it’s certainly true for whole numbers, from the basic definition of multiplication
in terms of counting. Recall that mn is the total number of objects in m groups
each containing n objects. If the total number is 0, clearly either there are no
groups (m = 0) or the groups contain no objects (n = 0) – or both. I’ll let you
think about why the Zero Product Rule is also true for decimal-point numbers.

Problem 2.40 Use the Laws of Arithmetic to prove that if ab = ac for some a ≠ 0,
then b = c.

Problem 2.40 indicates another type of cancellation – you can “cancel” the
common factor a in the equation ab = ac (if a ≠ 0). As previously, I recommend
against using “strike-out” in this situation, as it can cause confusion. For
example, suppose you obtain the equation 18x = 6 when solving some math
problem. You write this as So what is x? Better to leave off the
strike-outs, and note that 6 × 3x = 6 implies 3x = 1, so x = 1/3.
Problem 2.40 makes use of a mathematical principle that we use frequently,
without being aware of it. The principle is

For example,

(the operation is: multiply by 3). Similarly,

(the operation is: add c). The case in Problem 2.40 is

Now, I don’t want to bore you with further technical details. But a very
perceptive reader might ask, how do we know, for example, that the formula for
multiplying fractions

is actually valid when a, b, c, d are themselves fractions? The explanation of this

formula given earlier assumed that a to d were whole numbers. The answer is
that it just does work out this way in general. The proof is not very complicated
(but skip it if you wish).
Let Then according to Eq. 2.31 we have a = bx and c =
dy. Therefore
Proof complete.
Dear reader, if this proof looks like gobbledygook to you, don’t fret. It took
me a while to think it up, and I’m a professional. (We will encounter more proofs
like this in later chapters, but hopefully they will then be less confusing.)

Absolute value
The absolute value (or magnitude) of a real number x is the positive distance of
x from 0, on the number line. This implies that

For example |− 7.5| = −(−7.5) = 7.5.

We also have

For example, |− 8 − 5| = 13, which is the distance between −8 and 5.

Another useful formula is:

Here, for any a > 0 the expression designates the square root of a; this is
defined as the positive number b whose square b2 equals a. Thus
 For example because 52 = 25. (Many students think that
but this is not how the symbol is used in math.)
Note that Eq. 2.38 says that finding the square root of a positive number is
the reverse operation to finding the square of a positive number. Reverse
operations are common in mathematics.

2.4 The real number system

The rational number system consists of all numbers m/n, where m and n are
whole numbers (n ≠ 0), together with the negatives of these numbers. Every
rational number has a decimal representation, which either terminates or repeats
endlessly. Decimal numbers that neither terminate nor repeat are therefore
irrational numbers. Infinitely many irrational numbers exist, since any
nonrepeating pattern produces such a number. The set of all decimal numbers,
repeating or not, is called the real number system. The real number system
corresponds exactly with the real-number line, since every real number can be
located, in principle, at a specific point on the line. In practice, we can only locate
an irrational number approximately on the number line, partly because of
measurement limitations, but also because calculating infinitely many
unpatterned digits “at once” is impossible. For example, it is known that is an
irrational number (this is proved later on). An algorithm exists for calculating
the decimal expansion of to any desired precision, but we can’t compute, or
list, all the digits in a finite amount of time. The same is true for the famous
number π (“pi” – see Chapter 5).
The real number system satisfies all the Laws of Arithmetic listed at the end
of the last section. For example, it is true that

and this even makes sense if you don’t know what the symbols mean. The real
number system is the basis for much of higher mathematics. For example, the
calculus is entirely based on the real number system.

Repeating decimals
Every repeating decimal is equal to a rational number, that is, a fraction m/n. But
how do you calculate the fraction? Consider the example

Write x = 0.132132 …, and consider

because the parts after the decimal point are identical.

This implies that 999x = 132, or

We have found the fractional form of x. (A quick check on your calculator will
confirm this.)
Do you see the general method? Can you find the fraction corresponding to
0.4545… in your head, now? Answer: 45/99 = 5/11. Work this out more
carefully, if you don’t see it. Note that you multiply by 100 in this case. In
general, if the repeating group has n digits, you would multiply x by 10n.
In general, a repeating decimal may begin repeating only after some number
of nonrepeating digits, as in

To use the above method, we first isolate the repeating part:

These examples indicate how any repeating decimal could be expressed as a

fraction.

Problem 2.41 Express as fractions: (a) 0.7575… ; (b) 1.00110011….

Factors and prime numbers

Let us review the definitions of factor, and prime number. If m is a natural
number, any natural number n that divides evenly into m is called a factor of m.
In other words, n is a factor of m if m = nq for some natural number q. For
example, 3 is a factor of 12 because 12 = 3 × 4. The list of all the factors of 12 is: 1,
2, 3, 4, 6, and 12, because each of these numbers divides evenly into 12.
Factors of m other than 1 and itself are called the proper factors of m. Thus
the proper factors of 12 are 2, 3, 4, and 6. Any number that has no proper factors
is called a prime number (often prime, for short). For example, 7 is a prime. So
is 101, though this takes some checking. In fact, how would one decide whether
101 is a prime? One has to check every possible number n < 101. Well, in fact,
you only have to check the primes n, and only those up to i.e. the primes
less than 10. (I’ll explain this later.)
By inspection, 2 is not a factor of 101. Also, 3 is not a factor, by the rule of
three. And 5 isn’t a factor, by inspection. The only other prime < 10 is 7, and we
can divide 7 into 101, showing that 7 is not a factor, either (101 = 14 × 7 + 3).
Therefore, 101 is a prime.

Problem 2.42 Which if any of 151, 153, 157, 159 are primes?

In checking whether m is a prime, why do we only need to consider prime

factors? And only primes less than First, suppose for example that 6 divides
m. Then the prime factors of 6, namely 2 and 3, must divide m too. So we only
need to look for prime factors. Second, if p divides m, then m = pq for some q. If
p > then q < (otherwise we would have pq > = m which is not
true). Once we have determined that m has no prime factors smaller than ,
we know that it can’t have any bigger than , because any such would pair
with another factor smaller than . Try this for m = 29. We have = 5 plus
something, so the only primes that need to be checked are 2, 3, and 5. None
works, so 29 is a prime.
 If you are a computer programmer, you may like to devise a code to print out
all the primes up to 105, say. There are about 1,100 of them. This raises a couple
of questions.
1. Is there a formula for generating primes?
2. How many primes are there?
The answer to question 1 is no, there is no known formula for producing
primes. The answer to question 2 is, there are infinitely many primes. This was
known to the famous Greek, Euclid, and here is his ingenious proof (Euclid was
big on proofs). Start by listing the primes, in order:

Suppose this list goes up to a certain prime N. Write

that is, Q is the product of all primes up to N ,plus 1. Then Q is not evenly
divisible by any of the primes up to N , because there is always a remainder of 1.
Therefore, either Q is itself a prime, or it has a prime factor greater than N. In
any case, no matter how many primes we start with, there’s always another,
larger prime.

Factoring
Given a natural number n we can, in principle, determine which primes divide n.
For simplicity we start by considering the smallest primes 2, 3, 5, etc., reducing
the size of the factoring problem whenever we discover a prime factor. For
example

Since 19 is a prime, this is as far as we can go. We say that 684 has now been
factored as a product of primes. We might write it more compactly as
where 22 = 2 × 2, etc.
A natural number that is not a prime is called a composite number. The
composite numbers are 4, 6, 8, 9, 10, etc.
This trial-and-error algorithm can be used in principle to factor any given
whole number. Therefore

In practice, however, factoring a large number can be extremely difficult. For

example, try to factor 73,097. Is this a prime? We’ll discuss this again later.

Irrationality
We next consider a fact about whole numbers that was known to the early Greek
mathematicians. The fact in question is that numbers like √2 are not expressible
as a fraction m/n of whole numbers. This fact was known to the school of
Pythagoras, and was at first considered to be an embarrassment, more or less on
superstitious grounds. The Greeks were familiar with fractions, which seemed a
good enough number system. Thus was a “bad” number, although it could
easily be realized geometrically, as the hypotenuse of 45° right triangle (see
Chapter 5).
Let m be any whole number. What can we say about the prime factorization
of m2? Consider two examples:
 As these examples indicate, any prime that occurs as a factor of m2 must
occuran even numberof times as a factor. For example, 202 has four factors of 2,
while 212 has zero (zero is an even number).
We can now give Euclid’s proof that is irrational. Suppose on the
contrary that is rational, in other words that = m/n for certain whole
numbers m and n. Then 2 = (m/n)2 = m2/n2, and this means that

If we now consider the prime factorizations of m2 and n2, we see that m2 has an
even number of factors of 2, but 2n2 has an odd number of factors of 2. Therefore
we cannot have m2 = 2n2 after all. The assumption that is rational
has led us to a contradictory conclusion. (This type of argument is called
“reductio ad absurdum” in Latin – reduction to an absurdity.) Therefore we
must conclude that is irrational.
To check whether you followed this argument, try modifying it to prove that
is irrational. What happens if you use the argument to try to prove that is
irrational? Also, prove that (the cube root of 2) is irrational. Here’s how to
start: suppose = m/n. Then 2 = (m/n)3 = m3/n3, so that m3 = 2n3. Why is this
an absurdity?
The Greek mathematicians were also interested in proving, or disproving,
that π is rational, but this problem defeated them. It was finally proved in the
19th century that π is irrational. One sometimes reads in the paper the “amazing
fact” that the decimal expansion of π will never repeat. This is true (because π is
irrational), but don’t the journalists realize that the decimal expansion of also
never repeats? The latter fact is fairly easy to prove– indeed, we have just proved
it.

Errors in arithmetic
We have finally completed our review of Arithmetic. The subject is quite
complicated. Getting through Arithmetic takes many years of schooling. Having
completed Arithmetic, students next study Algebra. Unfortunately, errors are
often made in Algebra. These errors, I believe, are usually the result of an
incomplete mastery of Arithmetic. Algebra is direct follow-on from Arithmetic,
so any gap in one’s training in Arithmetic is likely to cause errors in Algebra.
 At this stage, you may wish to look at the two final sections of this book. Try
the Diagnostic Test, and read about Common Errors. These sections can help
you decide whether you still need to study parts of Chapters 1 and 2 carefully,
before going on. Likewise, if any of Chapter 4 (Algebra) is not entirely clear to
you, come back and study Arithmetic meticulously.

Aspects of understanding mathematics

Math Overboard stresses the importance of understanding the mathematics that
you’re learning. Three questions related to understanding a certain math concept
or result are:
– Why is it appropriate, or true?
– How does it calculate (algorithm)?
– How does it relate to other math concepts or results?

For example, consider the method for adding fractions:

Why is this true? The answer is, first that we can write both fractions using a
common denominator:

Second, fractions having the same denominator are added simply by adding the
numerators. Hence the above formula is valid.
We also remember that this argument can be run backwards, as a way of
checking. For example, if asked to add you would get and this can be
checked immediately:

Is this related to anything else? Well, subtraction is the same:

and this can also be checked by “undoing it”‘
It’s also worth noting here that the formula for multiplying fractions is
completely different:

Knowing all this should prevent anyone from making certain types of mistakes.
Here is another question related to understanding a math formula:
– Are there interesting special cases?
In the multiplication formula, suppose C = D. Then, because for sure,
the formula says that

This is the Cancellation Law, which is therefore a special case of the

multiplication formula. (This line of thinking might also remind you about the
“false cancellation law” in which is thought to equal a dangerously false
equation.)
When learning new things in math, I have always made it a habit to look for
special cases that I already understand. I find that this helps greatly in learning
the new material. Similarly, if I am trying to solve some math problem, and can’t
see how to proceed, I find it extremely helpful to first simplify the problem by
looking at a special case. Obviously if I can’t solve the simpler problem, I will be
unlikely to solve the harder one.

2.5 Review problems

1. Calculate (a) 63 − 27 − 58; (b) 63 − (27 − 58).
2. Write in increasing order: 14, −8, 9, 0, −3.
3. Express in symbols: subtraction is the inverse of addition.
4. Prove that if a > b then a − b > 0. Which Laws of Inequality did you use?

5. Calculate (a) 3 × (−2 + 9)−6; (b) (11−1)×(1−11); (c) 5×(−6×(7−8)).

6. Remove brackets and simplify: (a) 3x − 2(x + y); (b) (a − b)(a − b).

7. Divide, and obtain the quotient q and the remainder r: (a) 207 by 2; (b) 71 by
11; (c) 1897 by 222. (Use a calculator.)
8. Show that if n divides m, and m divides p, then n divides p (divides” means
“divides evenly into”). Start by considering an example.

9. Find the prime factorization of (a) 186; (b) 1611 (this is divisible by 9). Use a
calculator if this helps.

10. Specify where the following numbers are located on the number line: (a)
83/6; (b) 105/15.

11. Find the decimal expansion of 4/9 by long division. Check by calculator.

12. Use your calculator to determine which reciprocals 1/n (n = 2, 3,…, 20) have
terminating decimals, and which have repeating decimals. Can you guess the
general rule?

13. Express as a single fraction in simplest terms: (a) 35/8 + 7/12; (b) 20/7 − 13/4.

14. Calculate (a) (4/7) × (2/3 − 1/2); (b) (7/6)/(1−8/9).

15. Decide which of the expressions a × b/c and a/ b × c is ambiguous, and
explain why.
16. Why is the expression a/ b + c not ambiguous?
17. Is the expression a/b/c ambiguous? Why?

18. Express as a single fraction: (22 − 7/3) ÷ (81/19 − 3).

19. Explain what the phrase “a goes into b, j times” means. Start by considering a
simple example.
Solution 2.1 (a) 9 − 8 = 1 means that 9 = 8 + 1, correct; (b) 92 − 45 = 47 means
that 92 = 45 + 47, correct; 28 − 13 = 14 means that 28 = 13 + 14, incorrect.
Solution 2.2 For example, 3 − 9 = ? This would mean that 3 = 9 + ?, but you can’t
add anything to 9 that could give the answer 3. (Of course, we will soon learn
how to do this.)

Solution 2.3 The problem is to calculate m − n, where m > n. The algorithm is:
1. Write m above n, aligning digits on the right, as usual. Draw a line.
2. Start with the units digit.
3. Subtract the lower digit from the upper digit, if this is possible, and write the
result below the line, in the same column. If, however, the lower digit is
bigger than the upper digit, “borrow” 1 from the next upper digit, and add
10 to the current upper digit. Now subtract as before. The “borrowed” 1
decreases the next upper digit by 1 in the next step.
4. Move to the left one column, i.e., to the next digit.
5. Repeat steps 3 and 4 until finished.
Observe the typical algorithmic nature of these steps: starting instruction, a
repeated procedure, and a stopping instruction. See Section 1.2. An example:

Here we couldn’t subtract 7 from 4 directly, so we borrowed 10 from 60. (The

reverse addition, 37 + 27 = 64, would involve a carry over, which exactly reverses
the borrowing.) Make up a couple more subtraction exercises yourself, and check
by addition.
Solution 2.4
Solution 2.5 −8.8 < −8.3 < −2.1 < 0 < 5.6 (If you didn’t get the same answer,
think about locating each of these numbers on the number line. In particular,
note that −8.8 is further to the left than −8.3.)

Solution 2.6 (a) −13; (b) 3; (c) −3; (d) 0. You get the same answers after changing
the order.
Solution 2.7 (a) −11.6; (b) 3.7; (c) −6.2.

Solution 2.8 (a) −20; (b) −24; (c) −102.

Solution 2.9 (a) −3; −17; 41; 0; (b) −22.80; −7, 128.1; 49.4; 0.67.

Solution 2.10 (a) −1.8−(−7.4−7.6) = −1.8−(−15.0) = −1.8 + 15.0= 13.2. (b) −(4.3
− 6.1) + (2.7 − 7.0) = −(−1.8) + (−4.3) = 1.8 − 4.3= −2.5. (c) 3.5 − (2.6 − (4.2 −
5.8)) = 3.5 − (2.6 − (−1.6)) = 3.5 − (4.2) = −0.7.

Solution 2.11 First, we used the convention that juxtaposition of symbols

implies multiplication: ab = a × b. Second, we used the convention that
operations inside brackets are performed before operations outside brackets:
a(bc) means first multiply b × c, then multiply the result by a. Third, we used the
convention that multiplications have precedence (i.e., they’re done first) in any
expression involving both multiplication and addition.
Solution 2.12 (a) 46; (b) 7; (c) 42.

Solution 2.13 (a) −c + 3d;(b) xy − 2xz; (c) w2 − w − 6.

Solution 2.14 (a) 10; (b) −20; (c) −20; (d) −34. (If you failed to get any of these
answers, re-read the box on Conventions regarding the minus sign.

Solution 2.15 (a) You could check, by a pen-and-paper calculation, that 563 × 48
= 27, 024. (b) The numbers that divide into 30 are n = 1, 2, 3, 5, 6, 10, 15, and 30.
The corresponding values of q = 30 ÷ n are q = 30, 15, 10, 6, 5, 3, 2, and 1. For 31,
the only values of n are n = 1 and 31, with corresponding q = 31 and 1.
Solution 2.16

Solution 2.17 n divides evenly into m if and only if the remainder r equals zero.
Example: 3 divides evenly into 24, because 24 = 3×8 (zero remainder).

Solution 2.18 9300 = 67 × 138 + 54.

Solution 2.19 We have 896 = 54 × 16 + 32. The crew members each get 16
doubloons, and the captain gets 48.

Solution 2.20 220 = 2 × 2 × 5 × 11.

Solution 2.21 which is located to the right of 5;

which is located to the right of 11. Return to the
main text to learn how to do this calculation. (By the way is equal to see
below.)

Solution 2.22

Solution 2.23

Solution 2.24 which doesn’t

allow any cancellation (so can’t be simplified.)

Solution 2.25 Factors are: (a) 36 and 8; (b) no factors; (c) no factors; (d) 36 and
(x + y), or you could say 36 and x + y; (e) 36, x and y. (Note: some of these
expressions could be factored further. For example, 36×8= 25 ×32. But the
“explicit” factors are as listed here.)
Solution 2.26 8/53 is larger.

Solution 2.27
Solution 2.28 (a) .833 … (repeating); (b) .22 … (repeating).

Solution 2.29

Solution 2.30
Congratulations if you got this right.

Solution 2.31

Solution 2.32

Solution 2.33

Solution 2.34 (a) 39/8; (b) 27/80;

Solution 2.35 (a) Try x = 1. Then Correct version:

(b) Try u = 0. Then There
is no way to simplify the given fraction. (Many other counterexamples could be
used here.)

Solution 2.36 Let Then by Eq. 2.32 we have By

inspection we see that Therefore

Solution 2.37 (a) 9/4; (b) 4; (c) −33/7; (d) 2/7.

Solution 2.38 15/44.

Solution 2.39 6.937037…. The repeating group is 370.

Solution 2.40 We assume that ab = ac for some a = 0. Then

Solution 2.41 (a) 25/33; (b) 10, 010/9, 999.

Solution 2.42 151 and 157 are primes. ; 153 = 9 × 17; 155 = 5 × 31.
Chapter 3
Using Elementary Mathematics
In many ways, applying mathematics is more difficult than the math itself. This
book tries to convince you that math really is useful. But you may have to work
hard to ‘get it.’ As always, persevere, and go slowly and carefully over parts that
you don’t fully understand at first.
Memory-learning is all but futile in applied math. There are just too many
possibilities. What one must learn is an understanding of certain basic principles.
For example, the proper handling of units is essential in many applied problems.
Unfortunately, working with units is often not stressed in school. In this chapter,
we pay special attention to using units. Other basic principles, such as scale
effects and proportionality are also discussed.

3.1 Calculator math

This section applies to the most basic type of calculator, which does little more
than add, multiply, subtract, and divide. However, the methods explained here
apply equally to a scientific calculator, if you happen to own one. Scientific
calculators are discussed in Sec. 5.5.
Familiarization
First, check the function of the keys on your calculator. Turn it on (unless it is
light-operated, in which case it is always on). Try some simple calculations. For
example, to add 32 and 59, key in ‘32 + 59 = ‘. You’ll get 91.
For your calculator, what is the maximum number of digits that the display
can hold? What happens when you try to enter too big a number? What if you
try to multiply two acceptable numbers, when the product is too big to display?
Most calculators will stop functioning in these situations, probably displaying E,
or Error. To reactivate the calculator, push or (Clear), maybe twice. If
that doesn’t work, try turning it off then on again.
Next, try to divide by zero (key in ‘1 ÷ 0 = ’). You should see the error
message again.
Find out how to enter a negative number, say −17.5. There should be a
key, or a (change sign) key.
If you have other keys, such as or ignore them for now.

Addition, multiplication, and subtraction

You can add a list of several numbers as follows: key in ‘19 + 65 + 22 = .‘
Subtraction is also easy: 21.8 − 13.2 − 5.6 = ? Try it; you’ll get 3. Also try 86.3 ×
7.7 × 12.3 = 8, 173.5 (remember, you should normally round off to maintain
similar accuracy to the input numbers).
Combinations
Now try to calculate 13 × (68 + 28). If you get 912, you’ve made an
understandable error. You probably keyed ‘13 × 68 + 28 = ’. Can you figure out
why this is wrong? If not, look up ‘brackets’ in the index of this book, and read
about them. Do you remember that bracketed expressions must be calculated
first? So how should you calculate 13 × (68 + 28)? Some calculators allow you to
use brackets, but many cheap ones don’t. If not, try this: 68 + 28 = (giving 96),
then ×13 = (giving the result 1,248). Another method is to use the calculator’s
memory – more on this later.

Problem 3.1 Find (a) 31.2 × (16.2 − 2.7). (b) 16 ÷ (21 − 7). (This may be
confusing.)
Avoiding errors
By now you realize how easy it is to make mistakes while using a calculator. And
this can be important. Sometimes I’ve been presented with outrageously wrong
bills by store clerks who made calculator errors. (Usually I don’t consider these
errors to be deliberate. One exception was a transaction involving a foreign
currency – always a confusing situation.) The two most common errors are:
entering a number incorrectly, and doing the wrong calculation (as in some of
the above examples). To avoid the first error, be sure to check each number on
the display, as soon as you’ve entered it and before pressing any operation key.
To avoid doing the wrong calculation altogether, first make sure that your
method is right. If in doubt, try it on a simple similar example, as a check.
Another extremely useful way to check your calculator’s results is to do a
quick, approximate mental calculation (Sec. 1.5). This is how I usually know
when someone gives me a wrong bill. Example: 21.55 ÷ (11.2 + 18) is
approximately 20 ÷ 30, which equals 2/3 or about .7. You’d know for sure that a
(mis) calculated result like 19.92 is dead wrong.
Using memory
Almost every hand calculator has a single-register memory, where you can store
a single number, which can later be recalled. As an example, consider a
calculation like (x + y) × (z + w), where x, y, etc. are certain numbers. There’s no
obvious way to key in this calculation on most cheap calculators. However,
memory comes to the rescue. The algorithm is:
calculate x + y and store in memory
clear the display (this does not clear memory)
find z + w
multiply (×)
recall memory
 equals

Try it on (66 + 18) × (39 − 17) = 1, 848. (Here is the calculation as done on
my small TI calculator: First, clear anything in memory by pressing
twice. If the memory is not clear, an ‘M’ is displayed on the LED. Now key in
This stores 66 + 18, or 84, in memory. Next key in
giving the answer, 1,848. Here, the key
stands for ‘Memory Recall,’ which brings the memory value 84 onto the screen
for multiplication.)
You will need to read your owner’s instruction book (if you have one) to
familiarize yourself with using memory. Again, using memory is normally easier
with a more advanced calculator. In fact, if you’re serious about upgrading your
math skills, I recommend not spending too much effort mastering a dime-store
calculator. Instead, get a scientific calculator – but first read Sec. 5.5.

Problem 3.2 Using memory, calculate (a) (70.3 − 47.6) ÷ (18.7+14.6). (b) (2.5 −
16.3) × (16.1 − 5.5). Do quick mental approximations to check.
Calculator accuracy
Try this calculation: ‘1 ÷ 7 = ×7 = ’. You may not get the exact answer, 1.
Calculators can only keep track of a finite number of decimal places, and a
number like 1/7 is a repeating, infinite decimal. My own dime-store calculator
gives (1÷7)×7= 0.9999997, but my HP Scientific calculator gives (1 ÷ 7) × 7=
1.000000000. Nevertheless, the HP calculator also sometimes does given slightly
inaccurate results – it’s inevitable. Anyway, be glad that you’re smarter than your
calculator, in some ways. You know that (1/7) × 7 = 1, but a calculator doesn’t
necessarily know this.

Problem 3.3 Try 1 ÷ 4 × 4 and 1 ÷ 8 × 8 on your calculator. How come?

Square roots
Your calculator probably has a square root key, labeled Find (first you
key in 9, then ). You may get 3. But possibly you’ll get something like
2.999997, which is the calculator’s approximation to 3. If you’ve forgotten what
means (it is read as ‘the square root of 9’), this calculation should remind
you. What is the definition of in general?
Answer: is a positive number which, when squared gives back x. Thus
= 3 because 32 = 9. Likewise = 5 because (what?)
Next, find and check (on the calculator) that Key in
My calculator gives the result 6.9999999.
Also, try it the other way around: find The answer is 7.5 (which
is what my calculator gives), but once again you may get an approximate result.
These approximations again result from the calculator’s rounding off of decimal
numbers.
Next, try to calculate (key in ). This should produce an
error message. Be sure that you understand why. If not, read the above definition
of again: is a (positive) number whose square equals x. But the square of
a number is always a positive number, so there’s no way for it to be −4. Many
students find this point confusing. If you still do, I suggest you write down,
without looking at the book, an explanation that is completely convincing to you.
Another point of possible confusion is that some students think that
(‘plus or minus 2’). It is possible that this is what you were taught in
school. It is incorrect, however. The correct statement is that and there
is a good reason for this – namely, mathematics always avoids ambiguity. The
same for calculations: if the calculator didn’t ‘know’ whether to say that
calculations could not be completed at all. This is why the basic
definition (above) says that is a positive number whose square is x. (By the
way, is true; I should have said that is a non-negative number whose
square is x, in order to include this case.)
Final point: Assume that x ≥ 0. Then (a) = what? (b) = what?
Answer: (a) x;(b) x. Check for yourself that the answers to (a) and (b) follow
directly from the basic definition of .
 Quiz question: Find Answer: 5. Check this. First you get (−5)2 = 25,
then √25 = 5. This example shows that the equation = x is correct if we know
that x ≥ 0tobegin with. If x< 0 the correct equation is = −x. Be sure you
understand this.

3.2 Per cent

‘Per cent’ means just that (in Latin, actually): out of one hundred. Thus 17%
equals 17/100, or .17. Mathematically speaking, we could do without per cent,
but there are many places where it is useful.
In practice, per cent virtually always refers to percent-of something else. If
sales tax is 11%, then on an item costing $50 you will pay 11% of $50 = .11 × 50 =
$5.50 sales tax. If you get 16% on your math test (shame!) it means that your
score was 16% of a perfect score. For example, 8 marks out of 50 would be 8/50 =
.16, or 16%. If the bank pays annual interest on savings certificates, and you
buy a $500 certificate, you will receive of $500 = .045 × 500 = $22.50
interest after one year.
Your calculator may have a % key. Mine works like this: key in ‘500 × 4.5%’
which gives 22.5 or $22.50. However, I find it easier to convert to decimals: ‘500
× .045 = .’

Problem 3.4 Your restaurant check is $62.50. Exactly how much is a 15% tip?
What do you get by quick approximation?

Problem 3.5 You purchased 100 shares of Axiom Corporation. In the first
month the price of Axiom shares went up by 50% of your purchase price. In the
second month the price fell by 50% from it’s one-month level. Did you gain or
lose, overall? How much?

3.3 Areas
The basic formula related to area is the rectangle formula:
where A denotes the area of a rectangle which has length l and width w.

(A rectangle is a four-sided figure, with each corner being a right angle, i.e., a 90
deg angle. Look up ‘right angle’ in Chapter 5 if you are not familiar with this
concept.) Of course, the length and the width are measured using the same units
(the same ruler, in other words). For example, if you actually measure the above
rectangle with a metric ruler, you will see that it has length l = 3 cm andwidth w
= 2 cm. The area is A = 6 square centimeters, which we usually write as A = 6 cm2
[cm2 should be read as ‘centimeters squared,’ or else as ‘square centimeters’].

The next figure shows why we know that A = 6 cm2 in the above example.
The 3 × 2 rectangle can be ‘cut up’ into exactly 3 × 2, or 6 one-by-one squares,
each having an area of 1 square centimeter.

In other words, the problem of finding the area of a rectangle is just an example
of the relation between multiplication and counting, as discussed in Chapter 1.

This argument works perfectly well for rectangles of any size l × w, at least in
the case that l and w are whole numbers. However, the basic formula A = lw is
valid for all lengths l and widths w, not just whole numbers.
Problem 3.6 Make a sketch to illustrate the fact that the area of a square with
sides cm is A = 1/4cm2. Suggestion: start with a ‘unit’ square, having sides 1 cm.

The example in Problem 3.6 can be used to argue that the area formula for
rectangles, Eq. 3.1, is valid for any values of l and w. For example, if a wall is 8
feet high and 14 feet 3 inches long, its area is square feet.
Area of a triangle
We’ll start by having you do an experiment.

Problem 3.7 Using a ruler and pencil, draw a rectangle on paper (draw the
corner angles ‘by eye,’ or by using the end of your ruler). Also draw a diagonal, as
shown.

Now use scissors to cut out the rectangle. Also, cut it along the diagonal.
What can you say about the two triangles you get? If you were careful, the two
triangles are identical.
Now consider a right-angled triangle, as in the experiment.

What is its area A? Just ½ the area of the original rectangle, right? Let us label
the sides of the triangle which form the right angle, as b (for base) and h (for
height). Then the area of the rectangle is bh, so that the formula for the area of a
right-angled triangle is

Equation 3.2 in fact holds for any triangle, not just right-angled triangles. The
‘height’ h means the perpendicular height above the base:

Problem 3.8 Using your ruler, measure the base b and height h for the triangle
shown above. Find the area A.

How can we be sure that Eq. 3.2 holds for an arbitrary triangle? Let us
imagine slicing the given triangle into very thin slices parallel to the base. Next,
slide all the slices to the right so that they butt up against a perpendicular line, as
shown. The new triangle is a right-angled triangle, which has the same base, the
same height, and the same area, as the original triangle. (Pause to make sure you
agree with this.) Since for the right-angled triangle, it must also be true
that for the original triangle.

This idea works also when the original triangle looks like this:
Complete the sketch for this case. How does one specify h, the height of the
original triangle? (Answer: one first extends the base line b to the right.)

Problem 3.9 Another way to show that for any given triangle is to cut
the triangle into pieces, somehow. Can you elaborate?

Problem 3.10 What is the formula for the area of a parallelogram (a four-sided
figure having its opposite sides parallel to each other)? How do you know?

3.4 Scale effects

An important math concept in Science is the idea of scale effects. For example,
consider two solid objects, made of the same material, which are exactly alike in
shape but different in size. If the second object is, say, twice as large (in
dimensions) as the first object, how are their weights related? Try to guess, then
try to explain.

The answer is that the second object weighs 23 = 8 times as much as the first
object. (Recall that 23, which reads as ‘two cubed,’ means 2 × 2 × 2, which is 8.) If
the scale factor is k, rather than 2, then the second object weighs k3 times as
much as the first one. By ‘scale factor’ we mean the ratio of the dimensions of the
two objects.
 If you own two dogs, you can check the scale effect:

What would you predict Fido’s weight to be? Answer: 33 lb ×(30/18)3 = 153 lb.
The actual weight could differ slightly from this, of course.

The general rule for the geometric effects of scaling is shown in the next
diagram.

In words, under scale change (with scale factor k), length scales as k, area as
k , and volume as k3. (In these pictures, the new object is a scaled-up version of
2

the old object. This means that k> 1. However, the formulas are correct also for
scaling-down, in which case k< 1.)

Problem 3.11 An elephant is about 4 m tall, and a mouse about 4 cm. If the
mouse weighs 3 g, about how much should the elephant weigh? (See Sec. 3.7 for a
discussion of weights and measures.)

Now for the explanation that volumes increase as k3. First consider a cubic
volume, with side s units. Its volume is s3 cubic units (and this is why we read s3
as ‘s cubed’). For example, a cubic box with side 5 inches has volume 53 = 125
cubic inches. Now consider a second cube, with side ks, where k is the scale
factor. It volume is (ks)3 = k3s3 (see Chapter 4). Thus the ratio of the two volumes
is k3s3/s3 = k3.

How can we extend this argument to arbitrary shapes?

Well, any solid object can be imagined to be made up of a large number of

tiny cubes. The picture shows an example in two dimensions, but the idea is the
same in 3 dimensions, as well.

When we scale up the smaller object by the factor k, all the tiny cubes [or
squares, in 2 dimensions] get scaled up also by k, so their volumes go up by k3 [k2
in two dimensions]. Therefore the whole volume increases by k3 [the area
increases by k2, for a two-dimensional object].

This is the scale effect for volumes [and areas].

Example. If a small child falls over and hits it’s head on the ground, he may cry,
but the injury is unlikely to be serious. It can be a different matter if you fall over
and hit your head. The reason is that the energies involved in the impact scale up
by k5,where k is the scale factor (relative heights) of the two persons. For
example, a 6-foot man’s head will hit the ground with 25 = 32 times the energy of
a 3-foot boy, or 35 = 243 times as much as a 2-foot toddler.

The factor k5 comes from basic Physics, namely where E is

energy, m is mass, and v is velocity. As we have seen, mass scales by k3. Also the
velocity of an object falling to the ground scales by k = height of fall. This makes
E scale up by k3 × k2 = k5. I just mention this example to illustrate the importance
of scale effects in Science.
Here is a completely different example.
An application of scale to traffic congestion
A certain city is expected to increase by 50% in population over the next 20 years.
How much will the city expand geographically, as a result? We also want to
forecast how traffic conditions may change.
To begin with, let us make some simplifying assumptions. Suppose that as
the city grows larger, its population density remains the same. Also, suppose that
the city expands uniformly in all directions, retaining the same shape (at least
approximately). Let k denote the scale factor for the dimensions of the city. Then
the new and old areas are related by

By assumption, Anew = 1.5Aold. Therefore k2 = 1.5, so that k = = 1.22. The

linear dimensions of the city will increase by 22% when the population increases
by 50%.
Next, consider the traffic situation. There will be 50% more vehicles, but also,
because the city is now 22% larger, driving distance per trip will become larger,
perhaps by the same factor of 22%. In other words, when the size (linear
dimension) of the city increases by the factor k, area increases by k2, and traffic
volume by k3. (Traffic volume is measured in vehicle miles per day.) Consider the
following two possibilities: (a) city planners increase the road system in
proportion to population size, that is, by the factor k2; (b) they increase the road
system in proportion to the volume of traffic, that is, by the factor k3.
In case (a), traffic congestion (cars per mile of road) increases by the factor k.
As the population continues to expand, congestion becomes saturation, and
traffic flow stagnates. In case (b), the road system eventually takes over virtually
all of the available level space in the city. The city planners are faced with an
impossible dilemma!
Here is a numerical illustration of case (b). Suppose the city originally has a
population of 100,000; call this P0. Let P be the population at some later time,
when the scale factor equals k:

Similarly, A = k2A0. Also, if R denotes the area covered by roads, then in case (b)
we have:
The proportion of the city’s total area A covered by roads equals

For example, suppose that originally, roads covered 20% of city’s area, i.e., R0/A0
= 20% = .2. Here is what happens as the city grows:

As the table shows, it is impossible to prevent traffic congestion by building

more roads. This is why freeway systems don’t work.
What other options are available to accommodate the increased need for
transportation? Public transit systems, to remove private cars from the roads, is
the usual proposal. The problem is that exactly the same scaling argument
applies to transit systems. Either (a) they become steadily more overcrowded
(Tokyo), or (b) they use up an increasing proportion of the city’s land area, or
both. (In addition, transit systems do not accommodate commercial vehicles.
This is why trucks take over the road system in increasing numbers, as the city
grows.)
The scaling concept is exceptionally powerful in studying phenomena
associated with the size of objects. Unfortunately, scaling is seldom stressed in
school math courses. Scientists and engineers learn about scaling in their
university courses, but students in other fields don’t usually take such courses.
Many economists, in particular, seem not to understand scale effects. The same
applies to politicians. These influential people see perpetual economic (and
population) growth as desirable, but they fail to recognize that many current
socio-economic problems may be inevitable scale effects of the ever increasing
size of our human population.
Mathematical models
The above discussion of traffic problems is an example of what is called a
mathematical model. Mathematical models are increasingly used today,
especially in the life and social sciences – biology and economics, for example.
This book can’t give you a complete course on mathematical models (how do
you make them, and what can you do with them?). However, here are some of
the features common to most mathematical models.
1. A mathematical model is a description, in mathematical terms, of some
real-world phenomenon, or situation. (For example, ever-worsening traffic
congestion.)
2. The model is a deliberately (and artfully) simplified description of the real-
world problem. A good model captures something essential about the real
world, and often gives surprising and unexpected insights into the modeled
phenomenon. (I invented the traffic congestion model for this book, as an
illustration of the idea of scaling. I was then surprised by the prediction:
traffic congestion to the point of stagnation may be an inevitable
consequence of scale effects in large cities.)
3. The model can be used for ‘thought experiments.’ Are the predictions made
by the model strongly dependent on the underlying (and perhaps hidden)
assumptions? What if these assumptions are modified? For example, I
assumed that the average commuting distance increased in proportion to
the scale factor k. What if it increases proportional to or some other
expression? Doubtlessly you could detect other tacit assumptions of the
model.
4. The model can be made more incisive by gathering data from the real
world, and adjusting certain aspects of the model to concur with the data.
For example, cities of various sizes in North America could be studied in
terms of total road area, population density, commuting distances, and
congestion factors.

Mathematical modeling is an interesting, creative, and useful activity. This

book will discuss other mathematical models on occasion, but usually as
illustrations of elementary mathematics, rather than as detailed studies in their
own right. One thing is certain – to appreciate, not to say to invent and analyze,
mathematical models, one must be highly adept at all aspects of basic math.

3.5 Proportionality
Proportionality is a basic concept frequently encountered in mathematics and
science. Some everyday examples are:
1. The amount of gas used on a driving trip is proportional to the distance
traveled.
2. The cost of a bunch of bananas is proportional to their weight.
3. The time needed to mow a lawn is proportional to its area.

Often the clue to recognizing proportionality in a certain situation is the

realization that if one of the two quantities is doubled (for example), then the
other quantity will also be doubled. For example, twice as big a bunch of bananas
will cost twice as much. Also, three times as big a bunch will cost three times as
much, and so on.
In mathematical terms, two quantities U and V are said to be proportional to
one another if there is a constant c such that

for all values of U and V . The constant c remains the same no matter what
values U and V have. In a practical situation, c usually has a meaningful
interpretation. The units for the constant c are given by

This is required to make the units on both sides of the equation U = cV match.
The symbol ∝ is used to denote proportionality:
 Thus U ∝ V means that there is some constant of proportionality c such that
U = cV.
For example, consider the cost of bananas. Let Y be the cost of a bunch of
bananas, and let W be weight of the bunch. Then Y is proportional to W , so that

In this case the constant of proportionality c is the price of bananas. For example,
if bananas cost 45¢ per lb, then c = 45¢/lb. To calculate the cost of 5 lb of
bananas, we have Y = 45¢/lb × 5 lb = 225¢, or $2.25.
Every proportionality problem is basically the same as this example.
As a second example, consider the question of gas consumption for a certain
automobile. Then D ∝ G, where D is distance driven and G is gas consumption.
Thus D = mG for some constant m. One could write instead that G ∝ D and G =
cD, but I want the constant of proportionality to have its usual interpretation,
mileage. In American units, D is measured in miles, and G in gallons. Therefore
the units of m are miles/gallons, or miles per gallon.
To determine the mileage for your car, first fill the tank and set the trip meter
to zero. The next time you fill the tank, record the amount of gas you bought, G,
and the distance on the trip meter, D. For example, suppose G = 8.3 gal and D =
225 mi. Then from D = mG we have m = D/G = 225 mi/8.3 gal = 27.1 mi/gal.
Repeating the experiment will probably give a slightly different value for m,
depending on driving conditions.

Problem 3.12 Continuing with the above example, suppose that freeway driving
is 50% more efficient than city driving, in terms of mileage. The figure 27.1
mi/gal applies to city driving. If gas costs $2.35 per gallon, what is the gas cost for
a trip of 1500 mi, using the freeway system?

(Notice especially how the act of writing down the units, mi and gal, helps
ensure the right calculation. More about this topic in Section 3.7).
(In metric countries, the gas-consumption figure is usually given in terms of
liters per 100 kilometers. For example, a consumption rate of 8 liters per 100 km
would be the same as 100 km/8 liters, or 12.5 km/liter. If you find this confusing,
you’re not alone.)

Problem 3.13 You are waiting in the doctor’s office, along with 5 other patients.
You notice that it takes 12 minutes for the first two patients to be called. Estimate
how much longer you will have to wait. Also, express this situation in terms of
proportionality.
Multiple proportionality
Here is a typical proportionality problem from school days. If 2 people take 4
hours to paint 30 feet of fence, how many people are needed to paint 150 feet of
fence in an 8-hour day? You can solve such problems as a kind of fill-in puzzle,
like this:
2 people take 4 hours to paint 30 feet
1 person takes 8 hours to paint 30 feet
5 people take 8 hours to paint 150 feet

The answer is 5 people.

Many people find such problems a bit confusing. What’s really going on?
Let’s see if we can set this example up in terms of proportionality. First, twice as
many people will paint twice as much fence, in any given period of time. Thus L
∝ N , where L = length of fence and N = number of people. Also, a given number
of people will paint twice as much fence in twice as much time. Hence L ∝ T ,
where T = time. We have ‘double proportionality.’
In such a situation, we have

for some constant c. In other words, L is proportional to the product NT . Check

that this works: doubling N doubles L (right), and also doubling T doubles L
(right).
Next, we calculate c from the given information, which is that L = 30 feet if N
= 2 persons and T = 4 hours. Therefore

or c = 30 feet ÷ (2 persons × 4 hours) = 3.75 feet/person/hour. This makes sense

– c is the rate at which the fence gets painted, in feet per person per hour.
Finally, the answer to the problem is obtained by substituting L = 150 feet
and T = 8 hours into the proportionality equation:
Solving for N gives

Double proportionality, with equations like U = cV W , is common in science

and math. Many examples will be encountered in this book.

Problem 3.14 The cost of painting an 8 ft × 12 ft wall with primer at $5.00 per
quart is $17.50. Find the cost of painting four such walls with paint costing $9.50
a quart.

To summarize, the sequence of steps used in solving a proportionality

problem (whether single or multiple proportionality) is:
1. Introduce symbols for the quantities in the problem.
2. Set up a proportionality equation, and do a mental check on it (twice the U
implies twice the Y , etc.)
3. Use the given information to calculate the constant of proportionality.
4. Finally, use the equation to calculate the desired quantity.

Problem 3.15 Six cats together eat 35 cans of cat food per month, at $1.69 a can.
How much does it cost a menagerie to feed 100 cats for a year?

With a little practice, you can perhaps solve proportionality problems in your
head, without worrying about equations and constants of proportionality. In
science or math, however basic laws are often expressed as proportionality
equations. Familiarity with the mathematics of proportionality is therefore
important.
Other types of proportionality
Other types of proportionality than those discussed so far are common in
science. Two general examples are:

In the case we say that U is inversely proportional to V.

Here is one particular example, discovered by Isaac Newton. The
gravitational force of attraction (F ) between two bodies of mass m1 and m2 is
proportional to both masses, and inversely proportional to the square of the
distance (r) between them. In symbols

where G is a constant called the universal gravitation constant.

You will encounter many other examples of proportionality in this book. It is
a basic concept in science, and in math.

3.6 Speed and distance

The speed of a moving object is defined as the distance traveled divided by the
time taken:

Expressing this in terms of symbols,

where S is speed, D is distance, and T is time. Equation (3.4) is the basic math for
this section of the book. Sometimes speed is called velocity, but for now we stick
with the word speed.
For example, suppose a man walks to work, a distance of .5 miles, in 15
minutes (1/4 of an hour). His speed for the journey is thus

This is pretty slow walking – normal walking is about 3 mi/hr. Perhaps he was
slowed up at street crossings. The table lists several familiar examples, using both
American and metric systems of units. (The question of converting between the
American and metric systems is discussed in the next section.)
Situation Speed (American) Speed (metric)
Person walking 3 mi/hr 4.8 km/hr
Olympic marathon runner 13 mi/hr 21 km/hr
Car on a freeway 65 mi/hr 104 km/hr
Passenger jet 610 mi/hr 980 km/hr
Speed of sound 1089 ft/sec 332 m/sec
Speed of light 186,000 mi/sec 300,000 km/sec

The first thing to notice about the speeds shown in the table is that speed is
always specified in appropriate physical ‘units,’ such as mi/hr, km/hr, mi/sec, and
so on. The expression mi/hr is read as ‘miles per hour,’ and is sometimes
abbreviated as MPH. Speed limit signs in America often use this abbreviation
(other countries use KPH, kilometers per hour).
Passenger jets do not travel as fast as the speed of sound (unless they are
supersonic, like the Concorde). Military aircraft, however, are often capable of
exceeding the speed of sound. If you read that some new fighter jet flies at Mach
3, this means it flies at 3 times the speed of sound.
Notice how great the speed of light is, 186,000 miles per second. No space
craft will ever fly faster than this, because of the basic principle of physics which
says that nothing can move faster than the speed of light.
Let’s return to the basic equation, 3.4:

This equation has three ‘variables,’ S, D, and T . By ‘variable’ we mean a symbol

representing some quantity, where the quantity can have different numerical
values. Equations containing variables are very common in math and in Science.
Another example from basic math is A = lw, the formula for the area of a
rectangle. This formula also has three variables.

Solving the equation

Given an equation with (say) 3 variables, it is always possible to solve the
equation for any one of the variables if the values of the other two variables are
known. This procedure is common in science.
Please look at the next problem. If you haven’t a clue how to solve it, read on.

Problem 3.16 (a) Suppose you drive at 60 mi/hr for hours. How far have you
driven? (b) The sun is 93 million miles from the earth. How long does it take
light to travel from the sun to the earth?

In each part of Problem 3.16 you are given the values of two of the variables,
S, D, and T , and are asked to calculate the value of the third ‘unknown’ variable.
For example, in part (a) you’re told that speed S = 60 mi/hr and time T = 2.5
hours. How do you find the unknown distance D, using S = D/T ?
To find D we first solve the equation S = D/T for the unknown D:

Why is this true? And how can you remember it?

Dear reader: please pause at this point, and check whether you can explain
for yourself why Eq. 3.5 is true. One explanation is ‘because you can cross-
multiply.’ In other words, is equivalent to ac = b. (As a matter of fact, the
latter statement is the very definition of division – see Eq. 2.31.) Changing the
symbols, means that D = TS (which equals ST ). If you have any trouble
with this, you need to go back and read section 2.4. Another possible explanation
is that you can multiply both sides of the equation by T , giving ST = D.
This is OK – it’s really just a different way of expressing the basic definition.

Problem 3.17 Solve the equation S = D/T for T , and explain.

The result of Problem 3.17 is that:

Do you need to remember all three of these equations? No, just remember
the basic equation S = D/T and realize that the other two equations can be
deduced from this. By the way, if you have trouble memorizing S = D/T , think of
an example, such as 20 miles (D) in one hour (T ) equals 20 miles per hour (S =
D/T ). Also, you can remember that ‘speed = distance over time.’
 Now, returning to Problem 3.16 (a): given S = 60 mi/hr and T = 2.5 hr, we
have

In this calculation, notice that the numbers are multipied (60 × 2.5 = 150), and
the units are combined The latter operation of combining
units is extremely useful and important. Most practical applications of
mathematics involve the proper combination of units of measurement.
Returning to Problem 3.16 (b): given D = 93×106 mi and S = 186×103 mi/sec
[speed of light], find T . First, S = D/T implies that T = D/S. Therefore the
solution to Problem 3.16 is

(because 8 min = 8 × 60 sec = 480 sec).

To learn more about units of measurement, read Section 3.7. To learn more
about solving equations, read Chapter 4.

Problem 3.18 A light-year is the distance that light travels in one year. Use your
calculator (and scientific notation) to find the number of miles in one light year.
Average speed
Suppose you commute from home to work. For the first 10 km your speed is 100
km/hr, but then you get into heavy traffic for the next 10 km, where the speed is
only 40 km/hr. What is your average speed for the whole trip? It’s not 70 km/hr,
as one might first think.
To calculate average speed for a trip in which actual speed varies, we still use
the basic formula

where D is the total distance, and T the total time of the trip. For the stated
problem, D = 20 km. But what is T , the total time? To find T we have to consider
the two segments separately. Let’s write T1 and T2 for the two segments, and also
S1,S2 and D1,D2. Then, from Eq. 3.6,

Therefore and the average speed for the trip is

Note that this is considerably slower than the naive (incorrect) guess of 70
km/hr. Slow traffic has a big effect in lowering your average trip speed, as most
commuters are aware.

Problem 3.19 You fly from Dallas to Houston, a trip of 244 miles, in a jet that
flies at 490 mi/hr. You take 30 min getting to the Dallas airport, 45 min waiting
before departure, and 30 min waiting for your luggage at Houston. Counting the
waits at both ends, what is your average speed for the trip? Total time? (This
might be useful knowledge if you want to compare flying with driving.)

3.7 Units of measurement and their conversion

The metric system, sometimes called the SI system (the International System of
Units; SI comes from the French rendition, Système Internationale d’Unités), is
used throughout Science. It is now the standard system of measurement in most
countries, the main exception being America. The accompanying table lists the
most common metric measures, and gives conversion factors between metric
and American units.
For quick, everyday use, approximate conversion factors can be used in
mental calculations. For example,

Automobile drivers crossing from the US to Canada or Mexico can

memorize the following approximate values:
 To convert elevations (for example of mountain peaks) from metric to
American, I use the approximate algorithm

For example, 1800 m ≈ 5400 + 540 = 5940 ft (exact value is 5910 ft).
A similar mental algorithm converts masses:
For example, an 8 kg bag of sugar weighs 16 +1.6= 17.6 lb(which is exact).
Temperature
The temperature scale used in most countries (other than the US) is Celsius. The
Celsius temperature C and its Fahrenheit F equivalent are related by the (exact)
formula

For example, the freezing point of water is 0 deg C, or by Eq. 3.8, 32 deg F .
Similarly, the boiling point of water is 100 deg C, or 212 deg F .

Problem 3.20 There is one temperature at which the Celsius and Fahrenheit
numbers are the same. What is it? (Use Eq. 3.8).

For everyday use, an approximate rule is that F equals 2×C plus 30 deg. This
pretty good for normal outside temperatures, though it goes a bit awry for hot
days:

Problem 3.21 (a) Solve Eq. 3.8 for C. (b) Use part (a) to calculate the Celsius
equivalent of normal body temperature, 98.6deg F .
Units and dimensions
You have no doubt heard the statement ‘you can’t add apples and oranges.’ But
what can you add? Well, first we can certainly add (and subtract) quantities
expressed in the same units:

We can also add quantities expressed in different units, provided that the units
can be transformed into one or the other:

because 1 cm = .01 m.
What we cannot add or subtract are quantities that have different
dimensions. For example, 5 yards plus 20 seconds is just nonsense. Some
examples of basic physical dimensions are:
Basic SI Units
Length (L) m (meter)
Mass (M) g (gram)
Time (T) s (second)

There are other basic dimensions, for temperature, electricity, and so on, but
we will not consider these here.
We can add or subtract two physical quantities that both represent length, or
mass, or time, but we cannot add a length to a mass, etc. If the physical quantities
are expressed in different units, we can add them after transforming to a single
unit.

Problem 3.22 Using the conversion table on page 135, find: (a) 5.3 tonnes plus
850 kg; (b) 3 hrs, 25 min, and 18 s (in terms of seconds); (c) 9.5 ft minus 2.7 lb.

Although quantities having different dimensions cannot be added, they can

sometimes be usefully multiplied or divided. We have already seen one example,
speed = distance/time. In basic SI units, the dimension of speed is therefore m/s
(meters per second).

Problem 3.23 A bandit is running towards the rear at 10 ft/s, on top of a rail car
travelling west at 40 mph. Find the bandit’s speed relative to the ground.
Mass versus weight
(You can skip this discussion unless this question interests you. But it does
illustrate the idea of physical units, or dimensions.) What is the distinction
between mass and weight, if any? I recently read in a school math text that, to
convert mass to weight you multiply by 9.8. This is utter nonsense – mass and
weight are completely different concepts, at least in Science. (In everyday life, we
often use ‘weight’ when the scientifically correct term would be ‘mass.’ If the
butcher tells you that the meat weighs 450 grams, he should say that its mass is
450 g. Its weight is actually 4.4 Newtons. Tell that to the butcher!)
The mass of an object is the same no matter where it is located. But the
weight depends on circumstances. Astronauts in an orbiting spacecraft are
weightless, but not massless. The weight of a moon rock located on the moon is
not the same as its weight on earth, but its mass is the same.
Weight is actually a force. Your weight as you sit in a chair is equal to the
force that the chair exerts on you to counteract the pull of gravity. (This is the
same force that your body exerts on the chair.)
The great 16th century mathematician Isaac Newton discovered the basic law
of force:

where F is the force, m is mass, and a is acceleration. This equation is one of the
keystones of Science, every bit as important (and famous) as Einstein’s E = mc2.
(The meaning of ‘acceleration’ is discussed below.)
Let’s not discuss the full significance of Eq. 3.9 here. For now let’s just use
this equation to explain weight W , for a mass m at rest on earth. Here we use a =
g, the acceleration of gravity on earth. The value of g is 9.8 m/s2 (see below). By
Eq. 3.9 we have, since weight is a force

Consider the example m = 450 g. Then

That’s right, the basic SI unit of weight has been called the Newton in honor of
Sir Isaac. Weight is measured in Newtons. One Newton (symbol N ) equals 1000
g m/s2 (gram meters per second squared), or 1 kg m/s2.
You have actually observed your own weight changing rapidly on many
occasions. Get on an elevator. When the elevator starts to go up, your weight
increases momentarily, according to Newton’s formula,

where a is the acceleration of the elevator. Likewise, when the elevator slows
down to stop at a floor, your weight goes down, because a becomes negative. Try
the next problem.

Problem 3.24 Here’s how you could measure the acceleration of an elevator (if
you wanted to do that). Take some bathroom scales, and stand on them in the
elevator. When the elevator starts up, your weight increases briefly, according to
Eq. 3.11. For example, suppose the scales first register your ‘weight’ as 80 kg. (a)
This is not correct scientifically – what does it actually mean? (b) Now suppose
that as the elevator starts up, the scale briefly indicates 100 kg. Calculate the
acceleration a of the elevator.
Acceleration
What exactly is acceleration? You probably realize that acceleration has
something to do with a change of speed. Specifically

where ΔS (read as ‘delta S’) is the change in speed, taking place in time ΔT . For
example, it is said that the Mercedes-Benz CLK55 can go from 0 to 100 km/hr in
5 seconds. What is the acceleration?
First, let’s transform 100 km/hr to meters/second. Since 1 km = 1,000 m and
1 hr = 3,600 s, we get

You pay a lot of money, and you don’t even get as much acceleration as gravity,
which is free!
Notice that, with speed measured in meters per second, acceleration (the rate
of change of speed) must have dimensions (meters per second) per second, or
m/s2. The phrase ‘meters per second per second’ is often used, because it reminds
one that acceleration is the rate of change of speed.
This discussion doesn’t quite explain why the number g used in calculating
weights has anything to do with acceleration. The connection is that, if a body is
allowed to fall freely under gravity, it will in fact accelerate at g = 9.8 m/s2 (on
earth). For example, if the body starts from rest, then after one second it will be
falling at 9.8 m/s, and after two seconds at 19.6 m/s, and so on.

3.8 Review problems

1. You sell your home for $140,000. The agent’s fee is $8,400. What was her
 percentage?

2. Write down the formula for the area of a triangle. How is this formula
derived?
3. How would you find the area of a quadrilateral (4-sided figure)?

4. Suppose that a polygon (n-sided figure) is scaled up, using scale factor k. What
happens to its area? Its perimeter? (The perimeter is the length of the
boundary of the polygon.)

5. A golf ball has diameter 1.7 inches, and a baseball 3.0 inches. Find the relative
masses and surface areas of a golf ball and a baseball.

6. The earth rotates around the sun once a year. How fast is the earth moving in
space as a result of this motion? (The length of a circuit around the sun is
about 584 million miles, or 940 million kilometers.)

7. Express the speed 65 mph in terms of feet/sec.

8. (a) Show that if X is proportional to Y , and Y is proportional to Z, then X is

proportional to Z.

(b) The gravitational acceleration for a given planet is proportional to its mass
(Newton’s law of gravitation). If Neil Armstrong weighs 98.8% less on the
moon than on earth, what is the ratio of the moon’s mass to the earth’s?

9. The human population of the world was about 7 billion in 2010. The rate of
increase was 1.5% per year. (a) How many extra people will be added to the
world’s population per year? (b) What will be the world population in 2020?
[Don’t consider exponential growth yet – see the next Chapter.]

10. The energy E of a moving object is proportional to its mass m and to the
square of its velocity v. (a) Express the relationship in symbols. (b) The
constant of proportionality is in fact (no units). What units is energy
measured in, using the SI system?

Solution 3.1 (a) 421.2. (b) 1.14 approx. (With a basic calculator, you may have to
do this in steps. First find 21 − 7, which is 14. Remember, or write down 14. Now
calculate 16 ÷ 14. A little later you will learn how to use your calculator’s
memory for problems like this.)

Solution 3.2 (a) 0.682. (b) −146.3.

Solution 3.3 These calculations are probably exact. The reason could be that
1/4= .25 and 1/8= .125 are finite (i.e., terminating) decimals.
Solution 3.4 15% of $62.50 is $9.38. I would do a quick approximation like this:
take 10% ($6.25), add half of this (about $3.10) to get $9.35. Leave a tip of $9.50
or even $10.00.

Solution 3.5 If the starting price was p,the one-month price was 1.5p (that is, p +
50% of p), which next became .5 × (1.5p)= .75p. You lost 25%. (A mathematically
naive person might think that if the stock market goes up 5% on Monday, and
down 5% on Tuesday, it balances out. But, in fact this would mean that the
market lost 1/4% over the two days – agree?)
Solution 3.6

This diagram shows that the unit (1 × 1) square consists of 4 smaller

squares, each of the same area, which we call A. Hence 4A = 1, or A = 1/4. (The
square has been magnified.)

Solution 3.7 If you were careful, the two triangles should be exactly the same –
one overlaps the other perfectly. (Such triangles are called ‘congruent’ triangles –
see Chapter 5.)

Solution 3.8 A = (½) × 3.9cm × 2.0cm = 3.9cm2.

Solution 3.9 Cut the triangle as shown. You get two right-angle triangles A1 and
A2,with bases b1 and b2. Therefore Add these
together: where b =
b1 +b2 is the base of the whole triangle. (Note the use of the distributive law here;
see Chapter 1.)

Solution 3.10 The formula is A = bh, where b is the base and h the height. One
way to prove this is to cut the parallelogram along a diagonal, obtaining two
congruent triangles with base b,height h. Another way is to slice it horizontally,
and slide the slices to form a rectangle.

Solution 3.11 In this example we have k = 4 m /.04 m = 100. Thus k = 1003 =

1,000,000 (one million). The elephant’s weight should be about 3,000,000 g, or
3,000 kg.
Solution 3.12 First, mileage on the freeway is 1.5 × 27.1 mi/gal = 40.7 mi/gal.
Gas consumption for 1500 mi is G = D/m, or

The cost of gas for the trip is 36.9 gal × $2.35/gal = $86.72.

Solution 3.13 At 6 min per patient, and with 3 patients ahead of you, you can
expect to wait another 18 min. For a proportionality representation, let N be the
number of patients called in T minutes. Then you assume that T = cN . Since T =
12 min when N = 2 patients, we have c = 6 min per patient.
Solution 3.14 Let C = cost of painting area A using paint costing $P per quart.
Then C = kAP (twice the area means twice the cost, as does twice the price). Now
C = $17.50 if A = 96 ft2 and P = $5.00. This gives k = C/AP = $.0036 per square
foot [store this value in memory in your calculator]. With A = 4 × 96 ft2 and P =
$9.50 we get C = kAP = $133.00. Quick check: paint is nearly twice as expensive,
and the area 4 times as large, so the cost should be about 8 × $17.50 = $140.00.
Close.

Solution 3.15 $11,830.

Solution 3.16 (a) 150 mi. (b) 8 min 20 sec. (The explanation occurs later on.)

Solution 3.17 T = D/S. Explanation: first multiply both sides of S = D/T by T ,

getting ST = D (as before). Then divide both sides of this equation by S, getting T
= D/S. (This explanation could be worded in various other ways.)

Solution 3.18 Use D = ST = (186,000 mi/s) ×1 year, and convert 1 year into
seconds:

In words, light travels 5.9 trillion miles per year. Again, notice how the various
units ‘cancel.’

Solution 3.19 The time spent actually flying is Tf = D/S = 244 mi/490 mi/hr =
.498 hr, call it The extra times add up to 105 min, or 1 3/4 hr, so your total
time for the trip is 2 1/4 hr. Therefore the average speed, including the waits, is S
= D/T = 244 mi /2.25 hr = 108.4mi/hr.
Solution 3.20 −40 deg C or F . To get this, set C = F in Eq. 3.8 and solve:
gives F = −40 deg. (Read Section 4.5 if you need help in solving
equations.)
Solution 3.21

(Read Sec. 4.5 to learn how to solve equations.)

(b) If F = 98.6 then C = 37.0. This is normal body temperature stated in Celsius.
Solution 3.22 (a) 6.15 tonnes (or 6,150 kg); (b) 12,318 s; (c) cannot be calculated.

Solution 3.23 First we express the bandit’s running speed in mph:

(Note again how the units cancel, just is if they were mathematical symbols.)
Therefore the bandit’s net speed is 40 − 6.9 = 33.1 mph, in the westward
direction. (Sorry for the silly ‘real-world’ problem, but the idea was to illustrate
two important concepts, conversion of units, and addition of quantities having
the same units.)

Solution 3.24 (a) The scale is really indicating your mass, 80 kg. Your weight (at
rest) is 80 kg ×9.8 m/s2 = 784 N; (b) Your weight during the acceleration is 100
kg ×9.8 m/s2 = 980 N. According to Eq. 3.11, W = m(g + a). Therefore

Hence a = 12.25 − 9.8= 2.45 m/s2. This is the acceleration of the elevator.
Chapter 4
Algebra
There is admittedly something of a problem in teaching algebra – why should
anyone learn algebra, anyway? Basic algebra doesn’t have many immediate
everyday uses. It all seems quite abstract and meaningless to many students.
Moreover, algebra is often quite finicky. But being adept at algebraic calculations
is essential in later subjects, such as trigonometry and calculus.
Fortunately, algebra involves little more than the use of the basic rules of
arithmetic discussed in Chapters 1 and 2. The main novelty is that we now work
with general expressions, such as ax2 − b, rather than with numbers only. The
letters in an algebraic expression represent numbers, in the sense that the letters
can be replaced by numbers.
Here is an example, which we will discuss in greater detail later:

(Recall that a2 = a × a, and a2 is read as ‘a squared’. Thus 62 = 36, etc.) This is an

algebraic equation, or to be more precise, an algebraic identity. What this latter
term means is that the equation is true for every choice of numbers a and b. For
example, if a = 7 and b = 5, the left side of Eq. 4.1 equals

whereas the right side is

Problem 4.1 Verify identity (4.1) for a = 1, b = 9. Also for a = −4, b = 2.

[‘Verify’ means check that the equation is true.]

An important aspect of algebra is the conciseness and precision of the

mathematical notation. The Greek mathematicians, for example, never
developed algebra. They did study equations, but were only able to write them in
words. Imagine having to learn that
‘The difference between the product of one number with itself and that of a
second number with itself is the same as the product of the sum of the two
numbers and their respective difference.’
Believe it or not, this is the same as identity (4.1), but expressed in words.
The fact that mathematical notation is so precise means that you have to
learn to write it accurately. (The same holds in spades for computer
programming – any carelessness whatever will prevent the computer from doing
what the programmer wants it to do. ‘Garbage in, garbage out’ is the salient
phrase.) Keep this in mind, especially in studying this chapter. Try for 100%
accuracy, not 98% or 60%!
The replacement of a symbol in an algebraic expression by a number is called
numerical substitution. For example, if we substitute a = 3 in the expression a2 −
4 we obtain 32 − 4, or 5. Thus an algebraic identity is an equation that is true for
every numerical substitution. The basic laws of arithmetic, such as a + b = b + a,
are algebraic identities. Many other important examples of algebraic identities
are discussed in this chapter.

4.1 Exponents
In Chapter 1 we considered the powers of ten, 10n = 10 × 10 × ···× 10 (n times),
which were so important in the positional notation for expressing numbers in
base 10. Similarly, for any number x we define xn

The expression xn is read as (‘x to the nth’, or ‘x to the power n’), except for n
= 2 or 3; x2 is read as ‘x squared’ and x3 as ‘x cubed.’ In Eq. 4.2 x represents any
number. Thus, for example 34 = 3×3×3×3(which equals 81). Similarly, (−2.7)2 =
(−2.7) × (−2.7) = 7.29. Also, in Eq. 4.2, n represents any whole number, n = 1, 2,
3, etc. (Later we will explain what xn means for other values of n.)
In the expression xn the number n is called the exponent of x. Thus, in x5, the
exponent of x is 5. (The number x is sometimes called the base of the expression
xn, but we won’t use this word here). Similarly, in a8 the exponent of a is 8.

Problem 4.2 If you have a scientific calculator, the yx key can be used for
calculating exponentials. Try it on 3n for n = 2, 3 and 4, and check the results by
hand calculation. Now calculate 3n for n = 10, 20, 30. What do you conclude?
Also calculate (.7)n for n = 10, 20, 30.

The following rules of exponents are logical consequences of the basic

definition, Eq. 4.2:

Rules of exponents

Errors in applying these rules are often made by math students. As usual, a
little understanding can prevent such errors. You will never forget the rules if
you understand why they are true from the outset. Please read the explanations
carefully.
It’s worth emphasizing again that much of mathematics depends on basic
definitions of terminology and notation. Thus, rules 4.3 – 4.5 are direct
consequences of definition (4.2), as you will see.
To explain rule (4.3), we refer to the basic definition in Eq. 4.2. Thus xm
equals xx ··· x [m factors of x], while xn equals xx ··· x [n factors of x]. Thus xmxn =
(xx ··· x)(xx ··· x) where the first bracket has m factors of x and the second has n
factors of x. Altogether there are m + n factors of x, which means that xmxn =
xm+n.
An example may make this clearer:

The second rule, (4.4), is also easily explained. First xm = xx ··· x [m factors of x],
so that (xm)n = (xx ··· x)(xx ···x) ··· (xx ··· x), with n groups of (xx ···x). Counting
up all the x’s, there are mn of them, which shows that (xm)n = xmn. An example:
Problem 4.3 Simplify (a) x2x6; (b) (x2)6.

Rule 3 also follows directly from the definition, Eq. 4.2. We have

An example of this rule: (3a)2 = 3a · 3a = 9a2.

Repeated exponents
What does mean? Try to simplify it. The answer is x8, not x6 as one might
expect. This is a consequence of the following convention:

For example, calculate Answer (by calculator): 1,953,125.

Negative exponents
Let us look again at the basic definition, Eq. 4.2:

Here x can be any real number, but n has to be a positive integer, n = 1, 2, 3, etc.
Otherwise ‘n factors of x‘ is meaningless. However, we can expand on this
definition so as to make sense of xn for all integers n, whether positive, negative,
or zero. The main point is to organize things so that the laws of exponents, Eqs.
4.3–4.5 always remain valid. Keeping math simple and consistent has always
been a goal of mathematicians.
First, what does x0 mean? We can’t say that x0 = xx ··· x with zero factors of x.
This simply doesn’t make sense. But if Eq. 4.3 has to remain true for n = 0 we
must have

For this to be correct, we must have x0 = 1. Therefore we define

This definition may seem a bit strange at first, but it is necessary for
mathematical consistency. (Don’t worry about the side condition x ≠ 0, i.e., x not
equal to zero. I will explain this later.)

Problem 4.4 Write out the values of 4n for n = 3, 2, 1, and 0. Do the same for 10n.

Next, we can use a similar argument to find the meaning of x−n. First,
remember that we want Eq. 4.3 to be true for both positive and negative integers
(and zero). Let n = −m in Eq. 4.3:

But we have just defined x0 = 1 – see Eq. 4.6. Therefore

which tells us that

 (Can you explain this last step? Given two numbers a and b such that ab = 1,
what can we say? Well, we can say that

right? Check back to Chapter 2 on division, if you need to.)

An example: and so on.
Next, let us show that

An example will help to explain why this is true:

The same applies to the general case, the n factors of x in the denominator
cancel n factors in the numerator, leaving xm−n (assuming that m> n). Equation
4.8 is also true if m < n, as the next example indicates:

so that x3/x4 = x3−4 = x−1.

Problem 4.5 Express using a single exponent: (a) z8z−4;(b) t2/t6;(c)(w3)2÷ w3w2.

To summarize the discussion of exponents, we have defined xn for three

cases: n positive (Eq. 4.2), n equal to zero (Eq.4.6), and n negative (Eq. 4.7). Also,
we know that the first rule of exponents, Eq. 4.3 holds for all integer values of m
and n. It also happens that the second rule, Eq. 4.4, holds for all values of m and
n. I’ll let you prove this for yourself, if you’re interested.

Problem 4.6 (Optional) Prove that (xm)−n = x−mn. Use Eq. 4.7 and Eq. 4.4 for
positive values of m and n.
 Similarly, it can be shown that the third rule of exponents, Eq. 4.5, is also true
for a negative exponent:

(xy)n = xnyn for negative n (as well as positive n, or n = 0)

You should be able to prove this for yourself – try it for an example, such as (xy)
−5
= x−5y−5. Obtain this by using the equations already proved. Thus, first

Can you complete this?

The end conclusion is that the three Rules of Exponents (Eqs. 4.3–4.5) are
valid for all integers m and n, positive, negative, or zero. There is one exception,
however. First, Eq. 4.7 is not valid if x = 0, because the right-hand side,
if x = 0. But division by zero is meaningless! You might want to
write (x ≠ 0) beside Eq. 4.7.
Another related point: what is 00? This raises a possible inconsistency:

The first of these equations would suggest that 00 = 0, whereas the second
equation suggests that 00 = 1. Which is it? To avoid any inconsistency, we must
decide that:

(As a check, try 00 on your calculator – it will probably give you an error
message.)
To remind yourself of the basic ideas behind the rules of exponents, I suggest
you do the next problem as a brief review.

Problem 4.7 (a) What is the definition of xn when n is a positive integer? (b)
How does this definition imply the formula xmxn = xm+n? (c) Why is (xm)n = xmn?
(d) Explain why we define x0 as 1. (e) Explain why
 Finally, to check your grasp of exponents, try the next problem. Work
carefully.

Problem 4.8 Simplify these expressions: (a) (b−2)−1; (b) (x3)−1x5; (c) (x3y−2)4; (d)
(a2/b)3.

Exponents other than integers

Can we have an exponent that is not an integer? What does 52/3 mean? or 81.57?
A scientific calculator will give values for such expressions, so they must mean
something. We will take up this question in Section 4.8.

4.2 Exponential growth and compound interest

You have probably heard the phrase ‘exponential growth.’ What does it mean?
(To confuse matters, exponential growth is sometimes called ‘logarithmic
growth,’ and other times ‘geometrical increase.’ I will just refer to it as
exponential growth, which is the most commonly used term nowadays.)
Let’s start with an important instance of exponential growth, namely
compound interest. Imagine that you invest $1,000.00 in a government bond that
pays interest at 7% per annum, compounded annually. The bond matures in 10
years. The bond doesn’t pay an annual dividend, but rather just accumulates
interest over the entire 10-year period. (Each year, the past year’s interest is
added to the value of the bond. The current year’s interest is then calculated at
7% of the new bond value.)
What will the bond be worth at maturity?
One way to figure this out is to calculate the bond’s value year by year.

Year, t Value of bond at beginning of the year, Vt Interest for the year, It
0 $1,000.00
$70.00
1 1,070.00
 74.90
2 1,144.90
80.14
3 1,225.04
85.75
4 1,310.80
91.76
5 1,402.55
98.18
6 1,500.73
105.05
7 1,605.78
112.40
8 1,718.19
120.27
9 1,838.46
128.69
10 1,967.15

Here, the symbol Vt is read as ‘V sub t,’ or just ‘Vt;’ it is interpreted exactly as
stated in the table. Thus the value of the bond at maturity is $1,967.15, or an
increase of 96.7% over the initial $1,000 investment.
To explain how this table was calculated, note that the interest It in any year t
equals 7% of the bond’s value Vt at the beginning of the year:
This annual interest is added to the current value of the bond, to determine the
next year’s value Vt+1:

This calculation is repeated for each year, resulting in the above table of values.
(Computer spreadsheets are a quick way to do such calculations, but they can
also be done using a hand calculator.)
But do we have to go through the whole calculation, if we only want to know
the bond’s final value V10 after 10 years? The answer is no. To explain this, we
first combine Equations 4.10 and 4.11:

(be sure you see how I got the last line). Next we use this equation, Vt+1 = (1.07)Vt
year by year, by putting t = 0, then t = 1, etc.

and so on.
Try to write down the formula for the general case. It’s Vt = (what?). Can you
see that it is

Check that this is correct for V4.

Equation 4.13 is the equation of compound interest for the special case
where the annual rate of interest is 7% or .07. More generally, if the annual rate
of interest equals i, then the compound interest formula is
This formula gives the future value Vt at the beginning of year t, corresponding
to an initial investment V0 at the beginning of year 0, when the annual rate of
interest is i, provided that the annual interest payments are left to accumulate.
The derivation of Eq. 4.14 is exactly the same as for the special case (i = .07) in
Eq. 4.13.
Obtaining mathematical equations, as in this example, is the very essence of
applied mathematics. One starts with a real-world problem, such as the future
value of an investment. Then one adopts various simplifying assumptions, such
as a fixed, unvarying interest rate and no annual withdrawals. The consequences
of these assumptions are expressed mathematically, as in Eq. 4.14. The resulting
equation is said to be a mathematical model of the original problem.
Once the model equation is obtained, it can be used for calculations. In our
example, we can calculate the value of the bond at any future year. As I will
explain shortly, this model can also be used for other purposes.
Among the characteristics of a useful model are the following.
1. The model includes various ‘parameters’ (symbols that can represent
different possible numerical values). Our bond model, for example, includes
the parameters i (the annual interest rate), V0 (the initial investment), and t
(thetimeto maturity).
2. The model can be extended, or widened, to deal with more complicated
situations.
 Indeed, our bond-value model, Eq. 4.14, is the basis of the mathematics of
finance. It applies to loans, bank deposits, bonds, mortgages, annuities, and other
kinds of financial investments. It also explains why stock and bond markets
respond sharply to changes in the interest rate. I can’t discuss all these topics in
this book, but I will describe a couple of them. But first, let’s consider Eq. 4.14 in
a bit more detail.

Problem 4.9 Using Eq. 4.14 and your calculator, find the value of the $1,000.00
bond discussed earlier, at the end of 20 years, and 30 years. Compare these values
with the result of simple interest. (Simple interest is not compounded, and does
not accumulate to earn interest on itself.)

It is interesting to examine a graph of the value Vt of the bond, versus time t –

see figure.
This graph shows that the bond value increases more and more steeply as
time progresses, even though the proportional increase is always the same, 7%
each year. As the value of the bond increases, the amount of the annual increases
grows larger.
This feature of an ever increasing increase is the characteristic of exponential
growth. Exactly the same feature applies to population growth, for example a
laboratory population of bacteria or fruit flies. Until their food supply runs out
(or the experiment is terminated), such a population will undergo exponential
growth. Population growth is considered later.

The bond market

Next we consider our bond model, Eq. 4.14, from a different perspective. You
may know that financial markets (bonds, stocks, mutual funds) react to changes
in the interest rate as specified by a country’s central bank, such as the U.S.
Federal Reserve Board. The rule is that when interest rates go up bond prices go
down, and vice versa. This effect is direct, immediate, and highly predictable.
(Stock prices also respond to interest rates, but not necessarily so directly,
because of the speculative component of the stock market.)
To explain this phenomenon we use Eq. 4.14, in a slightly modified form.
Consider, for example, a bond that will mature in 10 years, at which time the
bond will be redeemed at par for, say $10,000. Thus V10 = $10,000, and Eq. 4.14
becomes

In this equation we consider i and V0 as parameters, or ‘variables.’ We wish to

know how the initial value V0 of the bond changes as the interest rate i is
changed.
We can solve Eq. 4.15 for V0, obtaining (see Section 4.5)

The value V0 is called the present value of the $10,000 ten-year bond. Thus V0
is what you would pay for the bond today (or what you could sell it for today),
given that the interest rate is i. The following table shows V0 for a range of values
of i.
 Notice that the present value of the bond, V0, decreases as the interest rate i is
increased. The decrease is not minor. Relatively small changes in the interest rate
cause quite substantial changes in bond prices.

Problem 4.10 Suppose the current Federal Bank interest rate is 6% per annum.
The Bank announces an increase of 50 basis points (100 basis points is 1%). By
what percent will bond prices decrease, for 10-year bonds?

(This simple model overlooks several important aspects of bond prices. For
example, investors may believe that interest rates may change in the future. If so,
the market price will reflect investors’ expectations of future interest rates.)
To conclude this brief introduction to the mathematics of finance let us just
record the general case of Eq. 4.16:

This equation, which follows by solving Eq. 4.14 for V0, expresses the present
value V0 of a future payment Vt which is due in t years time, given that the
annual interest rate is i. This equation is fundamental in Economics and
Business. It is sometimes called the discounting formula, the idea being that the
future value Vt is discounted to produce the present value V0.
Information provided by bond dealers usually specifies the yield for each
bond. The yield is the interest rate i such that Eq. 4.17 holds for the given price
V0 (and the given date t and value Vt at maturity). Any annual dividends are also
included in the calculation of yield. Thus the yield is just the rate of interest that
the bond purchaser will realize.

Problem 4.11 Fir trees are harvested 90 years after planting. An acre of fir trees is
expected to be worth $100,000 in 90 years time. How much would a logging
company evaluate an acre of newly planted fir trees, if the interest rate is 5% per
annum? 10%? (What additional assumptions did you make?)

Population growth
Exponential growth also occurs in biological populations. Suppose, for example,
that a scientist is growing a bacteria culture. She places an initial sample of P0 (P
for ‘Population’) bacteria in a Petri dish with plenty of bacteria food. Every 20
minutes each bacterium divides into two bacteria. How many bacteria will there
be after one hour? One day? One week?
Let Pn represent the number of bacteria after n generations. Then

This is another example of exponential growth.

In one hour (n = 3) there will be P3 = 8P0 bacteria. In one day (n = 24 × 3)
there will be Pn = 272P0 = (4.7 × 1021)P0. A single bacterium can produce an
astronomical number of descendents in a short time, given enough food. For
example, bacterial diseases can strike quickly.
After one week (n = 7×24×3 = 483 generations) our model predicts that P483
= 2 P0 = 2.5 ×10145P0. Something must be wrong with this, because cosmologists
483

have estimated the total number of elementary particles in the universe to be

around 1080 – much less than P483! What’s wrong with the model, of course, is
that exponential growth can’t continue forever unabated. Long before the week is
out, the Petri dish will be saturated with bacteria and population growth will
cease.

Problem 4.12 The human population on Earth in 2010 was about 7 billion. The
current rate of increase is about 1.5% per annum. Calculate the population 2000
years in the future, assuming this growth rate is maintained. Is the calculated
population possible?
 The general equation of exponential growth is

In this equation, or model, t is a time variable (t = 0, 1, 2,…), Xt represents the

magnitude of some quantity at time t, and a is the proportional growth rate per
time period. The duration of one time period is chosen appropriately for the
situation at hand. If a > 1, the quantity Xt increases over time, ever more steeply,
as shown graphically below.

The larger the growth parameter a, the more rapidly the quantity Xt grows, as
shown in the figure.

In some applications, Eq. 4.19 has a < 1. In this case, Xt decreases over time,
approaching zero – see figure. This process is called exponential decay. Physical
examples include the temperature of a hot object while cooling down, and the
decay of a radioactive sample.

Problem 4.13 [Optional] Radioactive carbon, carbon-14, has a half-life of 5,730

years. In other words, after 5,730 years a sample of C14 decays to one-half of its
original size. In symbols, X5730 = .5X0. By Eq. 4.19 we have a5730 = .5. Now we do a
little exponential algebra (see Section 4.8):

Your problem: how much of the original C14 sample would be left after 10,000
years? 20,000? 100,000?
 I hope this section has shown you that quite elementary mathematics can be
extremely useful. The models of exponential growth and decay are used
throughout science and technology. The exponential function (discussed in
Chapter 9), which is closely related to exponential growth, is also widely used.

4.3 Operations with fractions

Cancellation
Recall the law of cancellation from Chapter 2:

Remember, any factor that is common to the numerator (top part) and the
denominator (bottom part) of a fraction can be cancelled. In Eq. 4.20, the symbol
a is such a common factor. (Of course, a ≠ 0 is assumed.)
Two examples:

Think carefully about the second example. In fact, you can’t cancel anything here
– why not? You might imagine you could cancel 3, but no – 3 is not a factor of 3
+t. (Look up ‘factor’ in the index if this is confusing.) Improper cancellation is a
hallmark of weak math students, and is a leading cause of failing grades in
college math courses.
One more example. Simplify

One way to do this is to first multiply out, then cancel:

(Check this; what got canceled?) However, most people do the cancellation first,
then the multiplications:

This works fine – if you’re careful. But it’s easy to make mistakes this way, and
it’s hard to check, too. I recommend the first method – multiply, then cancel.

Problem 4.14 Simplify using cancellation, if possible:

Correct cancellation is an important aspect of algebraic calculations, so be

sure you understand how it works. If you like, make up and solve additional
exercises yourself.

Addition and subtraction

Recall the basic method of adding fractions having the same denominators (see
Chapter 2). For example

You can also combine fractions having unequal denominators, but first you must
rewrite each fraction so that they do have common denominators. For example

Read Section 2.4 if you have forgotten how to do this.

The same method applies to adding fractions that are algebraic, rather than
numeric. For example
 Here, the two fractions have the same denominator, x2, so they can be added
directly (or subtracted directly).
The general format for the case of equal denominators is

where A, B, and C represent arbitrary algebraic expressions. This is exactly the

same rule used for numerical addition; see Eq. 2.11. The reason that the rule is
the same for numerical and algebraic cases is that algebraic expressions always
represent numbers.
Next, how do we combine (i.e., add or subtract) algebraic fractions if the
denominators are not the same? There is one method that always works,
although it may produce an unnecessarily complicated result. This method is

This is the same as Eq. 2.12.

Please pause to make sure you understand both steps in Eq. 4.22. Why is
And why did we make this change? Finally, why is the second equality
in Eq. 4.22 valid? You should re-read Sec. 2.4 if you can’t answer these questions.
Here’s an example:

Notice that there is a quick way to check the answer in Eq. 4.23. Namely, just
split the result into two parts, and use cancellation:
which equals the original expression. This check is always worth doing (I
invariably do it in my own research), especially since you can do it quickly in
your head. Try it again for yourself in Eq. 4.23. Also, observe that the check
basically just reverses the procedure used to obtain the result in the first place.
Carefully checking algebraic calculations is always worthwhile. Anyone can make
a slip while doing algebra, and immediate checking usually catches any mistake.
Train yourself to do this regularly. (I will give you some helpful hints on good
techniques for checking algebra, as we proceed.)
It is customary to use alphabetic order (when convenient), in expressions
such as 3x2y. However, it would be correct to write 3yx2 instead. What about
writing y3x2? While still correct, this would be quite unorthodox. Most readers
would suspect a misprint.

Problem 4.15 Combine the following fractions, and check your results. (a)

The method of adding fractions, given by Eq. 4.22, sometimes produces an

overly complicated result. This was explained for numerical fractions in Chapter
2. For example, using Eq. 4.22 gives

This answer is correct, but it can be reduced to Similarly, the method of Eq.
4.22 can be used in the example

This answer can be simplified:

This simple answer could be obtained directly by using the method of lowest (or
least) common denominator, or l.c.d. Namely, we find the simplest expression
that contains both of the original denominators, x3 and x2, as factors; this
expression is the l.c.d. In this example, it is x3 (since both x3 and x2 are factors of
x3). Using the l.c.d., we have

This calculation is simpler and more direct than the previous one.

Problem 4.16 Find the l.c.d., and then combine:

A method for finding the l.c.d. in complicated cases in discussed in Part 2.

Multiplication and division

The two basic formulas for multiplying and dividing fractions are:

These formulas were discussed, for numbers, in Chapter 2. (Remember that Eq.
4.25 is often thought of as: ‘to divide by a fraction, invert it and multiply.’) These
formulas are also correct if A, B are any algebraic expressions. Thus, for example

Problem 4.17 (This will test your skill!) Simplify

 The basic algebraic calculations discussed in this section are important in
later work, especially in applications of math. It is therefore vital that you master
the skill to do such calculations quickly and correctly. Checking every calculation
is always desirable, whether by repeating it, or by ‘undoing it’ (e.g., when adding
fractions). The next problem will give you a little more practice, but if you still
feel at all shaky, I recommend borrowing a school book on algebra and doing
drill exercises (boring as that may be!).

Problem 4.18 Simplify by reducing to a single fraction. Check all calculations.

4.4 Polynomial algebra

Dear reader: the remaining sections of Chapter 4 are at a somewhat more
advanced level than previous sections. I have put all this Algebra in a single
chapter for convenience, but in school you probably studied Geometry before
these Algebra topics. Feel free, if you wish, to study Chapter 5 before returning to
complete Chapter 4. Ultimately, however, everything in this chapter is needed
later.
The example

will indicate what we mean by a polynomial. In this example, we have a

polynomial in the variable x, but we can have polynomials using any symbol.
Thus b2 − 2b + 1 is a polynomial in b.
A polynomial in x, then, is an expression consisting of a sum of terms, each
being of the form

where a is some real number, called the coefficient of xn, and where n is a non-
negative integer (n = 0, 1, 2, etc.), called the degree of the term (or the power of x
in that term). Thus the above polynomial has 4 terms, 2x3, −7x2, x and −12. The
term −12, which doesn’t involve x, is called the constant term in this polynomial.
(Remember that x0 = 1, so that −12 could be written as −12x0. We never really do
this, but it shows us that constant terms are bona fide terms of the form axn, with
n = 0.)
The degree of a polynomial is the highest power of x (or whatever the
variable is) in the polynomial. The above example 2x3−7x2+x−12 therefore has
degree 3. Polynomials of degree 3 are usually called cubic polynomials – because
x3 is called ‘x cubed.’ However a second degree polynomial is never called a
‘square’ polynomial, but instead a quadratic polynomial. For example, 6x2 +4x +
9 is a quadratic polynomial.
A polynomial of degree 1, such as 3x − 2, is called a linear polynomial. The
reason for this term is that the graph of a linear polynomial is a straight line; see
Chapter 6.
To familiarize yourself with these terms, try the next two problems.

Problem 4.19 One of the following three expressions is a polynomial, the others
not. Explain.

Problem 4.20 Consider the polynomial (a) What is its degree? (b)
List the terms, and the coefficient of each term. What is the constant term?

It is customary to write polynomials either in decreasing order of exponents,

or increasing order. Thus would usually be written as
(or the reverse order).
Before proceeding, let us be more specific about the meaning of a polynomial
expression. For example, what is the value of the polynomial

when x = 4? Try this – you should obtain 49. If so, you performed the
calculations like this:
But did you notice that there is a tacit precedence rule involved here? The rule is

Thus 3x2 + 1 actually means (3(x2)) + 1. First we find x2, then multiply by 3,
then add 1.
Without the precedence rule, such an expression would be ambiguous. For
example, 3x2 + 1 might be incorrectly interpreted as (3x)2 + 1. Though few people
actually make this mistake, it is important to realize that there is an accepted
precedence rule behind the correct interpretation.

Addition of polynomials
Again, an example will indicate how two polynomials (in the same variable) can
be added:

This procedure is called ‘collecting like terms’ – here we first collect the x3-
terms, then the x2-terms, and so on.

Problem 4.21 Simplify: (2x2 − 1) + (x2 + x +3) − (x − 5).

How can you check your work in such a calculation? The method I use is just
to re-do the problem, but in my head. Consider the above example,

By inspection you can first pick out the highest order terms, x3 +2x3 = 3x3. Then
go to the next highest order terms, −2x2 +3x2 = x2. Next, 4x and finally 5. The
result, 3x3 + x2 +4x + 5 agrees with the original calculation.
With a little practice, you can always add polynomials using the mental
method. In fact, you may be less likely to make mistakes (for example, copying
errors) with the mental approach. But you should always check, by re-doing the
calculation. Try it on:

The answer is 4x4+2. [Remember that the factor 3 for the second polynomial
applies to each term: 3(x4 +2x2 − 1) = 3x4 +6x2 − 3.]

Multiplication of polynomials
Multiplication of polynomials can be done in several ways, including mentally.
There is lots of room for error, so one has to be careful. Unless you consider
yourself really adept at algebra, I suggest you read this section slowly and
carefully!
Here is an example, which we first do longhand:

Be sure you understand each line of the calculation; you might want to re-do
the problem yourself on a new sheet of paper, looking at the above lines to check
your attempt.
On the first line, for example, we use the Distributive Law (Sec. 1.3), which
implies that

for any expression A. Here A is (2x2 − x − 5). Then, on the second line, we use
the Distributive Law again: x2(2x2 − x − 5) = 2x4 − x3 − 5x2 and so on. (We also
use the law of exponents, xmxn = xm+n here.) Finally, we collect like terms.
Problem 4.22 Multiply out longhand. (a) (2x3 − 5x2 + x +1)(x − 2); (b) (x2 − x −
3)(4x2 + 1). If your answers don’t agree with the solution, re-do the calculations
more carefully.

What about mental checking of multiplication? I want to show you my own

method, which I actually use for doing the calculation in the first place. This
method is certainly worth learning if you plan to take advanced math courses.
First consider the general pattern in (A + B)(C + D) = A(C + D)+ B(C + D)=
AC + AD + BC + BD. Let’s display this:

The answer has 4 terms. Each is the product of one term from the first
bracket and one from the second, as indicated. You can just read off the answer
in this way, without any intermediate step. The same works no matter how many
terms there are. For example

Note here that (3 terms)× (2 terms) gives (6 terms).

Problem 4.23 Read off the answers directly: (a) (U + V )(X + Y + Z); (b) (A + B +
C)(D + E + F ).

The rule is, when multiplying two sums together, choose one term from each
sum and multiply, then add up every possible such product. If there are m terms
in the first sum, and n terms in the second, your final result will consist of mn
terms.
When multiplying two polynomials, there is the additional complication of
collecting like terms. If you do the mental multiplication in a certain order, you
can collect terms ‘on the fly,’ so to speak. Consider this example:

First, decide what powers of x can occur in the answer. Clearly x4 and lower
powers only. To get x4 you have to choose the x2-term from each bracket, getting
2x4:

Next, to get x3-terms, you have to use an x2 and x term, and you can do this in
two ways:

Can you see how many ways there are to get x2-terms? Three! Look for them,
getting

Pause to be sure you follow the last step. (Of course, in actuality you wouldn’t do
this in the above discursive fashion, which is only for explanation. You do it all at
once. But it’s a good idea not to combine the coefficients in the first go through,
however, since this can easily cause an error.)
Completing the calculation, we get

There are two checks you can make. First, there should be 3 × 3 = 9 terms,
before collecting coefficients. Count them: 9, right! At least we haven’t forgotten
a term. Next, check that both sides agree when x = 1. Put x = 1 in the original
problem. You get 2 × 2 = 4. Now put x = 1 in the answer. You get 2 + 7 − 9+7 − 3
= 4. Wow! If you don’t get the same number, your answer is wrong for sure. If
you do get the same number, your answer is very likely right (though not
absolutely certain).

Problem 4.24 Multiply, using both checks (a) (2x +1)(x2 − 3x − 2);(b) (3x3 − 5x −
3)(x4 +2x2 + x − 1).

The square of a binomial

The following equation is encountered so often that you should memorize the
pattern (but be sure also that you can derive it):

To obtain this, we just do the multiplication:

(a + b)2 = (a + b)(a + b)= a2 +2ab + b2.

Problem 4.25 Find the squares. (a) (x − 3)2;(b) (2w + z)2;(c) (x2 +3)2.

Division of polynomials
Division of polynomials closely resembles long division of integers. For example,
long division gives

This can be checked by adding the numbers on the right side:

The general form is

where N stands for Numerator, D for Denominator, Q for Quotient, and R for
Remainder. In Eq. 4.27, the numerator is sometimes called the dividend, and the
denominator D is the divisor. We must have

Check these symbols, and condition (4.28), against the above example. Question:
what is the remainder R in the case that the denominator D divides exactly into
the numerator N ? (Answer: zero.)

A parallel situation holds for division of polynomials:

Let me pause to explain the meaning of expressions like N (x), and so on.
First, N (x) isread as “N of x.” It represents some polynomial in x. In any
particular case, N (x) would be specified, for example N (x)= 2x2 +5. This still
contains a variable, x, which itself represents some real number. The notation N
(3) then denotes the value of N (x)for x = 3:

Notice that the expression N (x) involves a double abstraction – x is abstract,

and N (x) is more abstract. But x and N (x) both represent real (i.e., decimal-
point) numbers. We can expect to encounter such symbols added, subtracted,
multiplied or divided – as in Eq. 4.29 for example.
As a quick check of comprehension, suppose we specify another polynomial
M(y)= 3y − 9. (There’s no rule limiting what letters we can employ!) What is
M(5)? Answer: 6.
In Eq. 4.29, N (x) and D(x) represent given polynomials, and we wish to
calculate other polynomials Q(x) and R(x), called the quotient and remainder
respectively, so that Eq. 4.29 is true. In analogy with Eq. 4.28 we also require that
That is, the remainder must have smaller degree than the denominator.
Here is an example:

Here we have Q(x) = 2x − 8 and R(x) = 31. (How to obtain this will be explained
shortly.) As before, this can be checked by combining fractions on the right side:

Before I explain my method for dividing polynomials, note that our basic
defining formula, Eq. 4.29, can be rewritten (by multiplying both sides of the
equation by D(x)) as

The problem is, given the dividend N (x) and the divisor D(x), calculate the
quotient Q(x) and the remainder R(x).
The algorithm I recommend for dividing polynomials is called MWTFU –
the Method of Wishful Thinking and Fixing it Up. Here is the example, (2x2 − 1)
÷ (x + 4). First, make sure that both polynomials are written in order of
decreasing powers of x, as they are in this example. Now, start with the first term
in the numerator, 2x2. I ‘wish’ this was 2x(x +4) [so that I could divide it by x +4].
I write

where −8x ‘ fixes it up’ (check that this is correct). Therefore

 Next I repeat this step with the next term, −8x:

Thus

Finished! We now have

Thus the quotient is (2x − 8) and the remainder is 31. Note that condition (4.30)
is met – the degree of the remainder (zero) is less than the degree of the
denominator (one). To check that the result is correct, we just multiply out on
the right side, which gives (2x − 8)(x +4)+31 = 2x2 − 1, as required.
Let’s try a slightly more complicated one, (x3 − 3x2 +4) ÷ (x2 − 9):

This is the required answer, using the form of Eq. 4.31. What is Q(x)? R(x)? And
how do we know to stop at the last line? Well, we stop at the last line because it’s
impossible to continue; we can’t hope to write 9x = (?)(x2 − 9) because degree
(9x) < 2. This is exactly the condition (4.30) for the remainder R(x). All this
should become clear when you try the next problem.
Again, we can check the answer by multiplying it out:

which is the original numerator.

Problem 4.26 Divide and check: (a) 2x4−3x2+5 by x2−2; (b) x3−x2+x−1 by x − 1.

As you have doubtlessly noticed, ‘MWTFU’ is just another algorithm.

Starting with the first (highest order) term of the numerator N (x), we use
Wishful Thinking plus Fixing it Up on this term. This calculation changes some
of the rest of N (x). Then we do the same thing to the next term, and so on. We
stop when the degree of the next term is smaller than the degree of D(x). Try
another example of your own to confirm that this description is correct.
If you’d like more practice you can make up your own problems, and check
the answers. I suggest you do this now.
An alternative way of dividing polynomials is given in Sec. 4.8.

Extracting common factors

How could you simplify the polynomial

Note that each term has a factor of 2. Therefore

The polynomial x2 − 5x + 4 may be easier to work with than the original

polynomial. In this example, 2 is said to be a common factor of the polynomial
2x2 − 10x +8.
Another example: 24x3−18x = ? Try to simplify this as much as possible by
factoring. The answer is (do it yourself before peeking) 6x(4x2 − 3).

Problem 4.27 Simplify, by canceling common factors:

(Be sure you understand why no further cancellations are possible in either case.)

Factors of a polynomial
Suppose you divide a polynomial N (x) by another, D(x), as in the above
examples. What must happen for D(x) to ‘divide evenly’ into N (x)? The answer
is that the remainder R(x) must be zero. This is exactly the same as for whole
numbers. For example, 5 divides evenly into 15 because the remainder is zero: 15
= 3 × 5. Similarly, 5 doesn’t divide evenly into 17 because 17 = 3 × 5 + 2, with a
remainder 2.
Here’s an example: does x + 2 divide evenly into x2 + x − 2? Well, we have

Yes, x + 2 divides evenly into x2 + x − 2. We say that x + 2 is factor of x2 + x − 2.

In general, given a polynomial A(x), if we can write

for polynomials B(x) and C(x), these polynomials are called factors of A(x). They
are called proper factors of A(x) if their degrees are smaller than the degree of
A(x).
For the above example,

we see that (x +2) and (x − 1) are proper factors of x2 + x − 2.

Problem 4.28 (a) Show that 2x + 1 is a (proper) factor of 2x2 +7x +3; (b) For
what value of c is x − 5 a factor of x2 − 2x − c?

Factoring a quadratic
Consider the multiplication (x − 4)(x +1) = ? The answer

will have three terms – an x2-term, an x-term, and a constant term. The x2-term
comes from multiplying the x-term from each factor; here you get x2. The x-term
comes from combining two products, which I picture as the ‘outer’ product x ×
1, and the ‘inner’ product −4 × x. These produce −3x for the x-term. Finally, the
constant term is −4 × 1= −4:

Teach yourself to perform such calculations in your head, by combining terms in

this way.

Problem 4.29 Multiply out, using the suggested method. (a) (x+2)(x+4); (b) (x −
3)(x − 4); (c) (x +1)(x − 6); (d) (2x − 3)(4x +1).

Next, can we reverse this procedure? Given the answer, e.g. x2 +6x +8, can we
find the factors? In other words, can we factor a given polynomial?
In general, factoring polynomials is a difficult problem. There is a general
algorithm (only recently discovered) that can be programmed into the computer,
and is now available in software packages for use by scientists and engineers.
Here we only look at quadratic polynomials – it can be useful to be able to factor
these polynomials by inspection. This topic will come up again in later sections
of this chapter.
Consider the example

If this can be factored at all, it must be

because the term x2 can only come from x × x. The constants (?) must give 6
when multiplied. What possibilities are there? 6 and 1, or 3 and 2 – any others?
Yes, −6 and −1 or −3 and −2. We use trial-and-error. Do any of these choices
give 5x for the x-term?
Well
which is not right. However,

We’ve done it! (This is why you learned to multiply out in your head.)
The method is this: To factor x2 + Ax + B, when A and B are integers, first
find the possible factors of B. Try each possibility. Example:

What factors of 10 add up to −7? Answer, −5 and −2. Therefore

which you check by multiplying out. Be sure you follow this example, then do
the next problem.

Problem 4.30 Factor: (a) x2 +7x + 12; (b) y2 − 7y + 12; (c) x2 − 8x + 12; (d) t2 −
11t − 12.

By now you get the idea, I hope. The overall pattern goes like this:

Looking at this in reverse, if given the quadratic x2 + px + q with p and q being

integers, we try to find integers a, b such that ab = q and a + b = p. For example,

What integers a, b have ab = 30 and a + b = 11? Aha? 5 and 6:

and this checks.

Now try this one: x2 − 8x − 30 = ? Got it? I hope not! Nothing works. You
probably tried −30 = − 6 × 5 = −10 × 3 = −30 × 1, but these don’t give the right
coefficient −8 for x. Having exhausted the possibilities, we can conclude that the
given quadratic can’t be factored, using integers.
But is there some automatic way (other than trial and error) to factor a
quadratic? Yes – we’re coming to that. It’s called the quadratic equation.

An important special case

Try factoring x2 − 4. You should get (x +2)(x − 2). Do you see why this works?
(Multiply out.) The two x-terms, 2x and −2x add up to give 0.
What is the general case? It is

This is true for any real number a and any value of x. This extremely useful
little formula is worth memorizing (but you’ll always remember why it’s true –
the x-terms ax and −ax add to zero). Also, compare this with Eq. 4.1.

Problem 4.31 Factor (a) x2 − 16; (b) x2 − 2; (c) x2 + 16 (Careful!).

Problem 4.32 (Review).

(a) Define these words: polynomial; term of a polynomial; coefficient of a term;
degree of a polynomial.
(b) What is meant by proper factor of a polynomial?
(c) Add: (2x3 − x +5)+3(x3 − 3x2 + x − 1).
(d) Multiply: (x2 +3x − 2)(2x3 + x2 + 1). Check by substituting x = 1.
(e) Divide x3 +3x2 +7 by x2 − 2. Check.
(f) Factor x2 − 16. Also factor x2 +6x − 16.

Problem 4.33 (a) Add First, factor the denominators,

then use the least common denominator; (b) Simplify, by first factoring:
4.5 Linear and quadratic equations
Many problems in advanced mathematics require the solution of a polynomial
equation. To solve an equation, for example

means to determine all values of x for which the equation is true. Thus x = 2
happens to be a solution of the above equation, because

(There could be other solutions to this equation; more on this later.)

Problem 4.34 Show that x = 1 is not a solution to the above equation.

The solutions of a given polynomial equation are sometimes referred to as

the roots of the polynomial.

Linear equations
A polynomial equation of degree 1, for example

is called a linear equation (because the graph of a linear polynomial is a straight

line – see Chapter 6). To solve the above example, we go through the following
steps:
Check this solution: Correct.

You need to learn to solve linear equations quickly and accurately in your
head. Just go through the above steps, mentally.

Problem 4.35 Solve for x. (Try to solve mentally.) (a) 2x − 6 = 0; (b) −5x +2 = 0;
(c) 3x − 1= x +5.
The step of adding 5 to both sides, in the previous example, is sometimes
called transposing. For example, 4x − 9 = 0 becomes 4x = 9 upon transposing −9
to the right side. You change signs when transposing (because you are really
adding 9 to both sides of the equation).
Look again at Problem 4.35, and think about transposing – this should make
the mental calculation easier. In part (c) you do two transpositions: x from right
to left, and −1 from left to right. Make up some more examples for practice.
To summarize, a linear equation Ax + B = 0 (where A and B are given
numbers) always has the unique solution x = −B/A, provided that But,
better than memorizing this as a formula, just remember the transposition
method:

Linear inequalities
Many students who have no difficulty solving linear equations (Ax + B = 0) seem
to find linear inequalities (for example Ax + B < 0) confusing. Fortunately, the
same solution technique works for both cases, with one additional twist for
inequalities. Here’s a first example:

Therefore x satisfies the original inequality 2x − 3 < 0 if and only if

 The operations used in this example are justified by the Laws of Inequalities
listed in Sec. 2.3. For example, the law

implies that we can add any number (positive or negative) to both sides of a
given inequality. The new inequality will be true if and only if the original
inequality is true. This is just the same as for equations. Also, we can think of
‘transposing,’ exactly as in the case of equations. Take your pick – add 3 to both
sides, or transpose −3 to the right side, changing sign.
In the same way, the law

implies that we can multiply both sides of an inequality by any positive number.
This is where solving inequalities differs from solving equations. It is extremely
important to understand and use this operation correctly.

Problem 4.36 What does happen if you multiply both sides of the inequality a <
b by a negative number c? (Try an example if that helps.)

Consider the inequality 4 − 3x < 0. To solve for x, we have

The solution could also be written as x> 4/3.

Problem 4.37 Solve the inequalities (a) 3x − 8 < x − 2; (b) 7 ≤ 2x − 5. (The laws
for ≤ are the same as for <.) (c) 4 − 3x> −1.

In summary, to solve a linear inequality, or a linear equation, (in the

unknown x, for example), first use transposition to combine and isolate the x-
terms on one side of the inequality (or equation), and the constant terms on the
other. For simplicity, arrange it so that the coefficient of x is positive. Then
divide through by that coefficient. Example:
 Thus the solution is x ≥−2. Re-do Problem 4.37 using this method, if you had
any trouble with that problem.

Quadratic equations
Consider the quadratic equation, in general form

Here the coefficients a, b, and c are given numbers, and x is the ‘unknown.’ To
solve this equation means to find all values of x for which the equation is true.
If a = 0 in Eq. 4.33, the equation becomes bx + c = 0, which is a linear
equation. We just learned how to solve linear equations, so we can henceforth
assume that a ≠ 0. It then turns out that the quadratic equation (4.33) has two
solutions x, which can be calculated from the formula

where the symbol ± means that you use + for one solution and − for the other. If
you plan to study college-level math, you will need to memorize this formula,
which is called the quadratic formula. Later I will explain in detail how the
quadratic formula is derived. But first, let us become more familiar with the topic
of quadratic equations.
Here is an example:

The values of the coefficients are a = 1, b = −7, c = 6. Thus b2 − 4ac = 49 − 24 =

25, and (Remember that always signifies the positive square
root.) Using the quadratic formula, Eq. 4.34, we get that the solutions are

We can check that these are correct: 62 − (7 × 6) + 6 = 36 − 42 + 6 = 0, and 12

− (7 × 1) + 6 = 1 − 7+6 = 0.
Perhaps you noticed that the polynomial x2 − 7x + 6 could be factored:

Using this factorization, we could have solved the polynomial equation x2 − 7x +

6 directly, without using the quadratic formula:

Here we are using a very basic and useful rule (see Section 2.4):

Here A and B denote any real numbers (and therefore A and B could be any
algebraic expressions, because such expressions always represent numbers). This
rule, if you think about it, is hardly surprising. If you multiply two non-zero real
numbers A and B together, you either get a positive number AB (if A and B have
the same sign), or a negative number (if they have opposite signs). To get AB = 0
you must have at least one of the numbers A or B equal to zero.
The zero-product rule is used regularly in the solution of equations. For
example, how would you solve 4(x − 1) = 0? There’s a hard way and an easy way.
The easy way is to use the zero-product rule, which tells us that x − 1 = 0
(because, for sure 4 ≠ 0). Thus x = 1 is the solution. (What is the hard way? 4(x −
1) = 4x − 4 = 0; therefore 4x = 4; therefore x = 1.)
 In attempting to solve a given polynomial equation P (x) = 0, suppose we are
first able to factor P (x) as

i.e., (x − a) is a factor of P (x). Then P (x) = 0 implies that either x = a or Q(x) =

0. In other words, one solution of P (x)= 0 is x = a. Any other solutions of P (x) =
0 are therefore solutions of Q(x) = 0. (We will discuss the relationship between
roots and factors of polynomials in greater detail in Sec. 4.8.)
For the case of a quadratic polynomial P (x) = ax2 + bx + c, if this quadratic
can be factored, then the roots of the quadratic can be obtained by inspection.
However, most quadratics cannot be factored by inspection. But the roots can be
found by using the quadratic formula. These roots can then be used to factor the
polynomial.
For example, consider the equation

No obvious factoring works here. Using the quadratic formula, we obtain the
roots

These roots can be used to factor the polynomial:

If you wish, you can do the algebra to check that this is correct.
This connection between roots and factors is always true: if x1 and x2 are the
roots of the quadratic equation ax2 + bx + c = 0 then (x − x1) and (x − x2) are
factors of this polynomial, so that

Problem 4.38 (a) Solve the equation x2 − 2x − 8 = 0, either by factoring or by

using the quadratic formula; (b) Solve the equation x2 − 2x − 7= 0.
Problem 4.39 Solve by using the quadratic formula (but don’t factor): (a) 3x2 −
5x − 1= 0; (b) 2x2 +2x − 2= 0.

Look again at the quadratic formula, Eq. 4.34. This formula contains a
square-root expansion What happens if b2 − 4ac < 0?
For example, consider the example x2 − 2x + 2 = 0. Using the quadratic
formula, we obtain the solution

But what is By definition of the square root, this is a number whose

square is −4. There is no such real number! (Remember, the phrase ‘real number’
refers to numbers that can be expressed in terms of decimals, possibly with
infinite decimal expressions; see Chapter 2. The square of any real number is
always ≥ 0.)
Is the quadratic formula wrong, then? No, it’s correct; the given equation has
no real solutions. In general the quadratic formula indicates that

A simple example that will remind you of this possibility is the equation

This equation obviously has no real solutions, because x2 ≥ 0 for any real number
x. Therefore x2 +1 ≥ 1, so that x2 + 1 = 0 is not possible for any x. What is b2 −
4ac for this example? Answer −4, so the situation described in Eq. 4.35 prevails.
In Volume 2 we will show that, by allowing for ‘complex numbers,’ we do
obtain solutions to the quadratic equation in the case that b2 − 4ac < 0. This
procedure of extending the real number system to the complex number system is
in the same spirit as extending the natural number system to the system of
integers, which allows one to subtract numbers with no restriction.
Problem 4.40 (a) For what values of c does the equation x2 +3x + c = 0 have real
solutions? (b) Show that if a and c are of opposite sign, then ax2 + bx + c does
have real solutions.

A final note: what if b2 − 4ac = 0? In this situation, the expression

i.e. 0. Thus the quadratic equation only gives us one
solution x = −b/2a. This is correct –this is the only solution for this case.
Example:

Here b2 − 4ac = 16 − 16 = 0, and the quadratic formula gives as

the single solution. By factoring,

and this shows why the equation x2 − 4x + 4 = 0 only holds if x = 2.

In summary, the three cases are: Given the quadratic equation ax2 + bx + c =
0, if:

The expression b2 − 4ac is sometimes called the discriminant for the

quadratic ax2 + bx + c.

Problem 4.41 Calculate the discriminant and determine the number of real
solutions, but don’t solve: (a) x2 − 9x − 1; (b) 1.1x2 +0.8x +2.7; (c) x2 − 10x + 25.

Roots and factors

Equation 4.36 specifies the conditions under which the quadratic polynomial ax2
+bx+c has 2, 1, or 0 real roots. This also determines how the quadratic
polynomial can be factored:

We have discussed examples of each of these cases, but you may wish to
check the further examples

Determine the roots of these equations, and also factor the given
polynomials, where possible. Note that the equations have 2, 1, and 0 roots,
respectively.
In the second case, where the factored form of the polynomial is a(x − x1)2,
we sometimes say that x1 is a double root of the given equation. Whenever a
quadratic equation has only one root, this is a double root. We will discuss this
situation in more detail later.

Completing the square

Here is a method for solving quadratic equations without using the quadratic
formula. Consider the example

Note that the first two terms are the same as the first two terms in (x+2)2 = x2
+4x+4. So let’s ‘fix up’ the original quadratic to look more like (x+2)2:
Our given equation now becomes

or

Therefore

or

(Please check that you would get the same result by using the quadratic formula.)
The above method is called ‘completing the square.’ One starts with

(for simplicity we temporarily assume that a = 1). Then one recalls that

Comparing this with x2 + bx + c, one sees that

Therefore

and this ‘completes the square’ and allows us to solve the quadratic equation x2 +
bx + c.
Try another numerical example.

Problem 4.42 Solve by completing the square: x2 − 2x − 9= 0.

We can now derive the quadratic formula, by transforming the general

quadratic equation, in several steps as follows. (Compare this calculation with
the foregoing example.)

You should read over this derivation carefully, and make sure that you
understand each step. Then close the book and try to write out the derivation
yourself. Could you do it stranded on a desert island? It is admittedly a little
complicated, but any mathematics student should be able to do it. As a review,
solve the equation 2x2 −x−5 = 0 (a) by completing the square, and (b) by using
the quadratic formula. (By the way, you may wonder why I wrote in
this calculation. If a < 0 we have but the ± sign allows for this
possibility.)
Equations that can be written as quadratic equations
Consider the equation

To solve this equation, we first multiply through by x − 1:

The latter equation can be solved by the quadratic formula, giving

A similar example: 1/(x +2) − 4/x − 1 = 0 This can be rewritten as x2 +5x + 8

= 0. (Check the algebra!). The latter equation has no real solutions, so that the
original equation also has none.
Next, consider the fourth-degree equation

We have not discussed such equations in general, but perhaps you can see how to
solve this particular example. Think about it for a minute.
Did you realize that the polynomial can be factored?

Next question: how do we solve this equation? Recall the zero product rule. We
must have

The solutions are therefore

More generally, consider the equation
This can be reduced to a quadratic equation, by writing y = x2. This gives

We can find the solutions y of this equation, either by factoring, or from the
quadratic formula. Then gives the solutions to the original equation.
Some, or all of these solutions may not be real numbers, however. For example,
what are the solutions of

Answer: there are two real solutions x = ±2.

Problem 4.43 Solve (a) 2/x − 1/(x +5) = 1; (b) (x +3)/(x +2) = x/2; (c) x4 − x2 − 12
= 0.

Extraneous solutions
In solving a given equation, we typically carry out various transformations of the
equation. We may re-arrange and collect terms, multiply both sides by a certain
constant, square both sides, etc. The logic behind the method is that any solution
of the original equation will also be a solution of the transformed equation. By
solving the transformed equation, we therefore solve the original equation. Here
is an example:

Thus x = 2 is the solution, and the only solution, of the given equation. The
final equation x = 2 is equivalent to the original equation 3x − 4= x, in the sense
that x satisfies one equation if and only if it satisfies the other. We know that this
is true because it is true for each step. For example, if Eq. (1) holds for x, then (2)
holds for the same x. Conversely, if Eq. (2) holds for x, then Eq. (1) holds also.
Similarly, Eq. (2) holds for x if and only if Eq. (3) holds for the same x, and so on.
The transformations used in going from one step to the next are reversible.
 For some kinds of transformations this reversibility may not hold. For
example, consider the following steps:

If we now try substituting each of these values of x back into the original
equation, we find that x = 1/4 is a solution, but x = 1 is not a solution, of the
given equation. In other words, the original equation and the transformed
equation (for example, (4x − 1)(x − 1) = 0) are not equivalent. What happened?
See if you can discover which of the transformations is not reversible.
The answer is that the transformation from Eq. (2) to (3) is not reversible. It
is true that if Eq. (2) holds for x, then Eq. (3) also holds for the same x, but not
conversely. Indeed, if Eq. (3) holds then we will have (by taking square roots)

This is not the same as Eq. (2). Thus the transformation from Eq. (2) to (3) is not
reversible.
Whenever one performs an irreversible transformation, it is possible to
obtain ‘extraneous’ solutions to the original equation, that is to say, numbers that
are not solutions to the original equation. Therefore, it is necessary to check each
final solution by substitution back into the original equation. In the above
example, the extraneous solution x = 1 was introduced in going from Eq. (2) to
(3), as you should check.
As an example, solve by squaring both sides,
simplifying, and squaring again. What are the solutions? Are any of them
extraneous?
The answer is that there is just one solution x = 5/4 to the final equation, and
it is not extraneous. In this example, no extraneous solutions arise.
Another example of a problematic transformation is multiplying or dividing
both sides of an equation by some expression involving the unknown, say x. For
example, in solving x2 = 2x, students sometimes divide both sides by x, obtaining
x = 2. They then state that the solution is x = 2. Can you see what’s wrong with
this conclusion? The answer is that x = 0 is also a solution of the given equation.
Dividing by x, which might be zero, gets rid of this solution. In this case, the
transformation eliminates one solution, rather than introducing an extraneous
one.
Transformations such as squaring both sides, or multiplying both sides of an
equation by an expression, are useful in solving equations, but care must be
taken regarding extraneous or eliminated solutions.
As an example solve

and check that the solution is correct. (Answer: x = 1 or 4; both check out as
valid solutions. The value x = 2 is not a solution, so multiplying by (x − 2) does
not affect the solution set.)

4.6 Inequalities
The statement a < b means that a lies to the left of b, on the real number axis
(assuming that this axis points to the right):

Given the decimal representations of a and b, we can determine which is

smaller by inspection. If a, b are both positive, we use direct comparison, as in
2.707 < 2.712. If a, b are of opposite sign (or one is zero), we use

as in −2.5 < 1.5, or −3.4 < 0. Finally, if a, b are both negative, then the order is
reverse to the magnitude, as in −3.2 < −3.1.
The laws of inequality
The laws of inequality follow from the above real-number-axis characterization
of inequality.

Problem 4.44 Show that if a, b are positive numbers with a < b, then 1/b < 1/a.

In addition to the symbol < we have the symbols

Of course, a > b means the same as b < a. We could thus dispense with >, but
it is often a convenience. Next, a ≤ b means just what it says, namely either a < b
or a = b. Inequalities of this kind are useful in various contexts later in this book.

Problem 4.45 Which of the laws of inequality remain valid if < is replaced by ≤
throughout?

Linear inequalities were discussed in the previous Section. We now consider

other examples. First, see if you can solve the inequality

It is tempting to first multiply through by x + 1, obtaining 1 < 5(x +1), which

gives x > −4/5. However, this is not correct – do you see why? Answer: the
calculation is wrong in the case that x + 1 is negative.
A good strategy for problems like this is to first simplify algebraically,
without multiplying through:

Now, a fraction is negative if and only if the numerator and the denominator
have opposite signs. There are two possibilities:
Case 1: −4 − 5x < 0 and x +1 > 0.
Thus x> −4/5 and x> −1. Hence x> −4/5. (Why? See below.)
Case 2: −4 − 5x> 0 and x +1 < 0.
Thus x < −4/5 and x < −1. Hence x < −1.
(In Case 1, note that to say x> −4/5and x> −1 is exactly the same as saying
just that x> −4/5. Be sure you understand this. Also, check case 2 again now.)
The conclusion from these two cases is that the given inequality holds if

This is the solution. (In particular, no number between −1and −4/5satisfies the
inequality. Try an example, x = −9/10. This gives 1/(x + 1) = 10, which is not <
5.)
Most students find such examples a bit confusing, because of the logic. Read
the solution again before trying the next problem.

Problem 4.46 Solve the inequalities (a) −2/(x +1) < 5; (b) 3/(x − 1) < −4.

Absolute values
A particularly useful type of inequality is
Here a and b are given real numbers, and x is a variable. We assume that b > 0.
Recall from Section 2.5 that |x − a| equals the distance between x and a on the
real-number line. For example, |5 − 2| = |3| = 3 is the distance between 2 and 5,
which is the same as |2 − 5| = |− 3| = 3.
Thus inequality |x − a| <b says, in words, that the distance between x and a is
less than b. This means that x lies between a − b and a + b;

Be sure that you understand this point:

Thus the inequality |x − a| <b means that x lies inaninterval of length 2b,
centered at x = a.
For example, let us find the solution of the inequality

First, divide through by 2:

Now use Eq. 4.40

This is the desired solution. As a check, note that at the end-points of this
interval, x = −4 or +1, we have |2x +3| = 5 in both cases. This makes sense – the
given inequality |2x +3| < 5 is true for x between these values, but not beyond.
We should expect to get equality right at the ends of the interval. (A sketch of the
graph of y = |2x +3| further confirms this point – see Chapter 7.)

Problem 4.47 (a) Solve the inequality |3x − 6|≤ 8. (b) Find an inequality
involving absolute values, corresponding to the interval −3 <x < 7. What is the
general formula?

Quadratic inequalities
Consider the inequality

This is an example of a quadratic inequality. To solve it, we proceed as follows;

The general method (algorithm) is:

Step 1. Complete the square, obtaining an inequality of the form

Step 2. Take positive square roots, if possible:

Step 3. Use Eq. 4.40:

Two comments regarding step 2: first, recall that for any real
number Q. Therefore Second, to complete step 2, we
must have B > 0. If this is not the case, then the original inequality has no
solution. Here is an example:

Step 1.

We can see, without further ado, that the latter inequality can never be true. The
given inequality has no solutions x.

Problem 4.48 Solve the inequalities (a) 2x2 ≤ 6x +3; (b) x2 − 6x +10 > 0. In both
cases, check your answer by using the quadratic formula.

4.7 The binomial theorem

I trust you are now familiar with the equation (a + b)2 = a2 +2ab + b2, but let’s do
the calculation again, in longhand form:
 Can we detect a general pattern in this? Could we predict the answer for (a +
4
b) from the pattern?
First, notice that lines (1) and (2) have the same sequence of coefficients 1,2,1
which were the coefficients for (a + b)2. Also, line (2) is shifted one position to
the left. Keeping track of only these coefficients, the pattern is:

What would be the pattern for (a + b)4? See if you can figure it out, before
peeking.
The next pattern will be

Thus

Besides the coefficients 1, 4, 6, 4, 1, notice the regular pattern in the exponents of

a and b here. Describe this pattern in words. Also note a similar pattern in (a +
b)3 above, and indeed, in (a + b)2.

Problem 4.49 Find the coefficient pattern for (a + b)5. Use it to write out the
expansion of (a + b)5.

B. Pascal (1623-1662) figured out a neat way to organize the pattern of

coefficients, which is now called Pascal’s triangle:
 The algorithm for constructing Pascal’s triangle is: start each new line with a
‘1,’ then obtain each coefficient by adding two adjacent coefficients from the line
above, as in

Finally, end with another ‘1.’ Looking back at how the coefficients 1, 4, 6, 4, 1
were obtained earlier from 1, 3, 3, 1, you can see that Pascal’s triangle does the
same thing, but in a tidy way.
The numbers on line n in Pascal’s triangle are called the binomial
coefficients in the expansion of (a + b)n. The written out expression is often
referred to as the expansion of (a + b)n.

Problem 4.50 Continuing Pascal’s triangle, find the expansion of (a + b)7.

Pascal’s triangle has the disadvantage that, in order to expand (a + b)n one
must figure out all the coefficients for all exponents up to n. Fortunately, there is
a single formula for these coefficients, which I will now explain. Let C(n, k)
denote the kth coefficient in (a + b)n. (You should read C(n, k) as ‘C of n and k,’
or just ‘C n k.’) Thus the entries on line n in Pascal’s triangle are

We have, therefore

In math, 3 dots (··· ) are used to indicate the continuation of an indicated pattern.
For example, 1 + 2 + ··· + n means the sum of the integers from 1 up to n. This
type of notation can only be used when the interpretation is obvious and
unmistakable. In reading examples of this notation, you should pause long
enough to understand exactly what the pattern is. For example, mentally insert
the first missing term.
Here is the formula for C(n, k), as I will explain:

First what does n! (called ‘n factorial‘) mean? By definition

That is, n! is the product of the integers from 1 up to n (n ≥ 1), and 0! is defined
separately by 0! = 1. For example,

Let us check that Eq. 4.41 does give the correct coefficients for n = 5.

and so on. Seems to work.

Problem 4.51 Check that the C(n, k) values are correct for n = 6 as given by
Pascal’s triangle. You only have to calculate C(n, k) for k = 0, 1, 2, 3 (why?)

Some properties of the binomial coefficients that can be seen either from
Pascal’s triangle, or Eq. 4.41, are:

The last equation is a symmetry condition – see Pascal’s triangle.

Now, how could we prove in general that the binomial coefficients calculated
using Pascal’s triangle, and the numbers C(n, k) given by Eq. 4.41 are the same
for all n and k? Remember how Pascal’s triangle is calculated by adding adjacent
entries on one line to get a number on the next line. For example,

Here, for example, C(4, 2) + C(4, 3) = 4 + 6 = C(5, 3), as indicated.

In general, we need to show that the numbers C(n, k) in Eq. 4.41 satisfy

Here is the algebra:

 On the second line above, we used the fact that which follows
because (k +1)! = (k +1)k!
Therefore the numbers C(n, k) given by Eq. 4.41 satisfy the condition (4.43)
that determines Pascal’s triangle. Also, C(n, 0) = 1, so the 1’s at the ends of each
line are also correctly given by Eq. 4.41. Hence these numbers C(n, k) must be
the same as the Pascal triangle numbers. In other words, C(n, k) are indeed the
binomial coefficients.

Problem 4.52 (a) Write out the first 4 terms in the expansion (a + b)10, using Eq.
4.41; (b) Show that Also What is the
general case?

Summation notation
The symbol sum Σ (Greek capital sigma) is used in math to designate
summation, as in

In other words, equals the sum of the values qk, with subscript k going

from 0 up to n. A numerical example:

(which happens to equal 14). The expression is read as ‘Sigma qk for k = 0

to n,’ or ‘Sum of qk for k = 0 to n.’

The summation index can be any letter, provided that the terms being
summed are expressed using the same letter. Thus
(Sometimes the index of summation, whatever it is, is said to be a ‘dummy
index,’ to indicate that the actual value of the sum (14 in this example) does not
involve the index, k, j,or whatever.)

Problem 4.53 Find to two decimals, using a calculator.

The summation notation provides a compact formula for (a + b)n:

Problem 4.54 (a) To be sure you understand Eq. 4.44, write it out fully for the
case n = 3; (b) Write down the binomial theorem for (u + v)p. Use x as the
summation index.

Problem 4.55 Expand (x − y)5, using any method you like for the coefficients.

4.8 Fractional exponents

In Section 4.1 we studied exponents in an expression xn, when the exponent n
was an integer (positive, negative, or zero). What about non-integer exponents?
Let’s start with an example:

Your calculator will give to three decimals. Now raise this number
5
to the 5th power; you’ll get (1.762) = 17. In other words, is the 5th root of
17.
In general, we define
which means that is the nth root of x. We sometimes write

where the symbol is read as ‘the nth root of’. When n = 2 we get the usual
square root, which is always written without the ‘2.’ The expression
is sometimes called a radical.

Problem 4.56 Find by inspection (no calculator needed!):

Note carefully that is the positive nth root of x. For example,

4
even though (−2) also equals 16. The reason for this is the need to
avoid ambiguity in all mathematical expressions.
Taking the nth root of a number is an example of an inverse operation.
Other examples of inverse operations that you are familiar with are subtraction
and division. Thus

As you can see, inverse operations are very common in mathematics.

Equation 4.45 implies that

For example, when n = 2 this says

Problem 4.57 (a) For what values of x is it true that For what values
4
of x is this not true? (b) Solve for x: x = 16 (find all real solutions).
Problem 4.58 Find by using a calculator. Check.

Our basic definition of in Eq. 4.45 raises some questions:

1. Why do we make this definition?
2. What about other exponents a in xa?
3. Are the three rules of exponents (Sec. 4.1) still valid?
Let’s consider each question in turn.
First, recall the rule (xm)n = xmn. If we wish to maintain this rule for
we need

This is exactly what Eq. 4.47 says. Thus the definition of is designed to keep
the math consistent and simple. (A similar argument was used in defining
see Sec. 4.1.)
Next, for an exponent (where p, q are integers) we define

Once again, this definition is motivated by the laws of exponents. It now turns
out that all three rules of exponents hold for arbitrary exponents. Here are the
rules again:

Here a and b can be any rational numbers (positive, negative, or zero), and x
and y are positive real numbers.
Should you, as an average math student, be able to explain (and remember)
exactly why each of these rules is true in general? I don’t think so. But you should
certainly remember why the rules hold for positive integer exponents (Sec. 4.1),
and then just keep in mind that they also hold for any exponents. This will
ensure that you never make errors in using exponents.
A partial proof of the rules for exponents is discussed at the end of this
section.

Problem 4.59 Use your calculator to check that 30.7 × 31.6 = 32.3.

The following additional laws can be deduced from the above laws of
exponents:

We discussed these laws for integer exponents in Sec. 4.1. The fact that they are
valid for arbitrary exponents is in line with the general consistency of
mathematics.

Problem 4.60 Show that for x, y > 0:

Suggestion: These are just special
cases of the laws of exponents.

Simplifying radicals
Consider the number We can simplify this as follows:
 A more general example of the same idea is

The perfect-square factor x2 under the sign can be removed from under the
sign, taking its square root x (if x> 0).
A similar calculation applies to nth roots. For example,

Another example of simplification is

This works because 32 = 25.

Problem 4.61 Simplify, if possible:

An example of simplifying a square root sometimes occurs when using the

quadratic formula. For example, let us use the quadratic formula to solve 3x2 −
2x − 3= 0:

This particular simplification occurs whenever the coefficients a, b, c are

integers, with b even. Try another example, x2 − 8x − 3 = 0; the solutions are
Rationalizing the denominator
Examine the following calculation:

Be sure you see how this works. You could also check the result numerically on
your calculator. (Before calculator days, was recognized as being a lot
easier to calculate than ) The calculation is an example of ‘rationalizing the
denominator,’ because is irrational, whereas 2 is rational.
Another useful example is

Again, the final result would be easier to calculate by hand than the original
expression.

Problem 4.62 Simplify

For any reader dying of curiosity, here’s the proof of Eq. 4.51. First, take the
case where q is an integer. We have to prove that For q =
2, we would be proving that

(remember, ). This particular formula is often used in algebra.

In fact, let’s just prove Eq. 4.55; the proof of Eq. 4.51 for other values of q is
pretty much the same. So how do we prove that We can only
use (i) the definition of and (ii) any formulas proved earlier in this book.
The equation something means that xy = something2, by definition (Eq.
4.45 for n = 2). So we have to prove that Here it is:

This completes the proof that

Now, let’s prove that, for example, which is another special
case of Eq. 4.51. Here’s the proof:

See? Math is very tightly organized and logical. Some people love mathematics
for this very reason. Others hate it. You can’t please everyone.
Most math teachers are careful to point out that no number of special cases
are sufficient to prove a general result. However, sometimes a well-chosen special
case can indicate how the general result can be proved. That is true for the
example just discussed, as the next problem shows.

Problem 4.63 (Optional) By emulating the proof that prove

Eq. 4.51 in general. (Start with the case etc.)

Now, while it is true that special cases do not establish a general theorem, the
study of special cases is nevertheless often worthwhile. Three advantages of
looking at special cases before tackling the general problem are:
1. Being easier to understand, special cases can clarify your ideas and help you
to grasp a more general principle. If you can’t understand the special case,
surely you have little hope of understanding the general result.
2. Mastering special cases can build confidence.
3. In original research, you may not know in advance what the general result
is. By looking at special cases you may be able to eventually discover the
general rule.
 For a student working on a difficult problem, or a confusing section of the
text, inventing and solving special cases may be the best way to approach general
understanding. However, it is always important to also master understanding of
the general case.

Roots of negative numbers

What is It should be a number whose cube equals −1. But (−1)3 = −1, so

More generally, for any positive number a, we have

because cubing the number gives −a.

The general result is that, for a> 0

However, does not exist if n is even.

Example. Calculate (−27)2/3.
Your calculator may produce an error message here. It is programmed to do
so whenever you try xy with negative x. However, [(−27)1/3]2 = [−3]2 = 9. Does
this mean that the calculator is wrong? No – it has a certain algorithm for xy,
which doesn’t work if x < 0. To use the calculator on such a problem, just keep
track of the minus sign separately:

whereas

The general situation is that (−a)m/n makes sense if n is an odd integer, but not if
n is an even integer. (In my opinion, this fact is an unimportant oddity.
However, it does sometimes occur on math tests.)

Problem 4.64 Calculate (if it exists) (a) (−3.6)–3/4; (b) (−1.7)–2/3.

4.9 More about polynomials

Another format for division of polynomials

We now carry out the division

in two ways, first using MWTFU, and then using a format that some people
prefer, called ‘long division.’

(1) MWTFU

The quotient is Q(x)= x2 +5x + 15, and the remainder is R(x) = 40.

(2) Long Division

 Both methods can also be used for dividing more complicated polynomials
As I’ll explain later, in fact the two methods are actually just different ways
of writing out exactly the same sequence of calculations.
The long division algorithm is:
(1) Write N (x)and D(x) in decreasing powers of x. Explicitly include any missing
terms in N (x), by using 0 coefficients (see the example).
(2) Use the format as shown.
(3) Divide the first term of D(x) into the first term of N (x), and write the result
above the line. (In the example, x into x3 gives x2.)
(4) Multiply this term (i.e., x2) by D(x), and write the result on the next line
down. (Here, x3 − 3x2.) Draw a line.
(5) Subtract the result from N (x), and write this result below the line. This gives
a new polynomial N1(x) of lower degree than N (x). (In the example, 5x2 − 5.)
(6) Repeat steps 3-5 for N1(x).
(7) Continue repeating the calculation until the degree of Nk(x) is less than the
degree of D(x). (In the example, N3(x) = 40 has degree zero, which is the first
case with degree smaller than one, the degree of x − 3.) Now Nk(x)= R(x), the
remainder. Also, the quotient Q(x) is now at the top of the calculation.
If you look carefully at both MWTFU and Long Division in the example, you
will see that they are really the same calculation, but differently organized. For
example, we first get the term x2, which we multiply by (x − 3). Then we ‘Fix it
Up’ by subtraction, getting N1(x) = 5x2 − 5 in both methods. Next we do the same
thing with N1(x), and so on.

Problem 4.65 Divide 3x4 − 2x3 + x by x2 + 2, using both methods. Check by

multiplying out, as usual. Watch how the two calculations parallel one another.
Which do you prefer?

Synthetic division
Consider the same example as above:

Compare this with:

Note that the bottom line contains the coefficients of the quotient Q(x)= x2
+5x + 15, and the remainder R(x) = 40. This is called synthetic division. The
algorithm is as follows.
To divide anxn + an−1xn−1 + ··· + a0 by x − a:
1. Write the coefficients anan−1 ··· a0 (including any zeros) on the top line.
Leaving space for a second line of numbers, complete a half-box, as shown.
 Write the value a on the left.
2. Copy an to the third line.
3. Starting with the left-most column, repeat the following steps, up to the
final column on the right:
1. (a) Multiply the number on the third line, current column, by a, and
enter on the second line, next column.
2. (b) Moving to the next column, enter its sum on the third line.

(This is easier to do than to explain!)

Check that this algorithm was used in the above example. To explain why
this method works, compare the two calculations for the example. Synthetic
division performs exactly the same arithmetic as long division, but omits all
unnecessary details. Try one or two more examples, both ways, to assure
yourself.
Note that synthetic division only applies to division by x − a, not by any
other form of divisor.

Problem 4.66 Divide (a) 2x4 − x +5 by x +3; (b) 3x3 − x2 − 8x − 4 by x − 2.

For additional practice, make up your own problems.

A fast check for the correctness of any particular synthetic division of P (x)
by (x − a) is that

where P (a) is the value of the polynomial P (x) when x = a. The example worked
out above had P (x) = x3 +2x2 − 5 and (x − a) = (x − 3). Thus P (3) = 33 +2 × 32 −
5 = 40, which equals the remainder R. You can check that Eq. 4.56 also holds for
the examples in Problem 4.66.
Equation 4.56 is known as the Remainder Theorem.
 To prove the remainder theorem, we use the basic definition of division of
polynomials (see Eqs. 4.27-4.29). Thus division of P (x) by x − a means that

where the remainder R is a constant. (Recall that, in general, deg R(x) < deg D(x).
Here D(x)= x − a, which has degree 1. Therefore deg R(x) < 1, so R(x) =
constant.) Substituting x = a in Eq. 4.57, we obtain P (a)= R, and this proves the
remainder theorem.

Problem 4.67 For the example P (x) = 4x2 − 1, and find R by synthetic
division, and check that P(a) = R. Do the same for a = 1.

For example, consider P (x)= 4x2 − 1, as in Problem 4.67. Here is a

root of P (x), because Also, is a factor of P (x), because
On the other hand, a = 1 is not a root of P (x),
because P (1) ≠ 0. And, sure enough, (x − 1) is not a factor of P (x).
To prove the factor theorem, we again use Eq. 4.57

First, suppose that a is a root of P (x), so by definition P (a) = 0. By the

remainder theorem, R = P (a) = 0. Therefore P (x) = (x − a)Q(x), i.e., (x − a) is a
factor of P (x).
 Conversely, if (x − a) is a factor of P (x), then P (x) = (x − a)Q(x). Therefore P
(a) = 0, i.e. a is a root of P (x).
The factor theorem is an ‘if-and-only-if’ theorem. This means that it is in fact
two theorems. First, if a is a root of P (x), then (x − a)is a factor of P (x). Second,
if (x − a) is a factor of P (x), then a is a root of P (x). The proof of an if-and-only-
if theorem must include both parts, as in this instance.

Solving higher-order equations

The solution of a polynomial equation of order higher than 2 is a difficult
problem in general. However, in some cases you can try to guess a solution. If the
guess is correct, the original problem can be simplified.
For example, consider the cubic equation

Can you ‘see’ a solution in your head? How about x = 1? Yes: P (1) = 0, so x = 1 is
a solution. Now apply the factor theorem, which says that (x−1) must be a factor
of P (x). By synthetic division we find that

Could there be other solutions of the equation P (x) = 0? If so, they must be
solutions of

Using the quadratic formula gives Thus we have found 3

solutions of the original equation. These are the only possible solutions.
Given any cubic polynomial equation, if we can find one solution, we can
find all the solutions. The method is the same as in the above example: first
factor P (x)= (x − a)Q(x), then solve Q(x) = 0 by the quadratic formula.

Problem 4.68 Solve x3 +3x2 − 10x − 24 = 0. Hint: x = 3 is a solution.

How does one come up with one root of a given cubic polynomial? We will
show in Chapter 7 that every cubic polynomial has at least one real root, which
can be found by numerical calculation. Hence any cubic equation can be
completely solved. There are at most 3 real roots.
How many roots can a given polynomial P (x) have? Here we consider only
real roots; ‘complex’ roots are considered in Part 2.
First of all, a polynomial may have no real roots. For example, P (x)= x2 + 1
has no real roots, because x2 +1 ≠ 0 for all real numbers x. On the other hand, P
(x)= x2 − 1 has two real roots, x = 1 and x = −1
In general, a polynomial of degree n can have at most n roots (real or
complex). To see this, note by the factor theorem that each root a of P (x)
corresponds to a factor (x − a). Thus, if a is a root, then

where Q(x) is a polynomial of degree n − 1. A second root of P (x), say b, must be

a root of Q(x), so

and therefore

This process can be continued at most n times, because each step reduces the
degree of the quotient by one. Therefore P (x) can not have more than n roots.

4.10 Review problems

1. Simplify (a) (a−3b2)−1;(b) (x2y−3)÷(x3y−2); (c) (ab2c3)4 ×(a3b2c)−4.
2. Combine fractions: (a) q/4r2 − s/2rt; (b) A2/6 − A/3; (c) 1/(x − y) + 1/(y − x).

3. Simplify (a) (1 + w/z) ÷ (1 − w/z); (b) (x2y/2 − xy2/3) ÷ xy.

4. Write as a polynomial (a) 3(x2 −2x−1) −(x2 +x−2); (b) (x+2)(x3 − x2 + x − 1);
(c) (x2 − 2)2.

5. Divide: (a) x4 − 3x2 +1 by x2 +1; (b) x5 − 1by x − 1.

6. Factor the polynomials: (a) y2 − 16; (b) y2 + y − 6; (c) 2y2 +5y +2.

7. Factor the denominators, then add fractions:

8. Solve for the unknown (a) 3 + y = 4y − 6; (b) (3x +5)/(x − 1) = 2 (first multiply
through by (x − 1).

9. Solve the inequalities: (a) 3x − 2 ≤ 2x +1; (b) 2.9y +4.1 > 7.3y.

10. Solve by completing the square: (a) x2 −2x−5= 0; (b)3x2+x−1= 0.

11. Use the quadratic formula to solve (a) x2+5x+1 = 0; (b) 9x2−2x−7= 0.

12. Solve by inspection (using factoring): x2 − 2x = 0; (b) x3 − 4x = 0.

13. Solve by reducing to quadratic equation: (a) x +4/(x − 1) = 6; (b) x4 − 2x2 − 3=

0; (c) 1/x2 +2/x +1 = 0.

14. Solve the inequalities: (a) 2x/(x−1) < 3; (b) |x+2|≤ 5; (c) x2 −4x < 12.

15. Expand by the binomial theorem: (a) (c2 + d2)3; (b) (x − 2)6.

16. Simplify

17. Calculate: (a) (81)−3/4; (b) (−125)2/3.

18. State the Remainder and Factor Theorems, and write out the proofs.

Solution 4.1 First, 12−92 = 1−81 = −80, while (1+9)(1−9) = 10×(−8) = −80.
Second, (−4)2 − 22 = 16 − 4= 12 [remember, (−4)2 = (−4) × (−4) = +16] and
(−4+2)(−4 − 2) = (−2) × (−6) = 12.

Solution 4.2 (Just notice that 3n becomes very large as n increases. On the other
hand (.7)n becomes very small for large n. More about this later.)
Solution 4.3 (a) x2x6 = x8; (b) (x2)6 = x12.

Solution 4.4 First, 43 = 64; 42 = 16; 41 = 4; and 40 = 1. Also, 103 = 1,000; 102 = 100;
101 = 10 and 100 = 1.
Solution 4.5 (a) z4; (b) t−4; (c)w.
Solution 4.6

Solution 4.7 If you’ve forgotten any of these points, go back and re-read the
corresponding discussion.

Solution 4.8 (a) (b−2)−1 = b2 by Eq. 4.4; remember that this equation is valid for
negative exponents; (b) (x3)−1x5 = x−3x5 = x2; (c) (x3y−2)4 = (x3)4(y−2)4 = x12y−8 by
Eqs. 4.5 and 4.4. The answer could also be written as x12/y8. (d) (a2/b)3 = (a2b−1)3 =
a6/b3. In general, (x/y)n = xn/yn;this is proved in the same way as this example.
Solution 4.9 After 20 years the value is V20 = (1.07)20V0 = $3, 869.68. Similarly,
the value after 30 years is V30 = (1.07)30V0 = $7, 612.26. Simple interest is interest
that is not compounded, but equals 7% of $1,000 each year. After 20 years, the
total simple interest would be 20 × .07 × $1, 000 = $1, 400, and the total bond
value $2,400. Similarly, after 30 years the simple-interest bond would be worth
$3,100. (Simple interest would apply to a bond that has annual dividends, in this
case $70.00 per year. The investor would presumably cash in these dividends
every year, perhaps reinvesting them, or perhaps using them in the meantime.)
Solution 4.10 We want to compare V0 for i = .06 with V0 for i = .065 (i.e. 6 ½%).
By Eq. 4.16 we have V0 = $5327.26 for i = .065, compared to V0 = $5583.95 for i =
.06. The decease in value V0 is equal to 4.6%.

Solution 4.11 Use Eq. 4.17 with t = 90, Vt = $100,000, and i = .05, or i = .10. You
get V0 = $1,238.69 and V0 = $18.82 respectively. (This calculation tacitly assumes
that there are no other expenses incurred in looking after the forest for 90 years,
such as taxes or fire insurance.)
This problem indicates that the present value of an asset that will only mature
many years in the future, can be very low. Low present values can be a serious
problem in renewable resource management.
Solution 4.12 The projected population is (1.015)2000×(7×109)= 6×1022. The
radius of the earth is about 6,000 km. Using the formula A = 4πr2 for the surface
area of a sphere of radius r we get A = 4.5 × 108 square km. About 1/4 of this area
is land, or 1.1 × 108 square km. Therefore the population density in 2,000 years
will be about 6 × 1022 ÷ 1.1 × 108 = 5.4 ×1014 people per square km, or (since 1
square km is 106 square meters), 5.4×108 people per square meter. In words,
there would be about 500 million people standing on every square meter of the
earth’s land surface. I don’t think so.

Solution 4.13 For t = 10,000 we get Xt = a10,000X0 = .298X0. Thus 29.8% of the
sample remains after 10,000 years. Similarly, 8.9% remains after 20,000 years,
and .0005% after 100,000 years. [Radio-carbon dating of archaeological items is
based on such calculations. The carbon content of a biological organism is
obtained from atmospheric carbon dioxide (CO2), and is a known mixture of
non-radioactive carbon and radioactive C14. Once the organism dies it no longer
takes up atmospheric carbon. By measuring the proportion of C14 in the
archaeological item, the time since its death can be calculated, as in this problem.
However, items older than about 50,000 years contain too little C14 for accurate
measurement, and other radioactive elements must be used.]

Solution 4.14 [First note that 6 + 9t = 3(2+ 3t), then cancel

the factor 3.] (c) No cancellation possible.

Solution 4.15 Did you check the

answers by splitting and canceling?

Solution 4.16 The l.c.d. is x2y2, and the answer is (Check this in the
usual way.)

Solution 4.17 (To get this, first add the fractions in the numerator, and
then in the denominator. Then invert, multiply, and cancel.)

Solution 4.18
(This can be simplified to by using Eq. 4.1.)

Solution 4.19 (a) is not a polynomial, because the term does not have
a non-negative exponent; (b) is a polynomial (the coefficients can be any real
numbers); (c) is not a polynomial because the term is not of the form axn for
some non-negative integer n.

Solution 4.20 (a) The degree is 5; (b) The terms are 2q (coefficient 2),
(coefficient ) and 7 (coefficient 7). The constant term is 7.

Solution 4.21 (2x2−1)+(x2 +x+3)−(x−5) = 2x2 −1+x2 +x+3−x+5 = 3x2 + 7 . Note

carefully that −(x − 5) = − x + 5. Read Sec. 2.2 if you are confused about this
point.
Solution 4.22 (a) 2x4 − 9x3 +11x2 − x − 2; (b) 4x4 − 4x3 − 11x2 − x − 3.

Solution 4.23 (a) UX + UY + UZ + VX + VY + VZ (or, you could have gotten UX

+ VX + UY + VY + UZ + V Z, which is the same thing in a different order); (b)
AD + AE + AF + BD + BE + BF + CD + CE + CF (or the other order, as in a).

Solution 4.24 (a) 2x3 − 5x2 − 7x − 2; (b) 3x7 + x5 − 13x3 − 11x2 +2x +3.

Solution 4.25 (a) (x − 3)2 = x2 − 6x +9; (b) (2w + z)2 = 4w2 +4wz + z2; (c) (x2 +3)2
= x4 +6x2 +9.

Solution 4.26 (a) 2x4 −3x2 +5 = (2x2 +1)(x2 −2) +7; (b) x3 −x2 +x−1= (x2 +1)(x −
1).

Solution 4.27

Solution 4.28 (a) By division, 2x2+7x+3 = (2x+1)(x+3). Therefore 2x+1 is a

factor of 2x2 +7x+3. (b) By division, x2 −2x−c = (x+3)(x−5)+15−c. Therefore the
remainder is zero if and only if c = 15.

Solution 4.29 (a) x2 +6x +8; (b) x2 − 7x + 12; (c) x2 − 5x − 6; (d) 8x2 − 10x − 3.
Solution 4.30 (a) (x +4)(x + 3); (b) (y − 4)(y − 3); (c) (x − 6)(x − 2); (d) (t − 12)(t
+1).

Solution 4.31
2
(c) x + 16
can’t be factored [Well, it could be factored using ‘imaginary numbers, but that’s
another matter and quite beside the point here.]
Solution 4.32 (a) and (b) Re-read the text to be sure you understand these
words; (c) 5x3 − 9x2 +2x +2; (d) 2x5 +7x4 − x3 − x2 +3x − 2; (e) x3 +3x2 +7 = (x +3)
(x2 − 2) + (2x + 13); (f) x2 − 16 = (x +4)(x − 4). Also x2 +6x − 16 = (x +8)(x − 2).

Solution 4.33

Solution 4.34 Substituting x = 1 gives x3 − 6x2 +2x +12 = 9, which is not 0, i.e., x
= 1 is not a solution.

Solution 4.35 (a) x = 6/2 = 3. (b) x = (−2)/(−5) = 2/5. (c) Use two steps here: first
step 2x = 6; second step x = 3. (Check this answer: 3 × 3 − 1= 8 and 3+5= 8 also.)
Solution 4.36 The inequality reverses direction: ac > bc if c < 0. An example: 2 <
3, but 2 × (−2) > 3 × (−2). Be sure you understand this example.

Solution 4.37 (a) x < 3. (b) x ≥ 6. (c) x < 5/3

Solution 4.38 (a) We have x2 −2x−8= (x−4)(x+2) = 0, so the solutions are x = 4

and x = −2; (b) Since x2 − 2x − 7 = 0 cannot be factored by inspection, we use the
quadratic formula. We have a = 1, b = −2, and c = −7. Thus b2 − 4ac = 32. The
two solutions are

Solution 4.39 In (b) did you

2
first factor out the 2, simplifying the equation to x + x − 1= 0?

Solution 4.40 (a) Here b2 − 4ac = 9 − 4c, so the equation has real solutions if and
only if 9 − 4c ≥ 0, i.e. if c ≤ 9/4. (b) If a and c are of opposite sign then ac < 0.
Therefore b2 − 4ac > 0, so the equation does have real solutions, by the quadratic
formula.
Solution 4.41 (a) b2 − 4ac = 85; two real solutions; (b) b2 − 4ac = −11.24; no real
solutions; (c) b2 − 4ac = 0; one solution (namely, x = 5).

Solution 4.42 x2 − 2x − 9= x2 − 2x +1 − 10 = (x − 1)2 − 10. Therefore the

solutions are
Solution 4.43
Solution 4.44

Solution 4.45 All the laws remain valid. This can be verified by considering
cases. For example, to check the transitivity law, now written as: If a ≤ b and b ≤
c then a ≤ c, we can consider the cases (i) a < b and b < c, (ii) a = b and b < c, (iii)
a = b and b = c. We get, respectively, a < c, a < c, and a = c, so in all cases we have
a ≤ c. This proves that transivity is valid.

Solution 4.46 (a) x < −7/5 or x > −1; (b) 1/4 < x < 1 (i.e., x is between 1/4 and 1).
(If you think solving inequalities is confusing, you are not alone!)

Solution 4.47 (a) −2/3 ≤ x ≤ 14/3; (b) |x−2| < 5. In general, A < x < B if and only
if |x − (A + B)/2| < (B − A)/2. Note that (A + B)/2 is the midpoint of the interval
from A to B.

Solution 4.48 To check, we solve the

2
quadratic equation 2x − 6x − 3 = 0 by the quadratic formula, giving
These are the end-points of the solution interval. (b) True for
all x. The quadratic formula gives i.e., no real solutions to the
quadratic equation.

Solution 4.49 The coefficient pattern for (a + b)5 is 1, 5, 10, 10, 5, 1. Thus (a + b)5
= a5 +5a4b +10a3b2 +10a2b3 +5ab4 + b5.

Solution 4.50 The 6th and 7th lines are 1, 6, 15, 20, 15, 6, 1 and 1, 7, 21, 35, 35,
21, 7, 1. Therefore

Solution 4.51 The values in Pascal’s triangle are 1, 6, 15, 20, 15, 6, 1, and these
equal C(6,k). (There is always a symmetry in the binomial coefficients, so for
example C(6, 2) = C(6, 4), etc.)
Solution 4.52 (a) (a + b)10 = a10 +10a9b +45a8b2 + 120a7b3 + ··· . (b) The general
case is

Solution 4.53

Solution 4.54 (a) For n = 3 we have, using Eq. 4.44 (remember that b0 = 1)

This is the correct expansion of (a + b)3.

Solution 4.55 (x−y)5 = x5−5x4y+10x3y2−10x2y3+5xy4−y5. (Theterms alternate in

sign because (−y)k = (−1)kyk = ±yk depending on whether k is even or odd.)

Solution 4.56 because 24 = 16; also because 53 =

125.

Solution 4.57 is true if x ≥ 0. On the other hand, consider the

2
example x = −2. Then x = 4 and In general,
is true if and only if x ≥ 0. Many people find this confusing, so be sure you follow
the reasoning here. (In fact we have for all values of x. Check that
4
this is true.) (b) x = 16 has two solutions, x = 2 and x = −2.

Solution 4.58 It checks out that (35.57)4 = 1.6 ×106.

Solution 4.59 Both equal 12.51 to 2 decimals, for example. Your calculator
should show equal values for 10 or so decimals.

Solution 4.60 this uses Eq. 4.51;

this uses Eq. 4.54.

Solution 4.61 (a) 9a2; (b) 81; (c) a+b (if a, b are positive; otherwise |a+b|);

Solution 4.62 (after multiplying numerator and

denominator by (c)27.

Solution 4.63 To prove that we must show that

according to Eq. 4.45. But by Eq. 4.5 and 4.47.
Finally by Eq. 4.47, the result just
proved, Eq. 4.5, and Eq. 4.48. (Notice that this proof is virtually identical to the
special case ).

Solution 4.64 (a) Doesn’t exist, because x1/4 does not exist for negative x. (b)
0.702.
Solution 4.65 3x4 − 2x3 + x = (x2 + 2)(3x2 − 2x − 6) + (5x + 12).

Solution 4.66 (a)

Therefore 2x4 −x+5 = (x+3)(2x3−6x2 +18x−55)+170. You should check by

multiplying out. Notice, as you do so, that you seem to be just reversing the steps
of the synthetic division. (b) Q(x)= 3x2 +5x +2, R(x)= 0.

Solution 4.67 By synthetic division, Thus R = 0.

Also For a = 1 we get 4x2 − 1= (x − 1)(4x +4)+3. Here R
= 3 = P (1).

Solution 4.68 By synthetic division, x3+3x2−10x−24 = (x−3)(x2+6x+8). The

quadratic can be factored: x2 +6x +8 = (x +2)(x + 4). Therefore the solutions are x
= 3, −2, and −4. The latter two solutions can be checked either by calculating P
(a), or the remainder R for each.
Chapter 5
Euclidean Plane Geometry
In presenting Euclidean geometry, we need to explain a number of basic facts.
Here are two of them:
1. In any triangle, the sum of the interior angles equals 180°.
2. In any right triangle, the square of the hypotenuse equals the sum of the
squares of the other two sides (Pythagoras’s theorem).

From these and several other basic facts we are able to develop much of
geometry and trigonometry. But how do we know that these facts are true?
Neither of the above statements is obvious in any sense. One can imagine two
approaches to establishing the validity of such statements as these, the
experimental method, and the theoretical method.
Experimentally, we could obtain measuring instruments (a protractor to
measure angles, and ruler to measure lengths), and check the truth of the
statements by taking measurements. You may wish to try such measurements of
a few examples, for the statements given above. (You will want a protractor in
this chapter, in any case.) Your measurements will probably verify that the
statements are at least approximately correct.
Mathematics, however, is not generally considered to be an experimental
science. Instead, all statements in math are supposed to be proved – then they
are called theorems. But proved how, and from what? From other basic facts,
perhaps, using accepted principles of logic. Then how are these more basic facts
themselves to be proved – from yet more basic facts? It was the strength of Greek
mathematics to realize that this process had to stop somewhere. One must begin
by accepting some statements as being true without proof. Such statements are
called axioms.
Anyone who has tried to read Euclid’s Elements knows how finicky it is to
deduce, by rigorous logic, all the results of plane geometry. This book takes a
more practical approach, explaining the basic facts of geometry in a convincing,
understandable way that, while based upon Euclidean axiomatics, avoids overly
technical detail.

5.1 Lines and angles

A fundamental notion in Euclidean geometry is that of a straight line. What
exactly is a straight line? According to Euclid, a line has length but no width; a
straight line is then a line ‘which lies evenly to the points within itself.’ Really!
But this is only Euclid’s lame attempt to explain what a straight line is, in the real
world. The current practice is to say that ‘straight line’ is an undefined term.
‘Point‘ is another undefined term. Any axiomatic system must start with
undefined terms, and in addition certain axioms regarding those terms.
Although the terms straight line and point are taken as undefined, we of
course have mental images of what these terms do refer to. These concepts are
related to figures that we can see, for example as drawn on a flat surface. But
these figures are not actually points and lines, any more than 7 is the number
seven.
For simplicity, we will henceforth use the word line to mean the same as
straight line.
Points and lines have the following properties (the statements being axioms).
Any two different points A and B determine a line, called the line through the
points A and B, or the line containing the points A and B. Any two lines
determine a point C (called the point of intersection of the two lines), unless the
lines are parallel. Given a line and a point P not on it, there is a unique line
through the point P , not intersecting the given line (the new line is said to be
parallel to the given line).
 To make intuitive sense of these axioms, we have to think of a line as being
extended indefinitely in both directions. For example, two finite line segments
need not intersect, or be parallel, but two infinite lines must either intersect or be
parallel. These three axioms seem obviously ‘true’ to most people, but it is not
quite clear why. The axioms make sense only for points and lines ‘in the plane,’
that is, lying on a given infinite, flat, two-dimensional plane (whatever all those
terms mean).
It may strike you as unsatisfactory that all of Geometry should depend on
assuming the truth of certain axioms that cannot themselves be firmly proved to
be true. You may still want to ask ‘but are they really true?’ Chapter 6 on Analytic
Geometry should help to set your mind at ease on this question. There we will
see, for example, that any line has an equation Ax+By +C = 0; parallel lines can
be identified by a certain property of the coefficients A, B, C, of this equation.
Also, it can easily be shown by algebra that parallel lines never intersect, and
non-parallel lines always intersect in a single point.
Contemporary mathematics is a big advance over classical Greek
mathematics.

Line segments
The part of a line lying between two points A and B is called a line segment, and
is denoted by AB. The length of a line segment is defined provided that a unit of
length has been specified. In practice, lengths would be measured in familiar
units such as inches, meters, or light years. In general, however, we can assume
some fixed unit of length without actually specifying what the unit is. The
distance between two points A and B is the same thing as the length of the line
segment AB.
Circles
A circle is defined as the set of all points that are equidistant from a fixed point,
which is called the center of the circle. The distance from the center to a point on
the circle is called the radius of the circle. Notice that, by this definition, a circle
is a curve, not a filled-in area. The filled-in area is called a circular disk.
An arc is a segment (connected piece) of a circle. The properties of circles are
discussed in Section 5.3.

Angles
A ray is a half-line, with an endpoint. An angle is a figure consisting of two rays
joined at their endpoints. The point where the rays meet is called the vertex of
the angle. The rays are called the sides of the angle.
 Two given angles are said to be equal angles if one of them can be moved
(without distortion) to coincide exactly with the other.

Three ways of naming an angle are shown below.

First, we can name the angle by its vertex, for example ∠V (‘angle V’).
Second, we can use three letters, for example, ∠ABC, (‘angle ABC’), where B is
the vertex, and A and C are points on the sides of the angle. Finally, we can use a
single letter, for example, α (‘alpha,’ the first letter of the Greek alphabet). The
letters used here are examples only; any letters can be used.
A straight angle is an angle whose two sides lie along a straight line:
 A straight angle can be partitioned into two equal angles, which are called
right angles. Two lines that meet at right angles are said to be perpendicular to
one another.

Problem 5.1 If line L1 is perpendicular to L2, and L2 is perpendicular to L3, what

can you say about L1 and L3?

The size of an angle

To specify the size of angles we begin by defining the size of a straight angle:

(180° is read as ‘one hundred and eighty degrees.’) The sizes of other angles are
determined by their relation to a straight angle. For example, a right angle has
size 90°, because two right angles make up a straight angle. Other useful angles
are 30°, 45°, and 60°. Thus three 60° degree angles make up

a straight angle, and two 45° angles make up a right angle. The 30°, 45°, and 60°
angles have special properties, discussed in Sec. 5.2.
 What about other angles? The protractor is a tool used to draw and measure
angles.

Problem 5.2 Draw two or three different right-angled triangles (i.e. having one
90° angle). Measure the other two angles, using a protractor, and write down
their sizes. What do you notice?

Complementary and supplementary angles

Two angles that add up to 90° are called complements of each other. For
example, 60° and 30° are complementary angles. In the figure, ∠PV Q and ∠QV
R are complements, assuming that ∠PV R = 90°. Similarly, two angles that add
up to 180° are called supplementary. Identify the supplementary angles in the
figure.

Two other terms regarding angles are these: an acute angle is an angle less
than 90°, and an obtuse angle is an angle between 90° and 180°.

If you can’t remember which is which, think of acute as sharp, and obtuse as
dull.

Opposite angles
Consider two lines intersecting at a point P . Angles directly opposite each other
at point P are called opposite angles. The two acute angles indicated in the figure
are opposite angles. So are the obtuse angles that are not marked.
A useful fact is that

You could check this experimentally for a number of examples, by using the
protractor. But keep in mind that experimental evidence is not considered
conclusive in math. We can give two explanations, one intuitive and the other
purely mathematical. The figure on the left indicates the first explanation. We
rotate one of the lines about point P until it coincides with the other line. The
rotation sweeps out the same angle on both sides, so the opposite angles must be
equal to begin with.

The mathematical explanation is indicated by the right figure, where three of

the angles are labeled α (alpha), β (beta), and γ (gamma). We want to show that α
= γ. But we have
 Therefore α + β = β + γ. Subtracting β from both sides, we conclude that α =
γ. QED.
The letters QED stand for ‘Quod erat demonstrandum,’ which is Latin for
‘that which was to be proved.’ Thus QED signifies the end of an argument, or
proof. (The abbreviation QED is a bit archaic nowadays. Many math books use
sign language, for example a block , to signal the end of a proof.)

Problem 5.3 Make yourself a glossary of terms introduced in this chapter so far.
Keep the list on hand, and add to it as you continue reading.

5.2 Triangles
A triangle is a figure with three line segments as sides. The corners of the
triangle are called its vertices. The vertices are often labeled using capital letters,
like A, B, and C.

In this case the triangle may be referred to as ΔABC (‘triangle ABC‘). The
angles at the vertices could then be called ∠A (‘angle A‘), and so on. Other ways
of naming these angles can also be used.
An acute triangle is one having three acute angles. An obtuse triangle has one
obtuse angle.(No triangle can have more than one obtuse angle, as we will see
shortly.) Finally, a right-angled triangle (a right triangle, for short) has one
right angle.
Parallel lines
We have defined two lines L1 and L2 to be parallel if they never intersect, no
matter how far they are extended in either direction. Another property of parallel
lines is that they have the same direction, in the sense that they make equal
angles with any third line L0 that cuts across the given lines. The line L0 is called a
transversal to L1 and L2 and the indicated angles are called corresponding
angles.

Here the phrase ‘and conversely’ means that if the two corresponding angles
shown in the figure are equal, then the lines L1 and L2 are necessarily parallel (i.e.,
they do not intersect). The boxed statement is an axiom.
Given a line L and point P not on L, we can construct a line L1 through P and
parallel to L, as shown below:

First, draw any line L0 through P , cutting line L. Then draw L1 through P ,
making the same angle with L0 that L does.

Problem 5.4 (Use your space intuition on this problem). Imagine you have two
lines in space. If these lines never intersect, does it follow that they are parallel?

Another pair of equal angles produced by two parallel lines is shown in the
next figure. These are called alternate angles. Can you see why alternate angles
are equal? It’s because the alternate angle is opposite to the corresponding angle
(check this in the figure). Since opposite angles are equal, as are corresponding
angles, the result follows.
 From this result we can now prove the following important theorem.
Theorem In any triangle the sum of the interior angles equals 180°.

Proof Let ABC be a triangle, with angles α, β, and γ as shown. Draw a line L
through C and parallel to line AB. Then line AC is a transversal to these two
parallel lines, so that as shown in the figure. Similarly, But
(since these three angles make up a straight angle at C).
Therefore α + β + γ = 180°.

Problem 5.5 Let ABC be a right triangle having ∠C = 90°. Show that ∠A and
∠B are complementary angles.

Problem 5.6 Let ABC be any triangle

with angles α, β, γ at the vertices A, B, C respectively. Let CD extend the side AC,
as shown. Show that ∠DCB = α + β. Suggestion: again, use the sum-of-interior-
angles theorem. (Think about this problem for a bit. You may not ‘see’ the
argument at first, but hopefully you will eventually. It might help to label the
angle BCD on the figures as angle X,say.)

Angle DCB is called an exterior angle to the given triangle ABC. The angles α
= A and β = B are called opposing interior angles to this exterior angle DCB.
Thus the statement of Problem 5.6 can be put in words: An exterior angle of any
triangle is equal to the sum of the two opposing interior angles.
By the way, note that Problem 5.5 can be viewed as a special case of Problem
5.6. Explain.
Did you figure out the explanation? If γ = 90°, then ∠DCB = 90° also.
Problem 5.6 says that ∠DCB = α + β. Therefore α + β = 90°, which is what
Problem 5.5 says.
Now for some advice. When reading this book, please read slowly and with
meticulous attention to details. Try to solve the problems yourself, before
looking up the solution. Also, take the time to answer questions. The problems
and questions are there to make you think. Thinking helps you to understand
and remember. Every math teacher knows that students need to think things out
for themselves, slowly and carefully. This is an essential part of learning
mathematics. It’s quite different from learning, say, Spanish, where you just keep
repeating new words until you memorize them.
 Even where there are no problems or questions, you can often pose questions
for yourself. Think of some examples. Ask if you understand why a statement is
true. Try to discover interesting special cases.
For example, take the theorem about the sum of the interior angles of a
triangle. Can you think of any special cases, other than a right triangle? Here’s
one: if all three angles are the same, how big is each angle? Or, if the other two
angles in a right triangle are equal, how big are they? Sketch these triangles.
Another question: suppose a person walks around a triangular field, always
keeping the field on her left. When she returns to the starting point, what total
angle has she turned through? Does this have anything to do with Problem 5.6?
If you’re studying this book on your own, you don’t have a teacher to ask
such questions. So be your own teacher – it will enrich and improve your
learning.

The parallel axiom

The above theorem, stating that the sum of the interior angles of any triangle
equals 180°, will be used repeatedly in the rest of this chapter, and also in
Chapter 8 (Trigonometry). The proof of this theorem depends critically on the
properties of parallel lines. But if space is really curved (as Einstein tells us), then
perhaps all lines are curved, and do eventually intersect, out in space somewhere.
If so, maybe Euclidean geometry is completely wrong!
Well, this line of reasoning seems totally confused, and it is. It confounds
mathematics and reality. Whether a given mathematical theory, or structure, or
axiom is ‘really true’ in nature is an undecidable question. Euclidean geometry is
based on certain axioms. Axioms are statements that are adopted as hypotheses.
They are not proved (if they were proved, it would be in terms of other axioms).
One might think that perhaps Euclid’s axiom system, including the parallel
axiom [parallel lines exist] was the only possible logically consistent system for
geometry. This hope was destroyed by two 19th century mathematicians, J.
Bolyai (1802-1860) and N. Lobachevsky (1792-1856), who showed
independently that a consistent theory of geometry could be developed without
the parallel axiom. Later, B. Riemann (1826-1866) extended the theory of non-
Euclidean geometry. Riemann’s geometry was what A. Einstein (1879-1955)
adopted for his general theory of relativity.
For everyday use, Euclidean geometry is perfectly adequate. It only breaks
down, as a model of Nature, at extremely great astronomical distances. (Whether
it may also break down at extremely small subatomic distances is not yet known.)

Proofs
Why should you bother learning proofs in math? Can’t you just memorize each
theorem and forget all about the proofs? This is a tricky question. I don’t want to
tell you that you MUST understand and remember every proof in math, forever.
Nevertheless, I do believe that taking the time and effort to carefully read and
understand the proofs in this book (there aren’t that many) will be well
worthwhile. Here’s why:
1. Reading and understanding the proof will ensure that you fully understand
all the concepts involved in the statement of the theorem.
2. The mental effort of understanding the proof will indelibly imprint the
theorem in your brain, even if you forget the details of the proof later on.
3. Understanding proofs will establish confidence in your overall mastery of
basic math. The subject will ‘hang together,’ rather than being a collection
of unconnected facts.
4. Understanding a proof can be a mental challenge, and overcoming mental
challenges is an essential aspect of learning mathematics. It’s the difference
between being an active participant and being a passive observer.
5. There can be a big difference between understanding a proof and
remembering it. Even if the proof is later forgotten, you will retain the
assurance when encountering the theorem later that you once understood
the proof.

Regarding some of the Problems in this chapter, you may be asking, ‘Why
should I do proofs? I’m never going to become a mathematician.’ It’s a fair
question; proofs can be hard. Today’s school math seems to downplay proofs,
but this is definitely not the case at the college level, where proofs, or ‘Show that
…’ questions often appear on exams. If you are in a college program or plan to
be in one, you need to learn how to do proofs. In any case, doing your own
proofs can be a powerful aid to learning mathematics.
So how should you tackle a proof, or ‘show that’ type of problem? It often
helps to write out the problem for yourself. What is given, and what is to be
shown? How could this relate to the material just discussed in the book? Even
quite simple proofs often take an inordinate amount of time to think up. If
you’re like me, you may tend to panic if you don’t immediately get the idea.
Well, don’t panic. Stick with it. Get some coffee. Try again, from the beginning. I
once spent a whole week on a Physics problem. The answer finally jumped into
my head while I was waiting for the bus to go to class. It was so obvious. (In fact,
I was the only student in the class to turn in the solution!)

Problem 5.7 What is the sum of the interior angles of any quadrilateral (4-sided
figure)? Suggestion: draw one and try to think up a construction that will reduce
the given problem to a known result.

Suppose ABC is a triangle, with vertices A,B, and C. It is sometimes convenient

to label the three sides using the corresponding lower case letters a, b, c as shown
here. Note that side a is opposite vertex A, and so on. We then also use a, b, c to
represent the lengths of these sides.

Now we come to perhaps the best known theorem in geometry. Pythagoras

(6th century BC) was one of the greatest of the Greek philosophers. (It’s
pronounced pie-thag-or-us.)

Theorem of Pythagoras Let ABC be a right triangle, with C = 90°. Let a, b, c

denote the lengths of the sides opposite vertices A, B, and C, respectively. Then
The side c opposite the 90° angle C is called the hypotenuse of the right triangle.
Pythagoras’s theorem is often stated in words:

‘In a right triangle, the square of the hypotenuse equals the sum of the squares
of the other two sides.’

Proof We begin with a construction. Arrange 4 copies of the given triangle, by

rotating it through 90° and joining up vertices as shown in the figure. Because
ABC is a right triangle, rotating it through 90° lines up sides a and b, as shown.
The corners of the large figure are all right angles, since they equal ∠C. Also, the
corners of the inside figure are also right angles, because they are formed by 90°
rotation. The resulting figure thus consists of two squares, one inside the other.
The small square has side c, and the large square side a + b. We have
 area of big square = area of small square + 4 × (area of triangle)

In symbols

Since the triangle has area (Eq. 3.2), this becomes

Now we need the algebraic formula (a + b)2 = a2 +2ab + b2; see Chapter 4.
Substituting into the preceding equation, we obtain

By subtracting 2ab from both sides, we conclude that a2 + b2 = c2.

Pythagoras’s theorem is used all the time, later in this book and in future
college courses. But probably few college students can recall the proof. Though
interesting, the proof is not germane to the applications of Pythagoras’s theorem.
Still, being so central to much of math, it seems to me that every student should
be expected to learn the proof at some time in their life. (I’ve seen several school
texts that don’t even bother to go through the proof of Pythagoras’s theorem
when presenting basic geometry. I find this intolerable.)
Problem 5.8 Find the length of the diagonal of (a) a square of side 6 cm; (b) a
rectangle of sides 3m and 4m.

Problem 5.9 There are many known proofs of Pythagoras’s theorem. Prove the
theorem on the basis of the figure shown here.

Problem 5.10 Re-read the proofs of the two theorems, sum-of-interior-angles,

and Pythagoras. Then close the book and try to recall the two proofs. You should
find that remembering the figures is the key to the proofs.

5.3 Similarity and congruence

You may recall learning about similar and congruent triangles in school.
Actually, the concepts of similarity and congruence apply completely generally to
any geometric figure. They apply to both two dimensions (plane) and three
dimensions (space), but here we will only consider plane figures.
Definition Two geometric figures in the plane are called similar if one is a
scaled-up or scaled-down version of the other. The figures may be located
anywhere in the plane.
Definition Two geometric figures in the plane are called congruent if they are
similar with scale factor k = 1.

In a sense, congruent figures are identical, except for location or orientation.

Similar figures are alike in shape, and differ only in scale (as well as location and
orientation).
These concepts can be expressed in terms of transformations. The types of
transformations allowed are:

Similarity: Congruence:
Translation Translation
Rotation Rotation
Reflection Reflection
Uniform scaling

These terms are best explained in terms of diagrams.

 Any transformation obtained by combining translation, rotation, and
reflection is sometimes called a rigid motion. Thus two geometric figures are
congruent if one of them can be made to agree exactly with the other by means of
a rigid motion. If you have a drawing program on your computer, there should
be buttons or handles allowing you to perform the various rigid motions. (There
is also a handle for scaling.)

Note carefully that the concepts of similarity and congruence allow for
reflection. The fact is seldom emphasized in school texts. You might imagine that
a rigid motion in the plane just means ‘sliding’ a figure around, using translation
and rotation. But this is not correct – reflection is also allowed. (Question.
Suppose you reflect a figure in one line, and then reflect it again in a second line.
Is the final figure ‘slideable’ to the original? Try to guess. The mathematical proof
requires advanced technique. The answer is yes.)

Congruent triangles
Congruent figures are identical in all respects (after a rigid motion). In the case
of triangles, if ΔABC and ΔA′B′C′ are congruent, then AB = A′B′, AC = A′C′, and
BC = B′C′. In addition, ∠A = ∠A′, ∠B = ∠B′, and ∠C = ∠C′. Thus, congruent
triangles have three sides equal, and three angles equal, one triangle to the other.
 Now let us consider the reverse problem. Given two triangles that have their
sides and angles equal (from one triangle to the other), are these triangles
necessarily identical in the sense of being congruent? The answer is yes, but we
can say much more. We actually only need to know that certain parts of two
triangles are the same, to be able to conclude that the triangles are in fact
congruent.
Let us formulate the question slightly differently. How much information do
we need to have about a triangle in order for this triangle to be uniquely
determined (so that any two such triangles would be congruent)? There are three
cases in which this is so:

We will consider these cases in turn. First let us agree on some conventions. The
triangle will be labeled ABC, going counterclockwise. Side AB is always given,
and will be drawn horizontally.
To get a feeling for Case 1, try the next problem.

Problem 5.11 (a) Using a ruler and compass, construct a triangle having sides 3,
5, and 7 cm. [Use inches if you prefer.] If you don’t have a compass handy, just
show how you would draw the triangle if you did have one. (b) Try doing this
with sides 3,4, and 8 cm. What goes wrong?

The point of this problem is to convince you that only one triangle (or its
reflection) can be drawn, given the three sides a, b, and c. The reason behind this
is that two circles (with centers at points A and B) can only intersect at two
points C and C′. Therefore, the construction produces only two triangles, which
are congruent by reflection. (Possibly the two circles fail to intersect at all, and in
this case there is no triangle having sides a, b, c. See Problem 5.11b.) This takes
care of case 1: Two triangles which have their three sides equal are congruent
because there is only one way to draw such a triangle (except for reflection).

Cases 2 and 3 are also quite easy to understand. (See the previous figure
showing the three cases.)

Problem 5.12 (a) Using a ruler and protractor, construct a triangle having AB =
5 cm, AC = 4 cm, and ∠A = 30°. (b) Construct a triangle having AB = 4.5cm,
∠A = 30°, and ∠B = 100°.

In Problem 5.12, as in the preceding problem, you see that the given
information, whether case 2 or case 3, can only produce one triangle, once side
AB has been drawn. (However, point C could be produced either above or below
AB, depending on how the angles are drawn, up or down. The two triangles that
could be drawn are, as before, congruent by reflection.)
Let’s summarize these facts about congruent triangles as a theorem. The
foregoing discussion should serve as an adequate proof of the theorem.

Theorem Let ABC and A′B′C′ be two triangles, with sides a, b, c and a ,b ,c
respectively. If any one of the following conditions hold, then the triangles are
congruent.
 1. (Three sides equal) a = a′, b = b′, and c = c′.
2. (Two sides and the included angle equal) b = b′, c = c′, and ∠A = ∠A′.
3. (One side and two angles equal) c = c′, ∠A = ∠A′and ∠B = ∠B′.

Check that these correspond to the three figures shown previously.

Regarding case 3, notice that if any two angles are given, then the third angle
is also known. This is true because of the fact that the three angles in a triangle
always sum up to 180°. For example, if ∠A = 52° and ∠C = 60°, then ∠B = 68°.
In each of the three cases, it is possible to draw one and only one triangle on
the basis of the given information (keeping in mind that two triangles that are
reflections of each other are considered to be congruent, or identical). Of course,
the given information needs to be consistent with the requirements of a triangle.
For example, if three sides a, b, c are given (case 1), no side can be longer than
the sum of the other two sides (Problem 5.11b). Or if two angles are given (case
3), their sum must be less than 180°.

Problem 5.13 Suppose that the three angles ∠A, ∠B, and ∠C of a triangle are
given. Is the triangle completely specified by this information? Why?

Similar triangles
Two triangles are defined to be similar if one is a scaled version of the other
(except for position and orientation). Now scaling does not change angles.
Therefore similar triangles have equal angles, between one triangle and the other.
Thus, in the figure we have ∠A = ∠A′, ∠B = ∠B′, and ∠C = ∠C′. Let us prove
the converse of this.

Theorem If in two triangles ABC and A′B′C′ we have ∠A = ∠A′, ∠B = ∠B′, and
∠C = ∠C′, then these triangles are similar.
Proof Move triangle A′B′C′ so that vertices A and A′ coincide, and side A′B′ lies
along AB. By assumption ∠A = ∠A′, so that side A′C′ lies along AC (possibly
after reflection).

Next, scale triangle A′B′C′ to make side A′B′ the same as side AB. This scaled
triangle has one side and two angles the same as triangle ABC, so these two
triangles are congruent, that is, identical. Thus ΔA′B′C′ can be scaled to agree
with ΔABC, which shows that the triangles are similar, by definition.

To repeat, if two given triangles have the same angles, then they are similar –
one is a scaled version of the other. (Of course, knowing that two of the angles
are the same is sufficient, because the third angles would then also be the same.)
What is the relationship between the sides of similar triangles? If triangles
ABC and A′B′C′ are similar, then the second is a scaled version of the first. If k is
the scale factor, we have a′= ka, b′= kb, and c′= kc. We can write these equations
in the form

Thus, if two triangles are similar, the ratios of their three sides are the same, and
all three ratios equal k, the scale factor. Conversely, if two triangles have their
sides in equal ratio, then they are similar:

Theorem Let ABC and A′B′C′ be two triangles, with sides a, b, c and a′,b′,c′
respectively. If
then these triangles are similar.

Proof The proof is similar to the preceding theorem; see the diagram there. We
first move triangle A′B′C′ so that A′ coincides with A, and A′B′ lies along AB.
Next, we scale A′B′C′ by the (inverse) scale factor k1 = a/a′= b/b′= c/c′. Then the
scaled triangle has sides a, b, c. By the theorem on congruent triangles, this scaled
triangle is congruent to ΔABC. In other words, ΔA′B′C′ can be scaled to coincide
with ΔABC, so that the given triangles are similar.

Problem 5.14 Suppose ABC is a triangle with sides a = 5, b = 12, c = 13 [note that
52 +122 = 132]. The angles are ∠A = 22.6°, ∠B = 67.4°, ∠C = 90°. If ΔA′B′C′is
similar to ΔABC, with a′= 8, find b′, c′and ∠A′, ∠B′, ∠C′.

Equation 5.3 is sometimes written as

This is read as ‘a′is to b′is to c′, as a is to b is to c.’ To repeat, this just means that
a′/a = b′/b = c′/c, as in Eq. 5.3. Then the theorem about similar triangles can be
stated as:

This colon notation is seldom used today, but you may encounter it in old
books.
To summarize, if two triangles ΔABC and ΔA′B′C′are similar, meaning that
one is scaled-up (or -down) version of the other, then
1. their angles are the same:

2. their sides are in the same ratio

Conversely, if either of these conditions hold for two given triangles, then the
triangles are similar.

Triangulation
Triangulation is a method of finding a distance without actually measuring it.
The method uses similar triangles. The next problem explains the idea.

Problem 5.15 Given that ∠A = 75°, ∠B = 65°, and c = 8 cm, draw the triangle
ABC carefully, using a protractor. Now suppose ΔA′B′C′is similar to ΔABC with
c′= 53 m. Find a′and b′by first measuring a and b on your diagram, and using
proportionality.

This method is used by surveyors. For example, to determine the distance to an

off-shore island, sightings from two points A and B on shore could be made. If
∠A, ∠B, and c are measured, then the distance AC can be obtained by
triangulation. (In actuality, there is a formula in trigonometry that allows one to
calculate AC without actually drawing a similar triangle on paper. The formula is
called the Law of Sines – see Ch. 8.)

Binocular vision is another example of triangulation. Muscle tension in the

eyes signals information to the brain about the angle between the lines of sight
from each eye to the visual subject. Using this information, the brain can use
triangulation to estimate the distance to the observed object (presumably using
experience, not trigonometry). This method is not very accurate, except at close
range. Other sensory inputs used to assess distance include the size and clarity of
the object, and the muscle tension needed to focus on it.
The brain’s calculation of distance can be fooled by optical illusions. Look at
the pattern. Now cross your eyes and then let them slowly

uncross, until adjacent lines overlap. Hold this position for a few seconds. The
pattern should ‘jump’ off the page and appear closer than it really is. Your brain
has used the angle information to re-calculate the distance to the lines. (For
fascinating discussions of mental computations, read How the Mind Works by
Steven Pinker, W.W. Norton, 1997).

A digression on logic and proof

Euclid’s books strongly emphasized proofs of mathematical theorems. What is a
proof? Indeed what is a theorem? First, most theorems have a general scope. For
example, the theorem on the sum of interior angles of a triangle is valid for every
possible triangle.
Mathematical theorems are often stated in the form ‘If P then Q,’ or (what is
the same thing) ‘P implies Q.’ Sometimes ‘P implies Q‘ is written as P ⇒ Q
(which is read as ‘P implies Q‘). In a ‘P ⇒ Q‘ theorem, P is called the hypothesis
and Q is the conclusion. To prove the theorem, one first assumes the hypothesis.
Then one gives a list of statements, following logically from the hypothesis (often
using earlier theorems), leading up to the conclusion - ‘QED.’ Here is an
example. Theorem. If x is an odd number, then x2 is also an odd number. Can
you prove this theorem without peeking ahead? Try it.
Here’s the proof. Suppose x is an odd number. Then x = 2n +1 for some
integer n (by definition of odd numbers). Therefore x2 = 4n2 +4n +1 so that x2 =
2m +1 where m = 2n2 +2n [check this]. Thus x2 is an odd number.
This is a quite typical (if simple) proof. First state the hypothesis. If there are
technical terms in the hypothesis, use their definitions. Then come up with some
kind of mathematical argument, here algebra, to lead to the conclusion.
If P ⇒ Q is a given theorem, then the statement Q ⇒ P is called the converse
of the first statement. The converse of a true statement is not necessarily a true
statement. It is extremely important to keep this in mind. (My collection of
school texts contains examples of this error, so it isn’t surprising if lots of people
don’t understand it.) Here’s an example. If n is divisible by 6 then n is divisible by
3. This is a true statement (agree?) The converse statement, if n is divisible by 3
then n is divisible by 6, is false.
How does one prove that a statement is false? Let us be careful here. Most
interesting theorems are general statements. They apply, for example, to all right
triangles, or to all integers, etc. Now a general statement is falsified if a single
example exists that contradicts it. Such an example is called a counterexample.
For example, n = 9 is a counterexample to the statement that if n is divisible by 3
then it is divisible by 6.

Problem 5.16 A student wrote on an exam Is this statement

true in general? For what values of x is it true?

Problem 5.17 Write down the converse of Pythagoras’s theorem.

It so happens that the converse of Pythagoras’s theorem is true, but this is by

no means obvious at this stage. See Section 8.3 for the proof of the converse.

5.4 Solving a right triangle using trigonometry

Consider a right triangle ABC. Suppose two of the sides are given. We can then
calculate the third side using Pythagoras’s theorem. This requires solving the
equation c2 = a2 + b2 for the unknown variable. For example, if a and c are given,
then
Problem 5.18 Find the third side of a right triangle (with hypotenuse c), given
that (a) a = 2.8, b = 4.5; (b) a = 3, c = 7.

Next, how can we determine the three angles of a triangle, if we know its
three sides? This is by no means obvious! The method of drawing the triangle
and measuring its angles can only give limited accuracy. We need an exact
method. It so happens that by using a scientific calculator you can calculate the
angles of a triangle, within the calculator’s accuracy – usually about 10 digits.
Such a feat was humanly almost impossible before the advent of electronic
calculators and computers (but trigonometric tables could be used for 3 or 4-
digit accuracy).
To use a calculator for solving triangles (to ‘solve’ a triangle means to
calculate the unknown sides and angles), you have to know some trigonometry.
A full treatment of trigonometry occurs in Chapter 8; here we will discuss the
important special case of right triangles. Consider a right triangle, as shown. One
angle is labeled θ (Greek letter ‘theta’). The sides of the triangle are called:

Hypotenuse (the side opposite the 90° angle)

Opposite (the side opposite the given angle θ)
 Adjacent (the side adjacent the given angle θ)

These names are an aid to memory only; later we rewrite everything using
standard math symbols.
The trigonometric functions of θ are defined as

Here sin is an abbreviation for ‘sine,’ cos for ‘cosine,’ and tan for ‘tangent.’
(Also sin θ can be read as ‘sine θ,’ or ‘sine of θ.’) These are the functions shown
on the keys of your scientific calculator. You will need a scientific calculator for
the rest of this section. (See Sec. 5.7.) The word-definitions in Eq. 5.4 are an aid
to memory only. Most people seem to remember sin θ, cos θ, and tan θ in this
way. If instead we use symbols as shown here, the definition becomes (check this
carefully).
 For example, suppose x = 6, y = 2. Then (to two
decimals), and we have

Also cos θ = x/r = .95 and tan θ = y/x = .33. (Your calculator will give these values
with much greater precision.)
You may be wondering what happens if the triangle shown near Eq. 5.5 is not
of a specific size. All the sides x, y, and r depend on the size of this triangle. Also
sin θ = y/r depends on y and r. Can you explain this possible confusion? What
would happen if two different right triangles (both having angle θ in the position
shown) were used – would two different values for sin θ come out?
The answer is that any two such triangles are similar, so that the ratio y/r is
the same for both triangles. Therefore, in the definition of Eq. 5.5 it is not
necessary to specify any particular right triangle, and any one will do. The
argument applies also to cos θ and tan θ.

Problem 5.19 For a right triangle with y = 1.9 and r = 3.2 find x, sin θ, and cos θ.

In this problem, you calculated sin θ = .59 for the given right triangle. But
how do you calculate the angle θ itself? Your scientific calculator does this.

On most calculators ASIN (or SIN−1) is an ‘alternate’ key – you have to press
some other key first, then ASIN. Each calculator is different – read the manual;
usually there is a color match between the alternator key and ASIN. By the way,
the official name of ASIN is ‘Arc sine,’ and SIN−1 is ‘sine inverse.’ These terms are
explained in Chapter 8.
As an example, enter .59 on your calculator, then activate ASIN (or SIN−1).
You should get 36.16°. Thus, given that sin θ = .59, the angle θ is 36.16°. (If your
calculator does not give this result, check that it has been set in DEG mode.
There will be a key combination that accomplishes this; again, see your manual.)
As a check, now push the SIN key; you should get .59 again.
To repeat,

The same applies to any (acute) angle θ:

And similarly for cos θ = z versus Acos z = θ, and so on. The functions sin
and Asin are called mutual ‘inverse functions’ because of this relationship.
Chapters 7 and 8 discuss inverse functions in detail. As a mnemonic device you
can think ‘angle whose sine is’ whenever you encounter Asin. For example, Asin
.5 is the angle whose sine is .5. Find this angle using your calculator. Surprised?
I’ll explain this in a moment.

Problem 5.20 (This problem will familiarize you with using a scientific
calculator to solve triangles.) Consider a right triangle with sides x = 5, y = 3.
Find the angle θ to the accuracy of your calculator, first by using tan θ = y/x and
the Atan key. Next, as a check, use cos θ = x/r and the Acos key. Finally, use sin θ
= y/r and the Asin key. The results should be exactly the same (or almost exactly
the same).

Problem 5.21 Try to calculate Asin 2. You will probably get an error message.
Can you figure out why? Suggestion: review the definition of sin θ, Eq. 5.5, to see
why sin θ = 2 would be impossible.
 The result of problem 5.21 shows that, for any acute angle θ (i.e., for 0 ≤ θ ≤
90°) we have

(Other angles θ will be considered in Chapter 8.)

Problem 5.22 Solve the following right triangles, to 3 digits accuracy. Make a
quick sketch of each triangle, to check that your answers are reasonable. (a) y =
2.8, r = 7.5; (b) x = 3.6, θ = 52°.

What about the values of tan θ? Use the ATAN key to find the angle whose
tangent equals 50, then 500. What explains this? You find that

which means that

Look at the figure above Box 5.5 and try to explain this. What would tan 90° be?
Now we would have x = 0, so is undefined. And for angles close to
90°,the ratio y/x becomes very large.
This phenomenon is typical for expressions containing x in the denominator.
These expressions are undefined if x = 0, and they usually ‘blow up’ if x is very
small. For example, use your calculator to calculate for x = .1,.01,.001,…, etc.
The reciprocal of a very small number is a very large number (and vice versa).
More on this in Chapter 7.
Please keep in mind that as yet, we have not discussed methods for solving
triangles other than right triangles. This topic is studied in Chapter 8.

Problem 5.23 (This problem involves drawing a graph. Skip it if you forget how
to draw graphs.) Using your calculator, make a table of values of sin θ for θ = 0°,
30°, 60°,…, 180°. Use this data to plot a graph of sin θ for 0 ≤ θ ≤ 180°.
5.5 Isosceles triangles
A triangle having two equal sides is called an isosceles triangle (pronounced eye-
saw-sell-ease). It seems evident from the figure that an isosceles triangle must
also have two equal angles, ∠A = ∠B. We can prove this (and also the converse)
by using the theorem about congruent triangles.

Theorem Let ABC be an isosceles triangle, with a = b (see figure). Then ∠A = ∠B.
Conversely, if ABC is a triangle having ∠A = ∠B, then it is isosceles with a = b.

Proof For the first assertion, we assume that a = b. We construct the line through
vertex C and bisecting the base AB, as shown. (‘Bisect’ AB means to cut AB into
two equal parts). Let D be the point of bisection. The triangles ADC and BDC
have three equal sides: (1) AC = BC by hypothesis (remember AC = b and BC =
a); (2) AD = BD by construction; (3) CD is the same for both triangles. Therefore
ΔADC and ΔBDC are congruent. Since congruent triangles have equal angles, it
follows that A = B, as claimed.
To prove the converse, we now assume that A = B. We now construct line CD to
be perpendicular to the base AB (it happens also to be the bisector of AB, but that
is not proved yet). Triangles ADC and BDC have two equal angles (why?), and
one common side. Applying case 3 of the theorem on congruent triangles, we
conclude that ΔADC and ΔBDC are congruent. Since congruent triangles have
equal sides, it follows that AC = BC, i.e., a = b.

There, that was relatively painless, I hope. You have now mastered four of the
most important theorems of Euclidean geometry. These theorems have
countlessly many uses, some of which we will see soon.

Problem 5.24 (a) Recall from memory the statements of: (i) the theorem on
interior angles of a triangle; (ii) Pythagoras’s theorem; (iii) the theorem on
congruent triangles (3 cases); (iv) the theorem on isosceles triangles. If you have
to look them up first, then try closing the book and writing them down from
memory. (b) Next, recall the proofs. The key to most of these proofs is the
construction that is used, so try to remember this. Ten years from now, you may
have consciously forgotten these proofs, but they will still be in your brain
somewhere - provided you take the trouble to understand them now. In the
future you’ll know that once upon a time you did know these proofs.

Problem 5.25 Prove that an equilateral triangle is equiangular, and vice versa.
(Equilateral means that all three sides are equal; equiangular means that all three
angles are equal.) Suggestion: use the theorem on isosceles triangles. What is the
size of an angle in this case?

In the proof of the above theorem about isosceles triangles, we constructed a

line CD from vertex C to the line AB. In the first part of the proof, the point D
was specified as bisecting AB, while in the second part, CD was specified as being
perpendicular to AB. But are these not the same thing? The answer is Yes, for an
isosceles triangle, but No in general.

Problem 5.26 Let ΔABC be isosceles, with a = b, and let D be a point on the line
AB. Prove that D bisects AB if and only if CD is perpendicular to AB. Suggestion:
think about the preceding proof of the theorem on isosceles triangles.

The line that is perpendicular to and bisects a given line segment AB is called the
perpendicular bisector of AB. Note that if O is the center of a circle that passes
through points A and B, then O must lie on the perpendicular bisector of the line
segment AB. This is because OA = OB (both equal to the radius of the circle), so
triangle OAB is isosceles.

Problem 5.27 Let AB be a chord to a given circle. (A chord is a line joining two
points on a circle.) Show that a radius drawn perpendicular to AB necessarily
bisects AB.

Two special right triangles

(1) The 45° right triangle. Consider a right triangle having one of its angles 45°.
The other angle is also 45° (because the two angles add up to 90° in a right
triangle). Therefore the 45° right triangle is isosceles. We can use this fact to
calculate sin 45°. Let the sides of the triangle be x = 1, y = 1; then by
Pythagoras’s theorem. Therefore
 (I’ve heard that the Boeing 707 jet was named after sin 45°, but this may be
apocryphal.) Check these values with your calculator.
The 45° right triangle is worth remembering, because there aren’t many
angles θ for which we can calculate sin θ, cos θ and tan θ ‘by hand.’

(2) The 30° − 60° right triangle. Consider a right triangle having one of its angles
60°. Thus the remaining angle is 30°. Now this 30° − 60° right triangle is half of
an equilateral triangle. Why? If we let the side of the equilateral triangle be 2
units long, then the sides of our 30° − 60° triangle are x = 1, r = 2, [Be
sure you see this.] From this we obtain
 Again, this triangle is worth remembering. It will be used often in the chapter
on Trigonometry.

Problem 5.28 Find sin 30°, cos 30°, and tan 30°. Suggestion: use the 30° − 60°
figure above, and use with θ as the 30° angle in that figure.
(Alternatively, you could redraw the triangle lying on its side.)

The last problem indicates a general pattern:

To see this, note that in the figure sin(90° − θ) = opposite/hypotenuse = x/r = cos
θ, where ‘opposite’ now means the side opposite to the angle (90° − θ). The
second equation follows in the same way.
 Equation 5.8 explains the word ‘cosine.’ The cosine of an angle θ is equal to
the sine of its complement, 90° − θ, and vice versa.

Problem 5.29 What is the formula for tan(90° − θ)?

We end this section with a theorem about circles. Remember that a circle is
defined to be the set of all points which are a fixed distance r (the radius) from a
given point O (the center).

Theorem Let AB be a diameter of a given circle. Choose any point C on the circle,
and consider the triangle ABC. Then C = 90°.

Proof First, draw the line OC. (I don’t draw this here, but you should make your
own drawing.) Then ΔAOC is isosceles, because AO = CO (both being equal to
the radius r of the circle). Therefore A = ACO (mark this on your drawing).
Similarly, ΔBOC is isosceles, and therefore B = BCO. Therefore C = ACO + BCO
= A + B.
Now A + B + C = 180° (sum of interior angles). Substituting C = A + B, this
becomes 2( A + B)= 180° or A + B = 90°. Therefore C = 90°.
 There are literally hundreds of theorems in Euclidean geometry that could
now be gone into. Most of them are not sufficiently useful to be included in this
book. The theorems that were discussed in this section will be used often later in
the book, however.

5.6 Circles and arcs

Recall that a circle is defined to be the set of all points in the plane, equidistant
from a fixed point O. The fixed point is called the center of the circle, and the
distance to points P on the circle is called the radius.
An arc is any segment of a circle. For example, in the figure, PP′ is an arc of the
circle shown.
Problem 5.30 Under what conditions are two circles congruent? Similar? (Re-
read the definitions of congruence and similarity, if necessary.)

The circumference of a circle

By the circumference of a circle we mean the length of the circle. The formula is

Here π (Greek ‘pi’) is a certain real number. Your calculator will display π =
3.141592654 (to 9 decimals).
Equation 5.9 raises some interesting questions:
1. Equation 5.9 says that the circumference C is proportional to the radius r. Is
this obvious from basic principles?
2. How is π calculated? Could the Greeks calculate it?
3. Is π = 22/7?

Let’s start with the ridiculous. I’m always surprised by the number of people who
think they remember that π = 22/7. Especially because it’s wrong! π is not equal
to 22/7. Your calculator will show you that 22/7= 3.1428 … whereas π = 3.1415
…. Thus 22/7 is an approximate value for π, valid to three digits only. This
approximation is good enough for many applications, but is not accurate enough
for every situation in modern science and technology.
Next, consider question (1) – why is C ∝ r? This is a matter of scaling. If we
scale a circle, with scale factor k, we get another circle, with radius and
circumference both scaled by the same scale factor k:
 Let the original circle have radius 1 unit and circumference C0 units
(whatever C0 may be). If the new circle has radius r, the scaling factor must be k =
r, so the new circumference is

This says that C ∝ r. Remember: the circumference of a circle is necessarily

proportional to its radius, by the scaling argument. The difficult question is, what
is the constant of proportionality? Question 2, on the numerical value of π, is far
more substantial. The Greeks and other early mathematicians used many-sided
polygons to obtain good approximations for π. For example, consider a regular
hexagon inscribed in a circle. This hexagon is made up of six equilateral triangles
with sides equal to r, the radius of the circle. The length of the hexagon is 6r, and
therefore C > 6r. This shows that 2π > 6, or π> 3.

Similarly, by considering a square of side 2r lying entirely outside the circle,

we see that C < 8r, and this implies that π< 4. As a first approximation, we
therefore conclude that 3 < π < 4. More accurate approximation can be obtained
by refining the argument, using n-sided regular polygons. The Greek
mathematicians hoped that eventually someone could show that π is a rational
number (that is, a fraction of two integers), but in the 19th century this was
shown to be false: π is irrational. Methods of calculating π to any prescribed
accuracy are known; at latest count, computers (using new algorithms) are
reported to have calculated π to over 4 billion digits. Your calculator has been
pre-programmed with the numerical value of π to around 10 digits. By the way,
the number π shows up in countlessly many situations in higher math. The ‘bell
curve’ used in statistics is just one example.

Area of a circle
If A denotes the area inside a circle of radius r, we know immediately from
scaling that A ∝ r2. Thus there is a constant k such that A = kr2. In fact k = π, the
very same number that comes up in the circumference formula, Eq. 5.9.

The Greeks knew this formula, and were able to prove it from Eq. 5.8. (See
Section 5.8.)

Problem 5.31 What is the area of the annulus shown here? The inner and outer
radii are 8 mm and 25 mm.

Problem 5.32 What is the area of a half-circle? A quarter-circle? Can you

generalize further?
Radians
One complete revolution equals 360°, being equal to two straight angles of 180°
each. Degrees are the standard unit for measuring angles, at least in surveying,
geography, and other fields. However, a second unit of angle measure, which is
common in science and math, is the radian, defined by

The word ‘radians’ occurs here in parentheses to indicate that it is usually

omitted. If, for example, θ is a right angle, we can write either θ = 90° or θ = π/2.
The degree symbol must be used when we mean degrees; when no symbol for
units is used, the measure is assumed to be radians. Angles given in radians have
no units of measurement – they are just numbers.
Here are some basic angles expressed in both ways:

Degrees (Radians)
360° 2π
180° π
90° π/2
60° π/3
45° π/4
30° π/6

Check that these are correct.

Problem 5.33 (a) Express 1 (radian) in degrees. (b) Express 1° in radians.

Arcs of a circle
An arc s is a segment of a circle. An arc is said to subtend an angle θ at the
center of the circle, in the accompanying figure.
The length of an arc of a circle of radius r, subtending an angle θ at the center, is
given by

Here, the angle θ is given in radians. (The simplicity of this formula is the
main reason for using radians in math.)
‘
Equation 5.12 is directly related to the formula C = 2πr for the circumference
of a circle. Namely, if we take θ = 2π, then the arc s is the entire circle. In this case
Eq. 5.12 becomes s = rθ = 2πr, which is correct. Since arc length s is clearly
proportional to the angle θ, Eq. 5.12 must be correct in general.

Problem 5.34 (a) Find the length of a 60° arc in a unit circle. What is the ratio of
this length to the length of the 60° chord? (b) Same problem for a 1° arc and
chord.

Equation 5.12, s = rθ, has three variables, and can be solved for any one
variable in terms of the other two.

Problem 5.35 (a) Solve s = rθ for r, and for θ. (b) An arc of a circle of radius 16
cm has length 8 cm. Find the subtended angle θ, in radians and degrees.

Example: Eratosthenes’ method. The geometer Eratosthenes (276-194 BC)

estimated the radius of the earth in the following way. He took sightings of the
sun at noon, from two cities, Alexandria and Syrene, both in Egypt, Syrene being
approximately due south of Alexandria. The distance between these cities is
about 800 km. The angles between the vertical and the direction to the sun
differed by about 7° between the two cities. This implies that the arc AS on the
surface of the earth between A (Alexandria) and S (Syrene) subtends an angle of
7° at the center of the earth. Using the formula s = rθ with s = 800 km and θ = 7°
= 7 × 2π/360 gives the value r = s/θ = 6, 550 km. The correct value for the radius
of the earth is about 6,000 km, so Eratosthenes was pretty accurate.

(The diagram assumes that the sun is directly overhead of Alexandria, which
is not the case. However, the directions to the sun from the two cities are parallel,
which still implies that the arc AS subtends an angle of 7° at the center of the
earth, O.)

Area of a sector
The region enclosed by an arc and the sides of the subtended angle θ is called a
sector of the circle. The area of a sector is
(where θ is given in radians). This equation can be deduced from the formula A
= πr2 for the entire circle. Namely, if θ = 2π then Eq. 5.13 becomes A = πr2, which
is correct. Since A is proportional to θ, Equation 5.13 is therefore correct for any
angle θ. (See Problem 5.31.)

Problem 5.36 If the radius of a circle is given in cm, what are the units of arc
length s and area A in Eqs. 5.12 and 5.13?

Problem 5.37 Find the area of the window shown here. The top of the window is
not a semicircle, but an arc of a circle of radius .6 m. Suggestion: first try to figure
out the area of the top part; this requires some thought.

Angular velocity
We now consider circular motion, as in a rotating wheel. The rate of rotation is
called angular velocity, and is often denoted by ω (Greek ‘omega’). Specifically,
where θ = angle turned, and T = time taken. Notice the analogy with speed, S =
D/T , as discussed in Chapter 3.
Angular velocity can be measured in various units, including degrees per
second, radians per second, or RPM (revolutions per minute). For example,
suppose a shaft is rotating at 1200 RPM. How many radians per second is this?
Answer:

i.e., 125.7 radians per second. Note again that radians are dimensionless, and
need not be explicitly mentioned.

Problem 5.38 Two meshed gears have 52 and 20 teeth, respectively. If the larger
gear is rotating at 175 RPM, how fast is the smaller gear rotating? What is the
general formula?

Velocity of a rotating object

Consider an object rotating around a central point – for example, a stone on the
end of a string, or a satellite in orbit around the earth. If we know the angular
velocity ω, and the radius r of the orbit, we can calculate the speed, or velocity, of
the object. We begin with Eq. 5.12

where s is arc length, subtended by angle θ (in radians). Dividing both sides of
this equation by time T ,we obtain
Here, s/T is distance/time, or velocity v (we use the term velocity here, rather
than speed, to avoid confusion between using s to denote speed, or to denote arc
length). Likewise θ/T is angular velocity, ω. Therefore the above equation
becomes

What are the units of measurement in this equation? Suppose, for example,
that r = 36 cm (as in a typical bicycle wheel) and v = 20 km/hr, which equals 20 ×
105/60 cm/min (check this). Then

and this means 926 radians/min. Since 2π radians equals 1 revolution, we can
write

Clearly there are several possibilities for the units in Eq. 5.15, depending on what
units are used for each of the variables v, r and ω.

Problem 5.39 A satellite is located 35,800 km above the equator. At what speed
must it move so as to remain stationary relative to a location on the surface of the
earth? The radius of the earth is approximately 6000 km.

Problem 5.40 A world-class golfer achieves a clubhead speed of about 130 mph.
How long does it take him to complete one swing (ignoring the take-away)?
Assume that the radius of the circle made by the clubhead is 6 feet, and that the
clubhead traces out an arc of 540°.

Tangent line to a circle

Consider a circle with center O, and let A be a point on the circle. There is a
unique line L passing through the point A but otherwise remaining entirely
outside the circle. Any other line through A would cross the circle, forming a
chord. The line L is called the tangent line to the circle at point A.

Now, because the circle is symmetric with respect to the line OA, its tangent
line at A must also be symmetric in this way. Therefore the two angles made by
the radius OA and the line L must be equal to one another. This means that each
angle is 90°, a right angle. We state this fact as a theorem.

Theorem The tangent line to a circle at a point A is perpendicular to the radius

OA.
The property of touching the circle at a single point (A) is customarily taken
as the definition of the tangent line at A. However, a more modern view (as
encountered in Calculus) is that the most important feature of the tangent line to
a curve, at a point A, is that the tangent line has the same ‘direction’ as the curve,
at point A. The figure shows the tangent line to a certain curve, at a point A. Try
mentally drawing another tangent line, at B.
Notice that this line crosses the curve at a second point, which shows that the
definition of tangent line as the line that meets the curve at only one point, is not
a valid definition for curves generally. One needs to study Calculus (which was
unknown to the Greeks) to fully understand the concept of tangent lines. See Sec.
7.6 for further discussion.
Next suppose we draw a tangent line to a circle, from a point P outside the
circle, with OP = d. What is the length PA of the line from P to the point of
tangency A? To answer this, note that ΔOAP is a right triangle (by the above
theorem), so that by Pythagoras’s theorem

Problem 5.41 A tangent is drawn from a point P located 12 cmfromthe center O

of a circle of radius 9 cm. Solve the triangle AOP , where A is the point of
tangency.
Problem 5.42 The shafts of two pulleys, with radii 15 cm and 5 cm, are located
30 cm apart. Calculate the length of a belt that fits snugly around both pulleys.
Suggestion: First, draw a 10 cm radius circle inside the larger circle, and use the
method of Problem 5.41. (This problem is a skill tester!)

Circles determined by triangles

Given three points A,B, and C, can we draw a circle that passes through the three
points? Yes we can, unless the three points are co-linear (lie on a single line). The
circle is called the circumscribed circle for the triangle ABC. Here’s how it is
constructed.

If O denotes the center of the circle, then O must lie on the perpendicular
bisector of line AB. This is because, if the circle passes through A and B, its center
must be equidistant from these points. For the same reason, O must lie on the
perpendicular bisector of BC. Theinter-section of these two bisectors determines
O uniquely, and O is indeed the center of the circle through A, B, and C. (Point O
is also automatically on the perpendicular bisector of AC – why?)
Problem 5.43 Explain how to construct the inscribed circle to a given triangle,
this being defined as the circle that is tangent to all three sides. Suggestion: First
prove that points lying on the bisector of a given angle are equidistant from the
sides of the angle, and conversely. (Use the congruent triangles theorem.)

Ruler and compass constructions

For historical reasons we next discuss some examples of ruler and compass
constructions. To the Greek mathematicians, these constructions probably
indicated that the concepts of Euclidean geometry were not mere abstractions,
but figures that could be realized in practice using very simple drafting tools, a
straightedge (ruler) and a compass.
A compass, of course, is a tool for drawing circles. The main property of
circles used here is that two circles intersect in at most two points. We will
assume that this is true, without proof (a proof is given in Chapter 6).
 Traditionally, ruler-and-compass constructions have certain restrictions. The
ruler can only be used for drawing straight lines connecting two given points.
You cannot use the scale on the ruler, or make new markings on it. The compass
can only be used to draw arcs of circles, with radius set equal to the distance
between two given points. (However, points can be drawn anywhere. This means
that you can just draw a line, or an arc, when needed.)
Two non-parallel lines intersect in a single point. Two non-tangent circles
either intersect in two points, or none.

Construction 1: an equilateral triangle Given a line AB, we wish to construct an

equilateral triangle ABC. To do so, first draw a circle centered at A, with radius r
= AB. Then draw a second circle at B,withthesame radius. These circles intersect
at point C (there are two such points). Then AB = AC (both equal the radius of
the first circle). Also, AB = BC (second circle). Thus ΔABC is equilateral.
Constructing an equilateral triangle is the way to construct a 60° angle, also.

Construction 2: bisecting an angle To bisect a given angle A, first draw an arc

centered at A, cutting the two edges of the angle at points B and C. Next, draw
arcs centered at B and C, with the same radius, meeting at point D. Then the line
AD bisects angle A. To see this, note that triangles ABD and ACD are congruent,
because they have three sides equal.

Construction 3: perpendicular bisector of a given line segment

Let AB be a given line segment. To construct the perpendicular bisector of AB,
construct the vertices C and C′of two equilateral triangles having base AB. Then
CC′ is the perpendicular bisector of AB, by congruent triangles ACD and AC′ D.

Other possible ruler and compass constructions include:

drawing a line perpendicular to a given line, through a given point

drawing a line parallel to a given line, through a given point, not on the
line
drawing a line tangent to a given circle, through a given point outside the
circle
dividing a given line segment into n equal subsegments

Geometry texts give the details, if you’re interested.

5.7 The Scientific calculator

Scientific calculators come in a wide variety of models and prices. Topof-the-line
calculators have extra capabilities, and are used by scientists and engineers. Even
the least expensive scientific calculators have amazing abilities, however. They
are so cheap that it is worth buying one as a first learning experience, even if you
expect to buy a more advanced type later. First, check that the calculator you buy
has all the features in the left-hand list.

Basic Scientific Calculator Advanced Calculator - Extra Features

Scientific notation option Graphics
Brackets Programmability
Multiple storage registers Complex numbers
Trigonometric functions Unit conversion
Logarithmic and exponential functions
Statistical calculations

You can quickly learn how to use your calculator by reading through the owner’s
manual and trying each feature. This book provides plenty of practice in using a
scientific calculator. Only a few pointers will be discussed here. Try them out.

Scientific notation. You can switch between regular display of numbers, and
scientific notation. The latter may be shown in Enotation, for example

Most scientific calculators display about 10 digits (plus the exponent). They
actually perform calculations to one or two more digits, resulting in high
accuracy. Nevertheless, minor errors in the last digit displayed inevitably do
occur sometimes, but this is virtually never a problem.
It is possible to control or ‘fix’ the number of digits displayed, and this is
definitely worth doing. The calculator always calculates in its maximum
precision mode, but then rounds off to your fixed display accuracy.

Brackets. Complex calculations are facilitated by using brackets. The calculator

may feature automatic closure of any open brackets. For example, entering ‘3 −
(7 ÷ (8 − 5 = ’ will actually compute

You need to use this carefully, to make sure that it’s what you want to do.
By the way, some scientific calculators use RPN – Reverse Polish Notation.
This is a method of calculation that dispenses completely with all brackets. I find
this vastly easier and more reliable than standard bracket-based calculation. But
it does take some getting used to.

Memory. The scientific calculator provides several memory registers, which are
useful in various ways. First, a complicated calculation can be broken into
simpler parts, with results stored in memory and recalled later. Also, you may
sometimes need to use the same value in several calculations. If so, you can store
this value in memory and keep using it. Most calculators will preserve values in
memory even when the calculator is turned off, which can be useful. (Storing a
new value in memory will erase the old one.)

Trigonometric functions. All the trig functions (sin, cos, tan) and their inverses
are built into the calculator. You have to ensure that the calculator is in the
proper mode, whether degrees or radians. (A third mode, grads, may also be
provided; this is used by engineers.)

Logarithmic and exponential functions. These are also built in; see Chapter 9.

Powers. A very useful feature. It instantly calculates yx for any values of x, y (but y
must be non-negative for most cases). For example, try calculating 73 = 343. The
sequence on my TI calculator is (There are also special keys for x3
and but not for powers higher than 3. To calculate you would
key in

Special keys. Try the keys for x2, 1/x, and π. To check your technique,
calculate π × (1/π) using these keys. The sequence is

Statistics. See Chapter 11.

Graphics. A graphic calculator draws the graphs of functions that you program in
(see Chapter 7). This could help a lot in math tests.

Other features. Your handbook will describe any other features of the calculator.
Enjoy your calculator – it’s one of the wonders of technology.

5.8 Area of a circle

For you readers who want to understand everything, here’s how to prove the
formula A = πr2 for the area of a circle. The circumference formula, C = 2πr is
our starting point. We know from the scaling argument that A = kr2 for some
constant k.
We consider two concentric circles, one of radius r and the other of radius r
+h,where h is a small number. The area of the thin strip is then
 Now, if h is very small, this strip can be straightened out into an approximate
rectangle, with width h and length C = 2πr (the circumference of the circle).
Thus the strip has area 2πrh, approximately. Equating this with the above
expression, we have

Divide this equation through by h, to obtain

Now, this last approximate equation holds no matter how small h is; in fact, the
smaller we take h the closer the approximation becomes. If we take h negligibly
small, we get

which proves that k = π. Therefore A = πr2.

This idea of using a small number h, which is eventually allowed to become
infinitesimally small, is the essence of Calculus.

5.9 Review problems

1. A parallelogram is defined as a four-sided figure (quadrilateral) in which the
opposite pairs of sides are parallel. Prove that the opposite sides of a
parallelogram are equal. Suggestion: draw a diagonal and use congruent
triangles.
2. Prove the converse to the theorem of Problem 1: if the opposite sides of a
quadrilateral are equal, then it is a parallelogram.

3. Express these angles in radians, and sketch them: (a) 3π/4; (b) 2π/3.
4. Three positive numbers a, b, c are given. What is the easiest way to tell whether
these numbers can form the three sides of a triangle? Explain. (Think of
drawing it.)
5. The figure depicts a 70° sector of a circle with radius 20 cm. Find (a) the arc
length s; (b) the area of the sector.

6. A parallelogram has sides of 2 m and 4 m, and one interior angle of 60°. Find
its area.

7. Find the area of an equilateral triangle of side s.

8. Find the area of an isosceles triangle having two sides of length s, and two
angles of size θ. Show that the answer to Problem 7 is a special case of this
result.
9. Two pulleys, of diameter 6 in and 2 in are joined by a belt. If the larger pulley
is rotating at 100 RPM, how fast is the smaller pulley rotating, and why? What
is the general formula?
10. Triangles ABC and A′B′C′ are similar. If AB = 6 and AC = 10, find the ratio A
′B′/A′C′. What is the general rule?
11. Two triangles are similar, with scale factor k. How are the perimeters related?
The areas? (The perimeter is the total length of the sides of the triangle.)

12. Suppose an arc of a circle is given, but the center is not known. Show how to
 locate the center. Suggestion: first show how to draw a line that must pass
through the center.

13. Find the angles in a triangle with sides 5, 12 and 13.

14. Solve the right triangle ABC (with C = 90°), given a = 8 and A = 38°. Check
by drawing a rough sketch.

15. Two radii OA and OB of a circle of radius 6 cm make an angle of 100°. Find
the length of the chord AB, and of the arc AB.

16. What angle above the horizon is the sun at mid-day on Dec. 21, at New York
City? Relevant information: On Dec 21 the earth’s axis is tilted from the sun
the maximum amount, NYC is located at 41° North latitude,
approximately. Suggestion: one way to think about this problem is to first
realize that, if you were located on the Arctic circle, at latitude
the sun would be right on the horizon on Dec. 21. Work
back from there.

17. Show how to calculate the length of the crossover tangent to two circles.
Suggestion: Use a method similar to that of Problem 5.42.

18. A rectangle R has the property that when cut into two equal smaller
rectangles (the cut being parallel to the short sides), the new rectangles are
similar to the original rectangle. What is the ratio of the longer to the shorter
side of R?

19. Rectangle R has length l and width w. If a square of side w is cut off, the
remaining rectangle is similar to the original rectangle R. Find the ratio l/w.
(This value of l/w is called the golden ratio. It is thought to determine the
most aesthetically pleasing shape for rectangles.)

20. Prove that the diagonals of a rhombus are perpendicular bisectors of each
other. (A rhombus is a parallelogram having four equal sides.) Also, state and
prove the converse.
21. The distance from the earth to the moon is about 384,000 km. The moon’s
disk subtends an angle of approximately 0.5° at the earth. Find the diameter of
the moon. Also, compare the volumes of the two bodies (the earth’s diameter
is about 12,000 km).

Solution 5.1 L1 and L3 are parallel to each other (or possibly coincident).

Solution 5.2 You should notice that the sum of the other two angles is always
90°, at least experimentally. Why this is true will be explained in Sec. 5.2.

Solution 5.4 No. Suppose L is a line drawn on the floor, oriented E-W. And
suppose Lis a line on the ceiling, oriented N-S. These lines will never intersect –
they ‘miss each other.’ But they certainly aren’t parallel. You can use your two
index fingers to illustrate the same idea – try it. Also, get two sticks of wood.
Hold one of them at some angle, and ask a friend to hold the other stick parallel
to yours, at a distance. He or she will have no difficulty following instructions;
there is a unique direction determined by your stick. But Euclid’s notion of non-
intersecting lines is not relevant here. Two lines in space are parallel if they ‘have
the same direction.’ The exact definition of ‘direction’ in space will be discussed
in Part 2.

Solution 5.5 By the theorem on interior angles of a triangle, we have ∠A + ∠B +

∠C = 180°. But ∠C = 90° by assumption. Therefore ∠A + ∠B = 90°, i.e., these
angles are complementary.

Solution 5.6 We have α+β +γ = 180°. Also γ + ∠DCB = 180° (a straight angle).
Therefore α + β + γ = γ + ∠DCB. Subtract γ from both sides, to conclude that α
+ β = ∠DCB.
Solution 5.7 The construction required is a diagonal of the quadrilateral, which
cuts it into two triangles. The sum of all the interior angles of both triangles is the
same as the sum of the interior angles of the quadrilateral [check this]. Therefore
the quadrilateral angles sum up to 180° + 180° = 360°.

Solution 5.8 (a) From a figure, you see that d2 = (6 cm)2 +(6 cm)2 = 72 cm2, so
that (b) Similarly, here d2 = (3 m)2 + (4 m)2 = 25 m2,
so that d = 5 m. This is the famous ‘3-4-5’ triangle, with 32 +42 = 52.

Solution 5.9 Note that the side of the small square is b −a. Equating areas
therefore gives c2 = (b − a)2 +4 × 1/2ab = b2 − 2ab + a2 +2ab = a2 + b2.
Solution 5.11 (a) First, draw side AB = 7 cm, for example. Next, place the
compass point at A and draw an arc of radius 5 cm. Finally, draw an arc of radius
3 cm, centered at B. These arcs intersect at point C; draw sides AC and AB. (This
triangle would be congruent, by reflection, to the triangle having C below AB.)
(b) You can’t complete this triangle because the two short sides aren’t long
enough.

Solution 5.12 (a) Following the given information you obtain the three vertices
A, B, and C. Joining B to C completes the triangle. (b) In this case, you wind up
with lines starting at A and at B. Since these lines are not parallel, they meet at a
point C, which completes the triangle.
Solution 5.13 No. Consider the two triangles shown. If BC is parallel to B′ C′,
then ∠B = ∠B′and ∠C = ∠C′(corresponding angles). But these triangles are not
identical (congruent).
Solution 5.14 The scale factor is 8/5 = 1.6. Therefore b′= 19.2 and c′= 20.8. The
angles are all the same: ∠A′= 22.6°,etc.

Solution 5.15 You should get a = 10.9 cm, b = 10.2 cm (approximately). The
scale factor between the two triangles is k = c′ /c = 662.5, and therefore a′= ka =
72.2m and b′= kb = 67.6m.

Solution 5.16 The statement is not generally true; x = 1 is one counterexample (

is not equal to 2). To find which x the statement is true for, square both sides,
obtaining 1 + x2 = (1+ x)2 = 1+2x + x2. This requires 2x = 0, i.e., x = 0. The
statement is true if and only if x = 0.

Solution 5.17 The converse is: If ABC is any triangle for which a2+b2 = c2, then
∠C = 90°.

Solution 5.18

Solution 5.19 You get (from Pythagoras: x2 + y2 = r2)or x =

2.57. Also sin θ = x/r = .59 and cos θ = y/r = .80.
Solution 5.20 θ = 30.963756532°. (Your calculator may give one or two fewer
digits, but all three calculations should agree except perhaps in the final digit.)

Solution 5.21 Equation 5.5 says sin θ = y/r. Now r > y because r =
Therefore sin θ< 1, so that sin θ = 2 is impossible.
Solution 5.22 (a) x = 6.96, θ = 21.9°, φ = 68.1° (where φ, called ‘phi,’ is the
complementary angle to θ); (b) y = 4.61, r = 5.85, φ = 38°. (Did you make the
sketches?)
Solution 5.23 The graph is a smooth curve starting at 0, rising to 1, then falling
back to 0 at θ = 180°. This is the first cycle of the famous ‘sine curve,’ (see
Chapter 8). If your calculator has graphing capability, you can even get it to draw
the graph of the sine curve.

Solution 5.25 First, let ΔABC be equilateral. Then a = b,so A = B by the isosceles
triangle theorem. Similarly, a = c, so A = C. Thus ΔABC is equiangular.
Conversely, if ΔABC is equiangular, then A = B,so that a = b. Also A = C, so a =
c. Thus ABC is equilateral. The size of each angle in an equilateral triangle is 60°
because the three angles must sum to 180°.

Solution 5.26 Proof. First assume that D bisects B. Then ΔADC and ΔBDC are
congruent (three sides equal). Hence the angles at D are equal, so both must be
right angles. This proves that CD is perpendicular to AB. For the converse,
assume that CD is perpendicular to AB. Then ΔADC and ΔBDC have two equal
angles and one side in common, so they are congruent. Therefore AD = BD, i.e.
D bisects AB.

Solution 5.27 First make a drawing, and label the center of the circle, as O, say.
Then ΔABO is isosceles, with a = b (both being equal to the radius of the circle).
Therefore, by Problem 5.26 on perpendicular bisectors of isosceles triangles, the
perpendicular OD bisects AB.
Solution 5.28 Either diagram shows that sin 30° = ½,

Solution 5.29 Using the same figure, we have tan(90° − θ) = opposite/adjacent =

x/y = 1/(y/x)= 1/ tan θ. (1/ tan θ is sometimes called cotan θ, ‘cotangent of θ.’)
Solution 5.30 Two circles are congruent if and only if they have equal radii. Any
two circles are similar to one another, with the ratio of their radii as the scale
factor.

Solution 5.31 To get the area of the annulus we subtract the area of the inner
circle from the area of the outer circle. Thus A = π(25)2 − π(8)2 = 1, 762 mm2, or
17.62 cm2 (since 1 cm2 = (10 mm)2 = 100 mm2).

Solution 5.32 Since the area of a circle is πr2, the area of a half-circle must be
(½)πr2. Similarly, a quarter-circle has area (1/4)πr2. The general result is that, if q
is a number between 0 and 1, then the area of a ‘q-circle’ equals qπr2. (See Eq.
5.13.)

Solution 5.33 (a) Using Eq. 5.11 we obtain 1 = 360°/2π = 57.3°. (b) Similarly, 1°
= 2π/360 = .0175.

Solution 5.34

(a) s = θ = π/3 = 1.05. The chord has length 1 (equilateral triangle), so the ratio of
arc to chord is 1.05. (b) s = θ = 2π/360 = .01745329 (see Prob. 5.33). For the
chord, we use some trigonometry: chord length = 2y = 2sin½° = .017453070. The
ratio is 1.000013. (The moral is that, for small angles, arc length and chord length
are nearly equal. The reason is that, for a small angle, the arc hardly curves at all.
It’s nearly straight.) The figure is exaggerated for easier viewing.
Solution 5.35 (a) r = s/θ, and θ = s/r; (b) θ = .5(radians) = .5 × 360°/2π = 28.6°.
Solution 5.36 Arc length s = rθ is in cm (because θ is in radians, which has no
units). Area A = 1/2r2θ is in cm2. Note that θ is always assumed to be given in
radians in both these equations. If not, it must first be transformed to radians.
Solution 5.37 The figure shows the whole semicircle, with center O. The line AC
is the top of the lower part of the window, so that AC = 1 m. Thus BC = .5 m, and
OC = .6 m (radius of the circle). From Pythagoras we get OB = .33 m. If α = BOC
we have sin α = BC/OC. Therefore α = .985. [Set your calculator to radians before
calculating α.] The area of the sector subtended by AOC = 2α is therefore A =
(½)r2(2α)= .355 m2. To get the area of the top part of the window, we have to
subtract the area of triangle AOC, 1/2bh = ½ × 1 × .33 = .165 m2. Hence Atop = .19
m2. The area of the window is 1.3+ .19 = 1.49 m2. (Though this solution may
seem a bit complicated, please note that it uses only basic geometry as discussed
in this chapter.)

Solution 5.38 One revolution of the large gear moves 52 teeth past the point of
contact. Hence 52 teeth of the small gear also move past this point. This results in
52 ÷ 20 = 2.6 revolutions of the smaller gear. Therefore the angular velocity of
the small gear is 2.6 × 175 = 455 RPM. To obtain the general formula, let n1,n2
denote the number of teeth on gear 1 and 2 respectively. Also let ω1,ω2 be the
angular velocities of the two gears. Then the number of teeth passing the contact
point per minute equals n1ω1 for gear 1 (because 1 revolution passes n1 teeth, so
ω1 revolutions pass n1ω1 teeth), and n2ω2 for gear 2. For meshed gears, these are
equal so that

This can also be written in the form

In the example, we had ω2 = 52 × 175/20 = 455 RPM.

Solution 5.39 The angular velocity of the earth’s rotation is ω = 1 rev/day =

2π/day. The radius of the satellite’s orbit is r = 41, 800 km. Therefore v = rω = 41,
800km × 2π/day = 262, 600 km/day, or 10,900 km/hr.
Solution 5.40 We have v = 130 mph, r = 6 ft. Now v = rω implies

This says angular velocity is 31.8 radians/sec, which is 31.8/2π,or 5 revolutions

per second. One revolution takes about 1/5 sec, so 1.5 revolutions takes about 0.3
sec. (This value underestimates the time for one swing, because the clubhead
does not move at 130 mph for the entire swing. A reasonable guess would be that
the average clubhead speed for the entire swing is ½ × 130 = 65 mph. This would
imply that the swing lasts about 0.6 sec.)

Solution 5.41 AOP is a right triangle with sides 12 cm, 9 cm and 7.94 cm. Thus
∠P = Asin (9/12) = 48.6° and ∠O = 41.4°.

Solution 5.42 Solving ΔAOP as before, we obtain

and ∠O = A cos(10/30) = 70.5°. Next, PABC
is a rectangle, because ∠A = ∠B = ∠C = 90°. Therefore the total length of the
two belt segments between the pulleys is 2AP = 56.6 cm. The part of the belt on
the large pulley subtends an angle of 360° − (2 × 70.5°) = 219° at the center, so
this part has length s = rθ = 15 cm × 219° × 2π/360° = 57.3 cm. Similarly, the part
of the belt on the small pulley has length s = 5 cm × 141° × 2π/360° = 12.3 cm.
The total length of the belt is therefore 126.2 cm. [It’s always worth doing a
rough mental calculation to check that your computed answers are more or less
correct. For example, the piece of the belt on the large pulley goes about 60% of
the way round, and the circumference is about 6 × 15 = 90 cm, so this piece has
length about 54 cm. It checks. Try this for the other parts.]
Solution 5.43 Following the suggestion, let line AP bisect angle A. Let PQ be the
perpendicular from P to one side of the angle, and PR to the other side. Then the
triangles AQP and ARP are congruent by case 3 of the congruent triangles
theorem (one side, AP , and two angles being equal). Therefore PQ = PR, i.e., P is
equidistant from the two sides of the angle A. The converse is proved quite
similarly.
Now the center O of the inscribed circle must be equidistant from the three sides
of the given triangle ABC. If we draw the angle bisectors to angles A and B (say),
the point O of intersection is equidistant from lines AB, AC, and BC. Hence O is
the center of the inscribed circle, as desired. (Point O also lies on the bisector of
angle C.)
Chapter 6
Analytic Geometry

6.1 The rectangular coordinate system

The great French mathematician René Descartes (1596-1650) introduced the
idea of a rectangular coordinate system into geometry. This invention achieved a
unification of geometry and algebra. Geometric objects such as lines, circles, and
ellipses had algebraic equations that reflected their properties. This
amalgamation of geometry and algebra is called analytic geometry.
A rectangular coordinate system in the plane consists, first of two real-
number axes, intersecting at right angles.

The point of intersection is the zero point on each axis. This point is called
the origin of the coordinate system. The two axes have identical number scales,
meaning that the unit distance on each axis is the same. These coordinate axes
are traditionally called the X-axis and the Y-axis, although any other letters
could also be used. Also traditionally, the X-axis points to the right, so that the x-
values increase to the right. Similarly, the Y-axis points upwards. Again, these
conditions may be changed in some applications.
Once the coordinate axes are specified, any point P in the plane is assigned
coordinates (x, y) determined as follows:

A vertical line (i.e. parallel to the Y-axis) meets the X-axis at position (real
number) x. Likewise, a horizontal line (parallel to the X-axis) meets the Y-axis at
position y. In the above figure we have x = 2.3 and y = 1.4, so that the coordinates
of P are (2.3, 1.4).
In this example, both x and y are positive numbers. Generally x and y can be
any real numbers, positive, negative, or zero. The two coordinate axes divide the
whole plane into four regions, called quadrants, ineachof which the signs of x
and y are given (+ or −). These quadrants are referred to as the first quadrant, the
second quadrant, and so on. Just remember that quadrant I has both coordinates
positive, and the other quadrants proceed counterclockwise.
Quadrant Sign of (x, y)
I (+, +)
II (−, +)
III (−, −)
IV (+, −)
Problem 6.1 (a) Which quadrant is each point in: (−3, −6), (2, −9)? (b) What is
special about the coordinates of a point on the X-axis? On the Y-axis? The
origin?

The terms ‘abscissa’ and ‘ordinate’ are sometimes used to refer to the x
(horizontal) and y (vertical) components of the point (x, y), but these terms are
used infrequently today.
It is important to remember that any point lying on the X-axis has y = 0. For
example, the coordinates of the point located at x = 5 on the X-axis are (5,0).
Similarly, points lying on the Y-axis have coordinates (0,y).
The distance formula
Now consider two points Pi having coordinates (xi,yi). Then x2 − x1 is the
distance between the numbers x1 and x2 on the X-axis. (See Chapter 1 if this is
not familiar to you.)

Similarly, y2 − y1 is the distance between y1 and y2. These distances form the
sides of a right triangle, of which P1P2 is the hypotenuse. If d denotes the distance
between P1 and P2, we can use Pythagoras’s theorem to conclude that

This is the distance formula for plane analytic geometry.

The distance formula is valid for all points P1,P2, not just points in the first
quadrant (as drawn in the figure). This is true because x2 − x1 is always the
distance from x1 to x2, regardless of whether x1 and x2 are positive, negative, or
zero. (The value of x2 − x1 will be negative if point P1 is to the right of P2. The
distance formula 6.1 is still valid in this case. Note that the distance d in Eq. 6.1 is
always ≥ 0. Also, for two points (xi, 0) lying on the X-axis, we obtain
which is the unsigned distance between these
points.)

Problem 6.2 Find the distance between the points (−3, 1) and (1, −3). Make a
quick sketch and check.

The midpoint of a line segment

Consider the line segment P1P2,where Pi = (xi,yi) for i = 1, 2. Define

Then (read as ‘x bar’) is the midpoint of the interval from x1 to x2 on the X-

axis. To see this, note that so that is
half the distance from x1 to x2. Similarly, y is half way between y1 and y2.
It follows that the point is the midpoint of the line segment P1P2.
To explain this, note that the two small triangles in the figure are congruent, so
that
For example, if P1 = (−1, 3) and P2 = (3, −5), then the midpoint is at (1, −1).
Make a sketch to show this.
Equation of a circle
Consider a circle, with center at (x0,y0) and radius r. By definition, all points on
this circle are at a distance r from the center (x0,y0). We now find an equation for
this circle. Follow the next argument carefully, as it is typical in analytic
geometry.

Choose an arbitrary, representative point (x, y) on the circle. Then the

distance between this point and the center (x0,y0) equals r, the radius of the circle.
Thus by Eq. 6.1

Squaring both sides of this equation, we obtain

You are now beginning to see the power of Descartes’ innovation. A

geometric object in the XY-plane (here, a circle) is represented by an algebraic
equation in two variables, x and y. Any point P on the circle has coordinates (x,
y) that satisfy this equation. Conversely, any pair of numbers x, y that satisfy the
equation (6.2) are the coordinates of a point P on the circle. There is a one-to-
one correspondence between points (x, y)on the circle and solutions (x, y) to the
equation. Geometry (the circle) has become algebra (the equation).
We will see later that a similar correspondence holds between algebraic
equations and other geometric objects, such as straight lines, ellipses, and
parabolas.
For the circle, if weare givenanequationwrittenin the form of Eq. 6.2, we
immediately recognize it as the equation of a specific circle, with center (x0,y0)
and radius r. For example,

is the equation of a circle with center at (3, −2) and radius 4. (Where does the −2
come from? Remember that y +2 = y − (−2), so y0 in Eq. 6.2 has the value −2.)

Problem 6.3 Identify and sketch the circle whose equation is x2+(y+3)2 = 9.
Check by substitution that the origin is a point on this circle. What is the lowest
point on the circle? The rightmost point?

Consider again the example (x − 3)2 +(y +2)2 = 16. Let us expand and
simplify:
or

Any circle equation, as in Eq. 6.2, can be expanded out in this way. The
general result will be of the form

where A, B, C are certain constants, depending on the circle.

Now we ask the reverse question: given an equation like (6.3), can we identify
the circle that the equation corresponds to? The answer is yes (but not always);
we use the method of completing the square. An example:
or
or
or

(Read ‘Completing the square’ in Sec. 4.5 if this isn’t clear.) Thus the given
equation is the equation of the circle with center (4, −½) and radius .
This calculation can be performed for any equation of the form of Eq. 6.3.
However, something may go wrong. Look at the term 69/4 on the right side of
the above example. What if this had turned out to be a negative number? Then
the radius r would not exist, because r2 must be positive.
What does Eq. 6.3 represent geometrically, in this situation? Nothing! That’s
right, nothing. The final equation (x − x0)2 +(y − y0)2 = −q2 cannot be satisfied for
any numbers x, y, which means that the given equation also has no solutions in
real numbers x, y. This is a perfectly reasonable outcome, for any quadratic
equation (recall Sec. 4.5). A simple, but typical example is x2 + y2 = −1; no real
numbers x, y can possibly satisfy this equation.
Can we tell by inspection whether a given equation of the form 6.3 is the
equation of a circle or not? Problem 6.5 goes into that question.

Problem 6.4 Complete the squares, to identify the circles (if any): (a) x2 + y2 +6x
− 2y +4 = 0; (b) x2 + y2 +6x − 2y +12 = 0.

Problem 6.5 Write down the relationship between the coefficients A, B, and C in
Eq. 6.3 that will ensure that this is actually the equation of a circle. Suggestion:
complete squares, as usual. What happens if you get r2 = 0?
The graph of an equation
We have seen that an equation such as (x − 2)2 +(y +1)2 = 9 corresponds to a
certain circle in the XY-plane. This relationship can be expressed in either of the
following ways:
(a) The equation of the circle with center (2, −1) and radius 3 is (x − 2)2 + (y +1)2
= 9, or
(b) The graph of the equation (x − 2)2 +(y +1)2 = 9 is the circle with center (2, −1)
and radius 3.

This connection between an equation and its graph is characteristic of

analytic geometry.

Definition The graph of a given equation in x, y is the set of all points (x, y) in
the XY-plane, such that x, y satisfy the equation. (‘Satisfy’ means that when you
plug the particular values of x and y into the equation, it becomes a numerical
equality.)

For example, (2,2) is one point on the graph of (x − 2)2 +(y +1)2 = 9, because
(2 − 2)2 +(2+ 1)2 does equal 9. Likewise, is another point on
the graph, as you could check for yourself.
You may be familiar with the process of ‘plotting’ a graph by first making a
table of x, y values and then plotting and joining up these points. Well, this is one
method of sketching the graph, but it is cumbersome and error-prone. Some
equations have graphs that are common geometric objects – lines, circles, and so
on. By recognizing these equations one can identify the graph directly, without
going through the point-by-point plotting process. However, plotting a few
judiciously chosen points whose coordinates are calculated from the equation, is
often useful. Examples occur throughout this and later chapters.
The important thing to keep in mind from now on is that the graph of an
equation is the set of all points whose coordinates x, y satisfy the equation
numerically.

6.2 Straight lines

The slope of a line

Let L be a (straight) line in the XY-plane. Let Pi(xi,yi) be two points on L (i =

1, 2). We write

(Δ is the Greek capital letter ‘delta.’ This letter is often used to denote
changes, or differences. Note that Δx, ‘delta x,’ is a single number, and not the
product of Δ and x).

Definition The slope of the line L is defined by

In figure (a) above Δx and Δy are both positive, so the slope m is positive.
Lines that slope up to the right have positive slope. In figure (b), Δy is negative
(because y2 < y1) while Δx is positive, so m is negative. Lines that slope down to
the right have negative slope.
 What about a horizontal line? In this case we have Δy = 0, so m = 0.
Horizontal lines have zero slope. Finally, what if the line L is vertical? Now Δx =
0, so that m in Eq. 6.4 is undefined. In summary:

line, L slope, m
slopes upwards m> 0
slopes downwards m< 0
Horizontal m = 0
Vertical m is undefined

(Sometimes one says that a vertical line has ‘infinite slope,’ but this usage is best
avoided, except in a colorful way. However, it is true that a line that is nearly
vertical has a very large slope, either positive or negative.)

Problem 6.6 Consider various lines L, all going through the point (1, 0). Find the
point P2 = (x2,y2) on L, given that x2 = 3 and (a) m = 1; (b) m = 0; (c) m = −2; (d)
m = 8. Sketch these 4 lines in a single coordinate system. Suggestion: show first
that y2 = 2m in this example.

Next, consider a line L with positive slope m. If θ is the angle between the
positive X-axis and line L, we have

This can be seen from the next figure, in which the Δx, Δy triangle is drawn with
Δx being along the X-axis. Recall that tan θ = opposite/adjacent, so that tan θ =
Δy/Δx = m.
 It happens that Eq. 6.5 also applies to negative slopes. In this case, θ is an
angle down from the positive X-axis. Such angles θ are considered to be negative,
and tan θ is also negative. Try finding Atan (−1) on your calculator. You will get
−45°. Negative angles are studied in Chapter 8.

Problem 6.7 (a) A line L has slope m = 1. What is the angle that L makes with the
positive X-axis? (b) Same, with m = 10 (use your calculator for part b). (c) Same
with

Example The gradient,or grade of a stretch of road is sometimes defined as

rise/run. This means exactly the same as Δy/Δx, so gradient (grade) means the
same as slope. On highway signs, the grade is often shown as a percent: 5% grade,
17% grade, and so on. These correspond to slopes of .05 or .17, respectively. The
latter, a 17% grade, is quite steep – trucks would have to gear way down. A very
steep mountain trail might have a 100% grade, say. What angle is that? 45°,
because tan 45° = 1 = 100%.
Parallel and perpendicular lines
When are two lines L1 and L2 parallel? Answer: when they have the same slope,
m1 = m2. For in this case, θ1 = θ2, so the lines are parallel by definition.
There is also a quick way to tell if two lines are perpendicular, namely m2 =
−1/m1. For example, if L1 has slope m1 = .2 then any line L2 perpendicular to L1
has slope m2 = Before reading why this is true in general, try the next
problem.

Problem 6.8 (a) Use a diagram to show that lines of slope m1 = 1 and m2 = −1 are
perpendicular to each other. (b) Find the slope of L1 if θ1 = 30°, and of L2 if θ2 =
−60° (i.e., L2 slopes down at 60°). Check that m1 = −1/m2.

The accompanying figure explains why m2 = −1/m1 for perpendicular lines L1

and L2.
First we draw a slope-triangle for line L1, having sides a, b, as shown. W e also
draw a slope-triangle for L2, with horizontal side of length b, as shown. Let a′
denote the length of the vertical side of this triangle. The fact that L1 ⊥ L2 then
implies that these triangles are congruent (see below). Therefore a′ = a, and we
have m1 = b/a, and m2 = −a/b (minus because L2 has negative slope). Thus m2 =
−1/m1.
To prove the congruence, label the two smaller acute angles in the triangles
as α1 and α2, respectively, for L1 and L2. Label the other L1 acute angle as β. Notice
that α1 + β = 90°. Also α2 +90° + β = 180°,or α2 + β = 90°. Therefore α1 = α2, and
this implies that the triangles are congruent (two angles and one side equal).

To summarize this information, we have

To put the conditions stated in Eq. 6.6 in words, first, parallel lines have
equal slopes, and second, the slopes of mutually perpendicular lines are negative
reciprocals of each other.
The equation of a line
Let P1(x1,y1) be a given point. We want to find the equation of the line that passes
through the point P1 and has a given slope m. The method is similar to that used
to derive the equation of a circle in the preceding section.

We choose an arbitrary, representative point P (x, y) on the line. Then, by Eq.

6.4, applied to the points P1 and P , we have

Cross-multiply by x − x1 to obtain

This equation holds for any point (x, y) on the specified line. Conversely, if x
and y are two numbers that satisfy Eq. 6.7, then (x, y)is a point on the line
through (x1,y1) with slope m. In other words, Eq. 6.7 is the equation of the given
line. We call this the point-slope form of the equation of a line, to indicate that
this equation applies when one point (x1,y1), and the slope m are given. Note in
particular that the coordinates (x1,y1) of the given point P1 satisfy Eq. 6.7.
For example, the line through (−2, 5) with slope −3 has the equation y − 5 =
−3(x + 2). This might be rewritten as 3x + y + 1 = 0, although the latter equation
no longer visibly shows the conditions that specified the line.

Problem 6.9 Find the equation of the line through (−4, −2) having slope ½. Then
simplify the equation by algebra. Does the origin lie on the line, and why? Sketch
the line.

Problem 6.10 (a) Review the derivation of the equation for a line, in point-slope
form, in your head. (b) What is the slope of the line passing through (1,1) and
(3,5)? Find its equation.

Problem 6.11 Find the equation of the tangent line to the circle x2 + y2 = 25, at
the point (4,3). Suggestion: first make a sketch.

The point-slope form of the equation of a line, Eq. 6.7, is useful for writing
down the equation of a line given one point and the slope. We next discuss
several other useful forms for writing the equation of a line.
First consider the problem of finding the equation of the line through two
given points (x1,y1) and (x2,y2). This line has slope m given by

Therefore, from Eq. 6.7, the equation of the line is

This is sometimes called the two-point form for the equation of a straight line.
It’s simpler to remember that this is just a special case of the point-slope form,
Eq. 6.7.
Consider the special case of Eq. 6.7 in which (x1,y1) is the point (0,b). We get
y − b = mx, or
 Note that (0,b) is a point on the Y-axis, with y-value b. We say that the line in
Eq. 6.8 has intercept b on the Y-axis, or that b is the Y-intercept of this line.
The slope-intercept form is especially useful for quickly sketching (or
otherwise recognizing) a line whose equation is given to begin with. For example,
consider the equation 2x − y = 6. By algebra, this can be written as

Therefore the given equation determines a line with slope m = 2 and Y-

intercept b = −6.

Problem 6.12 Rewrite these equations in slope-intercept form: (a) 3x − y +8 = 0;

(b) 4x +2y − 5= 0.

An important consequence of the above discussion is that

This explains why Ax + By + C = 0 is called a ‘linear equation.’ The equation Ax

+ By + C = 0 is called the general form of the equation of a straight line. Here A,
B, C are any given numbers.

Problem 6.13 Find the equation of the line parallel to x − 4y = 2, and having Y-
intercept 3.

Question: What can be said about the line given by Eq. 6.9 in the event that A
= 0? or B = 0? Answer: If A = 0 the line is horizontal. If B = 0 it is vertical. (If both
A and B are zero, Eq. 6.9 is not the equation of a line. To be absolutely correct,
the proviso that A, B are not both zero should be included in Eq. 6.9.)
Problem 6.14 Find the slope m and Y-intercept b, if possible: (a) 3x − 2y +1 = 0;
(b) 3y +1 = 0; (c) 2x − 7= 0.

Another useful form for the equation of a line is:

Here, a is the X-intercept and b the Y-intercept. To understand why, first let y =
0 in Eq. 6.10; then x/a = 1, or x = a. This says that point (a, 0) lies on this line. In
other words, the line cuts the X-axis (y = 0) at x = a. Similarly, (0,b) lies on the
line 6.10, so the line cuts the Y-axis at y = b. The figure shows an example with a,
b both positive.

For example, the equation 2x − 3y − 8 = 0 can be written as

Thus a = 4, b = −8/3 are the intercepts. (Notice that you have to get 1 on the
right side, for this to work. The equation has to be exactly as given in Eq. 6.10.)

Problem 6.15 Write the equation 8x +11y + 62 = 0 in (a) slope-intercept form,

and (b) two-intercept form. List the slope and intercepts.

Problem 6.16 A square has corners at (1, 0), (0, 1), (−1, 0), and (0, −1). Sketch
this square and find the equations of its four sides. What is the area of the
square?

Digression: The meaning of an equation

Confusion can occur if you are not aware of the different ways that equations are
used in mathematics. Consider the following examples:
1. 6 + 3 = 9
2. 3x +3 = 9
3. 3(x +3) = 3x +9
4. 3x + y = 9

Read these equations and think about what they mean.

Equation (1) is straightforward – it is just a true arithmetical statement. But
Eqs. (2)-(4) are more elaborate, in that they involve symbols. What’s more, the
role of the symbols is quite different in each of these equations.
Equation (2), for example, can be solved for x, giving x = 2. This is the one
and only value of x for which Equation (2) is true.
Equation (3), on the other hand, is true for all values of x. It is an algebraic
identity, in fact a special case of the distributive law. Equations like this are used
frequently in many kinds of situations. But you would never want to solve Eq. (3)
for x.
Equation (4) is different again. You can’t solve it – at least not uniquely, as in
Eq. 2. And it’s not an identity, either. Instead, Eq. (4) specifies a relationship
between x and y. If one of these variables is specified, say y = 6, then the other
variable is determined, and can be calculated by solving the equation; here you
would get x = 1.
We know that in fact Eq. (4) is the equation of a certain straight line in the
XY-plane. We can think of this in two ways, however. First, it is just the equation
for a certain geometric object – a line. Second, this line is the ‘graph’ of the given
relationship between x and y. The latter interpretation will be expanded upon in
Chapter 7.
For the present, the equation 3x + y = 9 will be interpreted as the equation of
a line. The important point is that this line consists of all points (x, y) whose
coordinates ‘satisfy’ the equation, meaning that when the values of x and y are
substituted into the equation, it becomes a true arithmetical statement.
Let us introduce the following terminology for equations with symbols (i.e.,
variables).

Identity: an equation that is true for all values of the symbol(s).

Conditional equation: an equation that is true only for certain specific values
of the symbols.

Relational equation: an equation that specifies a relation between two or more

variables.

Definitional equation: an equation used to define some concept.

For example, the equation is a definitional equation, which defines the

concept of slope m.
Many equations involve symbols with different interpretations. For example,
in the equation Ax + By + C = 0, the symbols x and y refer to the coordinates (x,
y) of a point. These symbols are considered to be variables, in the sense that they
apply to all points on the line. But the symbols A, B, C refer to specific but
unspecified numbers. Such symbols are called parameters. In any specific
example, Ax + By + C would have actual numbers for A, B, and C – for example x
− y + 8 = 0. By writing out the general form Ax + By + C = 0 we are able to
discuss all such equations at once.

Problem 6.17 For Eqs. 6.2, 6.4, 6.7, and 6.10 of this chapter, specify (a) the type
of equation, and (b) the variables and parameters.

Please don’t underestimate the importance of this discussion of equations.

With practice, you become adept at handling all kinds of equations without
consciously thinking about what types they are. But many students do experience
considerable confusion about the actual meaning of certain equations. Reading
the above discussion carefully may help clarify your understanding of equations.
Simultaneous linear equations
Suppose we have two linear equations in x and y, for example

Graphically, each equation corresponds to a line in the XY-plane, as shown.

Unless two such lines are parallel (which these are not), they must have a unique
point of intersection. How do we calculate the coordinates (x, y) of this
intersection point? The logic is this: the point of intersection must lie on both
lines. Thus x and y must satisfy both of the given equations. From this fact we
can calculate x and y. There are several ways to do this. Each method involves
combining the equations in some way. For the above example, let us first add the
equations – this means that we add the left sides and add the right sides:
(Note that this procedure has ‘eliminated’ the y. Elimination is one of the general
methods, as will be described below.) From 3x = 9 we obtain x = 3. Now the
equation x + y = 4 becomes 3 + y = 4, so that y = 1. Therefore the point of
intersection of the given lines is (3,1). Check that this agrees with the figure: the
point of intersection of the lines does seem to be approximately (3,1).
It is also worth checking that the values x = 3, y = 1 do satisfy the given
equations: 2x − y = 6 − 1= 5; x + y = 3+1 = 4; correct.
The above example is worthy of careful study. It is typical of the general
problem of solving simultaneous linear equations. Given two linear equations
in x and y, we can calculate the unique solution (x, y) of both these equations by
the method of elimination, unless the equations represent parallel lines. Let us
look at another example.

In this case, adding the equations doesn’t help. Instead, we first multiply the
first equation through by 2:

Now we can eliminate x by subtracting the second equation from the first:

Therefore y = −1. Substituting this into the original equation x − 3y = 5 gives x =

2. The solution is (2, −1). This can be checked against the given equations.

Problem 6.18 Solve for x and y: 3x +2y = −4; x − 4y = −6.

Any system of two linear equations in two variables can be solved in this way.
Operations that are used to carry out the solution are:

(a) Addition or subtraction of equations.

(b) Multiplication of an equation by a non-zero number.

Using these operations in combination, one follows these steps:
1. Eliminate one of the variables.
2. Solve for the other variable.
3. Back-substitute to determine the first variable.

Let’s do a final example.

To eliminate x we first divide the first equation by 2:

Next we multiply the top equation here by −5 and add it to the second equation
(this eliminates x):

Therefore y = −2.
Returning to the equation (*), we now have

The solution is x = 1, y = 2, and this is easily checked by substituting into the

original equations.
The sequence of steps used here is called Gaussian elimination,after the
great German mathematician C.F. Gauss (1777-1855). Gaussian elimination can
be used to solve any pair of simultaneous linear equations. It also works for 3 or
any number of equations. Simultaneous linear equations arise in many areas of
applied mathematics, especially engineering and operations research. Computer
software is available for the efficient solution of systems that may contain
thousands of equations.

Problem 6.19 Solve: 4u − 3v = 17, 3u +2v = 0.

Two simultaneous linear equations always have a unique solution, unless the
lines corresponding to these equations are parallel. What happens to the algebra
of the solution, in this case? Here is an example

Let us multiply the first equation by 5/3

The second equation is

If we now subtract the second equation from the first, we get

This false equation means that the original system of simultaneous equations
cannot be solved for x and y. Look back at these equations and observe that the
lines corresponding to these equations both have slope 1/4. In other words, these
lines are parallel, and never intersect. In this case, the given equations are said to
be inconsistent – they do not have any solution.
Analytic versus Euclidean geometry
Analytic plane geometry is much more concrete than Euclidean geometry. Here
are some comparisons:
Concept
Euclidean
 Analytic
point undefined an ordered pair (x, y)of real numbers
straight line undefined the solution set of a linear equation Ax + By
+ C = 0
point on a line undefined (x,y) satisfies the given linear equation
distance between two undefined distance formula, Eq. 6.1
points
two points determine a axiom Two-point formula, Eq. 6.7
line
parallel lines exist axiom lines having the same slope m are parallel
two non-parallel lines axiom the meeting point can be calculated by
meet in a single point Gaussian elimination, which fails if the lines
are parallel

We have not discussed things like triangles here, but of course the theorems
of Euclid remain valid in analytic geometry. Next we study circles in analytic
geometry.

6.3 Circles
Recall that the equation

is the equation of a circle with center (x0,y0) and radius r.

The intersection points of two circles

Two circles in the plane may intersect at two points, or at a single point (if the
circles are tangent), or at no points. Given the equations of the two circles, we
can determine whether they intersect by comparing their radii with the distance
between the centers. Consider the example

Completing the squares in the second equation gives

The distance between the centers is from Eq. 6.1, so the center of the second
circle lies inside the first circle, which has radius 5. The second circle has radius
3. Since is bigger than 5, the circles do intersect. (Make a sketch to check
this.)
How can we find the coordinates of the points of intersection? This takes a
bit of ingenuity. Here are the given equations:

We need to solve these as simultaneous equations. First, it is useful to subtract

equation 1 from equation 2:

What next? We can solve the last equation for y, getting

We then substitute this y-expression into x2 + y2 − 25 = 0, obtaining

At last, we have an equation without y; in other words, we have eliminated y. The

method used here works for any two circles; if in fact the circles don’t intersect,
then the x-equation will have no solutions.
Finally, how can we solve the above equation for x? After squaring out and
simplifying, we will get a quadratic equation, which we know how to solve by the
quadratic formula. Omitting the details, the result is x = 3.43 or 4.97,
approximately (two solutions for x). We canthencalculate y from the equation y
= −2x+21/2, which gives y = 3.64 or 0.56, approximately. Thus the two points of
intersection of the circles are (3.43, 3.64) and (4.97, 0.56). If you made a careful
sketch, it should agree approximately with these values for the intersection
points.
The principles involved in this solution are more important than the
numerical results. First, given any two circles, the method used here will always
lead to a quadratic equation in x. Quadratic equations have either two, or one, or
no real solutions. Thus the algebra proves that circles have two, one, or no points
of intersection – a fact that we took for granted up to now.

Second, the example illustrates how one might go about solving any two
simultaneous nonlinear equations. The key is to eliminate one of the variables
(by algebra), and then solve the remaining equation for the other variable.
Finally, use substitution to calculate the value(s) of the first variable. (In practice,
this may be difficult, or even impossible.)

Problem 6.20 Find the points of intersection of the line y = 2x with the circle x2
+ y2 = 16. Make a sketch and comment on the symmetry of the problem.

Problem 6.21 Find the equation of the line passing through (5,0) and tangent to
the circle x2 + y2 = 9. Suggestion: see Sec. 5.6 on tangents to circles.
How to solve math problems
How should you set about solving a math problem? There are two kinds of
problems, as far as you are personally concerned. Let’s call them (1) routine
problems, and (2) confusing problems. Routine problems are those that you
quickly see how to solve – you only have to perform routine steps to complete
the solution. An example would be solving simultaneous linear equations, or
finding the unknown sides and angles of a right triangle. Learning math is partly
learning to solve routine problems, but only partly. The real crux comes when
you face confusing problems. Then what do you do?
My advice is ‘play with it.’ Try various ideas, make sketches, write down
vague thoughts. Try to come up with a strategy that could work. In other words,
enter into a mental search mode. Be creative. False starts are better than no starts.
Let’s look at a stream-of-consciousness approach to solving Problem 6.21,
finding the equation of the tangent line to a circle. (1) Make a sketch. (2)
Remember, the tangent is perpendicular to the radius line through the point of
contact. (3) So, it’s a right triangle. What are the sides? Oh, 3 and 5. So the other
side is (oh, yes, the 3-4-5 triangle). (4) Now what? What about
angles? Can I find θ in the sketch? Then what? m = tan θ is the slope of the radius
line. Great – the tangent line is perpendicular to this line. It’ll work! (From here
on, the problem is routine.)
This description is much too neat – no false start, no hesitation, no feelings of
despair. But you get the idea. Your brain is searching through memory for
relevant information, trying to understand the problem. Assuming that you have
thoroughly understood the math you’ve learned up till now, the search should
eventually be successful.
Solving confusing problems in nature is what the human brain has evolved to
do. Math and science, and creativity in general, take advantage of this natural
ability. This is how we differ from computers.
Most of the problems in this book are fairly routine. But every now and then
there’s a confusing one. Don’t expect to obtain the solution to a confusing
problem instantly. You have to work on it. You’ll learn a lot from taking the
effort. Why don’t you try the next problem?

Problem 6.22 What is the radius of the circle inscribed in an equilateral triangle
of side s?

6.4 Transformations
As pointed out at the beginning of this chapter, analytic geometry melds
geometry and algebra. Here we see how this works for transformations such as
translation, rotation, reflection, and scaling. We only look at a few examples to
illustrate the ideas but not to give an exhaustive treatment.
Translation
A translation in the XY-plane shifts every point (x, y) a certain distance to the
right (or left) and a certain distance up (or down).

Let h = horizontal shift and k = vertical shift. Let (x, y) be an arbitrary point
and let (x′, y′) be the translated (shifted) point. Then

For example, if h = 3 and k = −1, each point (x, y) is transformed to the point (x
+3,y − 1), i.e. 3 units to the right and one unit down.
A given geometric object, such as a line, triangle, or circle, is also shifted by h,
k by the translation in Eq. 6.11. The translation leaves the shape of the object
unchanged – the translated version is congruent to the original.

This is probably quite obvious to you, but let’s just check that translation
preserves the distance between two points, for example. Let Pi = (xi,yi) be two
given points with translated positions If d′ denotes the distance
between we have
(because x2 + h − (x1 + h)= x2 − x1,etc. Thus d′ = d: distance is preserved by a
translation. The algebra proves that our intuition is correct.

Example The line L, with equation 2x + y = 4, is translated by (1,3) to get a new

line L′. Find the equation of L′ (in X, Y coordinates).
To answer this, let (x, y) denote some point on L′. Then (x, y)has been
obtained by translating some point on L, specifically the point (x−1,y −3).
(Check that you understand this.) This point satisfies the equation for L:
and this implies that

(As a quick check, the point (0,4) is on L, and gets translated to (1,7), which
satisfies the equation for L′.)

Problem 6.23 (a) Find the equation of the new line L′, obtained by translating
the line L: x − y = 5 by (h, k)= (0, −1); (b) Same for L : y = 5 and (h, k)= (2, 2) [do
part (b) in your head].

The technique used in the above example (and Problem) can be described as
follows. We wish to determine the equation, in x, y-coordinates, of a certain
curve (or line) C. To start with, we let (x, y) denote some point on C. Next we use
the given characterization of C to deduce an equation involving x and y. This is
the equation of C.
To mention a familiar example, suppose C is the circle of radius 5, centered
at the point (3,1). If (x, y) is a point on C, then the distance from (x, y) to (3,1)
equals 5:

This is the equation of the given circle C. We usually write it in the form

Problem 6.23 shows that lines are transformed into new lines, having the
same slope as before, by a translation. This implies in turn that angles are
preserved by a translation, because the slopes of the sides of a given angle remain
the same.
For example, a translation of a triangle would give a new triangle having the
same sides and angles as the original triangle. Translation is an example of a rigid
motion.
Rotation
Just for the record, here are the equations for a rotation about the origin:

Here θ is the angle of rotation, measured counterclockwise. (Don’t memorize

this!) Understanding these equations requires trigonometry, so we won’t go into
them here; see Chapter 10. It turns out that rotation preserves distances and
angles, so it also is a rigid motion. The point here, however, is that in analytic
geometry, rotation is expressed in terms of equations.
Reflection
Consider a reflection in the Y-axis. What are the equations? A glance at the
figure will convince you that

For example, the line y = mx + b gets transformed into the line y = −mx + b.
(Use the same argument as before to check this.)
It would be straightforward to show that distances and angles are preserved
by a reflection, which is therefore another instance of rigid motion.
Uniform scaling
The equations of a uniform scale change, with scale factor k (with k > 0), are

Problem 6.24 Show that uniform scaling with scale factor k multiplies distances
by k, but does not change the slope of a line.

Uniform scaling is not a rigid motion, but it does preserve angles, and hence
shapes. Recall from Ch. 5 that uniform scaling is the basis for the concept of
similarity.
What about nonuniform scaling? you might ask. An example would be

where h and k are different. This transformation preserves neither distances nor
angles. But we humans do recognize a certain similarity between an image and a
‘squashed’ version of it. An ellipse is a squashed circle. Aunt Harriet’s face seen
from an angle is a squashed version of the head-on view, but you recognize her
immediately. The brain instantly applies all kinds of transformations to visual
images, because recognition (or its lack) has important consequences But if
you’re like me

6.5 Conic sections

Ellipses, parabolas, and hyperbolas are examples of conic sections, so called
because they can be obtained by cutting through a cone with a plane. To get a
full, two-piece hyperbola, for example, you have to cut both nappes of a cone, as
shown. Here we do not use the conic aspect of these curves, but obtain them
from other considerations.
Ellipse
One way to obtain an ellipse is by stretching a circle nonuniformly. We apply the
transformation x′ = ax, y′ = by to the circle x2 + y2 = 1. The transformed circle has
the equation

To make a quick sketch of this curve, first find the intercepts by inspection.
Namely, for the X-intercept, put y = 0 in the equation, giving x2/a2 = 1 or x = ±a.
Similarly the Y-intercepts are at y = ±b.

Suppose that a> b. Then the line from −a to +a on the X-axis is called the
major axis of the ellipse, and the line from −b to +b on the Y-axis is the minor
axis. Of course if a = b, the ellipse becomes a circle.
Next, consider an ellipse centered at the point (x0,y0), with major and minor
axes as before. Its equation is
You can think of this as the translation by x0,y0 of an ellipse centered at the
origin.

Problem 6.25 Identify the curve whose equation is 4x2 +9y2 − 24x +18y + 9 = 0.
Suggestion: complete the squares.

We see from this problem that any equation of the form

is the equation of an ellipse, provided that A and C are both positive, and
provided that after rewriting in the usual form by completing the square and
dividing through by the appropriate constant, the number +1 appears on the
right side of the equation. (The other possibility is that −1 winds up on the right
side. In this case the equation isn’t the equation of anything. Why not?)
Hyperbola
Let us make one small change to the ellipse equation (6.13).

This is the equation of a hyperbola. Drawing the graph requires a little ingenuity.
First, for the X-intercepts, set y = 0. This gives x = ±a. For the Y-intercepts set
x = 0, getting y2 = −b2. But this is not possible – no value of y satisfies this
equation. Thus there are no Y-intercepts. The curve never crosses the Y-axis.
Next, we observe that the graph of Eq. 6.14 is symmetric by horizontal
reflection in the Y-axis. If (x, y) is a point on the graph, then (−x, y)is also on the
graph, because the only appearance of x in Eq. 6.14 is in the term x2.

Thus reflection in the Y-axis transforms the graph into itself, so the graph is
symmetric. The same argument shows that the graph is also symmetric by
reflection in the X-axis. (By the way, these symmetric properties also apply to the
ellipse of Eq. 6.13 – see that graph.)
To proceed, we next solve Eq. 6.14 for y in terms of x:
This gives

From this we see that we need x2 > a2 to get a value for y. There are no points on
the graph for x lying between −a and +a. Check the figure for this feature.
Finally, what happens as x becomes large? The figure shows that the
hyperbola gets closer and closer to one of two straight lines. The lines have
equation y = ±bx/a. These lines are called the asymptotes of the hyperbola. How
do we explain this fact? Intuitively, if x is large, then x2 − a2 ≈ x2 (the term a2 is
relatively small). The symbol ≈ is read as ‘is approximately equal to.’ Therefore
[More details on this point are given below.]
A quick way to obtain the equations of the asymptotes to the hyperbola x2/a2
− y2/b2 = 1 is to use the equation x2/a2 − y2/b2 = 0, which has the solutions y =
±bx/a, and these are the asymptotes. Why does this work? Roughly speaking, if x
and y are large, then the constant term 1 is relatively negligible, so x2/a2 − y2/b2 ≈
0.
To summarize, the curve given by Eq. 6.14 is a hyperbola. It has the following
properties:
1. The hyperbola consists of two separate branches, which open up to the right
and left, along the X-axis.
2. The X-intercepts are at x = ±a. The branches lie to the left and right of the
intercepts.
3. The hyperbola is symmetric by reflection in both the X and Y-axes.
4. As x becomes large (positive or negative), the hyperbola approaches one of
the two asymptotes

Problem 6.26 Make careful sketches of the following hyperbolas: (a) (x2/4) − y2 =
1; (b) y2 − x2 = 1. Suggestion for (b): review the above discussion as it would
apply to this slightly different case.

Here is the algebra needed to show conclusively that the hyperbola given by
Eq. 6.15 (with the + sign) does approach the line as x becomes large. For a given
value of x, the difference in the y-values on the line and the curve is
The last line results from multiplying numerator and denominator by
and simplifying. (You should check this.) Now the final
expression approaches zero as x becomes larger and larger. This is what we
wanted to prove.

Problem 6.27 Describe thecurve whoseequation is

Parabola
Consider the equation

This is a parabola. To get an idea of what a parabola looks like, do the next
problem.

Problem 6.28 Sketch the graph of y = x2, for −2 ≤ x ≤ 2. First make a table of
values of x and y, for x = 0, 0.5, 1.0, 1.5, and 2.0. Plot the points (x, y) and join
with a curve. Finally, use symmetry to complete the graph.

Problem 6.29 If you own a graphics calculator, try it on the preceding parabola,
and also on (part of the) hyperbola
The foci of an ellipse
We have described an ellipse as a circle that has been stretched in one direction.
Most textbooks define an ellipse as the set of all points in the plane, the sum of
whose distances from two given points is a constant. The given points are called
the foci (singular, focus) of the ellipse. Some rather finicky algebra is required to
establish this fact.

Let the foci be located at x = ±c, on the X-axis, as shown. Then the sum of the
distances from (x, y)to these points (−c, 0) and (c, 0) equals

Suppose the X-intercept of the ellipse is at x = a. Then the sum of the distances of
(a, 0) from the foci equals (a − c)+(a + c)= 2a. Therefore the constant equals 2a,
and our equation is

Taking the square of both sides, and at the same time expanding (x + c)2 etc.
gives

where refers to the original -expressions. By combining terms

and shifting some to the right-hand side, we obtain (after cancelling 2s)
We again square both sides:
Expanding leads to

Now quite a few terms cancel out, giving

or
Finally

This we recognize as Equation 6.13 of an ellipse, having b2 = a2 − c2. We have

thus proved the statement about the ellipse as the set of points such that the sum
of the distances from the two foci is constant. Moreover, we have determined the
position of the foci:

The ratio c/a is called the eccentricity of the ellipse, sometimes denoted by e.

A circle has zero eccentricity; a greatly elongated ellipse has eccentricity nearly
equal to 1. (See the next problem.) We always have e< 1, for any ellipse.

Problem 6.30 (a) A circle is an ellipse in which a = b. Where are the foci of a
circle? What is the eccentricity? (b) Roughly sketch the ellipse (x2/100) + y2 = 1.
Where are the foci? What is the eccentricity? (c) If e denotes eccentricity, express
the ratio of minor to major axis (b/a)in terms of e.

Example Planetary orbits. Johannes Kepler (1571-1630) discovered that the

orbits of the planets are ellipses, with the sun located at one of the foci,
approximately. This and other of Kepler’s discoveries were later explained by
Newton’s theory of gravity and motion.

Problem 6.31 The eccentricity of the earth’s orbit is e = .017. Find the ratio of the
largest and smallest distances from the sun to the earth, over a year’s cycle. (The
seasonal change of solar distance affects the earth’s weather, especially the
severity of winters in the Northern hemisphere.)
The foci of a hyperbola
A hyperbola can be defined as the set of points P in the plane such that the
difference in the distances from P to two given points (the foci) is constant. If the
foci are at x = ±c, and the X-intercepts are x = ±a where a < c,this definition
implies that

This equation can be simplified exactly as in the case of the ellipse. In fact, the
calculations are virtually the same; the final equation is exactly the same (because
the minus sign disappears in the second squaring):

However, we now have a < c, so this equation is better written in the form

This is the equation of a hyperbola, as in Eq. 6.14, with b2 = c2 − a2. The

eccentricity e is again defined as e = c/a. For a hyperbola we have e> 1.

Problem 6.32 Find the foci, and the eccentricity, for the hyperbolas (a) x2/4 − y2
= 1; (b) x2 − y2/4 = 1. Sketch the hyperbolas.

Where does the parabola fit into this scheme of foci and eccentricity? It
happens that e = 1 for any parabola, so that parabolas are somehow intermediate
between ellipses and hyperbolas.
The parabola can be defined as the set of points P in the plane that are
equidistant from a given point (the focus), and a given line, called the directrix.
To obtain the equation corresponding to this definition, we place the focus at
(0,c) on the Y-axis, and the directrix at y = −c. Then we have
(use a sketch to check this) so that
This reduces to

and this agrees with Equation 6.16 for a parabola, with a = 1/(4c).

Example A parabolic reflector is a surface obtained by rotating a parabola in

space, about its center axis. The parabolic reflector has the property that light
rays entering the reflector parallel to its axis are all reflected to the same point,
namely the focus of the parabola. This is presumably where the term ‘focus’ came
from. The simplest proof of this fact uses Calculus, so we omit the proof here.

Parabolic reflectors are used for both receiving and transmitting rays of many
kinds, including light, radar, TV and radio signals, and also sound. When such
rays are reflected by a surface, the angles of incidence and of reflection are equal.
In the parabolic reflector rays parallel to the axis are all reflected to the focus.
Example Most comets return periodically (like Halley’s comet), and have elliptic
orbits around the sun. The higher the eccentricity of the ellipse, the longer the
period between sightings of a comet. The eccentricity of some comet orbits,
however, is close to 1.0, and some are even greater than 1.0. These comets have
parabolic or hyperbolic orbits, and do not recur. Once they leave the solar
system, they never reappear. (The orbits of comets are actually not perfect
ellipses etc., because of the gravitational influences of the planets, which perturb
the elliptic orbit about the sun.)

Problem 6.33 Find the equation of a curve defined as the set of points in the
plane such that the ratio of the distance from a given line to the distance from a
given point is a constant, say q. Identify the curve, depending on q. Suggestion:
use the same setup as for the parabola.

We have now shown that an equation of the form

is the equation of some conic section (ellipse, hyperbola, or parabola) unless

there are no points (x, y) satisfying the equation. You may have noticed the
absence of the term Bxy in this equation. If this term were included, the equation
would be a general quadratic equation in two variables x and y. It turns out that
this general quadratic equation also corresponds to one of the conic sections,
which, however may be rotated through some angle, rather than having its axes
in the X or Y-direction. We do not study this situation here, as it requires
familiarity with rotations in the plane. See Chapter 10.

6.6 Review problems

1. Find the equation of the line perpendicular to 2x + y = 5 and passing through
(a) the origin, and (b) the point (2, −1).

2. Find the point of intersection of the lines 2x − y = 5, 3x +2y = 2.

3. Find the points where the circle (x−3)2 +(y −1)2 = 4 cuts the X-axis.
4. True or false? If line L2 has twice the slope of line L1, then the angle that L2
makes with the positive X-axis is twice the angle that L1 makes with the
positive X-axis. Explain.
5. Find the point on the line 2x − y = 4 that is closest to the point (1,1).
Suggestion: first find the equation of the line through (1,1), perpendicular to
 the given line. A sketch should indicate why this is relevant.
6. Find the equation of the line that passes through the points of intersection of
the circles x2 + y2 +2x − 6= 0and x2 + y2 − 6x − 2y − 8= 0. (Don’t try to find the
intersection points themselves.) Suggestion: simply subtract one equation
from the other. Then explain why the result is the answer.

7. Find the equation of the perpendicular bisector of the line P1P2,where P1 = (1,
4) and P2 = (3, 2).

8. What is the graph of the equation Sketch it.

9. The hyperbola x2/a2−y2/b2 = 1 and the hyperbola x2/a2−y2/b2 = −1 are called

‘conjugate hyperbolas.’ Sketch both in the same XY-plane to explain this term.
10. Find the area inside the ellipse x2/a2 + y2/b2 = 1. Suggestion: remember that
the ellipse is a stretched (or scaled) circle, but with scale factors a and b in the
X and Y-directions.

11. Sketch the hyperbola x2 − 4y2 − 6x + 5 = 0. Identify the intercepts, the

asymptotes, and the foci.

12. Find all points of intersection of the hyperbola x2 − y2 = 2 with the circle x2 +
y2 = 4. Use a sketch to check the answer.

Solution 6.1 (a) III, IV. (b) A point on the X-axis has y = 0. Thus the point has
coordinates (x, 0). Similarly, a point on the Y-axis has coordinates (0,y). The
coordinates of the origin are (0, 0).
Solution 6.2 This is the hypotenuse of a 45o right triangle with legs of
length 4, determined by the given points.
Solution 6.3 The circle has center (0, −3) and radius 3. The origin (0, 0) lies on
the circle, as we see either from the graph or by substitution: 02 + (0 + 3)2 = 32 = 9
as in the equation. From the graph we see that the lowest point on the circle is (0,
−6), and these coordinates also satisfy the equation: 02 +(−6+3)2 = (−3)2 = 9.
Similarly, the rightmost point is (3, −3), and this also satisfies the equation, as
you can check.

Solution 6.4 (a) Circle with center (−3, 1) and radius (b) Nothing.

Solution 6.5 Completing the squares in Eq. 6.3 leads to (x − A/2)2 +(y − B/2)2 =
(A2 + B2)/4 − C The condition needed to get a bona fide circle is therefore A2 + B2
> 4C. In the case of equality, A2 + B2 = 4C, we get r = 0, i.e. a ‘circle’ of radius
zero. Such a circle is in fact a single point, at the center (A/2,B/2). To summarize,
Eq. 6.3 represents:

Solution 6.6 First, use m = Δy/Δx with Δx = x2 − x1 = 2. This implies that Δy =

mΔx = 2m. Also, Δy = y2 − y1 = y2. Thus y2 = 2m. The answers are then y2 = 2, 0,
−4, 16 for cases (a)-(d).
Solution 6.7 (a) Since Δy = Δx in this case, the slope triangle is an isosceles, 45°
triangle: θ = 45°. (Also m = 1 = tan θ implies θ = 45°.) (b) Here θ = Atan (10)=
84.3°. Notethat a largeslope m makes θ nearly 90°. Try m = 1, 000 and 10,000 for
example. (c) θ = −60°.

Solution 6.8 (a) As shown in Problem 6.7, a line with slope m1 = 1 makes an
angle of 45° with theX-axis. In thesameway, a linewith slope m2 = −1 makes an
angle of −45° with the X-axis. These lines therefore make an angle of 90° with
each other. In other words, they are perpendicular.

(b) By looking at the Δx, Δy triangles for lines L1 and L2, we see that
Therefore

Solution 6.9 The equation is y +2 = which can be written as

(or else as and other possibilities). Yes, the origin (0,0) does lie on
this line, because these coordinates satisfy the equation of the line.

Solution 6.10 (b) Slope: m = 4/2 = 2. Equation: y − 1= 2(x − 1).

Solution 6.11 Recall that the tangent to a circle at a point P is perpendicular to
the radius through P. In this example, the radius line has slope 3/4. Therefore the
tangent line has slope −4/3. The equation of the tangent line is y − 3= −(4/3)(x −
4).

Solution 6.12 (a) y = 3x +8; (b) y = −2x +5/2.

Solution 6.13 The given line has slope 1/4, so the equation of the new line is y =
(1/4)x +3, or x − 4y +12 = 0.

Solution 6.14 (a) m = 3/2, b = ½; (b) m = 0, b = −1/3; (c) m is undefined; the line
is vertical at x = 7/2.

Solution 6.15
intercepts

Solution 6.16 The sides a, b, c, d have equations: a : x + y = 1; b : −x + y = 1; c :

−x − y = 1; d : x − y = 1. The square has sides of length so the area is A = 2.

Solution 6.17
Solution 6.18 x = −2, y = 1. One way to get this is to multiply the first equation
by 2 and add it to the second equation, thereby eliminating y.
Solution 6.19 u = 2, v = −3.
Solution 6.20 Substituting y = 2x into x2 + y2 = 16 gives x2 +4x2 = 16, so
The corresponding y-values are y = 2x,or The line
and circle are both symmetric by reflection in the given line itself, and also in the
perpendicular line y = −x/2.

Solution 6.21 The triangle shown is a 3-4-5 right triangle. The slope of the radius
line is therefore tan θ = 4/3 (opposite/adjacent, remember). Therefore the slope
of the tangent line is −3/4. From the point-slope form, the equation of the
tangent line is y = (−3/4)(x − 5). This can be written more neatly as 3x + 4y = 15.
(Congratulations if you figured this out – there are various ways of doing it. If
you didn’t succeed, how about reviewing all the math principles used in the
solution given here? Each sentence involves one or more principles.) By the way,
the other tangent line has equation 3x − 4y = 15.
Solution 6.22 First, make a sketch. Next, recall from Chapter 5 that the center of
the inscribed circle lies at the intersection of the angle bisectors. Therefore AOD
is a 30 − 60° triangle. Thus Since
AD = s/2, we see that and this is the radius of the inscribed
circle.

Solution 6.23 (a) If (x, y) is on L′ then (x−0,y+1) is on L, i.e., x−(y+1) = 5, or x −

y = 6. This is the equation of L′. (b) y = 7; you get the same result if you use the
method of part (a).

Solution 6.24 We have

Thus distances are multiplied by k. The line Ax + By + C = 0 transforms into

Ax/k + By/k + C = 0, and this has the same slope m = −A/B as the original line.
Solution 6.25 The equation becomes 4(x2−6x+9)+9(y2 +2y+1)−36 = 0, which can
be written as

This is an ellipse centered at (3, −1) and with semi-axes of length 3 and 2.
Solution 6.26 (a) X-intercepts at x = ±2; asymptotes y = ±(½)x have slope ±½;
(b) Y-intercepts at y = ±1; this hyperbola opens vertically; asymptotes y = ±x.

Solution 6.27 A hyperbola centered at (2, −1), with asymptotes as lines

through(2,-1) withslope ±2/3. The intercepts on the line y = −1 occur at x = 2 ±
3, i.e., at x = −1 and 5. In other words, the equation represents a translation of
the hyperbola x2/9 − y2/4= 1 to have center at (2, −1).

Solution 6.28 The table could be written like this

Plotting these points and joining them up gives the graph for x > 0.(Note that
one does not join the points with little line segments. The parabola is a smooth
curve, not a jagged one.)
The parabola does not have asymptotes – it continues to curve more and more
steeply upwards.
Solution 6.30 (a) By Eq. 6.17, c = 0. Thus the foci of a circle are both at the
center. (The sum of the distances from the foci in this case equals 2r where r is
the radius of the circle.) The eccentricity c/a is zero. (b) Here
Thus the eccentricity is c/a = 9.95/10 = .995 – a highly eccentric ellipse. (c) From
Eq. 6.17, c2 = a2−b2. Therefore c2/a2 = 1−b2/a2, from which we have
Thus b/a = 1 if e = 0 (circle), and b/a is nearly
0 if e is close to 1 (elongated ellipse).
Solution 6.31 From a graph of the ellipse with the sun at a focus, we see that the
ratio of longest to shortest distances equals

Solution 6.32 From b2 = c2 − a2 we have for both

examples. The eccentricity e equals for case (a), and
for case (b).

Solution 6.33 We have the equation which can be

2 2 2
written as x +(1 − q )y + Ay + B = 0, where A and B are constants involving c
and q. If q = 1 this is the equation of a parabola (as above). If q = 1, we rewrite the
equation as
for new constants A1,B1. This equation is an ellipse if q < 1 (making 1 − q2 > 0),
and a hyperbola if q > 1. (In fact, you can show that q = e, the eccentricity, for all
three cases. Thus we have a unified, single characterization of the three conic
sections. Also, the fact that the parabola is intermediate between the ellipse and
hyperbola now makes sense.)
Chapter 7
Functions and Graphs

7.1 Sets
Modern mathematics makes extensive use of sets. In basic math, for example,
various geometric objects, such as lines and circles, are defined as the set of all
points satisfying a certain condition. What exactly is a set? A set is any collection
of objects. Here we will talk about mathematical objects, such as numbers,
points, and so on, and not about physical objects, such as chairs, or dogs.

Notation
Let A,B,… designate sets of some kind. Then
1. x ∈ A means that x is a member of A, or an element of A. We often read x
∈ A as just “x is in A,” or “x belongs to A.” For example, if A is the set of all
even integers, then 2 ∈ A, but 3 ∉ A (“3 does not belong to A”).
2. A ⊂ B means that set A is contained in set B, or A is a subset of B.
3. A ∪ B (“A union B”), the union of A and B, is the set of objects that belong
to A or B, or both.
4. A ∩ B (“A intersect B”), the intersection of A and B, is the set of objects that
belong to both A and B.

Problem 7.1 Decide which of the following statements are true. A diagram may
be helpful. (a) A ⊂ B and B ⊂ C implies A ⊂ C; (b) A ⊂ (A ∪ B); (c) A ⊂ (A ∩
B); (d) (A ∩ B) ⊂ A; (e) A ⊂ B and x ∈ A implies x ∈ B; (f) A ∪ (B ∪ C)= (A ∪
B) ∪ C.

We also have
5. {x : Px} (“The set of all x such that Px.”) Here Px represents some statement
about x, and {x : Px} is the set of all x for which Px is true.

For example, {x : 0 ≤ x and x ≤ 1} is the set of all numbers x lying between 0

and 1, inclusively.

Problem 7.2 Identify the sets A = {(x, y): x2 + y2 = 16} and B = {x, y): y = x +2}.
(Interpret A and B as sets of points in the XY-plane.)

Note that, to make sense of the expression {x : Px}, the allowable choices for x
must be specified, at least tacitly. For example, in Problem 7.2, the notation
suggests that (x, y) is a point in the XY-plane. If there could be any doubt, the
possible choices for x should be stated explicitly. Thus {x : x is an integer and x ≥
5} is unambiguous, but {x : x ≥ 5} might be interpreted as the set of all real
numbers greater than or equal to 5. What is intended might be clear from the
context, but if not, the meaning should be stated explicitly.

Problem 7.3 The following are true statements. Read them until you understand
them. Then write them in plain English. (⇔ means if and only if.)
(a) For any x, x ∈ A ∪ B ⇔ x∈ A or x ∈ B.
(b) For any
(c) {x : x is a real number and x2 = 1} = {−1, 1}. (The notation {−1, 1} means the
set consisting of two numbers −1 and 1.)

Intervals
A closed interval [a, b] is a subset of the real line given by
Here a and b are two given real numbers, with a < b. The expression a ≤ x ≤ b
means that a ≤ x and x ≤ b; in other words x is between a and b, inclusively.

An open interval (a, b)is given by

Thus a closed interval includes its endpoints a, b, whereas an open interval

excludes them.
Half-open intervals, such as [a, b) = {x : x is a real number and a ≤ x< b} are
sometimes used.

Problem 7.4 Under what circumstances is [a, b] ∪ [c, d] itself a closed interval?
Suggestion: make a sketch.

Problem 7.5 Same question as in Problem 7.4, but for [a, b] ∩ [c, d].

If you didn’t get either or both of these problems, try them again carefully.
Use sketches.
We use a special symbol ∅ for:

Thus ∅ is the set that has no members. This seems a bit strange at first, but it
is quite useful. For example, case 1 of Problem 7.5 can be expressed as

In this case, the two intervals have no point in common, which means that
their intersection is empty. Exercise: A ∪ ∅ = ? A ∩ ∅ = ? Answer: A, ∅,
respectively. Review the definition of ∪ and ∩ if this is not clear.
Half-infinite intervals are:

and similarly for (a, ∞)and(−∞,a). Note that ∞ and −∞ are not numbers, but
merely indicators that the interval extends indefinitely to right or left. Also

In other words, (−∞, ∞) is the entire real-number axis.

Time out for another discourse on definitions in math. New definitions are
often confusing at first. Considerable time and effort may be needed before a
new definition “sinks in.” You absolutely must not skip on to the next topic, in
the hope that the new idea isn’t that important. Definitions are basic. You need to
understand and memorize them. Part of what makes math so concise (and so
useful, for that very reason) is that defined terms are stated and used with
absolute precision.

Problem 7.6 Make a list of all new terms introduced in this section, including
their definitions.

7.2 Functions
One of the most important concepts in mathematics is that of a function. The
function concept is so general that at first it may be a little difficult to understand
its significance. We will therefore begin with fairly simple examples. Here is the
basic definition.

Definition. A function is a procedure for transforming any object of a given set

A into a specific object in another set B. If the function is named f, then for any x
∈ A we write f(x) for the transformed object, which is an element of B. The
expression f(x) is read as “f of x.” This definition of function is universal
throughout mathematics. However, you might have encountered a different
looking definition in your school math. Later on I will explain how other
definitions are related to the one given here.

Example 1. A is the set of all real numbers. The function f is defined by

This is the way that functions are often encountered, being defined by a single
formula. (It is certainly not the only way, however!) In this example, what would
f(5) mean? Answer: f(5) = 3 × 52 = 75. One just substitutes x = 5 in the defining
formula. To check that you understand this, find f(0), f(−1), f(1/3), and f(u + v).
Answers: 0, 3, 1/3, 3(u + v)2. In all cases you replace x in the formula for f(x) by
the value in between the brackets of f. Note that, as stated in the definition of
function, in this example, f transforms any given object x in A (i.e., any real
number x) into a new object, f(x)= 3x2. The new object is also a real number.
Question: what would f(w) mean, here? Answer 3w2.

Example 2. Let N be the set of positive integers. For any n ∈ N let p(n) be the
number of different prime factors of n (not counting 1 as a prime factor). For
example, p(5) = 1, p(12) = 2, and p(140) = 3 (because 140 = 22 × 5 × 7). This
example shows that a function does not have to be defined by a formula.

Example 3. Let Δ be the set of all triangles. For any t ∈ Δ let A(t) be the area of t.
Is A really a function? Yes: every triangle t has an area. This example illustrates
the generality of the function concept.

Let us summarize. A (mathematical) function f is specified if (i) a set A is

specified, and (ii) a procedure is stated whereby each element x ∈ A is
transformed to a specific “value” f(x). The set A is called the domain of
definition, of the given function, or just the domain, for short. In the three
above examples, note that the domain of each function is explicitly stated. Read
these examples again, to check this.

Problem 7.7 For the functions given in Examples 1-3, find (a) f(−2); (b) p(64);
(c) A(t1)where t1 is the triangle with vertices (0,0), (0,1), and (2,3).

Problem 7.8 By trying them out, determine the domains of the following
calculator functions: sin, Asin
A constant function
Let f(x) = 2 for all x ∈ (−∞, ∞). Is f a function? Well, does the equation f(x) = 2
tell us how to determine the value of f(x) for every x? Yes it does. Therefore f is
indeed a function. It is called a “constant function.” Of course, a constant
function is not of much use (but they do come up, for example in Calculus).
Nevertheless, a constant function is a bona fide function.

Computer functions
(Skip this subsection if you’re not interested in computers.)
Every computer programming language not only contains built-in functions,
but also allows the programmer to define other functions. Here is an example
using the language C.

Let’s try to figure out what this function is, mathematically. The name of the
function is F. The domain of F is defined as “double,” which specifies double-
precision real numbers (about 17-digit precision). The other “double” on the first
line specifies that the values of F (x) will also be obtained in double precision.
Finally, the body of the function, contained in curly brackets, specifies the value
to “return,” which is 1/(1 +x * x), or 1/(1 +x2). Thus F (x)= 1/(1 + x2) in this case.
This example shows that the term “function” is understood in computers in
the same way as in math. A domain is specified, and a procedure is given to
calculate the function values. (Of course, functions programmed in C can be
much more complicated than this example.)

Functions, look-up tables, and ordered pairs

Consider the table shown here. Such a table can be used to specify y as a function
of x, which could be written as usual as y = f(x). In this case, the “procedure” for
calculating f(x) is simply to look up the value in the table. For example, here we
have f(1) = 5 and f(2) = 7, and so on. What would f(6) be? The answer is that f(6)
is undefined, since x = 6 does not occur in the given table. The domain of this
particular function f is the set of x-values in the table, namely {1, 2, 3, 4, 5}.

X y
1 5
2 7
3 3
4 2
5 7

Such look-up functions occur often in practice. For example, to identify a

taxpayer from his social insurance number, the government uses a look-up table
(stored on a computer). Similarly, to find the price of an item at a supermarket,
the check-out clerk looks in a price table. This may be done automatically by a
bar-code reader, which reads x = bar code number, and then looks up y = price
of the item.
Thus, look-up tables are functions, according to our basic definition. Now
consider the next table. Look at this table carefully. Do you see why this table is
not a function? What is f(3)? On the first line we would have f(3) = 4, but on the
fourth line f(3) = 2. Such ambiguity is not allowed in a function. To specify a
function f, it must be possible to find a unique value f(x) for every x in the
domain of f. Recall that the definition at the beginning of this section says that a
function is a procedure for transforming any object of a given set A into a specific
object of another set B. The above table cannot be used to do this, so it is not a
legitimate look-up table. Examples such as taxpayer lists and supermarket price
tables must be bona fide functions, with unique values y = f(x) corresponding to
each object x in the table. This uniqueness of function values f(x) is an essential
aspect of the mathematical concept of a function.

X y
3 4
5 0
1 5
3 2
5 7

Functions and ordered pairs

An ordered pair is just what it says, a pair (a, b) of mathematical objects, in the
given order. Thus (4,2) is an ordered pair of integers and (2,4) is a different
ordered pair. A familiar example of ordered pairs is the pair (x, y) of coordinates
of a point in the XY-plane.
If you studied “new math” in school, you may have learned that a func tion f
is a set of ordered pairs, with the property that (x, y) ∈ f and (x, z) ∈ f ⇒ y = z.
Like much of new math, many students found this pretty confusing.
However, this definition of function coincides exactly with the look-up table
idea. If the set of ordered pairs is written as {(x1,y1), (x2,y2),…, (xn,yn)} then the
look-up table has x1,y1 on the first line, x2,y2 on the second line, and so on. Thus
the first look-up table given above could be written instead as {(1, 5), (2, 7), (3,
3), (4, 2), (5, 7)}. The uniqueness condition is that if two x-values are the same,
then the corresponding y-values must also be equal. This condition is the same
for look-up tables as for sets of ordered pairs. It is also the same as saying that
f(x) must be uniquely determined, for each x in the domain of f.

Problem 7.9 (a) Consider the set G = {(−1, 2), (1, 3), (4, 2), (7, 1)}. Is G a
function? If so, find G(1). If not, remove pairs that result in ambiguity. (b) Same
for h = {(5, 0), (2, 1), (1, 0), (2, 5)}.

I hope you have not found this discussion overly confusing. It was necessary
to go into these matters, because different traditions of teaching have
emphasized seemingly different definitions of the term “function.” But as we
have seen, these differences are only apparent.
One point remains to be clarified, perhaps. How can a function given by a
certain formula, such as f(x) = 3x2, be considered a set of ordered pairs? Here, x
can be any real number. The set of ordered pairs in this case becomes an infinite
set, consisting of all the pairs (x, y) with y = 3x2, and x any real number. In set-
theoretic notation we have
In practice, we never actually think of the function f(x) = 3x2 in this roundabout
way, of course. (The “look-up table” for this f would have infinitely many entries
—a practical impossibility, but not a conceptual one.)
The look-up procedure works only for finite sets, such as G above. However,
most functions used in math and science involve infinite sets of ordered pairs.
It’s not possible to “look something up” in an infinite set, which is why the
definition given at the beginning of this section insists on there being a
procedure for calculating the values f(x).
We will not use the ordered pair characterization of functions in this book,
but you may encounter it elsewhere.

Other notations
Consider again the function f(x)= 3x2. Another way of writing this function is
sometimes used, namely

which is read as “ f transforms x to 3x2.” Different books use different ways of

expressing functions, but the approach discussed at the beginning of this section
is the most common one.

Problem 7.10 (a) Define a function M by

What is the geometric interpretation of M? What is its domain? How else could
the definition of M be expressed? , what is the “natural
domain” of f ? (Invent an interpretation of “natural domain.”)

The situation in Problem 7.10 (b) is common. A certain function f(x) is

defined by a formula, but there may be exceptional values of x for which the
formula is meaningless. All other, non-exceptional values constitute the natural
domain of f , that is, the set of all objects x for which the function can be defined
by the formula. Problem 7.10 (a) indicates that a function can involve any kind
of mathematical objects whatsoever. Another example from plane geometry is

Since a line is a set of points satisfying a certain equation, this could be written
explicitly as

(recall the point-slope form for the equation of a line, Eq. 6.7.) What about the
natural domain of L? The fact that x2 −x1 occurs in the denominator of the
equation of the line, alerts us to the fact that m is undefined if x1 = x2. Actually,
there are two cases here. First, if the given points coincide, so that (x1,y1)= (x2,y2),
then there is no uniquely determined line through these points; L is undefined in
this case. Second, if x1 = x2 but y1 ≠ y2, the two points lie on the vertical line
whose equation is x = x1. Therefore

Thus L is a bona fide function, with domain A = all pairs of distinct points (x1,y1),
(x2,y2), and with values equal to the straight lines whose equations are given
above (two cases).
This example again illustrates the broad scope of the function concept. Any
kind of mathematical objects can constitute the domain of the function, and any
kind of objects can occur as values. The only strict stipulation is that the value of
the function must be unambiguously specified, for each object in the domain.

Problem 7.11 Let C(x0,y0,r) be the circle of radius r, centered at (x0,y0). Express
C(x0,y0,r)in {···} form.

Additional terminology
By definition, to specify a certain function f , we must specify (i) a set D of
mathematical objects (D is called the domain of the function), and (ii) a
procedure for obtaining the values f(x) for any x ∈ D. In practice, the domain D
may not be explicitly stated, in which case some natural domain is taken for
granted.
The set Rf of all possible values f(x) of the function f is called the range of
values (usually just range, for short) of f :

For example, if f(x)= 3x2,with D = (−∞, ∞), the set of all real numbers, then Rf =
[0, ∞), the set of all non-negative real numbers. Why? Because, first, 3x2 ≥ 0 for
any x. Second, any number y ≥ 0 can be obtained as the value f(x)for some x ∈
D. Namely, has f(x)= 3x2 = 3y/3= y. This proves that Rf = [0, ∞).

Problem 7.12 Determine the range of f(x) = 2x +1, (a) with domain (−∞, +∞);
(b) with domain [0,1].

A further point of terminology. One often encounters statements like “let y =

f(x) be a given function,” or “consider the function y = 3x2.” Although such
phrases may not agree exactly with the way that we have defined the concept of a
function in this book, there is no difficulty in interpreting these statements.
Functions are so common in math and science that it would be unreasonable to
expect everyone to use exactly the same terminology.

Problem 7.13 Determine the domain and range of the function of Problem 7.9,
G = {(−1, 2), (1, 3), (4, 2), (7, 1)}.

Functions of several variables

Consider the function

This is a function of two variables, x and y. The natural domain D of f is the set of
all ordered pairs (x, y) of real numbers. The usual method of substitution applies;
for example
 (Here the dot “·” is used for multiplication.) What would f(y, x)be, in this
example? Answer: f(y, x) = y2 +4x2 +6yx – you substitute y for x and x for y.
Countlessly many practical examples could be cited. In meteorology, for
example, (x, y) could represent the latitude and longitude of a position on the
earth’s surface. Then T (x, y) could refer to the temperature at this position, and
P (x, y) the barometric pressure there. These functions are published in the daily
media, usually in graphic form. Since temperature and pressure are time
dependent, we should probably write T = T (x, y, t), etc., where t is the time. The
weather report on TV may show a dynamic prediction of T (x, y, t) for your state,
for the next 24 or 48 hours. Functions and graphs are everywhere!
Given a function of several variables, such as T (x, y, t), the input variables x,
y, t are sometimes called the arguments of the function T. This terminology is
used in computer programming for example.

7.3 The graph of a function

In this section we will consider functions f(x) defined for real numbers x, and
taking real values f(x). Some examples are f(x)= 3x2; g(x)= 2x +1; h(x)= 1/(1 + x2).

Definition. If f is a function having domain A, where A is a set of real numbers,

the graph of f is the set Gf of points in the XY-plane, given by

This definition means that the graph of the function f is exactly the same
thing as the graph of the equation y = f(x) (x ∈ A). Recall Section 6.1 on graphs
of equations.

Example 1. f(x) = 3x2 for x ∈ (−∞, ∞).

We know that y = 3x2 is the equation of a parabola with vertex at the origin, and
opening upwards. To sketch the graph accurately, we need only plot a few points
(x, y), to help pin down the parabola. Thus (1,3) and (−1, 3) are points on the
graph, because f(1) = 3 and f(−1) = 3. Likewise (0,0) is on the graph.
 What should we do if we wished to depict the graph of f(x)= 3x2 for a wider
range of values, say −10 ≤ x ≤ 10? Try it for yourself, first.

Because f(10) = 300, we need to show y-values from 0 to 300. If we use the
same scales on the X-and Y-axes, we get an unrevealing graph, as shown above.
However in depicting the graph of a function f(x), there is no need to use
identical X-and Y-scales. Rather, one should choose scales that are appropriate
for the situation. After all, graphing a function is not the same as doing analytic
geometry, where the scales have to be the same. The second graph above shows
f(x)= 3x2 with appropriate scales.
 Graphs of functions are used frequently in science, and then the x-and y-
values typically have physical meaning.
The next graph comes from my own research in modeling the foraging
(feeding) behavior of fish. The graph shows the average feeding rate f, as a
function of visual range r in a lake. The model takes account of various practical
components of foraging, including swim speed v, handling time h, and prey
density d. The function graphed is

where a and b are constants involving the parameters v, h, d. The shape and
position of this curve is useful for understanding the behavior of fish and their
prey species at different times of the day. (Note that this graph uses a
“logarithmic scale” on the horizontal axis. Such graphs are often used in
scientific publications.)
Whether one is considering a given function in math, or in some scientific
situation, displaying a graph of the function is an efficient way of conveying
information about it. What portion of the graph to display depends on the
context. In math, we are usually concerned with displaying the salient features of
the function, and graphic details should be chosen accordingly.
Example 2. g(x)= 2x +1

The graph of this function is the straight line y = 2x +1. Not much more need be
said here.

Example 3. f(x)= 1/(1 + x2).

You are probably not familiar with the graph of

Let us analyze this carefully. Many students learn only to “make a table of values,
then plot the points and join them up to get the graph.” This unintelligent
approach to graphing is almost worthless in later courses.

What can we say about the graph of y = f(x)= 1/(1+x2)? First, as x gets large, y
becomes small. For example, f(10) = 1/101 ≈ .01. The maximum y-value is f(0) =
1. Also, we see that a negative value of x gives the same y-value as does the
corresponding positive value of x. With these ideas in mind, let’s now make a
table.

x y = f(x)
0 1
±1 .5
±2 .2
±3 .1
±4 .06

Plotting and joining up these points gives the graph shown. But this is a bad
picture! Graphs of algebraic functions are always smooth, without sharp kinks.
What went wrong? I used a computer to draw the above graph – did the
computer screw up? No, I did. I should have told the computer to use many
more x values in its “table.” When I did that, the graph became smooth. You
could have guessed that the earlier sketch was wrong, if you remember that most
simple functions have smooth graphs. To draw a non-smooth graph you have to
know the reason why.

How do we know that the graphs of algebraic functions, such as 1/(1 + x2),
are smooth? This is a fact obtained from calculus. A brief discussion occurs later
in this chapter.

Problem 7.14 Sketch the graph of f(x)= 1/x for x> 0. Be careful about values of x
less than 1. Can you guess what the curve is?

Problem 7.15 If the graph of f(x) is as shown here, find the approximate values
of f(0), f(.5), and f(3). For which x does f(x) have the maximum value, and what is
the maximum value?

It is important to keep in mind the relationship between the graph of a

function f(x), and the values f(x) themselves. If the graph is displayed, then each
number x on the X-axis gives rise to exactly one point (x, y) on the graph, with y
= f(x). Thus the vertical distance (up or down) from the X-axis to the graph, as
measured by the Y-axis scale, at any value of x is equal to f(x). If f(x) > 0 then the
point (x, y) on the graph is above the X-axis, whereas for f(x) < 0 the point is
below the X-axis.
Properties of functions
1. Symmetry. We noted that the graph of f(x)= 1/(1 + x2) is symmetric by
reflection in the Y-axis. This occurs because replacing any number x by its
negative (−x) does not change the value of f(x). Any function f(x)with this
property is called an even function.

Here A is the domain of f.

Similarly, an odd function f is defined by

The graph of an odd function is symmetric by double reflection, first in the Y-

axis, then in the X-axis. This double reflection has the same effect as a 180°
rotation about the origin.
Problem 7.16 The graph of a certain function f is shown here for x ≥ 0. Sketch
the graph of f(x) for x < 0 if the function is (a) even, (b) odd.

Any polynomial function that contains only even powers of x is an even function
(because (−x)n = xn if n = 0, 2, 4,…). Any polynomial function that contains only
odd powers of x is an odd function (because (−x)n = −(x)n if n = 1, 3, 5,…).
Examples are shown above. Most polynomial functions are neither even nor odd.

Problem 7.17 Determine by inspection whether each of the functions is even or

odd, or neither. f(x)= 3x5 +4x; g(x)= (x2 − 1)2; k(x)= (x +1)/2x.
Do not attempt to draw the graphs.

2. Vertical asymptotes. Consider the function f(x) = 1/x2. Since division by zero
is impossible, f(x) is undefined for x = 0. The clue to graphing such functions is
this: division by zero is impossible, but division by a very small number produces a
very large number. For example 1/10−n = 10n (law of exponents). Here is a table of
values that makes use of this fact, for f(x)= 1/x2.
Since f is even, and f(x) > 0 for all values of x, the graph must look as shown.
We describe the property that f(x) becomes indefinitely large for x near 0, by
saying that this graph has a vertical asymptote at x = 0. The phrase f has an
infinite singularity at x = 0 is also sometimes used.
Vertical asymptotes occur in situations where division by zero would
otherwise be indicated. The explanation is the same as in the example: division
by a small number produces a large number. If the small number is positive, then
the large number is also positive. Or, if the small number is negative, then the
“large number” is actually a large negative number.

Example. f(x) = 1/(x2 − 1). This function is undefined at x = 1 and −1, so there
should be vertical asymptotes. If x > 1 then x2 − 1 > 0, so f(x) > 0. As x
approaches 1 (with x> 1), f(x) approaches +∞. Also, if x < 1, f(x) < 0, and f(x)
approaches −∞ as x approaches 1. Plotting a few points, such as f(0) = −1 and
f(2) = 1/3 pretty well fixes the entire graph as shown.
Problem 7.18 Graph f(x)= x +1/x. Suggestion: Note that, for large x we have f(x)
≈ x (because 1/x becomes small).

The solution to the preceding problems used the expression f(x) → +∞ as x →

0 (x > 0), which is read as “f(x) approaches (or goes to) plus infinity as x
approaches zero (x positive).” This is a succinct way of stating that the values y =
f(x) become progressively larger, without bound, as x becomes small (close to
zero), with x positive. In Problem 7.18 we could use a similar phrase f(x) → +∞ as
x → +∞ to indicate the behavior of the function (and graph) for large positive
values of x. How would you describe the graph for x < 0, using this kind of
notation? (Answer: f(x) →−∞ as x → 0 (x < 0); also f(x) →−∞ as x →−∞).
The symbol ∞ used here is, of course, not a number. We can’t add, multiply,
subtract, or divide ∞. Rather, the combined symbols →∞ are used as shorthand
for the phrase “becomes indefinitely large.”
Please observe how much information can be obtained about a graph from
“qualitative” considerations, such as symmetry and asymptotes. Often only a
very few actual numerical values (x, y) need to be located to complete a
reasonable sketch. If you own a graphical calculator, you can use it to confirm
the sketch. (The calculator may not show vertical asymptotes correctly,
however.)

3. Continuity and smoothness. A function f(x) is called continuous if its graph

is unbroken. A continuous function f(x) is said to be smooth if its graph does not
have sharp corners. Thus “smooth” means an unbroken graph without corners.
What kinds of functions are smooth? To begin with, the linear function f(x)
= ax + b is certainly smooth, since its graph is a straight line. Next, if f(x) and g(x)
are smooth functions, then the combined functions f(x)+g(x), f(x) · g(x), and
f(x)/g(x) are also smooth (except for points where g(x) = 0, in the latter case). We
cannot go into the detailed proof of these facts here, as the proof requires
techniques of calculus. However, the statement is hardly surprising.
It follows that any polynomial function, such as f(x)= 2x3 − x2 + 8, is smooth,
as is any quotient Q(x) = f(x)/g(x) of two polynomials, except for x-values at
which g(x) = 0. Looking at some examples, we have:
Example. The absolute value function |x| (“absolute value of x”) is defined by

Graphing this function shows that |x| is a continuous function (no breaks), but
not a smooth function, because of the corner at x = 0. We say that |x| is smooth
everywhere except at x = 0.

Problem 7.19 Sketch the graph of Comment.

4. Intercepts. The intercepts of a curve in the XY-plane are the points where it
crosses the axes. For the case of a function graph y = f(x), there is only one Y-
intercept, at y = f(0), assuming f(0) is defined. The X-intercepts, if any, are the
solutions of the equation y = 0, i.e. f(x) = 0. Solving this equation may be
somewhat difficult – see Section 7.7 for a computational algorithm. In some
cases, the equation f(x) = 0 can be solved by inspection or by factoring.

Example f(x) = x3 − 4x. This can be factored as f(x)= x(x2 −4), which equals 0 if x
= 0, 2, or −2. These are the X-intercepts. The Y-intercept is at y = f(0) = 0. Thus
the origin is both an X-intercept and a Y-intercept. Next, since f(x) is odd, we can
concentrate on x ≥ 0. If 0 < x < 2, then f(x)= x(x2 − 4) < 0, whereas f(x) > 0 for x>
2. Finally, for large x we have f(x) = x3 − 4x → +∞ as x →∞ (see below). The figure
shows all these features clearly. By the way, students of calculus can quickly find
the coordinates of the minimum point on this graph; they are (1.2, −3.1)
approximately. See Section 7.5 for details.

Problem 7.20 Sketch the graph of f(x)= (x2 − 1)/x.

5. Behavior for large x. Consider a polynomial function f(x)= anxn + an−1xn−1 + ···
+ a0. The behavior of f(x)as x → +∞ or x →−∞ depends entirely on the leading
term anxn (because the other terms are relatively small, compared to this term, for
large x). Now, as x → +∞ we have xn → +∞ also, so anxn → +∞ or −∞, depending
on the sign of an. Next, as x →−∞ we have xn → +∞ if n is even but xn →−∞ if n is
odd. The behavior of anxn as x →−∞ therefore depends on an and n.

Example f(x) = 2x5 − 10x4 + 5. Here f(x) → +∞ as x → +∞, and f(x) →−∞ as x →
−∞. The graph “wiggles” a bit (the wiggles can be figured out using calculus, as
explained in Sec. 7.5), but behaves as described, for x →±∞.
Problem 7.21 Determine the behavior of the following polynomials as x → +∞
and as x →−∞: (a) f1(x)= −x2 +8x + 1056; (b) f2(x)= 9x3 − 15x; (c) f3(x)= (x +2)5.

Problem 7.22 Sketch the graph of f(x)= x4 − 5x2 + 4. Suggestion: first find the
zeros of f(x) by factoring.

Example f(x) = x/(x2 + 1). As x →∞, f(x) ≈ x/x2 = 1/x for large x, because the
denominator term 1 is insignificant relative to x2 for large x. Therefore f(x) → 0 as
x →∞. We also have f(0) = 0, and f(x) > 0 for x > 0. What must happen is that
f(x)atfirst increases with x and then turns around and decreases to zero for large
x. A brief table of values is useful here.
Note also that f(x) is odd.

As you can see, many features of the graphs of algebraic functions can be
figured out from fairly straightforward principles. Fine details require either
numerical computation, or the use of calculus, or both. Section 7.5 discusses
some basic methods of calculus for polynomial functions, but a full course in that
subject is required to master these techniques in general. Of course, it is an
advantage to have a computer with graphical software, such as I used for all the
graphs shown here. But the computer isn’t much use by itself – you have to know
what you’re doing.

Problem 7.23 Sketch the graphs of the following functions. For each graph,
identify any symmetries, asymptotes, and intercepts. (a) f(x) = x3; (b)

Learning to sketch the graphs of functions is a worthwhile skill, even if you have
a graphing calculator, or computer. In many cases, making your own sketch can
be a lot quicker than typing the function into the calculator. In addition, gaining
familiarity with graphs of many kinds is useful in itself. For example, see if you
can think of a function f(x) whose graph looks like this:
One such example is f(x) = ax/(x + b), where b is a positive number. This graph
has asymptote y = a as x →∞, and also has f(0) = 0. By adjusting the value of b,
you can change the shape of the graph.

Transformation of functions
Suppose we have obtained the graph of a certain function f(x). What can we say
about the graphs of related functions, such as f(x) + 2, or f(x +2), and so on? We
will show that such graphs are obtained by transformation of the original graph,
specifically, by translation or scale changes.

Case 1. f1(x) = f(x)+ a. The graph of f1(x) has equation y = f(x) + a. Thus, if (x, y)
is a point on the graph of f(x), then (x, y + a) is a point on the graph of f1(x) = f(x)
+ a. In other words, the graph of f(x)+ a is just the graph of f(x)translated in the
Y-direction by the amount a (up if a > 0, down if a < 0).

For example, the graph of g(x) = x2 + 2 is just the parabola y = x2, translated
upwards 2 units.

Case 2. f2(x) = af(x). If (x, y) is a point on the graph of f(x), then (x, ay) is a point
on the graph of f2(x) = af(x). The figure shows an example with a > 1. (What
would this graph look like for a = ½? Or a = −½?) The graph of f(x) is obtained
from the graph of f(x)by scaling the Y-coordinate by scale factor a. If a < 0, this
also involves a vertical reflection in the X-axis.

Problem 7.24 Let f(x) = 1/(1 + x2) (discussed earlier). (a) Sketch the graph of f(x),
and of f(x) + 1 on the same XY-plane. What asymptotes are there? (b) Same for
f(x) and −2f(x).

Next we consider transformations involving the X-coordinate.

Case 3. f3(x)= f(x + a). (Study this case carefully, as many students find it
confusing at first.) Let (x, y) be a point on the graph of f3(x)= f(x+a). Then y = f(x
+ a), which is the value of f at x + a. This means that any point on the original
graph of f(x) is translated left by a units (if a > 0) to obtain a point on the graph
of f(x+a).
 For example, the graph of g(x)= (x +2)2 is a parabola with vertex at x = −2.
This graph is congruent to the parabola y = x2. Similarly, the graph of g(x) = (x −
2)2 is the parabola shifted right two units. If you understand and remember these
two examples, it will help you to keep in mind that the graph of f(x + a) is the
graph of f(x) translated a units to the left (if a > 0).

Case 4. f4(x)= f(ax). Let (x, y)be a point on the graph of f4(x) = f(ax). Then y =
f(ax), which is the value of f at ax. This means that the original graph of f(x) is
scaled in the X-direction, by the scale factor 1/a, to obtain the graph of f(ax).
For example, the graph of f(x)= (2x)2 = 4x2 is a parabola, obtained by
“squeezing” the parabola y = x2 by a factor of ½ in the X-direction. Likewise, the
graph of is the parabola obtained by stretching y = x2 by a factor
of 2 in the X-direction.

Problem 7.25 The figure shows the graph of sin x for 0 ≤ x ≤ 2π. (a) Redraw this
graph, and then superimpose the graph of (b) Same for sin (2x). (c)
Same for 2 sin x.
Problem 7.26 Given the graph of f(x), which of the following functions have
graphs that are congruent to that of f(x)? Similar? Explain. (a) f(x)+ a; (b) f(x +
a); (c) f(ax); (d) af(x).

7.4 Inverse functions

Let f be a given function, with domain A and range B. The inverse function f−1,
with domain B and range A, is defined by the following statement:

Here x ∈ B and y ∈ A. The expression f−1(x) is read as “f inverse of x.”

(Please don’t get hung up by the change in the roles of x and y. Until now we
have usually had x ∈ A and y ∈ B, with y = f(x). In this section we have to be
more flexible.)

Example 1. Let f(x) = 2x +1 (x ∈ (−∞, +∞)). What is f−1(x)? To answer this, we

apply Eq. 7.5: y = f−1(x) means that x = f(y) = 2y +1. We solve this for y, obtaining
Therefore
What is the procedure used in this example? Given a function y = f(x), how
do we determine the inverse function f−1(x) as a function of x? The method is
this:
1. Write that y = f−1(x) means that x = f(y).
2. Solve the equation x = f(y)for y in terms of x.
Try this yourself, in the next problem.

Problem 7.27 Given f(x)= x3, find f−1(x).

This all seems quite straightforward. However, in trying to solve x = f(y) for y,
two things may go wrong. The first is that there may be more than one solution
y, implying that y = f−1(x) has more than one value. This is not allowed: a
function (here y = f−1(x)) must always have a unique value. If the equation x =
f(y) has more than one solution y for some value of x, then the original function
f(x) does not have an inverse. (The second difficulty, actually solving x = f(y)for
y, will be discussed later.)

Example 2. Let f(x) = x2 for all x ∈ (−∞, +∞). Then y = f−1(x) would mean that x
= f(y), or x = y2. This would imply that an ambiguous (and therefore
incorrect) result.

The problem in this example, as in any such example, is that the given function
f(x) = x2 produces the same value y = f(x) for two different x’s. For example 4 = 22
= (−2)2, so we can’t solve 4 = y2 uniquely for y. Let us look at this example
graphically. The graph of f(x)= x2 is shown in the first figure. The equation we
want to solve for the inverse function is x = y2, shown in the second figure. For
each x> 0 there are two solutions for y, so the inverse function is undefined
(being ambiguous).

The difficulty, in this example, stems from the fact that the original function f(x)
= x2 is not “one-to-one.” A function f(x) is called one-to-one if, for each value of
y in the range of f, there is just one value of x in the domain of f, such that y =
f(x). This is not the case for f(x)= x2, as the upper graph shows. For a given
function f(x) to have an inverse function, f(x) must be one-to-one. Then the
equation x = f(y) has a unique solution y, which equals f−1(x)by definition. To
summarize:
 Now, how do we tell if a given function f(x) is one-to-one? The answer is
quite simple: a (continuous) function is one-to-one if and only if it is monotonic.
“Monotonic” means that f(x) is either entirely increasing, or entirely decreasing.
If you stop and think about it, this statement should be clear.

Certainly a monotonic function is one-to-one. And equally certainly a non-

monotonic function is not one-to-one.
For example, f(x)= x3 is monotonic (think of its graph), so it is one-to-one,
and has an inverse However, f(x) = x2 is not monotonic, not one-
to-one and does not have an inverse (as it stands).
 Finally what should one do if one wants to define an inverse function for a
function f(x) that is not monotonic? The answer is, restrict the domain of f so as
to make it into a monotonic function.

Example 3. We saw that f(x) = x2 is not one-to-one, at least if the domain of f is

(−∞, +∞). But if we take f(x) = x2 with domain [0, ∞), this function is one-to-one
(see the previous figure). Its inverse function is the usual square
root function. To show this, we apply our basic procedure of solving x = f(y), or
in this case, x = y2 (y ≥ 0). The solution is there is only one solution
because of the domain restriction on f(y), which says that y ≥ 0.
This method of restricting the domain of a function f(x) in order to be able to
define the inverse function, is used for the inverse trigonometric function Asin x
(same as sin−1 x), and so on. See Chapter 8 for the details.

Problem 7.28 Let f(x) = 4x − x2, for 0 ≤ x ≤ 4. Sketch the graph of f(x). Show
how to restrict the domain of f(x) so as to obtain a one-to-one function. Find the
inverse of this one-to-one function, and sketch its graph on the same XY-plane.

The graph of the inverse function

Consider a given one-to-one function f. Its graph has the equation y = f(x). The
graph of the inverse function f−1 has the equation y = f−1(x), which means that x =
f(y). Thus if the point (x, y) is on the graph of f, then (y, x) is on the graph of f−1.
Now the point (y, x) is the reflection of the point (x, y) in a 45° line through the
origin, as the figure indicates.
(As always, a numerical example can help in understanding this. Try (x, y)= (2,
1).)
Note that the scales on the X and Y axes are assumed to be the same here.

Consequently, the graphs of f and f−1 are reflections of each other in the 45° line, y
= x. The figure shows this for the example f(x) = x2 (x ≥ 0). Another example was
shown in the solution of Problem 7.28. Further examples will be encountered
later.

Problem 7.29 If for 0 ≤ x ≤ 1, find f−1(x) and sketch.

Problem 7.30 Continuing with the preceding problem, show that each of the
following functions is its own inverse, and sketch; (a) f(x)= 1/x (x> 0);

Example 4. Let f(x)= x3 for −∞ <x < ∞. Then also for −∞ <x <
∞. Note that

and also
This is entirely typical: for any one-to-one function f we always have

This is one way to check that you have calculated f−1(x) correctly. Note also
that the domain of f−1(x) equals the range of f(x) and vice versa.

Problem 7.31 Let Find g−1(x), and verify that g

−1
(g(x)) = x. Sketch the graphs of g(x)and g−1(x). What is the domain of g−1(x)?

To prove Eq. 7.6 we use the basic definition of an inverse function. Thus, to
prove that f−1(f(x)) = x, let z = f(x). Then x = f−1(z) = f−1(f(x)) as required.

Problem 7.32 Prove that f(f−1(x)) = x for a one-to-one function.

Equation 7.6 emphasizes the fact that the inverse function f−1 of a given function f
reverses the action of f. In the figure, the action of f−1(f(x)) starts with x, then
produces y = f(x), then produces x = f−1(y) again, as the arrows show. (What does
f(f−1(y)) do here?)

Some students find inverse functions a bit confusing. Perhaps a useful

example to keep in mind is the calculator function ASIN or sin−1 (see Sec. 5.4).
This is used for the purpose of solving the equation y = sin θ, when y is known.
In other words, ASIN reverses the procedure of calculating sin.

Inverse transformations
Consider the equations of a translation in the XY-plane
This translation is in fact a function, but of a somewhat different kind than the
functions f(x) considered above.

Denote this function (translation) by Ta,b. Thus

In words, the function Ta,b shifts the point (x, y) to a new point (x + a, y+b).
[If you find this example confusing, recall that in general, a function can
transform any kind of mathematical object, into another mathematical object.
Here the objects are points (x, y) in the plane.]
Can you figure out what the inverse function is? Look at the preceding
figure. Keep in mind that the inverse of a given function reverses the action of
the function. What is the reverse operation to translating (x, y) to (x + a, x + b)? I
hope it is clear that

so that The inverse function translates points in the

opposite direction, compared to Ta,b. If this is less than obvious to you, you can
check it by using the basic procedure for finding inverse functions. First, we
write that means that(x, y)= Ta,b(u, v)= (u+a, v+b). Therefore
u+a = x and v+b = y. This gives(u, v)= (x−a, y−b). Therefore T −1(x, y)= (u, v)= (x
− a, y − b)= T−a,−b(x, y). (The explanation is more complicated than the fact,
perhaps.)
Problem 7.33 Let Rθ denote the rotation of points (x, y) in the counterclockwise
direction about the origin, through angle θ. What is (Note: positive angles
are measured counterclockwise, and negative angles clockwise.)

7.5 Graphing sums and products

Consider the function A convenient method for sketching the
graph is to combine the graphs of y = x and y = 1/x: We first sketch these graphs.
Next for various values of x, we graphically add the two y-values. This method is
usually both faster and more revealing than attempting to sketch the whole graph
x +1/x at once. A little practice will help.
Problem 7.34 Sketch the graph of

You can also develop your skills at sketching graphs involving other
combinations. For quotients f(x)/g(x), with g(x) > 0, it helps to realize that 1/g(x)
< 1 if g(x) > 1 and vice versa. Also 1/g(x) → 0 if g(x) →∞. The figure shows the
graph of 1/(1 + x4) thought of as 1/g(x), where g(x) = 1+ x4.

Composition of functions
We next discuss another common way of combining functions, called
composition.

Example 1. Let f(x)= 1/(1 + x) and g(x)= x2. Then

To check that you understand this, try to find g(f(x)). (Answer below.)
Definition If f and g are two given functions, the composition f o g [which is
read as “f-oh-g”] is the function defined by
 Thus, in Example 1 we have f o g(x)= 1/(1 + x2). Also,

What are the natural domains of f o g and g o f, in this example? By

inspecting the formulas obtained for these compositions, we see that the natural
domain of f o g is the set of all real numbers, and the natural domain of g o f is all
real numbers except −1.
Problem 7.35 Let and v(x)= 1 −x3. Find formulas for uo v(x) and v o
u(x), and determine the natural domains of these functions.

We can picture the composition f o g like this:

In a sense, the function f o g goes “directly” from x to f o g(x). Given the

formula for some function, for example we don’t necessarily think of
this as coming from the composition of two functions. In calculus, however, you
will need to recognize expressions like in terms of composition
of functions. What would f and g be, here? Answer: and g(x)=
2 2
x . Alternative answer: and g(x)= 1+ x . Be sure that you understand
this point.

7.6 Polynomial calculus

Recall the concept of the slope m of a straight line:

The slope m measures how steeply the line rises (if m > 0) or falls (if m< 0). A
line with zero slope is horizontal.
What would we mean by the slope of a curve, such as y = x2? It should be clear
that the slope changes from one point to another. The slope m is now a function
of x, rather than a constant.

But what is the slope of a curve, exactly? And how can we calculate it? Here’s
the general method. We consider a given point P on the curve y = f(x), and wish
to determine the slope mP of the curve at point P. To calculate mP we consider a
nearby point Q. Let mPQ denote the slope of the line PQ; we can calculate mPQ
from the basic formula, Eq. 7.9.

Finally, we allow Q to approach P indefinitely closely. The limiting value of

the slope mPQ is defined as the slope of the curve at the given point P :

(This is read as “limit of mPQ as Q approaches P.”) Let us apply this definition
to the parabola y = x2. Let P be the point on the parabola with coordinates (x, x2).
Let Q be a neighboring point (x + h, (x + h)2). Then

As h → 0the point Q approaches P. Therefore the slope of the parabola at P is

To summarize, the slope of the parabola y = x2 at a point P = (x, y)is equal to

2x. To make sense of this result, do the following experiment.

Problem 7.36 Make a reasonably accurate freehand sketch of the parabola y = x2

for −2 ≤ x ≤ 2, using the same scale on both axes. Now estimate by eye the slope
of this curve at points P having x = −2, −1, 0,.5, and 1.5. Compare these estimates
with the exact slopes given by m = 2x.

The derivative of a function

We now reformulate the above definition of the slope of a curve, using the
notation of functions. To be specific, let f be a given function. Let P = (x, y) be a
given point on the graph of f; thus P = (x, f(x)). Let Q = (x+h, f(x+h)) be a
neighboring point on the graph. Then the slope of the chord PQ is

You should identify the sides of the right triangle shown in the figure; they
are Δx = h and Δy = f(x + h) − f(x). Therefore equals the expression
above. The expression on the right side of Eq. 7.10 is called the difference
quotient. The limit of mPQ as h → 0 is the slope of the curve y = f(x)at point P =
(x, f(x)). This quantity (the slope) is denoted by f′(x):

Here f′(x) is read as “f prime of x.” The function f′(x) as defined by Eq. 7.11 is
called the derivative of f(x).
To summarize, if a function f(x) is given, then its derivative f′(x) is defined by
Eq. 7.11. The derivative is also a function; it represents the slope of the curve y =
f(x), at each given value of x. For example, if f(x) = x2 we showed above that f′
(x)= 2x.

Problem 7.37 Show that if f(x) = x3 then f′(x) = 3x2. Suggestion: use the basic
definition, Eq. 7.11, and emulate the calculation for y = x2.

The definition of the derivative, Eq. 7.11, is basic in calculus. A college

calculus course will spend the entire first semester studying the implications and
applications of this basic concept. Historically speaking, the derivative was
introduced by Isaac Newton (1643-1727) and Gottfried Leibniz (1646-1716).
Perhaps more than any other mathematical concept, the derivative has proved to
be fundamental in modern Science.
In this book we will only show how to calculate f′ (x) for simple functions.
We also show how knowledge of f′(x) can assist you in understanding the graph
of f(x).

The tangent line to a curve

The tangent line to a curve y = f(x) at a point P0 = (x0,y0) is the line through P0
that has the same slope as the curve at that point. For example, the tangent line
to the parabola y = x2 at the point (1,1) has slope m = 2x = 2, so the equation of
this tangent line is
In general, to find the equation of the tangent line to a given curve y = f(x) at a
given point x = x0, weuse thefact that the slope m of the curve (and its tangent
line) at x0 is equal to f′ (x0):

Consequently the equation of the tangent line is y − y0 = m(x − x0) = f′ (x0)(x −

x0), where y0 = f′ (x0). We next explain how to calculate f′(x0) for polynomial
functions.
Before continuing, however, let us repeat the basic definition of the tangent
line to a curve. The tangent line, at a given point on the curve, is the line that
1. goes through the point, and
2. has the same slope as the curve, at that point.
I find that many of my calculus students, having studied tangents to circles at
school, have memorized the definition that “the tangent to a curve is the line that
touches the curve at a single point.” Well, this happens to be true for circles (and
a few other curves), but is not true in general. Worse, this definition conceals the
main feature of tangent lines, which is that they have the same direction as the
curve, at the point of tangency. This is what you need henceforth to keep in
mind: the same direction as the curve.

The derivative of a polynomial

We next obtain the formula for the derivative of f(x)= xn. First we calculate the
difference quotient

where the terms represented by ··· all involve a factor of h, h2,etc. (Please review
the Binomial theorem, Eq. 4.38, if you have forgotten it, or don’t understand the
last statement.) These terms (···) therefore approach zero as h → 0, and this shows
that f′ (x)= nxn−1. We emphasize this result.

For example, the derivative of x2 equals 2x, while the derivative of x3 equals
3x2, and so on. As a review, make a sketch of the graph of f(x)= x3. Is it obvious
that the slope of this graph is ≥ 0 for all x? What is the slope at the origin?
(Answer: 0.)

Problem 7.38 Explain the significance of Eq. 7.12 for the cases n = 0 and n = 1.

Let us show next that

and

To prove Eq. 7.13 we use the definition, Eq. 7.11:

The proof of Eq. 7.14 is similar.

Formulas 7.13 and 7.14 are among the basic rules of calculus. Here we will
only apply them to polynomials.

In other words, a polynomial function can be “differentiated term-by-term.”

For example, the derivative of f(x) = x3 − 2x is f′(x) = 3x2 − 2. We show below
how this information is used in sketching the graph of f(x).
The proof of the Theorem is an immediate consequence of Eqs. 7.12 to 7.14.
Thus Eq. 7.12 shows how to differentiate each power xj, while Eq. 7.13 shows that
the derivative of axj equals jaxj−1. Finally Eq. 7.14 says that we can add
derivatives.

Problem 7.39 Find f′(x) given that (a) f(x) = x2 − 4x +2; (b) f(x) = 3x4 +4x3 +8x2 −
16x + 100.

Problem 7.40 (a) Show by counterexample that (fg) ≠ f′g′. Suggestion. Try very
simple functions f and g. (b) The correct formula is (fg) = f′g + f′g. Verify this
formula for f(x)= x2 +2 and g(x)= x.

Max and min points

One of the most common uses of calculus is in solving maximization problems.
Imagine a smooth curve y = f(x). Suppose that this curve has a local maximum
value at x = x0 – see the figure. What can you say about the slope of the tangent
line at this point? Try to answer this before reading further.
 This condition allows us to determine local max or min points of f(x) exactly,
provided we can calculate f′(x) and solve the equation f′(x) = 0 for x. (Don’t let
the phrase “necessary condition” throw you. It just means that f′(x0) = 0 must be
true if x0 is a local max or min of f(x).)

Example f(x) = x3 − 3x. To sketch the graph of f, we first note that f is odd. Since
f(x) = x(x2 − 3), the graph intersects the X-axis at x = 0 and Also, f(x)
→ +∞ as x → +∞. Since f(1) = −2, the general shape of the graph is as shown. To
locate the minimum point, we calculate f′(x) = 3x2 − 3 = 3(x2 − 1). Thus f′(x)= 0
for x = ±1. The local min is therefore at x = 1,y = f(1) = −2. The point (−1, 2) is a
local max. The figure shows all these features.
 Another example, f(x)= x4 − 5x2 + 4, was considered in Problem 7.22. Let us
now find the local max and min points on the graph of f(x). We have

Thus f′(x) = 0 for x = 0 and Because of other features

of the graph (see Solution 7.22), we conclude that x = 0 corresponds to a local
max, and the values x = ±1.58 are local minima.

Problem 7.41 Determine the coordinates of the minimum point on the parabola
y = x2 − 4x + 1 (a) by completing the square; (b) by using the derivative.

Problem 7.42 The figure shows the graph of a certain function f(x). The function
is even, and has a horizontal asymptote at y = 2. Make an approximate sketch of f
′(x).

Problem 7.43 (a) Use the definition, Eq. 7.11, to show that the derivative of f(x)
= 1/x is f′(x) = −1/x2. (b) Locate the minimum point on the graph of g(x)= x +
(1/x)for x> 0.
 Recall from Eq. 7.12 that (xn)′ = nxn−1 for n = 0, 1, 2,…. This equation is also
true for negative integers, n = −1, −2,… Problem (a) covers the case n = −1: (x−1)′
= (−1)x−2 = −1/x2. You may wish to do the proof for the general case for yourself.
Write n = −m and put xn = 1/xm. Use Eq. 7.11 and some algebra.

Other notations for the derivative

If y = f(x), the derivative f′(x) is also denoted by y′, dy/dx,or(d/dx)f(x), to mention
the most common alternatives. These symbols all refer to the same thing, the
derivative of f(x). The derivative can be interpreted as the instantaneous rate of
change of y with respect to x, and this is important in Physics. Since we are not
studying calculus in detail here, we do not follow up on these important ideas.

Limits
The definition of the derivative uses the idea of a limit – see Eq. 7.11. But what
exactly is meant by the term “limit?” And how are limits calculated?
In general, this can be a bit complicated. (Indeed, nearly 200 years transpired
between the invention of Calculus, by Newton and Leibniz, and the development
of a rigorous theory of limits by 19th century mathematicians.) However, the
basic idea is that the statement

means simply that the values of f(x) approach the limit L, as x approaches x0.
More precisely, f(x) becomes arbitrarily close to L, provided that x is sufficiently
close to x0.
In some cases, the value of L is quite obvious. For example, what do you
imagine is the value of

Answer: 11, just what you get by substituting x = 3 here. This works because
f(x)= x2 + 2 is a continuous function.
Limits don’t always work out this easily. For example, what is

Substituting x = 2 now gives the meaningless result What does this say about
the limit? One possibility is that since 0/0 is meaningless, the limit does not exist.
However, let us consider some values of x near 2:

It looks as if the limit equals −1.0. Can you see how to explain this?
Perhaps you noticed that the denominator can be factored: x2 −5x+6 = (x −
2)(x − 3). Therefore we have

Here, after canceling the factors (x − 2), we get an expression for which the
limit can be found by substitution. This is a common situation.

Problem 7.44 Find

Note in the above example (and in the Problem) that the function F (x) in
is undefined at x = x0. In these examples, preliminary algebraic
simplification gets rid of the difficulty, after which the limit can be found by
substitution. Exactly the same situation arises necessarily in calculating a limit
involved in finding derivatives. For example, given g(x)= x2, we have
Here we cannot substitute h = 0, yet (why not?). Algebra comes to the rescue,
giving us

Problem 7.45 Let Find w′(x) on the basis of the definition, Eq.
7.11; comment on the “disappearance” of the troublesome h. Suggestion:
rationalize the numerator in the limit expression.

In all the above examples, direct substitution does not work because the
expression whose limit we seek is undefined at the limit value. This may seem
bizarre, but in fact it is the most common and useful case. In the derivative
formula

substituting h = 0 (where h = Δx) would correspond to taking Q = P in the

diagram for mPQ = slope of chord joining points P and Q on the graph of f(x).
This would clearly make no sense; instead we take Q ≠ P but let Q → P by sliding
along the curve. This is the same as letting h → 0. But we can’t begin by
substituting h = 0 (though this turns out to be OK after the algebra). Be sure you
understand this important point.
Problem 7.46 If find f′(x) from the basic definition, Eq. 7.11.

Problem 7.47 (Tangent to a circle.) Consider the function

(a) Sketch the graph of y = f(x), −a ≤ x ≤ a. Hint: first identify this curve. (b)
Find f′(x) (this is similar to Problem 7.45). (c) Show that the tangent line at the
point (x0,f(x0)) is perpendicular to the radius line through that point. (Recall that
we proved this earlier, see Sec. 5.6.)

7.7 Numerical solution of equations

Many problems and applications of math involve the solution of an equation f(x)
= 0, where f(x) is some given function. For example, Chapter 4 described
algorithms for solving linear equations (ax+b = 0), and quadratic equations (ax2
+ bx + c = 0). In addition, Chapter 4 discussed the solution of polynomial
equations by the method of factoring. This method is not a complete algorithm,
however, because not every polynomial can be factored by sight.
Given the function f(x), values of x for which f(x) = 0 are called the zeros of f.
Thus, to find all the zeros of f means the same as solving the equation f(x) = 0 for
all possible values of x.
Assume now that f is a continuous function of x; recall that this means that
the graph of f is unbroken. Suppose that we have determined two values, x1 and
x2 (with x1 < x2) such that
Then because of continuity, the interval (x1,x2) must contain at least one zero of f.
Thus we have already approximated this zero of f to within an accuracy of x2 − x1.
How can we improve on this approximation?
Let x′ = 1/2(x1 + x2) so that x′ is half way between x1 and x2. (Do not confuse x′
with the derivative of something; x′ is just a real number.) Calculate f(x′). There
are three possibilities:

(Check that these statements are true, by looking at the graph of f(x).) In the
first case, we have determined a zero exactly. In the remaining cases, we have
improved on the accuracy of the approximation by a factor of 1/2. By repeating
the process n times, we can approximate the zero of f to within (½n)(x2 − x1). This
tolerance can be made as small as we wish, by taking n sufficiently large. The
procedure is called successive bisection.
Let us express the process as an algorithm. Let t be the required tolerance, or
accuracy of the approximation. First, we determine (by trial and error) values x1
< x2 such that f(x1) < 0 and f(x2) > 0. [Alternatively it may happen that f(x1) >
0and f(x2) < 0; the algorithm is easily adjusted for this case. Or else you can
replace f(x) by −f(x).]
1. Let and calculate f(x′).
2. If f(x′) = 0, stop: x is a zero of f.
 3. Else: if f(x′) < 0 change x1 to x′, leaving x2 unchanged;
Else: if f(x′) > 0 leave x1 unchanged and change x2 to x′.
The new interval (x1,x2) contains a zero of f.
4. If x2 − x1 ≤ t, stop. Else go to step 1.
This algorithm is well suited to computers, or programmable calculators. In
fact, many advanced scientific calculators have built-in equation solving
programs, which may be based on this algorithm, or on some similar method.
Example Solve x3 + x − 1 = 0. Here f(x) = x3 + x − 1 and we see that f(0) = −1 and
f(1) = 1. Therefore there is a zero of f between x = 0 and 1. Also, f′(x) = 3x2 +1 > 0
for all x; this tells us that the graph of f(x) has positive slope at every point. Thus f
can have only one zero.
Here is the result of the algorithm with a tolerance of .01; only two decimal
places are shown.

After 4 steps we know that the solution x of our equation f(x) = 0 is between x =
.63 and x = .69. Two more steps locates the solution between .68 and .69. In other
words, x = .68, with further digits unknown as yet.
By continuing the algorithm, we could calculate the solution x to any desired
accuracy. From an HP Scientific calculator I obtained the approximation x =
0.68232780383, for which f(x) = 0 to 11 decimals. This calculator uses the
successive subdivision algorithm, and takes several seconds to calculate the
result.

Problem 7.48 Explain why the above algorithm works unchanged in the
alternative case f(x1) > 0, f(x2) < 0, provided we replace f(x) by −f(x).
Problem 7.49 Solve cos x = 2x for x> 0 (remember to use radian mode), using a
tolerance of .1.

7.8 Interpolation
Suppose y = f(x) is a given function. We suppose that the values of f(x) are given
only for certain values of x. We wish to find approximate values of f(x)elsewhere.
For example, suppose that a boy’s height H has been measured on his
birthday:

age, A height, H
12 yr 152 cm
13 yr 159 cm
14 yr 168 cm

How tall was he at age 12 yr 3 months? At what age did he reach 165 cm
height? We can answer these questions, approximately, by using interpolation
(more precisely, linear interpolation). This amounts to assuming, for the sake of
calculation, that the boy grew at a steady rate between ages 12 and 13, and at a
(different) steady rate between ages 13 and 14.
Can you do the first calculation in your head? At age 12 1/4 the boy should
have gained about 1/4 of his height increase for the year, which was 7 cm.
Therefore his height at at age 12 yr 3 mo was approximately 152 + (1/4) × 7 =
153.8 cm. To check that you followed this argument, try to calculate the boy’s
height at 13 yr 6 mos, and 13 yr 9 mo. Answer: at 13 yr 6 mo his height was
approximately 159 + (½) × 9 = 163.5 cm, and at 13 yr 9 mos it was approximately
165.8 cm.
 A similar calculation applies to the second question. Since 165 cm represents
6/9 of boy’s total height gain between ages 13 and 14, he reached this height at
about 13 + (6/9) = 13 yr 8 months.

Problem 7.50 Given that find an approximate value

for by mental calculation based on interpolation. How good is the
approximation?

To obtain the general formula for linear interpolation, we assume that f(xi)=
yi for two values x1 < x2. Then for x1 < x <x2, the formula is

Notice that if we write y instead of f(x), then Eq. 7.17 is just the equation of
the straight line joining the two points (xi, yi). The above examples of
interpolation used this equation, in a mental calculation. For example, in the
boy’s height problem we have (A1,H1)= (12, 152) and (A2,H2)= (13, 159). With A
= 12 1/4, Eq. 7.17 gives H = 152 + (7 × 1/4), as before.
Before the availability of hand calculators, students and scientists spent many
painful hours interpolating from the values printed in trigonometric and
logarithmic tables. Such tables listed values of these functions to 3 or 4 decimals,
but by interpolating, one could improve on this by one additional decimal.
Fortunately, no one has to do these tedious calculations nowadays. (It was the
practical impossibility of doing by hand the vast calculations needed in designing
the H-bomb, that led to the first “electronic brains” – now called computers.)
Nevertheless, interpolation is still sometimes useful. We end this chapter with
one example.

Solving equations by repeated interpolation

We again consider the problem of numerically solving an equation f(x)= 0. We
assume that two values x1,x2 have been determined so that

Recall that the interval bisection algorithm began by calculating f(x), where x
is the midpoint between x1 and x2. This allows one to narrow the interval
containing the solution of f(x) = 0 by one-half in each repetition. A more
efficient algorithm obtains x by interpolation. As the figure suggests, this will
usually yield a closer approximation to the zero of f(x), than the midpoint. The
algorithm then proceeds as before, by altering the interval [x1,x2] according to the
sign of f(x). This calculation is more efficient than interval bisection. Work out
the details for yourself, if interested.

7.9 Review problems

1. A set A has 6 elements, and B has 8. What is the largest and smallest possible
number of elements in the sets (a) A ∪ B; (b) A ∩ B; (c) B − A = {x : x ∈ B and
x /∉ A}?
2. Find the natural domains of the functions

3. Which of the following sets of ordered pairs is not a function? (a) {(2, 1), (3, 1),
(4, 1), (5, 1)};(b) {(1, 2), (1, 3), (1, 4), (1, 5)}.

4. For any finite set A of real numbers, define max (A) to be the largest element
of A. Is max a function? What is its domain?
5. Sketch the graphs, and identify any symmetry, zeros, and asymptotes. (a) f(x)
= 4 − x2; (c) h(x) = x3 − 3x2 + x. (Suggestion: use the
quadratic formula in finding the zeros of h(x).)

6. Find the coordinates of the vertex of the parabola y = x2 +2x +3 (a) by

completing the square; (b) by calculus.
7. Let f(x)= x3 − 6x2 + 9. (a) Find any local max or min points on the graph of f.
(b) Using the result in (a), explain why f has at least three zeros. (Do not try to
calculate these zeros, although this is possible by using the successive bisection
algorithm.)

8. The function f(x) has its graph as shown (the graph repeats to the right and
left). Sketch the graphs of (a) 2f(x); (b) f(2x).

9. Let f(x) = 1 − 1/x for x ∈ (0,∞). (a) Sketch the graph of f, and identify its
asymptotes. (b) Find the inverse function f−1(x), and sketch its graph. How do
the asymptotes of f−1 relate to those of f ?
10. Let f(x) = x2 +1/x (x > 0). (a) Sketch the graph of f(x) by adding the graphs of
x2 and 1/x. (b) Using calculus, find the local minimum point of f(x).

11. Use the definition in Eq. 7.11 to find f′(x) if f(x) = 1/x2.
12. Same for f(x) = 1/(x2 +2).
Solution 7.1 All are true except (c).
Solution 7.2 A is the circle of radius 4, centered at the origin, in the XY-plane. B
is the line with slope 1 and Y-intercept 2, in the XY-plane.
Solution 7.3 (a) For any x, x belongs to A ∪ B if and only if x belongs to A or x
belongs to B. (b) For any x, x does not belong to A ∪ B if and only if x does not
belong to A and x does not belong to B. (c) This says that the solutions of the
equation x2 = 1 are x = ±1.

Solution 7.4 For simplicity, assume a ≤ c (the opposite case is similar). A sketch
shows that there are 3 possibilities:

Solution 7.5 The same 3 cases arise:

Solution 7.7 (a) f(−2) = 12, obtained by substituting x = −2 in the formula f(x) =
3x2. (b) p(64) = 1, because 64 = 26, which has only one prime factor, 2. (c) A(t1)=
1.5, because t1 has base b = 1 and height h = 3.

Solution 7.8 sin and 10x have domain (−∞, ∞), the set of all real numbers
(actually the set of all numbers that the calculator is capable of working with). 1/x
has domain all real numbers except 0. has domain [0, ∞), the set of all non-
negative real numbers. log x has domain (0, ∞), the set of all positive real
numbers. Asin has domain [-1,1], the set of real numbers satisfying −1 ≤ x ≤ 1.

Solution 7.9 (a) Yes: G is a set of ordered pairs, and no two elements of G have
the same first number and different second numbers. Here G(1) = 3, and so on.
(b) No: h is not a function, because h(2) is not uniquely specified.
Solution 7.10 (a) M gives the midpoint of the line segment joining two given
points (x1,y1)and (x2,y2). Its domain is the set of all pairs of points in the XY-
plane. It could be defined by writing

(b) Here, f(x) is undefined if x2 − 1 = 0, i.e., if x = 1 or −1, but otherwise f(x) is

defined. The natural domain of a function f(x) that is defined by a formula is the
set of all x for which the formula gives a defined number. In this example, the
natural domain of f is the set of all real numbers except for ±1.
Solution 7.11 C(x0,y0,r) = {(x, y) : (x − x0)2 +(y − y0)2 = r2}. It is assumed that r ≥
0.

Solution 7.12 (a) Rf = (−∞, ∞), because any real number y equals 2x+1 for some
value of x, specifically x = (y − 1)/2. (b) Rf = [1, 3], because f(0) = 1 and f(1) = 3,
and the values of f(x) increase from 1 to 3 as x increases from 0 to 1.

Solution 7.13 The domain of G is the set {−1, 1, 4, 7} and the range RG is {1, 2, 3}.

Solution 7.14 First, as x becomes large, f(x)= 1/x becomes small. Also f(1) = 1.
Next, as x gets small, f(x)= 1/x grows larger and larger. Here is a representative
table:

The figure shows this graph. Note that the curve is asymptotic to the X-axis as x
approaches +∞. It is also asymptotic to the Y-axis, as x approaches 0. This may
remind you of a hyperbola, which is what this graph is. If you include the graph
for x < 0, you get both branches of the hyperbola. Try it.
Solution 7.15 f(0) = −.4, f(.5) = .2, f(3) = −.1. The maximum of f(x) occurs for x
= 1, with f(1) = .34. (All values are approximate.)

Solution 7.16

Solution 7.17 f(x) is odd; g(x)and h(x) are even; k(x) is neither.

Solution 7.18 There is a vertical asymptote at x = 0, and f(x) → +∞ as x → 0 (x >

0). Also, the graph is asymptotic to the line y = x, as x → +∞. The point (1, 2) lies
on the graph. Finally, f is odd, so a congruent branch exists in the 3rd quadrant.
(The curve is a hyperbola.)
Solution 7.19 First, for x ≥ 0, y = This is the upper part of the
2
parabola x = y . For x < 0, is the upper part of the parabola −x = y2. The
two parts join up at the origin, where the curve has a vertical tangent. An
infinitely sharp point like this is called a cusp. Atanyrate, f(x) is certainly not
smooth at x = 0.

Solution 7.20 f(x) is odd, with vertical asymptote at x = 0, and X-intercepts at x =

±1. For large x, f(x) ≈ x. (It’s another hyperbola.)
Solution 7.21 (a) f1(x) →−∞ as x → +∞ and as x →−∞; (b) f2(x) → +∞ as x → +∞,
and f2(x) →−∞ as x →−∞; (c) f3(x) → +∞ as x → +∞, and f3(x) →−∞ as x →−∞ (the
leading term of f3(x) is x5).

Solution 7.22 We have f(x)= (x2 − 4)(x2−1), so that f(x)= 0for x = ±2 and ±1.
Also, f is even, and f(x) → +∞ as x → +∞ and as x →−∞. (The minimum points on
the graph are at this can be shown using elementary
calculus – see Sec. 7.5.)
Solution 7.23 (a) f(x) = x3 is odd, has X-and Y-intercept at (0,0), and f(x) →∞ as x
→∞. There are no asymptotes. is even. After squaring
2 2
both sides of we get y − x = 1, which is a hyperbola. The graph of
g(x) is the upper half of this hyperbola. It has asymptotes y = ±x. (c) h(x) = x2/(1
+ x2) is even. Also h(x) → 1 as x → ±∞, so the line y = 1 is an asymptote. Both X-
and Y-intercepts are at (0,0). (d) k(x) is undefined if x2 ≤ 1, that is, for x ∈ [−1,
1]. It has vertical asymptotes at x = ±1, and k(x) → 0 as x → −∞. The function is
even.

Solution 7.24 (a) The graph of f(x) + 1 is the graph of f(x) shifted one unit up;
there is now a horizontal asymptote at y = 1; (b) In this case, the asymptote
remains at y = 0. The curve lies below the asymptote, reaching a minimum value
−2.
Solution 7.25 The graph of sin x in fact repeats cyclically for other values of x, so
that these transformed graphs also repeat cyclically. See Chapter 8.)
Solution 7.26 (a) Congruent; because this is the graph of f(x) translated
vertically; (b) congruent; graph translated horizontally; (c) neither congruent nor
similar; graph is scaled in the X-direction only, which changes its shape; (d)
neither congruent nor similar.

Solution 7.27 First, y = f−1(x) means that x = f(y) = y3. Therefore

Solution 7.28 The graph of f(x) is an inverted parabola (below). To obtain a

monotonic, one-to-one function, we restrict the domain of f to 0 ≤ x ≤ 2. The
equation of f−1 is then obtained from x = f(y)= 4y − y2. Solving for y by the
quadratic equation gives To get 0 ≤ y ≤ 2, we take the minus
sign. Thus

Solution 7.29 (a) The solution of is

−1
Therefore f (x)= f(x): the function f is its own
inverse. The common graph is that of a quarter-circle, centered at the origin;
note that this curve is symmetric by reflection in the 45° line.

Solution 7.30 (a) The solution of x = f(y)= 1/y is y = 1/x, so f−1(x)= 1/x = f(x) (x>
0). The semi-hyperbola y = 1/x (x> 0) is symmetric by reflection in the 45° line.
Case (b) is similar: f−1(x) = f(x). The graphs, shown below, are symmetric, as
before.

Solution 7.31 First, the range of g(x) is [2, ∞] (because x2 + 4 ≥ 4), so that the
domain of g−1(x) is [2, ∞]. Next, solving gives
 The graphs are parts of hyperbolas. Finally
and similarly for g(g−1(x)).

Solution 7.32 To prove that f(f−1(x)) = x, let z = f−1(x). Then x = f(z)= f(f−1(x)) as
required.

Solution 7.33 which is rotation in the opposite direction.

Solution 7.34

Solution 7.35 natural domain is x ≤ 1. Also, vo u(x)= 1 −

x3/2; natural domain is x ≥ 0.
Solution 7.36 (Your estimated slopes should agree roughly with the exact
values.)

Solution 7.37 We calculate

Taking the limit as h → 0, we conclude that f′(x)= 3x2.

Solution 7.38 For n = 0, Eq. 7.12 says that if f(x) = 1 = x0, then f′(x) = 0. The
graph of f(x) is the horizontal line y = 1, which has slope 0 for all x. For n = 1, Eq.
7.12 says that if f(x)= x, then f′(x)= 1. The graph of f(x) is the line y = x, which has
slope m = 1 for all x. Thus Eq. 7.12 does give the correct, familiar slope in these
special cases.

Solution 7.39 (a) f′(x)= 2x2 − 4; (b) f′(x)= 12x3 +12x2 +16x − 16.

Solution 7.40 (a) For example let f(x) = 1 and g(x)= x. Then (fg)′ = (x)′ = 1
whereas f′g′ = 0. (b) We have (fg)′ = (x3 +2x)′ = 3x2 +2 and f′g′ + f′g′ = 2x · x +(x2
+2) = 3x2 +2.

Solution 7.41 (a) Completing the square gives y = (x − 2)2 − 3, so that the
minimum point is x = 2,y = −3. (b) With f(x) = x2 − 4x +1 we have f′(x)= 2x − 4,
so that f′(x)= 0 for x = 2. The minimum point is at (2,f(2)) = (2, −3).

Solution 7.42 Using the fact that f′(x) equals the slope of the curve y = f(x) at x,
we see that f′ is odd; f′(0) = 0; f′(x) → 0 as x → +∞. Also, f′(x) reaches a maximum
at some point x1, which is where the given curve y = f(x) has maximum slope.
Here, x1 ≈ 0.6.

Solution 7.43 (a) We have

and this → −1/x2 as h → 0. Therefore f′(x) = −1/x2. (b) Here g′(x) = 1−1/x2 which
equals 0 at x = 1. Hence the minimum point is at (1, g(1)) = (1, 2). [The graph of
g(x) is shown at the beginning of Section 7.5.]

Solution 7.44 (a)

Solution 7.45 We have

Here the factor h has been cancelled at Step 3. (Note that the result
fits with the general rule, Eq. 7.12.)

Solution 7.46 We have

Solution 7.47 (a) Squaring both sides of the equation gives y2 =

a2 − x2, or x2 + y2 = a2. Thus the graph of f(x) is the upper half of this circle. (b)
After algebraic simplification (including rationalization), we obtain

Letting h → 0, we obtain (c) The slope of the tangent is

The slope of the radius line is These being
negative reciprocals, it follows that the lines in question are perpendicular.
Solution 7.48 Replacing the given function f(x)by −f(x)(which we now rename
as f(x)), we have f(x1) < 0and f(x2) > 0. Also the new f(x)equals 0 if and only if the
original f(x) equals 0. Hence the algorithm applies.

Solution 7.49 Putting f(x)= 2x−cos x, we have f(0) = −1 and f(1) = 1.5, so there is
a solution x to f(x) = 0 between 0 and 1. The algorithm becomes

Thus x is between .44 and .5.

Solution 7.50 The correct value (by calculator) is

2.45, so the approximation is good to about one decimal place. [Write out a table
similar to the boy’s height example, if you find this problem confusing.]
Solutions to Review Problems
Chapter 1
1. 104 = 10, 000 (ten thousand); 108 = 100, 000, 000 (one hundred million); 1010
=10, 000, 000, 000 (ten billion).
2. 846 × 1, 000 = 846, 000; 0.0372 × 1, 000 = 37.2.
3. See the text.
4. See the text.
5. First, recall that 10n equals 10 multiplied by itself a total of n times. Therefore
105 × 103 is 10 multiplied by itself (5 + 3) times, i.e. 108. In general, 10m × 10n
=10m+n for the same reason.
6. (a) The distance from Earth to the sun is about 150 million kilometers, or 1.5
× 108 km. Note how this answer is rounded off to two significant digits. (b)
The answer is still 150 million kilometers. The extra precision in the miles-
to-km figure doesn’t change the precision level of the given number, 93
million.
7. See the text.
8. (a) 693.5; (b) (3×100)+(1×10−1)+(4×10−2)+(1×10−3)+(6×10−4).
9. (a) 12.42 (or 12.4 after rounding off); (b) 237; (c) 11.34 × 109 (or 11.3×109
after rounding off). This could also be written as 1.13×1010.
10. 1,2,3,4,5,6,7,10,11,12,13,14,15,16,17,20,21,22,23. (Observe that octal notation
uses 8 digits, 0 to 7. Similarly, the hexadecimal system, with base 16, uses 16
digits, 0 to 9 and A to F. The hexadecimal number 6EA thus equals 6 × 162
+14 × 16 + 10, or 1,770 base 10).
11. 0.9, 0.99, 16.5, 30.7, 84.1
12. We have
Chapter 2
1. (a) −22; (b) 94.
2. −8, −3, 0, 9, 14.
3. a − b = c means that a = b + c.
4. Assume a > b. Then a +(−b) > b +(−b) by the first law of inequality.
Therefore a − b > 0.
5. (a) 15; (b) −100; (c) 30.
6. (a) x − 2y;(b) a2 − 2ab + b2.
7. (a) 207 = 2×103+1; quotient = 103, remainder = 1. (b) 71 = 11×6+5; quotient
6, remainder 5. (c) 1897 = 222×8+121; quotient 8, remainder 121.
8. If n divides m then m = nq. And if m divides p then p = mz. (q and z are
natural numbers.) Therefore p = mz = nqz, which says that n divides p.
9. (a) 186 = 2 × 3 × 31; (b) 1611 = 3 × 3 × 179.
10. (a) 83/6= 13 5/6 is 5/6 beyond 13; (b) 105/15 = 7.
11. 0.444 … repeating.
12. Terminating for n = 2, 4, 5, 8, 10, 16, 20; otherwise repeating. The general
rule is that the decimal expansion of 1/n is terminating if and only if n =2j ×
5k for some whole numbers j, k.
13. (a) 119/24; (b) −11/28.
14. (a) 2/21; (b) 21/2.
15. a × b/c is unambiguous, because (a × b)/c =(ab)/c and a × (b/c)= (ab)/c also.
However, a/b × c is ambiguous, because (a/b)×c =(ac)/b whereas a/(b×c)=
a/(bc). The same ambiguity applies to a/bc, which could be interpreted as
(a/b)c or a/(bc). The best strategy in all such situations is to always use
brackets, thereby eliminating any possible misunderstanding. The same
applies to an expression like a/b + c; write this instead as (a/b)+ c, even
 though the precedence convention implies that this is in fact the correct
interpretation.
16. Because of the precedence of division over addition, we have a / b + c =(a/b)+
c, and not a/(b + c).
17. Yes, a/b/c is ambiguous, because (a/b)/c = a/(bc)whereas a/(b/c)= (ac)/b.
Brackets must always be used to remove this ambiguity.
18. 1121/72.
19. For example “5 goes into 15, 3 times” just means that 3 × 5 = 15, so “a goes
into b, j times” must mean that j × a = b, which is the same as j = b/a. (Thus
the “goes into” concept is not needed in math. But feel free to use it if you
like it. Myself, I don’t gain anything by knowing that 6.23 goes into 1.7
approximately 0.27 times.)
Chapter 3
1. Her percentage is 8, 400/140, 000 = .06, or 6%.
2. The formula is , where b is the base and h the height of the triangle.
(See Section 3.3 if you’ve forgotten the argument behind this important
formula.)
3. The trick here is to draw a diagonal, thereby cutting the quadrilateral into
two triangles, for which we know how to find the area.
4. The area scales up by k2. The perimeter scales up by k, because lengths of
lines increase by the factor k under scaling by k.
5. The scale factor is k =3 ÷ 1.7= 1.765. Volume scales as k3 =5.5, so a baseball is
about 5 ½ times the mass of a golf ball (assuming they are made of similar
density material). Surface area scales as k2 =3.1.
6. From Eq. 3.4, S = D/T ,we have

or 67,000 mph. (If you were surprised to learn that the earth is racing
through space at 67,000 mph, think what Galileo’s contemporaries must have
thought of his claim that the earth rotates around the sun, rather than the
reverse!)
7.
8. (a) By Eq. 3.3, X being proportional to Y means that X = c1Y for some
constant c1. Similarly, Y = c2Z for some constant c2. Therefore X = c1Y = c1c2Z
= c3Z where c3 = c1c2, i.e. X is proportional to Z. (b) We know that
Armstrong’s weight is proportional to g, which is itself proportional to the
mass of the heavenly body in question. If his weight on the moon is 98.8%
less than on the Earth, this implies that the moon’s mass is 98.8% less than
 the Earth, so the moon’s mass is 1.2% of that of the earth.
9. (a) 1.5% of 7 × 109 is .105 × 109 = 105 × 106, or 105 million. This is the
approximate increase in population for one year. (b) To a first
approximation, 10 years growth will add 10×105 million, or 1.05 billion
people to the world’s population, which will therefore be a bit over 8 billion
in 2020. Of course, this assumes that the growth rate of 1.5% per year will
continue for the next 10 years, which may not be the case.

10. (a) E = kmv2 for some constant k. . Therefore units

2 2 2
of E are g (m/s) , or g m /s . (The basic unit of energy is the Joule, defined as
1 kg m2/s2.)
Chapter 4
1. (a) a3b−2; (b) 1/(xy), or x−1y−1;(c) a−8c8,or c8/a8.
2. (a) (qt − 2rs)/4r2t; (b) (A2 − 2A)/6; (c) 0.
3. (a) (z + w)/(z − w); (b) (3x − 2y)/6.
4. (a) 2x2 − 7x − 1; (b) x4 + x3 − x2 + x − 2; (c) x4 − 4x2 +4.
5. (a) (x2 − 4) + 5/(x2 + 1); (b) x4 + x3 + x2 + x +1.
6. (a) (y − 4)(y + 4); (b) (y +3)(y − 2); (c) (2y +1)(y +2).
7. (2x − 3)/[(x − 2)(x − 1)(x +1)].
8. (a) y =3; (b) x = −7.
9. (a) x ≤ 3; (b) y< 0.93.
10. (a) Completing the square gives (x − 1)2 = 6, so the solutions are
(b) Dividing through by 3 and completing the square gives (x
2
+1/6) =13/36. The solutions are
11.
12. We have x2 − 2x = x(x − 2), so the solutions are x = 0 or 2; (b) Here we have
x3 − 4x = x(x2 − 4) so the solutions are x =0, −2, or 2.
13.
14. (a) x < 1 or x > 3; (b) −7 ≤ x ≤ 3; (c) −2 <x < 6.
15. (a) c6 +3c4d2 +3c2d4 + d6; (b) x6 − 12x5 +60x4 − 160x3 + 240x2 − 192x + 64.
16.
17. (a) 1/27; (b) 25.
18. See text.
Chapter 5
1. Label the parallelogram ABCD, and draw the diagonal AC. By hypothesis,
lines AB and DC are parallel, as are lines AD and BC. Now BAC = DCA,
these being alternate angles for the parallel lines AB and DC. Similarly, DAC
= BCA, because lines AD and BC are parallel. We can therefore conclude
that ΔADC and ΔABC are congruent (two angles and one side equal). It
follows that AB = DC and

2. Use the same drawing as in Problem 1. The hypothesis is that AB = DC and

AD = BC. We therefore conclude that ΔADC and ΔABD are congruent,
because they have three sides equal. Consequently, BAC = DCA, and this
implies that lines AB and DC are parallel. (This uses the converse of the
statement that, if two lines are parallel then alternate angles formed by a
transversal are equal. Namely, if alternate angles formed by a transversal to
two lines are equal, then the lines are parallel. This converse is proved by
noting that corresponding angles are also equal, if alternate angles are equal.
Hence the given lines are parallel, by definition–they make equal angles with
the transversal.)
3. Use the fact that π = 180° to obtain 3π/4 = 135° and 2π/3 = 120°. These are
both obtuse angles.

4. Three line segments form a triangle if and only if the longest segment is less
than the sum of the other two segments. For example, if a ≥ b and a ≥ c, then
a, b, c form a triangle if and only if a < b + c. To explain this, imagine trying
to construct the triangle by drawing arcs of radius b and c centered at the
endpoints of segment a. These arcs will not intersect if b + c< a, but they will
 intersect if b + c> a. (The case b + c = a would give a “collapsed” triangle,
which is not usually considered to be an actual triangle.)
5. (a) s = rθ =20 cm × (70°×2π/360°)= 24.4cm.
6. First, if h is the height of the parallelogram, we have h/2 = sin60°, or h =1.73
m. Therefore A = bh = 6.93 m2.

7. As in the preceding exercise, we have Hence

8. We have h = s sin θ and x = s cos θ. Therefore

For θ = 60° this gives
as in Exercise 7.

9. 300 RPM. The general formula is d1ω1 = d2ω2,where di is the diameter of

pulley i, and ωi is its angular velocity. To explain this, note that the speed of
the belt on pulley i is vi = riωi. But the belt speed has to be the same for both
pulleys, so v1 = v2. Thus r1ω1 = r2ω2, which implies d1ω1 = d2ω2 (because di =
2ri).

10. 0.60. The general rule is A′ B′ /A′ C′ = AB/AC, and similarly for other ratios.
The reason for this is that, if k is the scale factor, then A′ B′ = kAB and A′ C′ =
kAC. Therefore A′ B′ /A′ C′ = kAB/kAC = AB/AC.
11. The perimeters scale as k while the areas scale as k2.
12. Draw any chord AB to the arc. Then the perpendicular bisector of this chord
 coincides with a diameter of the circle. Using two different chords gives two
different diameters, the intersection of which is the center of the circle.

13. Since 52 +122 = 169 = 132, this is a right triangle. For the smaller angle θ we
have tan θ = 5/12, which implies that θ = 22.6°. The complementary angle is
67.4°.
14. First calculate x from y/x =tan θ, or x = y/ tan θ =8/ tan 38° =10.2. Then
Also B = 90° − 38° = 52°. As a check, A being a
bit less than 45°, we should have BC slightly smaller than AC, which agrees
with the calculation.

15. We have x/r = sin 50°, or x = r sin 50° =4.6 cm. Therefore AB = 9.2 cm. The
length of the arc joining A and B is s =10.5cm.
16. Each degree down from the Arctic circle increases the sun’s angle above the
horizon by one degree. Since NYC is south of the Arctic circle, the sun
will be above the horizon there.
17. Draw a diagram similar to that used in the solution of Problem 5.42, except
that you now draw a circle with radius r1+r2, outside the larger circle. Let AP
be the tangent from this new circle, drawn from the center P of the smaller
circle. The crossover tangent BC has the same length as AP ,namely
[We need to have d > r1 + r2, to have a crossover
tangent.]

18. Let l and w be the length and width of R. Then the dimensions of the smaller
rectangles are w by l/2. Therefore l/w = w/(l/2). Hence l2 =2w2, or
19. We have l/w = w/(l − w). This becomes l2 − lw − w2 = 0. Divide through by w2,
and set x = l/w, obtaining the equation x2 − x − 1= 0. The solution of the
latter equation is (by the quadratic formula)
This is the golden ratio l/w. (The other solution is negative,
and not relevant.)
20. Name the rhombus ABCD as in the diagram, and O as the intersection of the
diagonals. Then ΔDAB ≡ ΔDCB (3 sides equal). Also, both of these triangles
are isosceles. Together, these facts show that the diagonal BD bisects the
rhombus angles B and D. Similarly, AC bisects A and C. Next we see
that ΔDCO ≡ ΔBCO (2 sides and enclosed angle). Hence COB = COD;
since these angles are supplementary, each equals 90°. Also OD = OB. This
 means that diagonal AC is the perpendicular bisector of diagonal DB.
Similarly DB is the perpendicular bisector of AC. (There are several
alternative proofs.)
The converse can be stated as follows. Let AC and BD be two line segments,
with each being the perpendicular bisector of the other. Then ABCD is a
rhombus. (Another formulation of the converse would be: if the diagonals of
a parallelogram are perpendicular bisectors of one another, then the
parallelogram is a rhombus.)
To prove this, note that all four triangles shown in the diagram are congruent
(2 sides and included angle). Therefore AB = BC = CD = DA,so that ABCD is
a rhombus.

21. By trigonometry we have d/2 = 384, 000 sin 0.25°, giving d = 3, 350 km for the
diameter of the moon. The ratio of the volumes equals (3, 350/12, 000)3 =
.022, i.e. the volume of the moon is about 2.2% that of the earth.
Chapter 6
1. (a) x − 2y =0; (b) x − 2y =4.
2. x =12/7, y = −11/7.
3. Solve the simultaneous equations (x − 3)2 +(y − 1)2 =4, y =0. The algebra is
. Thus

4. False. Almost any example will disprove this. Take m1 =1, m2 =2. Then θ1
=45° and θ2 =Atan m2 =63.4°.

5. The perpendicular line through (1,1) has equation x +2y = 3. The point of
intersection of this line and the given line 2x − y = 4 is (11/5, 2/5). This is the
required point on the line, closest to (1,1).
6. By subtraction we get 8x +2y +2 = 0, or 4x + y +1 = 0. This is clearly the
equation of a line. To see that this line passes through the points of
intersection of the circles, note that any such intersection point (x, y) satisfies
the two circle equations simultaneously. Therefore (x, y) also satisfies the
equation obtained by subtracting the equations, that is, (x, y) is a point on the
line 8x +2y + 2 = 0. Thus the line passes through both points of intersection
of the circles. (The circles do actually intersect.)
7. The given line segment has slope −1 and midpoint (2,3). Hence the equation
of the perpendicular bisector is y −3= 1(x−2), or y = x+1.
8. After squaring, we have y2 = x. By analogy with y = x2, this is the equation of a
parabola opening to the right. The original equation is the upper
half.
 9. Both hyperbolas have the same asymptotes y = ±bx/a. The first hyperbola has
X-intercepts at x = ±a, and opens to the left and right. The second hyperbola
has Y-intercepts at y = ±b, and opens up and down.
10. We imagine a unit circle x2 + y2 =1, with area A = πr2 = π, and we think of it
as being covered by tiny squares. Sketching this figure by factor a in the X-
direction and b in the Y-direction produces the ellipse x2/a2 + y2/b2 = 1, which
is covered by little rectangles having area ab times their original area. Hence
the area of the ellipse is A = πab. Note that this agrees with the formula for
the area of a circle, in the case that a = b.
11. By completing the square and simplifying we obtain the equation (x − 3)2/4 −
y2 = 1. This is the equation of a hyperbola centered at (3,0), with asymptotes
The foci are at distance from the
center, and are therefore located at The eccentricity

12. To solve these equations simultaneously, first add them together, giving
Then from y2 = x2 − 2 we get y = ±1. There are
four points of intersection and this agrees with the figure (note
the symmetry).
Chapter 7
1.

2. (a) all x ≠ 2; (b) all x;(c) all x except 0 <x < 1 (because x(x−1) < 0 if x lies
between 0 and 1).
3. (b) is not a function.
4. Yes, max is a function. Its domain is the family of all finite sets of real
numbers, as stated.
5. (a) The graph of f(x) = 4 − x2 is a parabola, opening downwards. f(x) is an
even function. The X-intercepts are at x = ±2. (b) The graph of
is the upper half of the circle x2 + y2 = 4. (c) First, h(x) =
x(x2 − 3x + 1). Solving the equation x2 − 3x + 1 = 0 by the quadratic formula
gives x =2.6or .4. Therefore the zeros of h(x)are x =0, .4, and 2.6
approximately. h(x)is neither even nor odd. The local max and min points
can be found using calculus. Namely, f′(x) = 3x2 − 6x +1, which equals zero
for x = .2 or 1.8,approximately. The point (.2,h(.2)) = (.2,.5) is a local max,
and (1.8, −2.1) is a local min.
6. (a) We have x2 +2x +3 = (x +1)2 +2. The min is at x = −1,y =2. (b) From
calculus, which equals zero for x = −1. This is
the same result as in (a).
7. We have f′(x)= 3x2 −12x =3x(x−4), which is 0 for x = 0 or 4. Thus f(x) has a
local max or min at (0,9) and (4, −23). Since f(x) →−∞ as x →−∞ and f(x) →
+∞ as x → +∞, the curve y = f(x) must cross the X-axis at least three times.
(Make a rough sketch to explain this.)
8.

9. The asymptotes for the graph of f are the

lines x = 0, and y = 1. The asymptotes of f−1 are y = 0 and x = 1. These
asymptotes, like the graphs, are reflections of each other in the 45° line.
10. The graph of f(x) has in inverted-U shape. By calculus, f′(x) = 2x − 1/x2 and
this is zero for x = The local min is at the point (.79, 1.9).

11. If f(x)= 1/x2, then

after simplifying. Letting h → 0, we obtain f′(x)= −2x/x4 = −2x−3. Note that

this agrees with the general rule for differentiating xn,with n = −2 in this case.
(You may have observed that, in this calculation we can not let h → 0 until all
the algebraic simplifications have been completed, and in particular, the
factor h in the denominator has been canceled with a factor h in the
numerator. Before this cancellation, you would obtain the fraction 0/0, upon
putting h = 0, but as you know, 0/0 is meaningless.)
12. A similar calculation shows that f′(x)= −2x/(x2 +2)2. (If you’re interested, you
may enjoy finding f′(x) in the event that f(x)= 1/g(x). The answer is that f′(x)
= −g′(x)/g(x)2. This is an example of the “chain rule” from calculus.)
Diagnostic Test
The purpose of this test is to help you identify any gaps or weaknesses in your
mathematical background. Presumably you already know whether you are weak
in topics such as trigonometry, or logarithms. If so, you will study the
appropriate chapters. But if you have been experiencing persistent difficul-ties in
working with mathematical expressions (for example, in simplifying an algebraic
expression correctly, or solving equations), your background in algebra or
arithmetic may be insecure. Therefore the test concentrates on those topics.
Write your answers to each question on a piece of paper. Also make an
indication if you found the question confusing in any way; did you guess the
answer?
Answers are given at the end of the test. If you chose the wrong answer, or
found the question confusing, you should study the section(s) of the book listed
beside the answer. Then, if anything discussed in that section is still unclear, you
should probably take the time to study the whole chapter in detail. An exception
to this would be a situation where only one or two words are unclear—in this
case, the Index will show you where the troublesome words are introduced and
discussed. Thus you can make use of all the resources in this book, to quickly
sharpen your understanding and skills in basic mathematics.
After doing the diagnostic test, you may wish to read the section on
Common Errors in Elementary Mathematics. This will also help you decide
where to start studying this book.

TEST

Part I. Arithmetic
Do not use a calculator for this part of the test.
 3. Simplify: 1.52 − 3(4.8 − 2.7)
(A) −0.62 (B) −20.98 (C) −15.58 (D) −4.78
4. Simplify: (15 − 3)/(5 − 2)
(A) 2/5 (B) 4 (C) 0 (D) 2
5. Find the decimal expansion of 15/8.
(A) 1.6969 … (B) 1.75 (C) 1.875 (D) 1.88 …
6. Find 13.7 × 1.8 rounded off to one decimal place.
(A) 11.9 (B) 15.5 (C) 24.6 (D) 24.7
7. Find the remainder on dividing 707 by 11.
(A) 0 (B) 3 (C) 7 (D) 64
8. Find 25% of $500.
(A) $125 (B) $12.50 (C) $525 (D) $275
9. Express the binary number 110110 in decimal form.
(A) 26 (B) 40 (C) 54 (D) 202
10. Write in decimal notation: 2.95 × 104.
(A) 2,950,000 (B) 295,000 (C) .000295 (D) 29,500
11. a(b + c)= ab + ac is called the
(A) commutative law (B) multiplication law (C) associative law (D)
distributive law
12. How long does it take a plane flying at 500 mph to travel from Denver to New
York City, a distance of 1575 miles?
(A) 1225 (B) 3 hrs 9 min (C) 3.75 hrs (D) 3 hrs 15 min

Part II. Algebra

2. Simplify: x(1 + 2x2) − 2(1 − 3x)

(A) 2x2 +4x − 2(B) 2x3 +7x − 2 (C) x3 +3x (D) 2x3 − 5x − 2
3. Factor: x2 − 6x +8
(A) (x−4)(x−2) (B) (x−6)(x+8) (C) (x+3)(x+5) (D) (x − 8)(x +2)

5. Expand: (1 − b)3
(A) 1 − b3 (B) 1 − 3b +3b2 − b3 (C) 13 − b3 (D) 1 − b − b2 − b3

8. When the system

 is solved for x and y, the value of x equals
(A) 7/12 (B) 19/11 (C) 14/11 (D) −5/12

9. The value of the binomial coefficient C(6, 3) [sometimes denoted as ] is

(A) 20 (B) 120 (C) 216 (D) 720
10. Simplify (x16z−4)−1/2
(A) x15.5z−4.5 (B) −x8z2 (C) x−8z2 (D) x12z−1/2
11. The remainder on dividing x3 − 6x2 +2 by x − 1 is
(A) −3 (B) x2 − 5x − 5 (C) −7 (D) −x2 − 5x +5
12. Upon completing the square, the quadratic expression x2 − 4x +2 becomes
(A) (x − 4)2 +4 (B) (x − 2)2 (C) x2 − 4x +4 (D) (x − 2)2 − 2

Part III. Functions

Use a calculator where necessary.
1. If f(x)= 2x +3x find f(.9)
(A) 4.26 (B) 4.48 (C) 4.55 (D) 4.81
2. If W (p)= p2 + 6, find (W (p + q) − W (p))/q
(A) 2q (B) 1 (C) q +6 (D) 2p + q

3. Define Then f−1(.6) equals

(A) 1.5 (B) .67 (C) .375 (D) 2.67

4. The range of the function f defined in Problem 3 is
(A) [0, ∞)(B) [0, 1] (C) [0, 1) (D) (0, 1)
5. If and g(y)= y2 − 1 find f o g(5).
(A) 2 (B) 4 (C) 4.9 (D) 24
 6. The maximum value of f(x)= 1 + 4x − x2 (−∞ <x < ∞) is
(A) 4 (B) 5 (C) 7 (D) 8

7. The graph of has vertical asymptotes at

(A) x =0 (B) x = ±2 (C) x =2 only (D) x = −2 only

8. Which of the following functions are odd? (I) f(x)= 1+ x,
(II) f(x)= x sin x, (III) f(x)= x2 sin x
(A) I, II, and III (B) II and III (C) II (D) III

Part IV. Geometry

Use a calculator if necessary.
1. Triangle ABC has A =45° and B =60°. What is C?
(A) 90° (B) 75° (C) 60° (D) 45°
2. The right triangle ABC has base 8 units and hypotenuse 12 units. Find the
area of the triangle.
(A) 96 sq. units (B) 57.7 sq. units (C) 48.0 sq. units (D) 35.8 sq. units
3. A slice from a circular pizza has side 6 inches, and angle 45°. Find the
approximate area of the slice.
(A) 18 in2 (B) 14 in2 (C) 12 in2 (D) 5 in2
4. A certain triangle has sides 5, 12, and 13 units. A second triangle, similar to
the first triangle, has short side 6 units. Find the length of the perimeter of
the second triangle.
(A) 36 units (B) 30 units (C) 25 units (D) 15.6 units
5. A right triangle ABC has legs (i.e, non-hypotenuse) 2 ft and 5 ft. What is the
smallest angle in the triangle?
(A) 66.4° (B) 23.6° (C) 22.9° (D) 21.8°
6. A tangent line is drawn to a circle of radius 6 units, from a point P at a
distance of 10 units from the center of the circle. Find the distance from P to
 the point of tangency.
(A) 11.7 units (B) 9.3 units (C) 8 units (D) 4 units
7. An angle of .5 is equal to
(A) 45° (B) 30° (C) 28.7° (D) 30°

SOLUTIONS
Part I
1. C (Sec. 2.3)
2. A (Sec. 2.3)
3. D (Sec. 2.1–2.2)
4. B (Sec. 2.3)
5. C (Sec. 2.3)
6. D (Sec. 1.5)
7. B (Sec. 2.3)
8. A (Sec. 3.2)
9. C (Sec. 1.4)
10. D (Sec. 1.6)
11. D (Sec. 1.7)
12. B (Sec. 3.6)
Part II
1. C (Sec. 4.3)
2. B (Sec. 4.4)
3. A (Sec. 4.4)
4. D (Sec. 4.5)
5. B (Sec. 4.7)
6. D (Sec. 4.3)
7. C (Sec. 4.3)
8. B (Sec. 6.2)
9. A (Sec. 4.7)
10. C (Sec. 4.8)
11. A (Sec. 4.9)
12. D (Sec. 4.5)
Part III
1. C (Sec. 7.2)
2. D (Sec. 7.2)
3. A (Sec. 7.4)
4. C (Sec. 7.2)
5. C (Sec. 7.5)
6. B (Sec. 7.6)
7. C (Sec. 7.3)
8. D (Sec. 7.3, 8.1)
Part IV
1. B (Sec. 5.2)
2. D (Sec. 5.2)
3. B (Sec. 5.6)
4. A (Sec. 5.3)
5. D (Sec. 5.4)
6. C (Sec. 5.6)
7. C (Sec. 5.6)
Common Errors in Elementary
Mathematics
Teachers and professors regularly encounter certain common types of student
error. Students who continue to make these errors cannot hope to succeed in any
science or technology program. Most of these errors result from a lack of
understanding of basic math, together with reliance on faulty memory.
A common symptom of lack of confidence in basic math is the use of bad,
sloppy handwriting. In doing mathematics, it is essential to write down your
work neatly and carefully. This also simplifies the task of checking your
calculations. I always recommend using a pen rather than a pencil. You can
delete a wrong step by drawing a line through it, which is better than erasing – it
sometimes turns out that you discover later that the erased step was correct after
all.
Misuse of the distributive law
The distributive law is

To understand this, you must know that a, b, c represent arbitrary real numbers,
and that juxtaposition ab means multiplication, and also that operations inside
brackets are to be carried out before other operations. Thus, for example, 3 × (5
+ 11) = 3 × 16 = 48. This checks out to be the same as (3 × 5) + (3 × 11), as in Eq.
1.
Common mistakes are: forgetting to do the bracketed calculation first, as in
x(y −2) = xy −2 [False]; or omitting necessary brackets, as in 2(w −3t+5
[Meaningless]. A third mistake is using the apparent form of the distributive law
where it does not apply. Examples of this error:

These errors all look vaguely like the distributive law. However, any student who
remembers why the distributive law is true, at least for whole numbers (see
Section 1.3), will be unlikely to make such mistakes. Question: explain how we
know that the distributive law is true for whole numbers. Hint: think of a
diagram, for a typical example. Read Chapter 1 if you have forgotten this.

Problem 1. For each of the above false equations (a) give a counterexample; (b)
write the correct equation, if there is one.
Misuse of cancellation
The cancellation law is

Here a, b, c represent arbitrary real numbers, with a and c not equal to zero. We
say that the factor a, which is a common factor in the numerator and the
denominator, can be cancelled from both:

I recommend that you do not indicate cancellation by striking things out,

because this becomes hard to read, and almost impossible to check later. For
example,

This is easy to read and check. Compare this with what happens if you strike
things out. Striking out can and often does result in errors.
Mistakes may also result from misunderstanding the cancellation law as in
the following examples:

Students who make such mistakes seem to be of the impression that any symbol
that occurs in the numerator and denominator of a fraction can be cancelled. Of
course this is incorrect: cancellation applies only to common factors, as in Eq. 2.
The common factor can be a complicated expression, as in

What is the common factor in this case?

 Indeed, what exactly is a factor in general? See if you can put this into words.
Answer: in any multiple product, such as abcd, the factors are a, b, c, and d.
Remember, the expression abcd is interpreted as a × b × c × d, and this is called
the product of a, b, c, and d.
Another question: after a factor has been cancelled from the numerator and
denominator, what’s left? Example: 2/(4x) =? Answer: the number 1 is left:

Problem 2. (a) Identify the factors, if any, in the expressions (i) 3(x − 2)(y + 4);
(ii) 3x − 2y + 4. (b) Factor the whole number 28 into its prime factors.

Problem 3. Decide whether the following calculations are correct, and explain:
Operations with fractions
Here are the rules for combining fractions

Once more, letters a to d represent arbitrary real numbers. Chapter 2 explains in

detail why these equations are valid.
I hope it goes without saying that the letters a, b, c, etc. in all of Eqs. 1 to 5 can
be replaced by any mathematical expression whatever, as long as those
expressions themselves represent real numbers. For example

because of Eq. 3 for multiplying fractions.

Mistakes often arise from misuse of these equations. For example, addition
and multiplication sometimes get confused, as in

Check that the correct result here is (2x + y)/2y.

Every math student should understand why Equations 3–5 are valid. Read
Chapter 2 if you have forgotten the explanations. Here I will just remind you
about Eq. 5, since the explanation is useful both for doing and checking
calculations. First, fractions that have the same denominator can be added
directly:

For example, 3/5+ 8/5 = 11/5. Second, for fractions that have different
denominators, we can first use the cancellation law to rewrite the given fractions
so that their denominators are the same:

This is the best way to remember how to add fractions, much better than
trying to remember Eq. 5 itself. For practice, try adding x/2+2y/5. You should get
(5x +4y)/10.
Furthermore, you can (and should) immediately check any such calculation
by reversing the steps:

You can perform this check mentally, if you prefer.

Problem 4. Write as a single fraction (a) 3/x − y/4; (b) (8x/3) ÷ (4x2/9). Note: it
is best to write these problems out in the usual vertical manner before
proceeding.

Negative numbers, square roots, etc

True or false: −x is a negative number. Answer: this is false in general. The
correct statement is that −x is a negative number if x is positive, whereas −x is a
positive number if x is negative. For example, −(−7) = 7. This can be a source of
error, as in

What is the correct result here?

Many students seem to find it confusing that “two minuses make a plus.”
Chapter 2 explains this point fully.
Square roots are often mishandled, as in
The symbol always refers to the positive square root of a, so that is the
unambiguously correct result.
Some students are mystified by the equation

where |x| denotes the absolute value of x. To show that Eq. 6 is correct, we
consider the cases of positive and negative x separately. For x > 0 we have
and |x| = x also, so the equation is correct. Example:
For x< 0 we have x2 > 0 and also |x| = −x in this
case. Example: Thus Eq. 6 is always correct (for x = 0,
as well).

Problem 5. Simplify
Summary
For any reader of this book who tends to make errors in mathematics, I
recommend the careful study of Chapters 1 through 4 before proceeding to the
later chapters. Also, try to adopt the following work habits rigorously:
1. Always write down your calculations carefully and neatly. A few connecting
words can help to make sense of your writing—for example: therefore …,
we have …, by substitution …, etc.
2. Consciously develop and use habits that help to avoid errors, and allow for
easy checking. For example, never use strike-outs to indicate cancellations.
3. Always double check every calculation before going on. Many useful
techniques for checking your work are discussed in this book.
4. Be sure you fully understood each technical term, such as factor,
denominator, radian, and so on. Use the book’s Index to look up such terms
when necessary.
5. Never be satisfied with mere memorization of something that is confusing,
or not fully understood. A firm background in mathematics requires
understanding of every definition, concept, argument, and formula. Exactly
what is meant by “understanding” is a complex psychological question,
which is addressed throughout the book.
6. Always exert the effort to understand the proof, or derivation, or
explanation, of every point in the book, or in class. This can be hard work,
but it will pay off later in terms of confidence in your understanding of
mathematics.

For further discussion of common errors in mathematics, see the website

www.math.vanderbilt.edu/schectex/commerrs/

Solution 1. (a) For example (2 + 3)2 = 25, not 22 +32, which is 13.
Counterexamples to the other false equations are also easily found. (b) We have
(a+b)2 = a2+2ab+b2. Also sin(a+b)= sin a cos b+cos a sin b (Chapter 8). The
other false equations do not have correct alternatives – they’re just false.
Solution 2. (a) The factors are 3, x − 2, and y + 4; (ii) There are no factors other
than the entire expression. (b) 28 = 2 × 2 × 7.
Solution 3. (a) Incorrect because p − 2 is not a factor of the numerator; (b) This
is correct.
Solution 4.
Solution 5. (a) |x − 1|, because (x − 1)2 = x2 − 2x + 1; (b) cannot be simplified.
Greek Alphabet

The Greek Alphabet

alpha α
beta β
gamma γ
delta δ
epsilon ε
zeta ζ
eta η
theta θ
iota ι
kappa κ
lambda λ
mu μ
nu ν
xi ξ
omicron o
pi π
rho ρ
sigma σ
tau τ
upsilon υ
phi φ
chi χ
psi ψ
omega ω
Mathematical Symbols
Symbol
How to read
Index entry

=
equals
equation

+
plus
addition

−
minus
subtraction

×
times
multiplication

÷
divided by
division

/
over
division
a over b division; fraction
—
( ), { }, [ ]
brackets
an a to the nth
exponents
<
less than
inequality
>
greater than
inequality
≤ (≥) less (greater) than or equal to
inequality
|x| absolute value (or magnitude) of x
absolute value
square root
square root

%
per cent
per cent

∝
is proportional to
proportionality
x2 x2, or x sub 2
subscripts
P(x) P of x
function
C(n, k) C of n, k
binomial coefficient
n! n factorial
factorial
sum, k equals 0 to n
summation
angle
angle

°
degrees
angle
end of proof
a′ a prime
Δ
triangle
triangle
perpendicular
perpendicular
sin
sine
sine
cos cos (or cosine)
cosine
tan tan (or tangent) tangent (function)
π
pi
circumference
(x, y)
xy coordinates; ordered pair; interval
Δx delta x
slope

∈
belongs to
sets

⊂
is contained in
sets

∪
union
sets

∩
intersection
sets
{x : …} the set of all x such that …
sets
[a, b] closed interval a, b interval, closed

∞
infinity interval, infinite

→∞
approaches
infinity asymptote
f transforms x to y
function
f -1 f inverse inverse
function
f′(x) f prime of x
derivative
f o g f oh g
composition
limit as x approaches x0
limit
(r, θ) r theta
polar coordinates
base of natural logarithms
loga log to the base a
logarithm
ln
l n
natural logarithm
Index abscissa, 285
absolute value, 51, 99, 196
absolute value function, 352
acceleration, 140
Acos (arc cosine), 252
acute angle, 227
triangle, 229
addition algorithm, 5, 52
decimal point numbers, 24, 51
fractions, 80, 161
polynomials, 167
whole numbers, 10
adjacent side, 249
algorithm, 5
addition, 6, 25, 52
of fractions, 80, 82
division, 65
division, long, 65, 95, 214
multiplication, 19, 58
order, 37
subtraction, 46, 49, 55
alternate angles, 231
angle, 224
acute, 227
alternate, 231
bisection of, 276
complementary, 226
corresponding, 229
exterior, 232
interior, 231
obtuse, 227
opposite, 227
 right, 225
size of, 225
straight, 225
supplementary, 226
angles, equal, 224
sum of, 231
angular velocity, 267
approximate calculations, 29
arc cos, 252
arc length, 265
arc of circle, 224, 260, 264
arc sine, 251
arc tan, 253
area, 114
of circle, 263, 278
of sector, 267
of triangle, 116
argument, 341
arithmetic, laws of, 34, 96
Asin (arc sine), 251
associative law, of addition, 10, 54
of multiplication, 14, 59
asymptote, 319, 348
Atan (arc tangent), 252
axes, coordinate, 284
axiom, 222
parallel, 233
binary arithmetic, 20
binocular vision, 247
binomial, 171
coefficients, 201, 202
theorem, 199, 205
bisection of an angle, 276
bisection, method of, 383
bisector, perpendicular, 256, 276
bonds, financial, 150
brackets, nested, 17
 rule of, 11
calculator, basic, 109
scientific, 276
calculus, 370
cancellation, 72, 159
errors in, 74
cancellation law, 73, 96, 159
carbon dating, 157
centre of a circle, 224
checking calculations, 162, 168
chord of circle, 257
circle, 224
arc of, 224, 264
area of, 263, 278
circumference, 261
circumscribed, 273
determined by 3 points, 273
diameter of, 259
equation of, 288
inscribed, 273
radius of, 224
tangent to, 270
circles, intersection of, 308
circumference of circle, 261
circumscribed circle, 273
closed interval, 331
coefficient, of polynomial, 165
coefficients, binomial, 201, 202
common denominators, method of, 81, 163
common factors, 175
commutative law, of addition, 9, 54
of multiplication, 14, 59
completing the square, 188
composite number, 68, 103
composition of functions, 369
compound interest, 150, 152
conclusion, 247
congruence, 239
congruent triangles, 240
conic sections, 316
constant term, 165
constructions, ruler and compass, 273
continuity, 351
conventions, for minus sign, 60
converse, 248
coordinate axes, 284
coordinates, rectangular, 283
corresponding angles, 229
cosine, 250
counterexample, 87, 248
counting principle, 10
cusp, 354
decay, exponential, 157
decimal point numbers, 22
full system of, 49
long division, 95
decimal system, 1
decimal, repeating, 79, 100
degree of polynomial, 165
denominator, 70
common, 81
least (lowest) common, 83, 163
rationalizing, 210
derivative of a function, 372
of a polynomial, 374
diameter, of a circle, 259
difference quotient, 372
digit, 2
positional significance, 1
significant, 33
dimensions, 136
directrix, 325
discounting, 155
discriminant, 187
disk, 224
distance, between points, 223
distance formula, 286
distributive law decimal numbers, 59
whole numbers, 15
dividend, 64, 171
division, 63
as reverse of multiplication, 63
by ten, 27
by zero, 88, 94
extended, 77
long, 65, 77, 95, 214
of fractions, 90, 163
of polynomials, 171, 214
of rational numbers, 90
synthetic, 215
with remainder, 64, 171
divisor, 64, 171
domain, 334
natural, 338
earth, size of, 266
eccentricity, ellipse, 323
earth’s orbit, 323
hyperbola, 324
parabola, 325
elimination, 304
Gaussian, 306
ellipse, 316
focus of, 321
empty set, 331
energy, 142
equation, conditional, 302
graph of, 291
linear, 180, 299
meaning of, 301
of circle, 288
of straight line, 292, 296
 quadratic, 183
relational, 302
solving, 179, 218
equations inconsistent, 306
numerical solution of, 382
simultaneous, 303
equilateral triangle, 255
Eratosthenes’ method, 266
errors in mathematics, 11, 105, 411
even function, 347
examples, generic, 17
exponential decay, 157
growth, 150, 157
exponents, 144
fractional, 205
laws of, 145, 207, 208
negative, 146
repeated, 146
rules of, 145, 207, 208
extraneous solutions, 192
factor, 13, 67, 101
common, 175
of polynomial, 175
factor theorem, 217
factorial, 201
factoring quadratics, 175
whole numbers, 68, 103
factorization, prime, 68
focus ellipse, 321
hyperbola, 324
parabola, 325
fractions, 69
addition of, 80, 161
decimal representation of, 77
division of, 90, 163
improper, 71
multiplication of, 83, 85, 163
 negative, 83
order of, 76
proper, 72
reducing, 73
subtraction, 83, 161
function, 333
argument of, 341
composition of, 369
continuous, 351
derivative of, 372
domain of, 334
even, 347
graph of, 341
inverse, 360
monotonic, 362
natural domain of, 338
odd, 347
one-to-one, 362
range of, 339
smooth, 351
transformation of, 357
Gaussian elimination, 306
gear ratios, 268
generic example, 16
geometric increase, 150
golden ratio, 281
googol, 6
grade, 294
gradient, 294
graph of equation, 291
of function, 341
of inverse function, 363
of sums and products, 368
greater than, 36, 50
growth, exponential, 157
population, 156
hyperbola, 318
 focus of, 324
hypotenuse, 236, 249
hypothesis, 247
identity, 302
inconsistent equations, 306
inequalities, 194
laws of, 38, 194
linear, 181
quadratic, 197
solving, 195
infinity, approaching, 350
inscribed circle, 273
integers, 47
intercept, 300, 352
interest compound, 150
interpolation, 385
intersection of circles, 308
of lines, 222
of sets, 329
interval, 331
closed, 331
open, 331
inverse function, 360
graph of, 363
operations, 206
transformation, 366
invert-and-multiply rule, 91
irrational numbers, 80, 100, 104
isosceles triangle, 254
l.c.d., 83, 163
larger & smaller, 36 see also: order, inequalities laws, of arithmetic, 35, 96
of exponents, 145, 207
learning mathematics, see mathematics least (lowest) common denominator, 83,
163
length of a line segment, 223
less than, 36, 50
like terms, collecting, 167
limit, 379
line (straight), 222
equation of, 296
intercept form, 300
point-slope form, 297
slope-intercept form, 298
segment, 223
slope of, 292
linear equation, 180, 299
inequality, 181
lines, parallel, 223, 295
perpendicular, 295
logarithmic growth, 150
logic and proof, 247
long division, 65, 82, 95
of polynomials, 171, 214
look-up table, 335
lowest common denominator, 83, 163
magnitude, 51, 99
orders of, 33
mass vs. weight, 138
mathematical models, 124, 152
mathematics, learning, 38, 233
solving problems in, 310
understanding, 7, 105
maximum, 377
measurement, units of, 134
mental blocks, 40
midpoint, 287
minimum, 377
minus, two minuses make a plus, 45, 56
using the minus sign, 60
mistakes, avoiding, 11, 74 see also errors models, mathematical, 124, 152
monotonic function, 362
multiplication by ten, 17, 27
decimal numbers, 28, 58
fractions, 83, 85, 163
 polynomials, 167
whole numbers, 13
MWTFU, 173
natural number, 5
negative number, 47, 83
nested brackets, 17
nines, casting out, 69
notation, scientific, 29
number, composite, 103
decimal point, 22, 49
irrational, 80, 100, 104
natural, 5
negative, 47, 83
prime, 68, 102
rational, 83, 87
real, 100
whole, 5
number line, 36, 47, 57
numerals, 2
numerator, 70, 171
numerical precision, 32
obtuse angle, 227
triangle, 229
odd function, 347
one-to-one function, 362
open interval, 331
opposite side, 249
order, 36, 50
see also inequalities for fractions, 76
of magnitude, 33
ordered pair, 336
ordinate, 285
origin, of number line, 47
of rectangular coordinates, 283
parabola, 320
focus of, 325
parallel axiom, 233
parallel lines, 223, 295
parallelogram, 118
parameter, 153
parentheses, 12
Pascal’s triangle, 200
per cent, 113
perpendicular bisector, 256
lines, 225
pi, 261
planetary orbits, 323
point, 222
polynomial, quadratic, 165
polynomials, 165
addition of, 167
derivative of, 374
division of, 171
linear, 165
long division of, 214
multiplication of, 167
quadratic, 165
roots of, 180
population growth, 156
positional significance of digits, 1
powers of ten, 3, 26
precedence rule, × over +, 18, 94
exponents over multiplication, 166
precision, numerical, 32
present value, 155
prime number, 68, 102
problems, how to solve, 310
product, 13
proof, 234, 247
by example, 16
proportionality, 125
inverse, 129
multiple, 127
protractor, 226
Pythagoras’s theorem, 236
QED, 228
quadrant, 284
quadratic equation, 183
factoring, 176
formula, 183
quotient, 64, 171, 214
radian, 263
radical, 205
simplifying, 209
radius of circle, 224
range, 339
rate of change, 378
rational number, 83, 87
rational number system, 100
rationalizing a denominator, 210
ray, 224
real number system, 100
reciprocal, 92, 296
rectangular coordinates, 283
reducing a fraction, 73
reductio ad absurdum, 104
reflection, 239, 314
reflector parabolic, 325
remainder, 64, 173, 214
remainder theorem, 217
repeating decimals, 79, 100
reversible transformation, 192
right angle, 225
rigid motion, 239
roots double, 188
of negative numbers, 212
of polynomial, 180
nth roots, 205
rotation, 239, 314
rounding off, 33
rule of nine, 69
rule of three, 69
ruler and compass constructions, 273
ruler, number, 23, 36
rules of arithmetic, 34
scale effects, 119
scaling, 239, 315
scientific notation, 29
sector of circle, 267
area of, 267
segment, line, 223
set, 329
empty, 331
SI system of units, 134
significant digits, 33
similar triangles, 243
similarity, 238
simultaneous linear equations, 303
sine, 250
singularity, 349
slash notation, 76
slope of curve, 371
of line, 292
smooth function, 351
solution, extraneous, 192
numerical, 382
of inequalities, 195
of quadratic equation, 183
of polynomial equations, 179, 218
solving triangles, 249
solving math problems, 310
speed, 129
average, 133
square root, 99
straight line, see line (straight) strike-outs, 75
substitution, 144
subtend, 264
subtraction, 43, 55
 algorithm, 46
and number line, 57
as reverse to addition, 45
of fractions, 83
summand, 13
summation notation, 204
symmetry, of functions, 347
synthetic division, 215
tangent (function), 250
tangent line, to circle, 270
to curve, 373
temperature, 136
term of polynomial, 165
theorem binomial, 199, 205
factor, 217
if-and-only-if, 218
Pythagoras’s, 236
remainder, 217
traffic congestion, 122
transformation, inverse, 366
of functions, 357
of plane figures, 239, 311
transitivity, 194
translation, 239, 311
transposing terms, 181
transversal, 229
triangle, 229
acute, 229
area of, 116
congruent, 240
equilateral, 255, 275
isosceles, 254
obtuse, 229
Pascal’s, 200
right, 229
similar, 243
solving, 249
triangulation, 246
trigonometric functions, 250
two minuses make a plus, 45, 56
understanding mathematics, 7, 105, 211
union of sets, 329
units of measurement, 134
variable, 165, 288
velocity angular, 267
of rotating object, 269
vertex, angle, 224
triangle, 229
vision, binocular, 247
weight, 138
whole number, 5
zero, 4
division by, 88, 94
of a function, 383
zero product rule, 97, 184