You are on page 1of 413

.

DOCUMENT FESUME
.

--ED, 143 514,

'

SE 022 991

lr

-,

.-

Herr iot, Slrah T.; And Others


Calc -ulus o (Elementary Functions; Part I.. Student,.
Text. Revi.se Edition,.
..'''
'Stanford Univ., Calif. School Mathematics Study

-1=1UTHO'R
TITLE
f

INSTITUTION

Gtou.p.

SPONS AGENCY
PUB DATE
N9TE

National Science Foundation, Washington, D.C.,


69

41315.; For related document s, see SE /022 -992-994;.

Cpntains occasional light and bloke

*Algebra; *Calculus; *InstructionallMaterials;


Mathematics; Number Concepts; Secondary Education;
*Secondary School Mathematics; *Textbooks
*School Mathematics Study Group

IDENTIFIER,5
,

ABSTRACT

type

MF:$0.83 C-$2209 plus'Posiage.

uRs' PRICE
SCRIPTORS

.
.

.This course is intended ,for students who .have a


.

thorough knowleZge of college prepdratoi.Y mathatics, including


ralytiO geometry.
This text, PartI, contains the first five chapters of the course and
two appendices. ChaptrerS in)Arded are: 41Pclynomial Functions; (2)
The p'erivative of a .Polynoial_Function; (3-) Circular Functions; (4)
Derivatives of'Circular Functions; and (5) Exponential, and Related
Functirs. The appendices are: (1) Functions and Their,
Repres ntations; and '(2) Polynomials. (RH)
.

-ia1gebra4, axiomatic geometry, trigonometry, and

11-

Z
4.1

.1 4

***********************************************************************
Documents acquire-by ERIC include many informal unpublished
*.materials not available from othet sourcei. ERIC makes every effort *
* to obtain the best' copy available. Nevertheless, items ofomarginal *
* _reproducibility are often encountered and this affects the quality *
* of the microficke and harddopy reproductions ERIC makes avaiAable
*
* xia.theERIC Documellt Reproduction Service (EDRS) . 'EDRS is AOt 1".
*
* regponsible fox the quality of the, origin4,dOcument. Reprochctions *
* supplied by- EDRS are the best that can be made from
*
e, original.
******gtx******1,**************************#******vitt** 4****v*********,'.

'*

^$

.SCHOOL
MATHEMATICS.

STUDY GROUP

.CA4CULUS
ELEMENTARY FUNCTIONS
Part
O

Student Text
.*(Revised Edition)

DEPARTMENT OF HEALTH
EDUCATIONS. WELFARE
HATION??,L INSTITUTE OF
EDUCATION

HAS. f!E E..

T.

SMSG

P E PRO

0.,(E0 Exac tcP, AS OECE,vE0 rROM


T.-4E PERSON OP ORGAN Zd707. OP &)N
OP OP,N,ON5
PO, rS OP
...C.,

,ratE0 00 NO* NE(EsSAR,Y PEPPE


r As NA! aNAL, .4,,ru7E OP

SEN.'

OP POL

.)

Ni

.
I,

.lz

`,l_)1,,,(

,t-f)

T.

' .11
'10

",

"T r;
Tr /

yr

w,

4.

t- CALCULUS OF
ELE71)1EIVTARI.FUATIONS.:
Part I
0, Student Text

)(Revised Edition).
7

The following is a list of all those who participated in the preparation of this
volume:
.
,

.
Sarah T. Hemet
Desmond T. JAkins
C. W, Leeds, III
George R Richatdson
Don'ald E. Richmond
.
Paul C. Shields

, Gunn High School, Palo Alto, Calif.


Palo Alto Senior' High School, Calif.
Simon's Rock, Great Barrington, Mass.
SMSG, Stanford University, Calif.
Williams College (Emeritus), Williamstown, Mass.
Wayrie State University, Detroit, Mich.

4
. .;

Financial support for 'the School Mathematics.Study- Group has been


provided by the National Science Foundatiop.

Permission to make verbatim use of material in this book must be secured


from the Director of SMSG.' Such permission will be granted except in
unusual circumstances. Publications incorporating SMSG materials must
include both an acknowledgment of the'SMSG copyright (Yale
Univer.
sity orStanford University, as the case may be)dnd a disclaimer of SMSG
endorsement. Exclusive license will not be granted save in exceptional
circumstances, and then only by specific aetlep Of the Advisory Board of
SMSG.

1968 and 1969 by The Boiird of Trustees of the Leland Stanford Junior University
All right reserved
.
Printed in the United States of America

..,

FOREWORD
.

The correspondence between graphs in the xy-plane and relations between

x and y

was one of the profound discoveries of mathematics.

In particular,

if no vertical line cameet a graph in more than one point, then the graph is
that of a function f
x -)y; that is, for .each first coordinate x of a
:

point on'the graph there is a unique number y such that

(x,y)

lies on the

graph. A central purpose of this 7.5:t is to study the'relatiOnship between

graphs of functions and'tbe expresSions which ,Define these functions.

We, shall

concentrate our attention on functions defined by polynomial, trigonometric,


exponential and logarithmic expressions,, or by combinations of such expressions.

These functions are usually referred to at elementary functions4

We should expect that properties of the graph of a function are related


to the expressioh which defines the functions.

F '

example, by analyzing the

functional expressiqn we should be able to determ ne the location.of high and

low points on the graph, and in addition answer q estions about the shape of
the graph (such as illitervals of rise or fall and .ow the graph bends).

Further-

more,.if the function is related to some physical roblem involving motion i7


seems reasonable that an analysis of its expressi

should enable us to deter-/

mine such aspects of the motion as velocity and ac eleration.

Likewise,, ry

might expect to be able to determine from the func iopal expression such/pr

perties as the average value of the function and,t e area of a region b un d ?ed
7.
.
by the function.
04.4: aim is to develop some of the concepts and techniqqes.w*ch/wi 1

enable us to obtain important information about gra


Theprimary et:Incept which we develop in Chapters 2,

tangent line at a point on the graph of a function.

of element ry functions.

6 is t

t of the

This tangent'line is des -

cribed as the'apraight Line which best fits the grap

'

and

In

near that/ ,point.

- particular, formulas are developed for finding the sl tes of tangent lines to
t14-graphs of various 'elemeneary functions.

Fora gi en function

formula defines a, new function called the derivative

In particular,

descrh bed.by the

derivative.' Furthermore, high and low points of the g =p


by,zeros of the derivative, while rise or fall can be d

such a

f. /Values of the

' derivative give 'a 'measure q the rate of change of the graph.

such aspects of .motion as velocity and acceleration ar

f,

of
e

can be V,iven

ined by thesign

of the, derivative.
.

.
,

0
0

,-,

Our fir4t task

unctions,

to analyze the simplest kinds of elementary

namely polynomials, the sine and cosine functions, and simple power, exponential and logarithmic functions.

These are discussed in Part One, while the

,more,difficult techniques associdted with various algebraic combinations


(sums, products, quoNents, composition) of these, basiC.functiorls are left

In additiond the concept of antiderivative

to Chapter 8 (in Part Two).

introduced in Part Two, the aRtideriptive of


derivative is

being a function whose

f. "We shall show how antideri4atives can be used to calculate

areas'and to solve such.prpblems as determining velocity given acceleration;


Furthermore, this relationship between antiderivatives and area provides us

with aowerful geometricIool for analyzing and approximating functions.


.

Chapters 1 and 2 discuss polynomial functions, with Chapter 1 Concearating on,definitions and simple algebraic and geometric properties:

The .concept

of tangent line at a point oncthe graph of polynomial function is introduced


in Chapter 2 and formulas for the derivative of a polynomial function are
.

Applications of the derivative to graphing (such as finding high

obtained.

and lbw points and intervals of rise or fall); its interpretation as velocity
;

or acceleration, and its use if gpproximationare also discussed .in Chapter 2.,
This same pattern is followed in the remaining four chapters of this volume.
Definitions and simple properties of the sine and cosine functions and the
power expohential and logarithm functions appear in Chapters 3 and 5, respec,.

tively, while Chapters 4 and 6 discuss derivative formulas for these respec,

tive classes,,as well as applications to graphing and approximations.


FOr their tlioughtful comments we are grateful to Frank F. Allen, Lyons's
.,
,

TOWnShip High

chool; Leonard Gillman, University of Texas; David-W. Jonah,

Wayne State Upiversityt; Albert W. Tucker, Princeton University; 'and the pilot

.^

_,

teachers and students at.Cubberley, Gunn, and Palo Alto High Schools in Palo
Alto; California; St. Mark's School.in Dallas, Texas; .and Simon's Rock in
.,

Great BariOngton, Massachusetts.

We are also indebted to previous SMSG writihg,teams,whose materials WO


t

considerable influence pn this text.' Many ideas and exercises were taken,

directly or ada'ped from two earlier MSG texts:


Calculus.

Elementary Functions and

In the first pgrt ofthis text. we borrow heavily from SNSG Elemen-

tary Functions; Appendix 2 of ,this text contains Sections 2-12; 2 -6, 2 -8, and
2 -9, of Elementary Functions, Appendifes '1, 3, 4, 5,, 6,.7, and 8 wgre adapted

6; A3-1, 2; 10-3, 4, 5, 6; 6-1, 2, 3,.4; A6-1, 2p

from SectionsA2 -1,

3-2, '3, 4,;A1-1,,,,,2; 5;3; and Alf-5 of SMSG Calculus..

ft

we%expresq appreciation to Nancy Woodman, whose function as a


'typist was transcendental..

' iv

Y.

*.

TABLE OF CONTENTS

Part 1

Foreword'

47.,""

Chaper 1.

POLYNOMIAL 'UNCTIONS

.)

C
ffntroduCtion'and Notation

1-1.

Quadratic FUnctions

_1:5.

7
24

Remainder and Factor Theorems

1 -7.

RatiOna.1 Zeros

1-8..

Approximatj.ng Zeros

1-9.

Degree oi'PolynOilial and Behavior of Graph

Polynomial Functions of Degree Higher Than One

39

56

.t

,75

THE DERIVATIVE OF A POLYNOMIAL FUNCTION

81

The Tangent at the y-Intercept. of a Grapb

82

The Behavior of the Graph Near an Arbitrary Point


.
.
The Slope as Limit of Difference Quotients, ,..

90

22-

.,

'The Derivative

105

Derivatives of GeneralyFolynomial Functions

Applications of the Derivative to Graphing

2.7.

Optimization Problems

2-

Rate of Change:

2 -5j,

The Secorid Derivative

98

. 2-1

: ..

111

123

136

e.

Velocity and Accelei.ation

148

162

..-. -...

Newton's Method _..

173

..

:.

2-11. Higher Derivatives and'Notation

'Chapter 3.

178
4,

CIRCULAR FUNCTIONS

"

187

)
3 -1.

The Sine and Cosine Functil6hs

3 -2.

Properties of the Circular Functions

188,

3-3.

Graphs of thecircular Functions

54.

Uniform

3-5
34.

The Addition Formulas

3-7.

Pure

72

2-2.

2-4:

65

2-1.

2-3.

49

Locating Zeros of Polynomial Functions.


,:-

,e_

1-4.
1-5.
1-6.

Chapter 2.

Constant andlinear Functi9ffs

1-2'.

207
217

'224

Circulir Motion
.

,,.

.\.

252.

Waves

239
....

Analie-Of General Wave.

Period

247

40,

Chipter 4.

DERIVATIVES OF CIRCULAR FUNCTIONS

4-1:-

The

angent at the y -Intercept'

4-2:

The

erivative as the Limit of a Difference Quotient

,..

253

253
263

4-3.

Line

Substitution

4-4.

'Velocity and Acceleration

273

279
.

4-5.

Higher Derivatives an; Approximations

287

0.

Chapter5.

EXPONENTIAL AND RELATED FUNCTIONS

303

Exponents

5-1.

.t.

303

5-2.

The'Exponentila Function, Growth and Decay:

5-3.

More'About.Ratiosal Exponents

5-4.-

Arbitrary Real Exponents

5-5.

Powers of the Base

5-6.

The Logarithm (Base 2)

5 -7.

Logarithms (General Base)

'i

Appendix 1.

308

313

...

319

as Powers
rs of

325

330
1.538

FUNCTIONS AND THEIR.REPRESENTATI04$

343
.0

343

A1-1.

Functions

A1-2.

CompositeFunctions

A1-31

Inverse Functiohs

A1-4.

Monotone Functiohs ...... ..

A1-5.

Polar Coordinates

359

Appendix 2.
A2-1.
A2-2.

364

,n5

, ..... ..... .
.,
...... ,.

372

377

POLYNOMIALS

382

SignkficanCe of Polynomials
t
Number of Zeros

38a395

.A2-3., ciomplex.Zeros

402..

--_

------'-'04,-

Part II

, _ ..- - -

.,..- ,.

eo ,Foreword

,
_.

.-,..-

,--

:-_-,,

_-7.,--.1..

Chapter

6. -45ERIVATiVEi.OF.E*NENTIAL AND,JRELATED FUNCTIONS


!!.

The Tangent ,line to-the Gr4S Of- ii-4aZ.:-ati,-,7(tta

6-2.

The Tangent at an Arbitrary Point

6-3.

ARplications of Exponential Functions

-6-4.

.s

"S'*:-T51:,,t,

6-1.

The Derivative of a Logarithmic Fun ction.).

6:5.

Taylor Approximations to the FUnction" x -,e

6-6.

The e-Powet Formula

6-7.

Approximations to Logarithmic,and Root Functions

gri

1.

--,

t.

Chapter 7,

.....

AREA AND INTEGRAL

7 -1.

Area Uhder a Graph ,

7-2.

The Area Theoret

7-3.

The Fillichmental Theorem of Calculus

7-4,

Properties of Integrals

7-5.

Signed Area

7-6:*

IntegrAion Formulas

C apter 8.
8'

-1.

DIFFERENTIATION THEORY AND TECHNIQUE

Differentiability

8-2:

Continuous Functions

8-3.

The'Mean Value Theorem

8-4.

Applications of the Mean Value Theorem

8.-5._ Sums and Multiples


8-6:

Products'

8-7,

Composite Functions

8-8.

The Chain Rule

''8-9.

The General Power and: Reciprocal Rules

8-10.

The Quotient Rule

8-11.

Inverse Functions

8-12.

IMplicitly Defined Functions

Chapter 9.
.

INTEGRATION THEORY AND TECHNIQUE

9-1.

The Method of Substition .0

9 -2.

Average Value

9 -3.

Volumes of Solids of Revolution

9=4.

Estimation of Definite Integrals

.7_12,5.

Taylor Apprqximations

tAppendix 3.
A3 -l.,

t.

MATHEMATICAL INDUCTION

The Principle of4lathematica1 Induction'

A3-2. 'Su/Ifs and Sum Notation

Appendix 4.

kuRTHER TECHNIQUES OF INTEGRATION

A4-1.

Substitutions of CircularFUnctions

A4 -2'.

Intagration'by Parts

Integration of Rational Functions


A4-4.

:4

Definite Integral;

vii.
9
r

Appendix

THE INTEGRAL FOR MONOTONE FUNCTIONS

A5-1.

Introduction

A5 -2.

Evaluation of an Area

It5-3.

Integration by Summation Techniques.

A5-4.

The Conceit ot%Integral.

A5 -5.

Elementary Properties of Integrals

Appendix 6.

Integrals of Monotone Functions

INEQUALITIES AND LIMITS

A6 -1.

Absolute Value'and Inequality

A6 -2.

Definition of Limit of'a Function.

A6-3.

Epsiionic Technique

A6:4,

Limit Theorems

Appendix 7.

-t

CONTINUITY THEOREMS

The Separation Axiom

AT-1.

Completeness of the Real Number System.

A7-2.

The Extreme Value and ,Intermediate Value Theorem& for Continuous

Functions
A7-3.

TheMean Value Theorem

A7-4.

Applications of the Mean Vdlue Theorem

Appendix 8.

MORE ABOUT ,INTEGRAIS

A8-1.

Existence of the Integral

A8-2.

TheIntegral of a Continuous Function

Appendix 9.

LOGARITHM AND pPONENTIAL'FUNCTIbNS AS SOLUTIONS


TO DIFFERENTIAL EQOATIONS

A9-1.

The Logarithm as Integral

A9-2,

The Exponential Functions


se

,V

iO
viii

Chapter 1
POLYNOMIAL FUNCTIONS

Here we An-

This is the first of two chapters.on p7olynomial functions.

centrqte on the deinitiati and simple algebraic and geometriF properties:of


polynomikil functions, leaving to the next chapter a discussion of tangents to
polynomial graphs.

between the'graph of a polynomial


Our.concern is with the relationship
.
-

function and the expres;ion which defines the function.

-i-

'

This relationships

Onstantand linear functions for these correspond to

easy to describe for

nonverttcal lines (Section 1-2).

For quadratic' functions the graph is a

parabola whose locatioh and general slope can be easily determined by using
the quadratic formula.
scale Change o

In fact any quadratic graph is just a transfationor

the graph of thesquaring function Section 1-3).

After discussing these familiar cases we turn to polynomial functions


of degree three or .larger.

Here the situation becbmes lass routin

thetic division serves initially'as a tool for plotting points (S


Later interpretations

Chapter 2) are more profound.

Syn-

tion 1-4).

'The general relation

between zeros end factors of the pblynomial is.discussd in Section 1-5.

In

Sections 1-6, 1-7, and 1-8 we discuss methods for locating, determining; and
.

approximatlnK zeros ofgpolynomial functions.

The final,section of this chapter

indicates some of the kinds of information about the graph of a polynomial

function which can be quick1ybbtained from its expression, includtng the


important result that the degree is'a bound for the number of iines its graph
crosses the x-axis.
ae

Further algebraic results are discussed in Appendix 2.

We begin our'studyith'polynomial functions because they are the simp


The theory, and techniques employed in

lest Cf,the elementary funci'ions.

Chapters 1 and 2 are fundamental to the rest of the text.


.

Not only will

ours approach to the analysis or.polynomial func9tons be useful as we de31

yith oder functions; but the polynomial functions themselves will serve as
,approximations to other functions 'we shstll study.

'

1-1

Introduction and Notation

1-1.

In this chapter weshall be concerned with functions that are defined by


expressions of the form
*IN

+ 8 X + 8 x
2
1

E1.4

where

+ anxn

is a non-negative integer and the coefficients

(i = 0, 1, 2.,

3,

n)

are real numbers.

Such expressions are called polynomials, and the

functions they define are called polynomial functions.


We commonly denote funttions by a single letter
to denote the value of

f(x)

at the point

is a function whose rule is given by

f(x) = a0 +a1x+

using the/symbol

f,

Thus a poiynorhial function

x:

-2

+anxn.

)1/4*

This notation is particularly useful when we wish to calculate various values

of ,f)

For example, suppose

is given by

..
..

= 2+ x.-

(0f(x)

,.4The values

and

f(0)., f(-1)

f(0)

are then given by

f(2)

+ 0-- 02 = 2

= 2

= 2 +. (-1)
f ( -)

x2.

= 2

1 2

- -

=o

(-1)2
= 2

(-2 )

.
.

We can sUf)stitute other letters or.expressile for

5c;

for example

f(t) = 2 + t - t2,

f(2 - y) = 2 + (2 -

ti

.3y

f(a + b) = 2 + (a

We may asso denote the function d


f

A
thus ,gtregsing thlit

a +

b.y -

4tined

In (1) by

x -*2 4 x - x

2
,

tv,

1,,is an opetation or association.

duceanother variable to stand for


graphing.)

1-

+ b - a2 - 2ab

= 2 +

(2)

(2. - y)2

f(x).

This is especially convenient for

For example we may rewrite (2) as

We frequently intro-

x -*y,

where

(3)

y =

2.+x-x2

y = x3

+x2

y = 1

y = 1 + 2x

'

Figure 1-la
The graph of a function,

(x,f(x))

is the set of pairs

as -wel.pieture

(In Fivre 1-la we sketch the graphs of

them on a plane, say the,xy-plane.

MuCh of our effort in this and the next chapter

thred polynomial functions.)

will be'directed toward quickly obtaining sue picturps.

The graph (a model)

'can help us t6 examine the behavior of a function (which may itself be an


Polynomial func-

tdealized matepatical alodel of some physical situation).

We give here two examples.

tiOns,often arise in applications.,


.
_

..

If we say "the volume of a sphere is a function of its

Exambl.o.1-1a.

,redius" we meah that if

is the volume function and

is the measure of

.1

is the measure of the volume.

where AT

the lius then-,f :,r -)11,

In

particular-we know that


.

.x..

(4)

.V

The expresion
volume ft5totion

-1;

nr

3.

is, of course , defined for any real number

nr

r.

The

r -41T

is, however, defined only when

r > 0.
.1

iadius is doybleid we can write


f(2r).= 1. n(2r)3 -

nr3

80- itr3),-.
3

which tells Us that doubling the radiudMultiplies the volume by eight.


S

A ball is thrown straightup with an initial velocity of

Example.1-1b.

64 ,feet per second so that its heiglit

feet above the ground after

seconds is given by tie function,

(5)

= 64t - 16t2.

(We shall later derive functions such as this as a consequence of various


This 2.unctin

physiCal assumptions about velocity, acceleration and gravity.)

can bnly serve as an idealized model of the physical situation over a particuliir

ihteval of values for

Since

t.

s = 0

when

t = 0

0 < t,< 4;

serves as a mathematical model over the interval


the air for

Or

4,

the function

the ball is in

To' find how many seconds it takes for the ball to

seconds.

reach its maximum height we can (by completing the square),write the function'
.

0'

in the form
t

The quantity ,-16(t -.2)2

64

and equals

)2 '4- 64.

reaches a maximum height of

64

Thus

cainot exceed

Therefore, we conclude that the ball

only when ft = 2.

64

t = 2.

is negative unless

feet after

seconds.

.p71'.

64 ft

height of
ball after
t
seconds

feet

ground

(i)

While we picture (Figure 1.3.b) the motion.funciion (5).as a parabola

we think,of the physical. motion of the ball itself as vertical (i).

1-1

'Exercises 1-1,
1.

.In Example 1-la


we expressed the volume of a sphere as "a function Of its
.
.

radius."
2.

txpress the volume of a sphere as Nt'function of its diameter'."

Suppose that a vellet is projected straight tap"and after a while comes


,

straight down via the same vertical path to the place on the grotind from

After

which it,was launghed.

seconda :the distance

feet of the

pellet above the ground is described by the equation


160t - 16t

2
,

which defines the/function

'1'
t -41.60t - 16t2.

when

Za)

What is the value of

(b)

Evaluate

(c)

How high above the ground is the pellet after

(d)

What is `the height of the pellet after. 6

t = 4?

f(6),
4

seconds?

seconds?
4N

Compare your answers for'parts (c) And (d).

(e)

Explain on physical

grounds.
f

How many seconds is -the-'pellet in the air?


(g)

'(h)

.A
'

How long does it take the pellet to reach, its highest point?
How high does the pellet go?

Tom is .standing on the top of a railroad car which is moving at a speed

of

32

ft. /sec. as it passes a. station.

As he passes Dick on the station

platfori Tom throws a-ball straight upward with an initial speed of


64

ft.isac;

After

seconds the ball is a horizontal distance of

feet and a vertical distance of


te,

The distances

ft. and

feet from.a ,point opposite Dick.

ft. are given by the-equations.

x = 32t
and

y. 64t
(a)

Does Tom have to move tocatah the ball?

(b)

What.is'the path Of the ball as Dick sees it from the platform?

(c)- Write

16t2.

y 'iriterms of

X.

IV

(d)

Name the curve that is the graph of your equation in part (c).-

(e)

Sketch the graph of

-(f)

For what 'values of .x

y = 2x
does

x2

a."
y = 0?

when x = 128?

(4)

What is

(h)

After how many' seconds does ToM catch the ball?

ci)

How far down the platform from Dick'does.Tom catch the ball?

I.

1J2

1-2.

Constant and Linear Functions


The simplest polynomial functions are cbnstaRt':futictions.

is any

which associates with every real number

real number, then the function* f

k,

the value .c,

If

fi

f : x -)c,',

is called a constant function.

The graph'of such'a'fUnction is a line parallel.


Some examples%of constant functions are

units from the x-axis.

to.aAd .jcl

graphed in Figure1-2a.

o4

y .0/

2,

Figu e 1-2a

X
Constant functions are quite simple, yet they occd?frequently in mathe-

matics and science.

A physical example of such a function is


f

re

Here the constant is the acceleration ..,


due to sravity,.that is2the constant
.-,,,

n distance is

amount by which the velopity of fall increases each secoAd.


da.

'32: at sea

'measured ill feet and tie in seconds, this constant is very nearly
level.

In other wordS7thevelocity of a falling body increases

feet

ter second every second.

A simple principle will later be useful when we encounter constant func-

tions:

%if

is a constant function and the value 1.(a)

is known,

(?-).

then

f(x)..-. f(a)

for all

is 6 constant' function and that

For example, if we know that


f(0) ;,-,10,

than we know that

restatement of the fact that if

x.

f(3)
f

is also

10.

Clearly, (1) is just a

is a constant function then all values

oP f`are the same.


./-

1 7

1-2

A linear function
that is,

mx + b;

all

is a function defined by .an expression of the form,

is linear if there are numbers

and

such that for

f(x) = rot + b,

(2)

If

m = 0

tlen

is a constant function.

crosses the Y-axis'at the point

'(0;b),

m f 0.

The graph of (2) is a line wihich

since

i(0) = b;
b

is called the yam- intercept of 'f.

The number

is called the slope of

FoUr linear'functions are

and gives'a measure of the steepness of the graph.


sketched in'Figure 1-2b.

f,

\
\

/
/

/
.

-4

\ \

X(

/
-1

/
x

y sx

= 2x + 1,

y = -x + 2

x
t

Figure 1-2b
4

is linearsadd

x
1

/ x2, then the slope

is given by

f(x2) - f(xl)

m -

( 3)

- x

'

that is,
of

f.

is the tsngent,(telgonometric ratio) of the angle of inclination

(See Figure 1-2c)..

(x2,f (x2)

- f(xl)

- xi

angle of inclinatibn

Figure l -2e
A

1-2

The ratio (3) is, of course, a simple consequence of (JI since


f(x

(mx 4- b) 2

- f(x1 )

b)

x2 - xi

o use a slightly diff rent.form o

It will often be convenient

This is c9ntain d/in the following:

expression (2).

The equatIon'of the 1 ne/through

Example 1-2a.
1

the

(h,k)

with slope

Find the quation of the line throu

(1,2)

and

*4.

The slope of this line


2
-

,8
9

3
3.-

(- F)

(1,2,

$o, using the point

the form (4)

y = 2 +

'

Ives the equation

8
-

1).

Using the point

(- 7

gives the

quation

-(

y - -

Simple algebra shows that these- two equations are just d fferent forms of the
equation
8
3r/;=

x +

10
.

The graph of this line s shown in Figure 1-2d.


1
y

Figure 1-2d
9

fi

19

1-2

A number of very generE4 concepts have sim-ple.and usefUl formulations for


V.

-0.f these we shall now discuss slope as velocity,

constant and lidear functions.

inverse linear functions, translation and scale change, and proportionality.

Slope as-relocity
For

The slope of d linear fUnction,hat various physical interpretations.

example, suppose a man walks north alOhg a long straight road at the uniform
rate of

At some particular time, say time

miles per hour%

t = 0,

man passed the milepost located One mile northoi Baseline Road.
before this, which we shall call time, t = -1,

this

An hour

he passed the milepost located

one mile south of. Baseline Road. (All hour after time

t = 0,

at time

he passed the milepost located three miles north of Baseline Road.

t = 1,

(See

Figure 1-2e.)

position
at t = 1

position at

.K

t =

BASELINE ROAD

position at

t =

Figure 1L2e
In

time

t 'hours, the man travels

,Singe he is at milepost

2t

at

he must be at milepost

2t + 1 at time t. Us4ng f(t) to


#
represent the directed distance (in miles) from Baseline Road at time t
t = 0,

hours, we seethat
f

t -of(t) = 2t + 1

The graph of

describes the man's moidft.

shown in Figure 1-2f thus plots

to

the man's positioil versus time:

20
I

10

1-2

.velbcity
(in mi/hr)

distance
(in !ages)

2t + 1

\.

slope = 2

g :t

time (in hours)

-1'

'Figure 1-2f

time (in hours)

..Figure 1-2g'

In Figure 1-2g we plot the man's velocity versus time. For all values

of t 'dining the time he,ls walking his velocity is


Hence if

g(t)

is his speed -at time


-g

that 'is, _13

is a constant function.

the position function

t
t

ti

miles perhour.

then
2,

In this case, the slope function 'g

can be interpreted as velocity.

'of,

We shall encounter

this relationship again.

Inverse Linear Functions


We recall that the rule for converting from Centigrade to Fahrenheit is
given by the formUla

(5)

C + 32

and the rule for converting from Fahrenheit to Centigrade is given by,tbff__

Vbrmula

(6)

32).

*That speed is the absolute value of velocity is spmetang that we shall


emphasize later.

2i

e0

'04

1-2

6s

`Since each of these formulas is useful, it is important to be eble,;to get,


say

6) from (5).

-then add '32.

in (5) to get from- C

Starting with

then divide by

we must', therefore, first,subtract

That is, to get

we first multiply by

.io

qi we first find

F - 32

and
324 and

then

1F

5/

t'

., N a.%
,.
whence
.)/i
?

'or- -1F
9

5.

,'

C = 2(F - 32) In our next example we shall study two


9
functions suggested by the foregoing formUlas.
a

Example 1-2b.

Consider the function's

and

32

5
g

5
x 7) -(x - 32)

land their graphs, sketched on one set of axes'in Figure

,We observe immediately that the slope of


g

is the reciprocal orthe slope of

and the slope of

is

f.

= f(x)
11P

100

Y = g,(x)
47

100

Figure 1-2h
Let us make some further observations to see how the functions
and their graphs are related.
-To-n,find the value of
---failid,then add

32;

Consider the equation


x

for a given value of

i.e.,

and

x + 32.

y = f(x) =
5

we first multiply

by

on the graph (Figure 1-2i) we.first"go up"


,

4!
re

2 x
5

1-2

to the Jotted line and then "garup"


32

more.

to get
y

Finally we "go across"

y =

9- x + 32. To go from
5

back to- x

we just reverse the

arrows of Figure 1-2i.


subtract

32

to get'

divide by-1 to get

''.

32

We first

i.e.,

x;

y - 32,,5

r
.

P*

9 x

5
,Let
.

.x

.//.

and then

x- and

exchange roles and


/

compare this form of the egtation

/
.

for

with the- function

g.

Figure 1-2i

The two functions


if

m /'0,

and -4,_a_re obviously closely related.

In general,

we say that the, linear

4.

is the inverse.of the linear function


f,: x -)mx + b.

The slopes of g

and f -are reciprocals of each other;-that is, the

product of the slopes is '1


The graph of
.

'g

can be easily obtained, from the graph of

s ' that (C,0 1 es on the graph of

f,

Suppose

f.

so that

d =f(c) = mc * b.
Solving for

c,

we have.
1

c=
Thus

(d,c)

lies on the graph of

d -

g.

.t

`similarly established.

g(d)

The converse of this statement can be


.

We,summarizepi.n (7) .'

If g is the inverse of
(7)

then

(c,d). :lies

on the graph of

f ,if and only if

(d,c). lies

on the graph of

g.

f,

er

miakr

1-2

.41. /

.
.

Thus the gphs oll and its inverse

are symmetric with respect to


ion, for it

This has a simple geometric interpret

the line given by y= x.

says that the graph of tie inverse

raph of

can be obtained from the


This s

merely by interchanging the coordinates of each poirit.

gests the
AN.

y = f(x)

trace the graph of

of

= g(x)

Merely-.

in slow drying ink and then fdld carefully along

(.Consider this mechanical procedure for the graphs of

f.

f.

The wet ink will then trace the graph of the inverse

y = x.

the line given

from that/of

following wayto obtairilthe graph of the inverse

and

y = f(x)

in Figure 1 -2h.

Translation and Stretch


Let

on

be'the line given by the equation

of replacing x by

(x - b).

21

x -axis at the point

(b,0):

y = x - b,

The new equation is

having the same slope as

represents.ths line

and consider the effect

y = x,

which

and ilterseeting the

Figure 1-2:j

4
Hence, replacing

by

x - b

translates or slides

right without changing its slope (Figure 1-2j).7

2-4
44,
11,

units to the

1-2

M
r

Now consider the effect on


equation is

y = mx,

of replacing

which represpnts the line

x
2"

by

mx,

m / 0.' The new

having slope

and

pgssing through the'origin.


0;

2"
4

Figure 1-2k

Note thaiif m > 1,

replacing "x

*4 steepens the slope of

Norizontal than
Thus, if
y = x

1,

while if

What happens if

2.

m > 1

by

mx

(or equivalently,

0 < m < 1

the new line

y
2" -is-ilore

m < 0?

the effect of replacing

is equivalent to stretching the ordinate

by
(y)

mx

in the equation

of each point on

fK

Figure 1-22
t

15 ze--\c:

1-2

"contracted." , Notf that conceptilally,


"shrinking each -0
ordinate is no different in this case from

each ordinate is

If*,0 < m < 1,

'"stretohing each

abscissa"; the resulting

2"

is identical,.

Y'

Either 1.47' we think about it-, the slope of

mation, and

by mx

2. is--c

aged under such a transfor-

by replacing
algebraically the stretch can be obt=iiined mereky

(or' y by 4)
m

in the equation of

2.

can be obtained
equation y = mx.+ b
The graph of any line given by the
stretching.' For example,
translation and/or
y = x by such
from the graph
may' first stretch each
ffrom y =
y 7 5x graph
of,
to obtain the
.
The equation
by replacing. x by 5x.
ordinate of, y = x by the factor 5
may
2:
shown in Figure 1-2n- Then
and its graph is 2'
becomes y.= 5x,
the equstx by x - 5 in
units to the right by replacing
be translated

tion of

2'.

26
16

1'
1-2

4- .The equpion bedomes


4

Y = 5(x 5 )

i or

y = 5x - 4

as desired, and its graph has been obtained from the


translating and stretching.

x.

by

and the graph is shown in Figure 1-2o.


5

by replacing

y =x by

A.4
.

Alternatiyely.fwe may first translate the graph

units to the right 'by replacing

a factor of

z$-

by

5x.

x,- 4.' The equatron.becomes

y = x - 4

Then we may "shrink" each abscissa by


The, new equation is

y = 5x - 4

_before; only our way of thinking about, the transformations is different.


4-.

t
o

as

1-2

9
x

.`$

Figure 1-26*

I
Proportionality

Ns

-,

.,,

I
The concept of proportionality is very useful in physics as well as other

or

parts
y,

science.

We shall use the idea frequently in this text.

is proportional to

means that there is a number

To say that

m' such that

y = mx

for all numbers

x._,The number

is called the constant elf proportionality.


4

Note that if
doubled.

is proportional to

x,

then

doubles when

The,same relationship 4o1ds for tripling; halving, etc.

the experimental observation that

is

In science

doubles, triplz1 halves, Ate., when

,7

does the same - 'usually leads to the hypothesis that

is proportional to

x.

Further observation is then'used to test such an hypOthesis and if no contra/

dietory evidence is found this proport4nality is usually stated as a,law and


,

th&elfter systematically used.

For exaMid1e, if air resistance is neglected

it is usually safe to use the assumption that the velocity of a freely falling
body is proportional to the time it

28

1.8

1-2

Example 1-2c.

Assume that -the velocity of a free falling body is directly

proportional to the time it fOas.

Suppose that a ball is dropped from the top,

of a building and attains.a velocity of, 64


fast

be falling after

,ground

then)?

sec,,we have

v = mt,

v =.64

t = 2,

we get

When

= 5,

t -) v = 32t.

velocity is

160

seconds.

Hoy

seconds (assuming that it,hasn't hit the

.'

Sincthe velocity

when

ft./sec. after

where

ft./sec. is'directly proportional-1 the time

is the constant of proportionality.

m = 32.

v = 160;

WA Otain the linear function


therefore,'after

ft./sec.

VAL

19

29

seconds the

If

1-2

Eke raises 1-2

1._ gee'r to Figure 1-2f .


,\

t -9 2t + 1?

(a)

What .is the /slope of the Linear function

(b)

What are'the units of "rise over run" in the graph of t -9 2t + L?

(c)

Compare the number and units.of parts (a) and (b) with the consilant
function and vertical units of Figure 1-2g.

0
2.

(a)

On separate sets of axes ,sketch graphs of the functions


g

t -9 32t

and

"g'

32.. Indicate vertical and horizontal

units appropriate to Example 1-2c.


(b)

What is the slope function of the linear function

(c)

What are -tie units of the slope of your "graph of

g
g

t -932t?
t -) 32t1

NO.er .

(d)

Compare the .vertical units for your graph of

gl

t -9 32

,..14th

your answer to part (c).


(e)

What word from physics is commonly asso.4.eted with the xatio. of


units you found in response to paii, (c)?

3.

Assume (as in Example l72c) that the velocity of a, free fallingbOdy is


directly proportional to the time it falls.

Suppose that a penny is

dropped from the top of a tower and attains a velocity of

second after 1

seconds.

hits the grounp4fier


4.

For the function

42

48

feet per

Determine the'impact velodity if the _penny


seconds.

x -92x + 1,

(a)

f(0)

(b)

f(1)

(c)

f(-1)

(d)

for h / 0,

f(x +'h)

(e)

for

f(x) - f(a)
x - a

find

- f(x)

5.

x / g,

Find the slope of the- graph of the function


numbers

x,

(a)

f(x) = 3x - 7

(b)

f(3c) . 6 - 2x

(c)

2f(x) =3 - x

(d)

3f(x) = 4x - 2

3020

if, for all.real

1-2

6.

Find'a linear' function


./(a)

whose graph has slope

and 'such that

-2

f(1) = 4

7.

8.

9.

(b)

f(0)

( c)

i.43),=

f(8) = -3

Find the slope of the graph of the linear function' f


(a)

f(0) = 4

(b)

f(2) = 3

(c)

f(5) = 5

(d)

f(6)

-13

if

f(1) = -3, and

/(

Find a function whose graph is the line joining the points

02,4)

(a)

P(1,1

(b)

P(-7,4), QC-5,0)

(c)

P(1,3),

Q(1,8)

(a)

P(1,9,

Q(-2,4)

e.

Find the linear function


coordinates
f

(-2,1)

tr

whose graph passes through the point With

and is parallel to the graph of the function

3x - 5.11,

10.

Given

x -

1'

graph of

(a)

P(0,)

(b)

P(-2,3)

(c)

1,(1,5)

-3x + 4,

find a, function whose graph is parallel to the

and passes through the point

(d)

11.

If

f '14 a constant function find .f(3)

(a)

f(1) = 5,_

(b)

f(8) = -3

(c)

f(0) :714

21

31

if

1-2

12.

Q(3,-1), and

e.points. P(1,3),

Do

S(7, -9)

all lie on a single line?

Pro e'your aasertion.


13.

The ,graph of a linear fuhction


and

Q(101,39)., Find

(a)

f(100.1)

(b)

,i(100.3)

(c)

f(101.7)

(d)

f(99.7)

f ,passes thrOugh the points, P(100,25)


I-

14 The graph of a linear function

P(3,25)

passes'"through the points

'and ,Q(54,:19).Find

15.

(a)

f(53.3).

(b)

f(53.8)

(c)

f(54.4)

(d)

f(52.6)

Find a, linear function with graph parallel to the line with equation

and passing through the point of intersection of the '

x - 3y.t'4 = 0

2x + 7y + 1 = 0, arid

lines with equatibns


16.

A(1,2),

Given the points

B(5,3),

C(7,0),

x -.2y + 8 = 0.
D(3,-1),

and

shoW tha

ABCD is a parallelogram.
17.

Find the coordinates of the vertex

if AC

of the parallelogram ABO

is a diagonal and the other vertices are the points:

18.

(a)

A(1,-1),

(b)

A(0,5),

If

it

B(3,4),

(a)

D(4,1)

B(1,-7),

If you graph the set of all Ordered pairs of th- form

tion
6

(t)

Find

f.

f(0)

(t

1, 3t + 1)

you will, obtain the gr ph of a linear func-

and ,-f(8).

_,,,

If yo

is on

x -)2x - 1.

for any real number


g

1)

is a real number, show that the point. Pct + 1, 2

the graph of
19.

D(2,3);.,

graph the set of all ordered pairs if the form

(t -

function

t2 + 1)
f.

for real

Find

f(0)

t,

and

you will btain the graph of a


f(8).

4 ).

32

22

1-2

20.

At what temperature do .Centigrade degrees. equa1 Yahrenbeit degrees?

21.. Ifsthe slope pf a linear function

1 <

22:

.then

f(\xi): >

"-

y 0

__that

if

f(x2)

Consider the linear,6nctions

that m

is negative

and

p = g(q)-

and

nix + b

if' q = f(p)

+ f3

p land

q.

23.

I40 f

x -)mx + b,

24.

What is the equation of the line perpendicular to the. line given by


.

25.

y =
If

+ b,

What is the relationship between

/ 0,

f(2x - 1) = 4x2

'0

find

and

the inverse of.

g,

such

for all real numbers


1.i?

f.

at the

- 8x + 3,

'find

f(2x).

tt
o

23

a.

0,5
4

1-3
Quadratic Functions.

1-3.

As we discuss the.behal0.or of polynothial functions


2

f : x -)ao + alx + a2x

conventional terminology: If

we shall use soc


of

f, is

11

we say that the degree

/ 0

while-

0,

f : x -'2 - 3x + x.

has.degree

For example

n.

has degree

+ anxn,

,71-

x5

This convention assigns no degree to the zero function

5.

(1)

x -)0.

Thezero function should not be confused with the zero of a fUnction.


We say that a-number
equation

f(x) = 0

if

r' is a zero of the function

f(r) = O.

For example,

x -)2x

- x

of of the

or a

is a zero of t e function

15

or a root of the equation


-

2x

-,x - 15 = 0,
xt

f(3) = 0.

since

If r

is a'zero of

then

(r,0) 'lies on the graph of

f crosses the x-axis at the point where

the graph of
4

f;

that

x = r..

Polynomial functions of degree

are known as quadratic functiOns.

us review some of the properties of quadratics.

The zeros of

f:x-,c+bx+ax 2, a/0,
are given by

/2
-b + O -Itac
(2)

If b
2

- 4ac < 0

and

-b -

2a

these roots are complex numbers.

- 4ac > 0.

34

2
-

-lac

-2a

Real roots occur if

Let

r
1 -3

If

---

ac = 0 then

b2

and we can factor toobtain


co.

where

r1
to obtain
,

If b

F2 =." 2a

- 4ac 150,

then

rl

and we can fbctor

ti

(4)

,x -q(x -

r2).

In the following three examples \ we illustrate the graph of f for each


of these cases. In the final example\ we review a method for graphing
quadratics

by translation and change of scale.

Example 1-3a.
\.

Graph

In this ,9ase

+ x + x2.

a = b = c = 1,

so
2

and

has no real zeros.

cross the x-axis.

---- 4ac < 0

We should expect that the'graph of

In fact, the graph of

doesn't

lies entirely above the x-axis.

We caA show this by "completing the square" to obtain


_

1 + x

5tnee

x)

> 0

unless

x2 =

x = -

f(x) >

+ x 4- x2)

1
2 '

we see that

- *I

-.%

'1

while

f
a

f(-

Thus the-graphof

= i..

lies above the line given by

y =

i,

touching

this line at the point


\ Al

value

x = --

the quantity

.2.- ,

3.-

(- Dv. Furthermore,
I

large. Also as

as

increases

the

,1
Vz.'4. x)2

becomes very large as ''x

decreases (to the left of

x = --

2:.

becomes.

) the quantity

(-1
.

)2 )2

inOreases, becoming very large.as we assign numerically large negative values


to x. Thus, without plotting points other than (- i,i) We,cat conclude
that the graph o'

appears a's shown in Figure 1-3a. 0; course, a more

25

r,

11

1-3
6

accurate picture can be obtained 'by plotting some points .(x,f(x)).


becomes large
as x moves
to the left
f
becomes large
as x moves
far .to the right

c6,

tpinimum(

74

'Figure l -3a..

Example. 1, Graph
In this case

and

x -41 + x + x2

x -4 4 - 4x + x2.

t=

so that

b2 - 4ac = ( -4)2 - 4

(1)(4)

b = -4,

c = 4,

x . 2.

has the single zero

= 0

Therefore, we can write

...

and

x / 2

f(x) > 0

so that we have

if

x / 2

(\

Therefore, the graph lies above the x-axis, touching this axis

f(2) . 0.

at the point

if

(x - 2)2 > 0

The quantity

(2,0).

As

x ='2

increases to the right of


.

decreases. tOthe left of

the quantity

x - 2)

or as

,,

increases, becoming

x moves far away frOm x = 2:' As in the previous example,


this gives us enough ,information to quickly sketch the graph of f, lhown
very large ,as

in-Figure 1-3b.

36
O

26

173

becomes large
as x moves
to the left

becomb; larg
as x moves
to the right

(2,0)

Figure 1-3b.

x -)4 - 4x + x2

Example 1-3c.

Graph, f : x -71

x - 2x

Since

b =4'1.1,

a = -2,

we have

c = 1,

'Minimum

2
.

b2 - 4ac =.(-1)2 - 4(-2)(1) = 9

The zeros of

are

r1 '=

we can write

-(-1)+

x -' -2(x + 1)(x -

The graph of
x < -1,

r2 - -(-2)4-

and

in the form

f'`

If

crosses the x-axig atthe two points

each of the quantities

'multiplying by

x + 1

1
x - -f

and

-2 ,we see that


f(x) < 0'

and

is negative.

if

x < -1.

Similarly, we could argue that

f(x) >0

(-1,0)

if _ -1 <x <

s,

f(x) < 0

if

x >

27

0 1

1
(IF, 0) .

Upon

1.-3
as

decreases to the'left of

Further arguments show that as

-1

or increases

to the right of
as

decrease, getting far below the x-axis

f(x)

the ualtie

x moves far to the right or left.;


-1 < x <

'In the interval

the graph of

lies above the x-axis.

We can "complete the square" to obtain

s,

#'

0
= -2(x2,+ 2 x

2)

= -2(x

.1

= -2(x + t)! + g

This expression has its greatest velue when

.1
E.

x =

9.
(- v,v

so that

is

"

the highest point on the graph ,of

(See Figur? 1-3c.)

f.

y
,e-

maximum
(1, 0)

(-1,0)
far below x-axis
moves left
as
lc
or right

Figure 1-3c.

- x - 2x

2
.

Translation and. Stretching of Parabolas

Just.as any line with Ositive slope may be obtained from the graph
y

x by appropriate translations and streAkhes, any parabola may be obtained

front the graph:Of

y = x2 ,by similar transformations.

y = 14 + 12x.+ 2x2

For example, to obtain the graph of

f'

of y = x

we first rewrite the equation as

from the graph

y = 2(x +.3)2 - 4

by completing

the square as follows:


10 + 12x -4- 2x

,t

6x '+ 7)
t

= 2(X2 + 6x + 9 - 2)

= 2(x + 3)a:- 4.
a

28

1 -3

In this form, the apOopriate translations and stretches are readily determined.
2
y = x

First, translate the graph-of


three units to the left by
.

replacing

by

.,

(x + 3).

.._...

The new
2

parabola has the equation

-5-f = (x + 3 )

. ,
i

(See Figure 1 -3d.)

'Then stretch" each ordinate of

to graph of
y

by X .
2

y = (x + 3)

by replacing
.2

'\

is shown in Figure 1-3e.

//

.'

The graph of r, y = 2(x + 3)2

-3

...

\..

Figure 1-3d
y = 2(x +

2:

2 i 1 0

/
/

/.k

e-

Figure 1-3e

'Figure 1-3f

Finally, die more translation,

Units down is required to obtain the

desired graph:, This is achieved by replacingy by


'Ty = 2(x 4/8)2ek See Figure 1-3fAfor the graph of

If the coe ficient of the

2
x

(y + Id

y-= 2(x + 3)2 - 4.

2
y =x1//- x - 2x ,

one more transformation

would have been 'required to obtain the graph from the graph of

x +

giving the graph of

5r= x2,

y = (x +

unit to the left by replacing


)2

Com7.

by

Then stretching each ordtnate by

4AV

y = x2.

y = -2(4 + .)2 + g and in this form we see we must

first translate the graph of


1

in the.equation

term inyte original equation had been

negative, as in'Example 1-3t, where

pletingythe square gi'es

-1

1-3

.r.
Translation
replacing

y, by

'produces

4.
the graph of

7,1

12
.

y = 2(x

elbw the extra transformation,.


rellnilag the graph 141-Che
x -axis by replacing
gives -the graph of

by 40,

y = -2(x +

teehis is physically equivalent to

f.44ding the graph along the x-axis

or'"clipping"

r51.

the graph about theme

Finally, translating this

exit.

graph bt units' -up by replacing

by;, y - 9g gives the desired


v.

rap&
In general, ..the graph of. any'
I.

a.

quaEattc function can be obtigped


`*2
by such
x
m the `graph of y
a sequence of trapslations, reflec-.
.tions and stretches.

quadratiC function

Hence, any
f : x

bx + c

represents merely a translated;


reflected, and/or stretched image
of the standard, parabola given by
2
the equatiob y = x .

Reflection

414

-1

1-3
Exercises 1-3
1.

Consider the function

Name the type of function

2.

(a)

-a = 0,

b =0,

c #0

(b)

a = 0,

b / 0,

c = 0

(c)

a =

0,

c #0

(0)

a / 0,

b = 0,

c = 0

(e)

a / 0,

b # 0,

= 0

(f)

a / 0,

b, # 0,

ax

+ bx + c.

is,/if

t1

Aristotle claithed that the speed of a free falling object depends on the
-r

weight of the object as well'as the length of time it falls.

Galileo

discovered that the need of a free falling object depends only on how
lo'ng it falls, and,in particula4r, that speed -v

proportional to'time' t
(a)

ft. /sec. is directly

:seconds.

A ballis dropped from the topof a building and attains a speed of


64

ft./sec. after

seconds.

How fast will it be Balling after

51..seconds?

(b) ,,,.4.11,,fe raft is dropped,from a helicopter and hits the water after
10

seconds.

If the raft is falling -at a speed of,,64

ft./sec. after

seconds,, determine how fast it is.going,aS it hits the,water.,


,.
'

Galileo:discovered that the distance traveled by a falling bogyHof

any,

weight depends only on the length of, time in which it has been falling.

Specifically it was discovereqrthat the number of feet fallen is directly


proportiohal to the square ofthe numbel- 'o,f seconds elapsed.%
(a)

Suppose we timed the fall of a ball


.feet'high and discovered ti
secondiTt

rom the top of a,builldiUg

the `x1/1

400

it the ground after ,5'

Find how long it Wouid take for the ball to hit the ground

if it'were dropped from a gUilding

144

feet high.

(b) ,Suppose that d ball is dropped'from a television tower and hits the
ground after

10

Seconds:

hit the ground after


400

'Previously we discoverg that the ball

Secongs,when it was dropped from a building

feet high.' How high is ,the television tower?


.

4.

31

it 1

1 -3

4.

For each of the following pSirs of functions,

(u,v)

is on the graph of

..,,

and

(a)

v < w.

v = w,

x --)2x2

g : x -)-2x
(b)

g. 'Determine which is correct:

is on the graph of

(u,w)

v > w,

1 x2

g : x -) 2x
(c)

5.

g : x

'2
-2x

On the sear set of axes sketch the graphsof the functions


x

6.

2
,

and

5x

2,

To x2.

Describe the location of the.points

y = ex2 ,
7.

42

f,: x

and

(p,q)

(-p,q)

on the graph of

relative to each other and the coordinate axes.

A pall is dropped from E; 47th story window of the Time-Life Buiiding in


New York City.
after

Its distance

feet above the Avenue of the Americas

seconds is described by the equation


s = 576 -- 16t

2
,

which serves to define the function


%

f
(a)

576 - 16t2.

Determine how many feet above the pavement the ball is after falling
the first

second_

(b)

WA/ high above'the ground is the ball after zero', seconds of;falli4g?

(c)

How'eigh above the-Avenue of the Americas is the 47th story window

*WO-

t'.

of the Time-Life Building from which we dropped the ball?

(d) 'If f
(e)

t .4576

16t2,

f(4)

evaluate

Four seconds after it is dropped 'from the 47th story window, how

Tar is the ball from the pavement?

for

lot

= 576

and

t > O.

( f)

Find the value.of.t

(g)

Determine how long it would take for a ball dropped from,e 47th

'which

story window of the Time-Life Building to hit the pavement below.

32

42

1-3
.

8.

A ball ii dropped from the top of the Fidelity Union Tower in Dallas,
Texas.' :After

seconds the height

ft. of the ball above the

ground is'given.by
1

s.= 400 - 16t2.

(a)

What is the height of Fidelity Union Toirer?

(b)

How long does it take for the ball to reach the ground:

9 The Woolworth Building in New York city is about

784

feet high.

A ball

is ,dropped from the top of the Woolworth Building so that its distance

feet above the ground after

seconds is described by the equation

s = at
(a)

+ c.

Relating your experience with other problems of this type to this


problem, try to determine appropriate values for

(b)

10.

and

c.

How long does it take for the ball to reach the ground?

The vertex of.the parabola given by


If

a > 0

y = ax

the graph of the function

x -*ax

+ c
2

is the point

+ c

and the vertex of the ,parabola whose equation is

opens

y = ax

(upward,downward)
+ c
is the

point.

( ighest,lowest)

The graph of the equation


number and

c > 0

y = ax

where

+ c,

is always a

is a, non-negative real

which is symmetric to the

, congruent to the graph of

x -*ax2

and

units
(above,below)

-Ae>pait'bola given by
11.

y = ax2.

A flowerpot falls from a'75th story windowsill of the Chrysler Building


in New York City.

We know, that,after

seconds the height

feet of

the floWerpot above the'grokind is given by the equation

's,= 1024 - 16t


(a)

2
.

How long does it take for the flowerpot to7hit the sidewalk at the
corner of,Lexington Avenue and Forty Second Street directly beneath
the window?

(b)
,

The distance from the 75th story windowsill to the roof of the
Chrysler Building is

22

feet.

33

How tall is the Chrysler Building?

Ae

1-3
*;

12.

Suppose that a ball is thrown straight up from the ground with an initial
'speed of

ft./sec. and is not acted upon by the force of gravity.

64

What

is its height above the ground after .10 ,seconds?


A ball is thrown from ground level straight up with an initial 'speed or
64

ft./see.

Its distance

feet above the ground after

seconds is

described by the equation


s = 64t

which specifies the function

c';

16t2,

such that

s = f(t)k..

.7(`1 ,Whatcis the,value df; s, when -t =,1? ,;

"

(b)

Evaluate

(c)

What is the height of the ball above the ground after


after

(d)

t(2).
3

seconds?

seconds?
s = 64t - 16t2.

Sketch the graph of

(e) *What is the path of the ball?


(f)

13.

What is the name of the graph-of the function

-16t2 + 64t?

Suppose that.a pellet is projected straight up and after a while comes


straight down via the same vertical path.to' the place on the ground from

After

which it was launched.

seconds the distance

ft. of the

pellet above the ground is described by the equation


s = 160t - 16t2,

which defines the function


f

(See Exercises 1-1, No. 2.)

160t - 16t2.

Sketch the graph of4 f

on the interval,

0 < t < 10.


14.

The product of two consecutive integers is zero.


integers be?

4A

What could the

15.

Suppcie that you are standing close to the edge on the top of a. building
80

feet tall.

You throw a ball upward wipl an initial speed of-:64 ft./

sec. in a nearly vertical path.

After

seconds the height

feet of

the ball above the ground is given by the function


t

=.8b +

Olt - 16t2.

(a)
(b)

,How long does it take for the ball to reach the ground?

How high above the building is the ball after one second?

after

three seconds?
(c)

The ball passes the edge of the top of the building from which it
was thrown as. it falls to the ground. ,After how many seconds does
.this occur?

16.

(d)

After how many seconds does the ball reach its maximum height?

(e)

How high above` the building does the ball go?

For each of the following pairs of equations, given that


the graph of the first equation and that

(u,w)

(u,v)s

is on

is on the graph of the


f".

second determine the values of


(a)

for which

y = 3(xi- 4)2

Compare the graph of

v = w,

v > w.

y = 3(x - 4)2

(b)

y = -3(x - 4)2
17.

v < w,

y = 3(x + 4)2
1,

x -,-0x - 3)

+ 2 ,with 'that of

1
x --).-

on the

listed characteristics by completing the following chart.


1
2
x -4-- x
2

Function
(a)

Media of graph

(b)

Opens (upward or ,downward)

1,
. 2
x -5 --pc - 3)
+ 2
2

.
.

(c)

Equation of axis.

(d)

Coordinates of extremum/point (vertex)

(e)

Type of extremum (minimum or maximum)

,
.

113.

Sketch the graphs of

x -, x

1,
. 2
x -52 --kx - 3)
,

and

of coordinate axes.

+ 2

on the same set


I

Check your answers for Number 17 against yur gra

35

4;

1 -3
,

19.

(a), Using one set of coordinate aXes plot the graphs of the following.
t

y = -x2

(ii)

y = -x

(iii)

y = -x, - 2x + 3

+ 4x + 5

On one set of,coordinate axes quickly sketch the graphs of

(b)

20.

(i)

(i)

(ii)

: x

(iii)

-x

+ 9
+ 4

+ 1)

Consider the functions

x -*ax2

and

x -4a(x

h)2 + k.

Let

be a point on the graph of, f.

(p,q)

We know that

(a)

-(x - 2)

Another equation relating

f(p) = q.

and

is

q =

We want to show that the point

(b)

(p + h, q + k)

lies on the, graph

of
4(c)

g. _Show that

g(p + h) _= q 4--:k.

To every point t(p,q)

on the graph of

point

(p + h,

on the graph of

there corresponds the


g.
o

(d) 'In particular,.we see that the vertex

4
..4.4en by .y = ax2

parabola given by

corresponds to the vertex


Y'"gliCk,,,/,h)

of the parabola

of the

+ k.
(..

,,,_

21.

Determine the coordinates of the vertex and the equation of the axis of
,---

the parabola given by each'of the foliOWiiig equatio'ni.

(a)

y = 2(x - 3)2 + 4

(b)

y = -2(x - 3)2 +4

(c)

Y = (x

(d)

y = - 1(x

(e)

Y = 3(x 4- 1)2

(f)

y =

1,(x -

3)2
-

1)2 - 1
2

2)2 - 3

5,

22.

Determine th

ext

whether it4ia

mum point'of each graph, in Number 21 and tell


or minimum.

36

46

1-3
.

For each of the following pairs of equations, given that

is

(lio')

on'the graph of the first equation and


second, determine the values of
(a)

(u,w)

for which

is on the graph of the

v <

= w,

y > w.

y = 2(x - 3)2. + 6

y = 2(x - 3)2 - 6

y = 2(x- 3)2 + 6

(b)

y = -2(x - 3)2 - 6
(c).

y'= 2(x

3)2

y = 2(x-+ 3)2 + 6
24.

JA

Write each of the following equations in the form y = a(x - h)2 + k.


2

(a)

y = x

(b)

y = 2(x2 - 6x + 9)

'(c)' y = 2x

- 6x + 9

- 12x + 18

(d)

y = 2(x2 - 6x + 9)

(e)

y = 2x2

12X+ 22----

(f)

y = -2(x2 - 6x + 9)

(g)

(h)

y = -2x

(i)

y = x. +

(j)

y = x

(k)

Y =

-2(x2 - 6x + 9) + 4
2

+ 12x,- 14

+9

-,2x + 1
I

- -2- x

+ x

x + x
(in)

y = x

(n)

+ 2x + 1

= 3X2

(0)

(p)

y = x

(q)

y =

6A + 9

2
3x + 6x + 11
2
- 44x + 4

2.-(X2

+ 4)

11

5
(r)

y =

5x

es*

11

37

47

1-3

25.

For the function


f (xi) = p(x2) = 0,

2
ax + bx + c,

f : x

a,41'

prove that if

0,

then

or$ X2 -

-b t

lac.

I
r
ti

Fd

Q.

zs,

mo

4-)

40

38

.or

1:h.

Pol nomial Functions of Degree Higher Than One


Most

al

f our initial work w4h polynomialrfunctions will be concerned wit})

two related problems:


4

Problem 1.

Given a function

and any number

in its domain,

Given a function

and any number

in its range, find

f(

Problem 2
all numbe s
of

x' such that

x 'for which

f(x) = y;

y = 0,

in particular, find those values'

i.e., the zeros of

f.

of these two problems.

Later we shall study the second

Soon we shall also develop techniques for determining maximum and minimum
points, intervals of increase and decrease and behavior for large values of

for polynomial functions of degreehigher than two.

As in the quadratic ex-

amples of the previous section these techniques will enable us to sketch


graphs of higher degree polynomial' functions very'quickly.
sider Problem 1.

To graph polynomial functions and find the sOlutions of

polynomial equations, it is important to evaluate a given


values of

For example, to graph

x.

For now we con-

we may want the values

fCx)

x -,)3x

for different

'

- 2x

+ x - 6,

x = 0, 1, 2, 3, etc.

at

f(x)

Of course,we may

obtain thtse values by dirrect substitution, doing all of the indicated multi-.
plications and additions.

For most.vplues this i& ;tedious.

Fortunately,

there isan easier way which we, shall_ialitsynthetic substitution.

To under-

,.

stand the method, we shall analyze a few easyi,examples.

.Example I

a.

Find the value of


f(x) = 2x

- x + 3-'at

x = 4.

We write
f(x) = (2x - 1)x +

When x

this becomes

4,

114

(2(4)

3 = 31.

Note that to evaluate our expression, we can"


i

40'

r.

Multiply

(i)

product to
(ii)

x2) by

(the coefficierit-of

(the coefficient of

-1

Multiply the result of (i)

and add this

x);

and add this product to

(the constant term).

Example l-4b.

Find

given

f(3),

f(x) = 2x3 - 3x2 + 2x + 5.


We may write

f(x)

as

(2x2 - 3x + 2)x + 5

or

3)x + 21x + 5:

i (2x
,

To find the value of this expression when


inside parentheses and

we may start with the

x F 3,

(i).

Multiply

.,

(the coefficient of

2
,

to

-3

x3)

by

(the coefficient of ,x2);

(ii,) Multiply the result of (1)-by


.

?
andedd this product:

a>

avid adiA this product to

coefficient of

x);

../

,'

,.

64

4...

(iii) Multiply the result of ( ii ) by

and add thiS produkto

(the constant term),

The result is

f(3) = 38.

-.
,.:,.

.-Ar,

These steps can be represented convenient


consists of the coefflcien
order:

f the successive

able whose fiFst row

owers, Of

ght is the pa

(The number at the far

.n depcending

,yalueii5f, x

being

substituted.)
2

-3

cy

(0?..3)

(3 '3) =

,/

'

(11 .3) =

,/

33

38

11

.4 so

When this tabular'arrangement is used, we proceed from left to right.


start the process by rewriting the first coefficient
Each entry in the second row is
preceding column.

2,

We

in the thiid rows".

times the entry in the third row of the

Each entry in the third row is the sur,s_of the tyg,entries

gbova it.' We note that the result

38,

can be checked by direct substitu-

tion.

40

50

1-4
Now let us consider the general cubic polynomial
.46

lit

f(x) = a3x3 + a2x2 + alx + a

When x

c,

0,

we have
4

f(c) = a c3 + a2c2 + ale +

0'

which may be written


f(c) = [(a c + a )c + 8 le +

'

0.

Again the steps employed in the procedure can be represented in tabulal form:
a2

a3

al

a3c + a2

c + a )c + a

(a

[(a3c + a2)c +

(a3c + a )c

a3c
a3

]e

f(c)

As in,/earlier examples, the number'teing substituted is written to the right


/

of the entire array.

Example 1-4c.

Given T(x) = 3x

- 2x

+ x - 6,

14*,

-6

18

determine

f(2).
1 '2

12

Now 12

substitution:

'

24 - 8 + 2 - 6 = 12.

f(2) =3(2)3 - 2(2)2 + 2 -

Example 1-4d.
that

= 0

This may be checked by direct

f(2).

is the result sought, namely

Given

f(x) =

3x

+ 2x - 5,

determine

f(3).

Note

and that this number must be written in,its appropriate place

as one of the detached coefficients in the firstrow.


1

-3

r5

9-

18

20

1,

60

55'

4,,oW

Thus,

T(3)i= 55,whish..as.before, may be checked by direct substitution,


With a little care and practice, the second line in' the above work cart.,

often be omitted-when

c -is,a small integer.

1-4

f(Z) = x4 - x3 - 16x2 +.4x + 48,

Example 1-4e. Given


for

f(x)

x = -3, -2,,-1, 0, 1; 2, 3, 4, 5.

In order to avoid confusion, it is sometimes

We detach the coefficients.

,,

evaluate

convenient to write them down at the bottom of a sheet of scratcli paper and
slide this down, covering pt` each steget446he work previously done.

As suggested

above, we omit thasecond line in each evaluation and write the value of x
we are using adjacent.to the answer.' The results appear in Table 1-4a.
0
e a table of f(x) and x. Note that the
The .last Two columns now bec
.

row that corm sponds to

as the sup entries as the, coefficient row.

x = 0

Table 1:4a
Coefficients of
1

-1

-4

) = x4 - x3 - 16x2 + 4x + 48

48

I- 4

16

-3

/..10

24

-2

-14

18

30

-1

48

36'

41.6

/
1

'

...3'7:'.

-2

-.---,,,-1

,1

I
-

-16

ON.,

-16

-12

-14.

-24

-10

-26

0 _
' -30

'1

- 4

-12

44

24

168

f(x)

''....

Themethod described and illustrated above is often called synthetic,


4

substitution or synthetic division in algebra books.

The word "synthetic"

14tera3.ly means "put together," so You can see how it is that "synthetic
_$9,1)stitution".,is appropriate here; later we shall illustrate'why the process

is also called "division."


evaluating .f(x),

The meth, gives a quick and efficient means ofTh

and we are now able to plot the graphs of polynomials more

easily than would_ be the case if the values of

f(x)

had,to be computed by

direct substitution.

52

..

/1.

Example 1-4f.- Plot the graph of the polynomial function


f

Sx2 - 12x + 13.

We prepare a table of values of x

and

by synthetic

f(x)

tion, and then plot the points whose coordinates

substiv/

(x,f(x)) .appear in Table

1-4b.
J

Coefficients of

Table 1 -4b'

f(x),---- 2x3 - 3x2 - 12x + 13

-3

-12

13

-9

15

-32

-3

-7

-2

-5

-7

20

-1

-3

-12

13

\-1

-13

-10'

-7

-3

45

it

From the table we observe that the points


(73,-32), (-2,9), (-1,20), etc.

(:c,f(x))

These points are located on a' rectangular

coordinate system Ers.shown in Figure 1-4a.

Note that we have chosen different

scales on the axes for convenience in plotting.

to be plotted are

11 WV

fi

, 1-4
f(x)

(4,45)

( 3,4)

ee

SOme Points_on Graph of


r

x _2 2x3 - 3x2 ; 12x + 13

Figure 148

Now'tht problem is how best to draw the graph.


graph has no breaks.

We shall assume thwt the

The question remains whether the points we haV'e already

plotted are sufficient to give us a fairly accurate picture of the graph, or


whether there may be hidden "peaks" and "valleys" not shown thus far.
.

We are

not in a position to answer this question categorically at present, but we


can shed "further light on it by plotting more points between those alrm4y
located.

By use of fractional values of, x

substitution, Fable 1-4b is

extend

and the method of synthetic

as shown in Table 1-4c

44

::

1-4
Q

Table 1-4c
Coefficients of

f(x) = 2x3 - 3x2 - '12x + 13

-3

-12

13

-8

-7

-6

-3t

-4

35

-10 -;r-

5
2
_

18

-2

4.

-13

-12

r5

5
2

20

t(x)

-7

When we fill in these points on the graph, it appears that if we connect the .
j

points by a smooth curve, we ought to have a reasonably accurate picture of


the graph of

in the interval from

-3

to

4.

This is shown in Figure 1-4b.

r,

e,

Exercises 1-4
1. Evaluate the followineggtYnomials for the given value% of x.

(a) f(x) = x 4 + x - 3;

x = -2, 1, 3
x -= -1, -3, 0, 2, 4

(b) f(x) = x2 - 3x3 + x 2


(c) g(x) = 3x3- - 2x

x=

(d)

r(x) = 6x3 - 5x 2 - 17x + 6;

(e)

s(x) = 6x3 - 29x2 +. 37x - 12;

2
1

3, 4

F(x, = 3x - 97x 3 + 35x 2 + 8x + 2; x =


(g)

G(x) =_x10 - 4x 3 + 10;

=2

2. If f(x) =2x3, - kx2 + 3x -' 2k,


3.

will f(2) = 4?

3x,

for what value of _ k will f(2) = 2?

Flind the value of each of the f011owing functioons when x =

(b)

(i)

x -3 2

(ii)
(iii)
(0)

x -3 2x = 7

01

x -3 2x

(c)

7 5x. 2 7

,..,

x -, 2x3

3x2 - 1.2x + 13

Plot the graph


of the function
-.
x -)

Ifa f X) = x3 + kX2

4.

for What value of

2x3

Rx2

+ 20',

Compare your:sketch of part Op with the graph of


f x 21.3 - 3x2 - 12x 4-13 in F,j.gtire 1-4b

(e) Find the value of each of the following functions when x = 1.


.

x `-3 2

x A 2x + 10
x

(iv)
(b)

2x

x -2x3

-x
3x2

- 12x + 13

Plot the graph of the function


x

ti

(c)

13

2x

+ 3x

- 12x.

Compare yotir sketch


Part (b) with the graph of
f
2x3 - 3 - 123c + 13 in Figure1-4.b.
,

1-4

6.

Plot the graph oft each of the following functions and compare eachgraph
with'the graph of

2x3

3x2 - 12x + 13

in Figure 1-4b.

x - -2x3 + 3x2 + iPx -.13


(b)

7.

2x3

3x2 - 12x

Plot the graph cd4each of the following functions.


(a)

x -)x3

(1>)

x )

AD

+ 4

x -)(x - 2)3
8.

9.

Plot the graph of eadli of the following functions.


x3 +' 3'x2 4: 4

(a)

(b)

x -) x3 + 3x

(c)

x**-4(x + 1)3 + 3(x +

1)2

Sketch the graph of each of the folldwing functions.


(a)

( b)

(c)

2x3 - 12x +

4 2x3

2 12x
3)3 - 12(x - 3)

10. 'Sketch,the graph oft x -)x


.

T2

over the interval

47

11:

if'o i,each

Plot the grap

.20,

of the folloling functions.

(a)

x-) x4 7,2x..- 5x2'+ 6x

(b)

x 74 -x4 + 2x3 +,5x2 - 6x

(c)

x -) x4 + 2x3 -.5x2 - 6x

For

f :: x-+44 + 4X - 13x2

;
-,

'4

12.

we give some of the functional

18x3 i 9x4

.1-

values in the following table.


-1 -2

f(x)

'y

-1

1 7

-304

'

44.

'.."5

'''''

-304

for which the function has

(a) From the table, estimate the vat,ue or- x


a mgXigium-value.'

-4'

7. .

1,

U < ,C<' lr

-4

'

1"--1

'

44

13.

'

44

"4.
4.

e, 'with special care given to the interval

(by Sketch the graph of

Approximate-0,e.maxiMtun value of the tunctipn


-

f' : x-)39 -

144

640x2

10;c.) -

640x4:

48v

53'

477

,
1N.77-

1-5
1-5.

Remainder and Factor Theorems


momentarily we shall turn away from graphing and take another look at the

process W5 rescribed in Section 1-4, in order to develop some theoremrthat4


will be useful in finding the zeros of polynomial functions.
substitUtion used,to determine

given

f(2),

x -)x3

The synthetic

7x2 + 3x - 2

will be the basis for this developtent, so let us examine it closely.


1 r

-7

-2

-kp

-14

-5

-7

-16

We'rewr/te the first row in the synthetic substitution as the givenmelra


nomial

by restoring the powers of

to each;entry in a given column.


1x3

x), and then attach the same power of

Thus we obtain
-7x2

+3x

2x
.

-5x

1)(3

-10x

-14

2
-7x

-16

The polynOmial in _the third row tis the Sum of the two&preceding polynomials.
S1hbe

f(X) = x3 - 7x2 + 3x - 2

and

1(2) ,;=. -S,

the above addition can be

written
.

f(x) + 2x

- 10x 1. 14 = x 3 - 5x
.

- 7x.'+ 1(2)."

By factoring, we May write


VV. y VV.!.
''

Solving for

4p f(x) + 2(x
f(x),

;#

i./e have

<

f(xl= X(x2.- 5x': 7) - 2


or

- 5x - 7) = x(x'" - 5x -,1) + f(2).

2_7 5x - 71-+c(2);

2)(X2 - 5x

f(X)

f(2).

The form of this expression may look fa4iar. 'It is, in fact, an
example' f the division algorithm:

Divldedd =(Divisor)(Quotien
our:example, if

(x

2)

+ Remaindel;.

is, the divisoi, ',then

q(xY=-x2

/i

49

-5 9

1-5,
is the quotient, and

is iWremdinder.

f(2)

This result may be generalized.

t is of sufficient importance to be stated as a theorem.

REMAINDER THEOREM.

If

n > 0

is a polynomial of degree

f(X)

a number, then the remainder-in the-division of

by

f(x)

and if

is

c' is

That is,

f(c).

f(x) = (x - c)q(x) + f(c),

where the quotient

n - 1.

is a polynomial of degree

q(x)

We shall prove the theorem only in the case of the general cubic,

Proof.

polynomial,

f(x) = 83x3 + a x2 '+ a x + a0.


1

Following the pattern of the previous example, to determine


the synthetic substitUtion
a

a3

a0

al

(83c + 82)c

(83c2, + a2c + a1)

(a3c + s2)

(83c3 + 8.2c

As before, writing in the appropriate pd ers of


8 x
3

2,

+41 x

we get
+ a

c + a )x
2

+ (a

4- a

c + a )cx

+ ain + a0)

+ a cx2

x,

r-

+ 8 X

x +
a3 x3

c2 + a2c + a

(a

(a3)

we trite

f(c)

+ a2c + a )c1

c + a )x + (a 63 + a
1

+ a c + a
1

0
_

.0,

We note that the polynomial in tilt third row is the sum of the two preceding
..

polynomials, that the polynomial in the first row is


Ta c3 +la c2 + a c + a )
2
1
0
1

f(x) + C(8 x2 + (a c + a )x'+ (a


x(8 x

+ (a c + a )x + (a

and that

f(c). .Hence we may write

is

f(x)

'

'+ a 2c +

c + a )] + f(c).
1
b

Thus we have
f(x) = (x - c)(8 x2 + (a c + a )x + (a c2 + a2c + s )] + f(c5
3

or

fr

f(x) = (x - c)q(x) + f(c).

50

60

1-5:
The process is the same for higher degree polynomials.
f(x) = (x
where

If the remainder
tient

c)q(x) +, f(c),

is a polynomial of degree

q(x)

f(c)

n - 1.

is zero, then the divisor

are factors of

q(x)

f(x).

It gives

x - c

and the quo- 11

Hence, we have as an immediate consequence

the Factor Theorem:

FACTOR THEOREM.
n

If

then

0,

is a zero of a polynomial function

is a factor of

- c

f(x),

of degree

and.conversely..

We know from the Remainder Theorem (applied to a cubic function)


there exists a polynomial, q(x)

of degree

n - 1

at

such that

f(x) = (x - c)q(x) + f(c).

-171k a

I
24-

zero of

then

f,

f.(x) 4 (x - c)ii(x),

y ,

Hence, x

is a factor of

Conversely, I

".

f(c) = O. and

4ik

is a polynomial,- q(x)

f(x),

by definition.

is a factor of

(x),

then by definition there

such that
f(x) =,(x - c)q(x).

we obtain

= c,

f(c) = (c - c)q(c),= 0,

and hence

is a zero of,

Example 1-5a.

f.

Find the quOtient and ,remainder


.f(x)

is divided by

=42:0 -

6x2 + x

-5

x - 3.-

Hence,

-2

q(x) = 2x2 + 1,
f(3) .= -2,

- -

and

, 2

2x3 - 6k
2x
,

+ x -

5=

(x - 3)(2x2 + 1)

51

2.

1-5
.

'

is a factor of

Example 1-5b.- Show that- x - 6


f(x) = x3

f(6)

q(x) = x

6x

+ x

q(x).

and find the associated

Here,

-6

J.1

-6

and

+ 1,

f(x) = (x - 6)(x2 + 1).


'..

In testing, for the divisibility of a3polynomial by


.

write

a / 0,

we

.
.

.ax + b .= a(x +

and see whether

if and only'if

b.

f(-

-a-)

= afx - (- --)]

--)

ax + b

By the FactoryTheorem,

= 0.

f(- 12-)

f(x)

ax + b,

(Note that

= 0.

b
-

is a factor of

is the root of,

ax + b = 0.)

In applying the Factor Theorem, it may sometimes be easier to compute


f(c)

by direct substitution, rather-than by the Method or synthetic substi-.

tution.

x - 1

Thus,, to show that

is a factor of

faft= )2;73

we note that
,

Evaluating

x37 = 1,

f(1) = 2 - 1 - 1
f(1)

by the synthetic substitution method would take considerably

longer.
er.

At this pointyou imay wonder what to do when confronted with a polynomial


such as

;5,

8x4

dor which you might like to factor.


device.

28x3 -62x2 + 7X + 15,,

Note that the Factor Theorem id only a testing

It does pot locate zeros of polynomial functions.

(Methods, other

than blind" guessing, for doing this will be developed. later.)


-7.

62
I

1-5
Exercises
Find

1.

q(x)

and

f(x) = (x - c)q(x) + f(c)

so that

f(c)
*N.N.

(a)

f(x) = 3x3 + 4x2 - 10x - 15

(b)

f(x) = x3 4 3x2 + 2x + 12 and

(c)

f(x) = -2x4 + 3x3 + 6x - 10 and


= 2x

(d)

- 3x

+ 5x - 2

and

C = 2

c = -3

and

c = 3

=2

Find the quotient and remainder when,

2.

(a)

x3 + 4x2 - 7x - 3

(b)

x3 + 3x2 - 4

(c)

3x3 + 4x2 -.7x + 1

3.A If

fn(x)

is divided by

is divided by g + 2
is divided by

is divided by

remainder ,-r(x)

x - 2

3x - 2

gm(x) # 0' so that a quotient

are obtained', what is the degree of

and a

q(x)

q(x)?

of r(x)?

Give a linear factor of each of the polynomials.

4.

5.

(a)

r(x) = 6x3 - 5x2 - 17x.+ 6

(b)

s(x) = 6x3

29x2 + 37x - 12

Consider the function


Determine

(a)

..(b)

qt.Ctor

ti

f : x -tx3 + 4x

f(-3), f(-2),

f(x)

+ x - 6.

f(0), f(1), f(2),

and

f(3).

over the integers.'

f(x) = 2x3 + x

6.'

and

.7.
8.

1
73.

Factor

- 5x + 2,

k(x)

If f(x) = x3 + 3x2 - 12x: k,


Find a value for

is exactlyedivisible by
Determine

f(1)

if

f(x)

at

x = -2, -1, 0, 1, 2,

find

so that f(3),=

so that
x3 -

9.

determine

over the integers.

+ kx

x - 3.

f(-1) = 0

and

x -)ax5 + ax4 + 13x

53

11x2 - 10x - 2a.

1-5
10.

Divide

(b)

Find

(c)

Determine

if

g(1)

+ 8x - 4

by

x - 1.

x -4.x4 + 2x1- 3x2 - 4x + 4.

a, 0,

x5 + x4

x2

5x3

x5 4..x4

(a)

and

5x3

if

x2 + 8x - 4*

+ r.

(x - 1)3
(d)

Determine. A, B, C, D, E, F

if, for all values of

x,

(x - 1)3(x + 2)2 = Ax5 + Bx4 + Cx3 + Ent2 + Ex + F.


0

Consider the function

11.

-)

x3 - 3x.

_successive synthetic divisions of


tients by

We submit a table to show three

f(x) = x3 - 3x

and, resulting quo-

x - 2.
1

-3

14

2
2

8
9
4

2
1

(a)

Determine g(x)

f(2) if

and

f(x) = (x - 2)g(x) + f(2)(


(b)

Determine

p(x)

and

g(2)

2)p(x) + g(2).

g(x) =
'(c)

Determine

q(x)

and,

if

p(2)

if
p(2) .',

(d)

What is

q(2)?

.(e)7 Show that,,forall

x,

we can-write

f(x) = (x - 2) {(x - 2)[(x

2)q(2) + p(2)] + g(2)) + f(2).

6 454

1-5.

In Chapier2 we shall find it useful (for analysis and translation)


to be able to express a given function in terms of

x - a.

We have

already developed the equipment necessary to do this for a simple


3
polynomial function such as f
x --4x
3x,
with a = 2. Using
:

the.results of,parts (a) through (e) of this problem determine


B, C

and

f(x) = x3

if, for all

x,

..3x = A(x - 2)3 + B(x - 2)'


2)2 + C(x --2) + D.

41114k

(g)

Sketch the graph of the function

f:
(h)

-)x3

3x.

Sketch the graph of the function


.F : x -4x3 +6x2 + 94 + 2.

4F

A,

1-6

1-6.

Locating Zeros of Polynomial Functions


FrQm the discussion in the previous section we khow.that a number

a zero of a polynomial function

if

f(r) = 0.

is

Furthermore we know that

the zeros of first and second degree polynomial _functions can be found by

solving linear and quadratic equations, fat which there are simple formulas.

We know how to find the zeros of polynomial functions of the first and
second degree.
If

If

x -)ax

mx + b,
2

then

/ 0,

+ bx + c,

f(-

then

0,

b.

= 0.

f(

-b t

c)

2a

Upon examining these solutions, mathematicians noticed that the zeros are
expressed in terms of the coefficients by formulas involving only the rational
operations (addition, subtraction, multiplication, division) and the extraction of roots 16f numbers, and believed that it might be possible to express

the zeros of functions of higher degree than the quadratic in the same manner.
In the first half of the sixteenth century such formal expressions for the
zeros of the third and fourth degree polynomial,functions were obtained by
Italian mathematicians.

Unfortunately, these formulas are too complicated to

be of.practical value in mathematical analysis.

Mathematicians usually find

it easier even in theoretical questions to work with the polynomial rather


than with any explicit expression for the zeros.
While these explorations prod, ced some significant, if largely peripheral,
results, they wer

later abandoned to be replaced by better procedures.

rejected the purs

t of formulas to solve equations of higher degree, mathe-

Having

maticians came to believe that perhaps the most fruitful path was to guess at
the solutions.

_ -

.:

Inspection of the x-intercepts of graphs we have sketched in earlier'


..,

sections enables us to approximate zeros of polynomial functions.

But plotting_

Inherent inAe pro-

graphs is,time-consuming, and there are better methods.


,

..

cess of preparing a table for graphing, however, is infoNmation that helps us


to make intelligent guesses about the zeros.
the following theorem.

d/

66

56

This information is contained in

4-6

THE LOCATION.THEOREM. If

i, E7i.polynomilaffunction and if

real numbers such th2t,e(a)

and

f(b)

is at least one zero of f between

are

b.

Geometrically this theorem means thAt'the graph of


(b,f(,b))

and

have oppositesigns, then there


and

from -(a,f(a))

to

intersects the x -axis in at least one point.

In Figure 1-6a we illustrate the


0

Location Theorem wit?! the observation


that

and

f(a)

f(b)

are' of opposite
f(b)

sign so that

must have'at least

one zerebetween- a

and

b.

The Location, theorem depends upon,

the fact that the g aU of n polynomial


ftnction has no "gas ", and hence can.

not have both positive and negative


values without crossing the x-axis
Figure 1.-6a
t

gtween.

makes

A complete proof of this

of a suitable formulation of the fact thqt the real line has no

:zr

"gaps" and 0.11 be discussed further in the appendices.

Since the Location

Theorem seems intuitively plausible we.shall,assume that It is true and con...

centrate on its consequences.

Example 1-6a,

Given that the polynomial function


'f

x - 12x3

8x2

* 14

has three real zeros, loCate each of them betweentwo-Consecutive integers.


..

We use theLoca'tion Theorem to search for values Of fM that are


Ak

opposite in sign.

It is convenient to do this in.a systematic, way by'synrm,_

thetic substitution, setting down the

or

as ip Table 1t6a.

t
me'

a,

'
57

O'

1-6
_-)

Table 1-6a

Loat g the Zeros of


12

-8

-21

14

12

-8

-21

14

8x2 - 21x + 14'1

x --)12x3

''"

Location of41.:7.-__

402ero
1 4

12

-17

-3

a.

40
12

.46

.111

36

12

28

63

203

12

-8

-21

3.4

12

-20

12

-32

15

-72

43

f(x)

WU=
The intervals that contain the real "zeros of

and

1,

are indicated

between

and

2,

and between-s2-,-,and

We hasten to add that it is entirely possible for

between

and

when

f(a)

the

Thus, we see that the real zeros .or...;'f

arrows at the right'in the table.

are located between

and

f(b)

to have zeros

have the same sign.

this possibility in Figure 1-6d.

We illustrate

f(x)

Figure 1-6u

Since the problem of locatirig zeros of a polynomial fUnctionf-s-eSseh

tially a matter of trial, We should ask the very practical questionHow


far sliould we extend the table of ac

tions of the-zeros of

and .f(x)

when we search for the loca-

fe,, In Elyample'1-6b this question arises.


'-

..-41
I.

58

1-6

Example f-6b.

Locate 'the real- zeros of

-.2x--- x

7 2x + 6.

We repeat a procedure similar to EVIIple 1-6a and compile Table 1-6b.-''Table 1,6b

Locating the Zeros of


2

-1

-1

- 2x

-1

2
.45

-2

2
.

2
2

'?Ib'

x --k2x

4.,;

3,

-2

1
^.

2,

-j

-5

8.

-io

.7'

19

-51

440

r(x)

The Locatiod Theorem tells us that there is at(leastoone real zero


between

-1

and

We can then write


f(x) =

where

has degree

2,

say

ex)
Depending upon the sign of

r)9(x)

2
x + bx + c.

- 4ac, .this will have two distinct real zeros,

one repeated Veal ''zero or tw.:,) complex zeros,'. Thus,:there are fir
(1),

there may be one, t.

interval between
(2)

, -or three real zeros, all contained in the

-1 'and

."

-2,

two zeros maybe complex, in which case there is only one,real


zero,
I.

(3)

one-or-two real zeros may be in some other inter;ial of the table

between successive-integral values of


(4)

x,

op

one or two real zeros may be in intervals outside the values of


shown in the table.
.

59

1-6

While it is difficult to rule out the'possibilityof complex ieros or


zeros between other entries Of the table we can at least show that no roots
-2 < x < 2.

can occur outside the interval

.Possibility (4) appears unlikely for the simple reason that when we
all the entries in the corresponding row of Table 1-6b

f(2) = 14,

evaluated

were positive.

They will be still greater for ereater values of, x;

table shows this for

greater than

f(x)

x > 2,

it appears that for

a zero of

and you can check it yourself for

xr- 3,

the

x = 4.

'Thus

must be positive, so that there cannot be

We shall prove this, as well as the fact that

2.

9agrxe cannot be a zero of the given polynomial less than

by application

-2,

. of the following, theorem.

SuppcAel f

UPPER BO,UND THEOREM FOR THE ZEROS OF A POLYNOMIAL FUNCTION.


f(a) > 0

polynomial function with

If

and if all the coefficients of

(x - a)q(x) 4-.f(a)

f(x)

a > 0.

for

poitive, then all the real zeros of

are less than

is a

q(x)

are

Wethen

a
*

3 call

an upper bound for the zeros of

Proof, For
x_T - a,

q(x)?

x*.-. a,

f(a)

and

and all real zeros of

f,

x > a,

For

f(x) = f(a)- > O.


are e-all Positive.
f

f.

x > a

Thus,

must be less than

by hypothesis,

is not a zero of

p.
.

..../

Now you will 'see from Table 1-6b that


.

of the
given polynomial.
....

is an Upper bourfd of the zeros

We really did not need to. evaluate

f(3Y.
v..

What about a lower bound for the zeros?


,

f(x) = 0

is a positive root of

the positive roots of


0

f(-x) '= 0,

f(-x) = 0,

Since any negative root of

its negative will be a lower bound for

...

if we find an upper bound for


t....

'theilegat.t.ve roots of

f(x) = O.

Let us "apply this test to our example.


.

From the given polynomial


Olt
2

f(x) = 2x3 - x

- 2x -I- 6,

we find that
f(-x) = -2x3.- x2 + 2x

6.

Since we are,trying to find the roots -of the equation, f(-x) f 94

it will

ble less confusing to multkay each' member of this equation by


to have a

in order

tive coefficient for the third degreeterm. Irhus,wehave


-f(-

2x3

4-

.-r

10
4 :v

-1

6o

x2 , 2x.- 6.

Using synthetic substitutioni we4obtain the results. shown in Table 1-6c

for positive values df

x.

Table 1-6c
Ewalluatlig

-f(-x) = 2x3 t.x2 - 2x - 6

2'

-2

-6

-2

-6

-5

'5

10

19

51

-f(-x)

44.

This table tells us two things.


occurs between

Occurs between

-1

and

ay

and

-2,

Firdt, a poSitive root of ..-f(-x) = 0

which means that a negative root of

is an upper bound for the roots of .-f(-x) = 0,

lower bound for the roots of


have been searching.
evaluate

-f(-x)

f(x) = 0.

and gence',

Secondly,
-2

isa

This is,the conclusion for which we

in actual practice, however, it is unnecessary to

to find a lower bound for the 'zeros of .f.

Table,1-6b that thesYnthetic(substitution for


signs for the coefficients of, q(x)
Suppose a negative number
If the coefficients of

f(4.:-=0

as previously shown in Table 1-6b.

and

x(= -2

f(-2).

is substituted (synthetically) in

q(Z) concluding with the number

sign, then all of the real zeros of

Notice in

gives, alternating

are greater than

is a lower bound for the zeros.


..
1
In -EAmple 1-6b, we have found that

f(a)
a.

f(x).

alternate In
We say that

a.

is an upper bound and

lower bOund'for the real zeros of the given function.


zeros of

Hence, all the real

are contained in the interval -2< x < 2 and we have found,


-1 ,and

Methods for showing that, ip fact, If


know'must lie in'the interval

has only one real zero (which we

=2 < x < -1)

section.

61

is a

that one zero lies- between

-2

11

are beyond the scope of this

1-6
4,

_Exercises 1-6
1.

Find intervals tetween consecutivevintegers that oontain tne real zeros


of

given that:

f,

f(x)

(b)

f(x) ='3x3 + x2 + x = 3

9 - x - x2 - x

(c). r(x)

2.

3x2 + 3'

(a)

(Hint:

3x 4, 1

(d)

f(x) = 3x3

(e)

fM) =2x3 - 5x2'- x + 5

(r) r(x)

3x2 + 6x - 9

(g) f(x) - x4

6x3 + x2 + 12x - 6
k

Determine the values of

f(2).)

evaluate

f(x) = x

far which

2x

+ 3x - k

has 'at

*least one real zero between


(a) CI

and

and

(b)

=3.

In Example 1-6b we located at least one zero of


.

f : x

between

-2' and

(Table 1-6b).

71

2x + 6

x2

While that example ser ed priMarily

as a vehicle for the development of larger considerations we afford you

the satisfaction of completing it here.


(a)

Evaluate

(b)

Divide

(c)

For what values of

(d)

How many times does the graph of

x +

2x3'- x2 - 2x + 6 Eby
1

2x2

does

+ 4 = 0?

- 4

x -42x3

x2 - 2x + 6

cross' the x-axis?


(

(e)

How many real zeros has the function

Cr)

What are the zeros of

.72
.
.

62

2x

f : x .242x3 - x2 - 2:fe+_6?

- x2

- 2.x

5+ 6?

Locate real zeros of each of the following functions.

(a)

(i)

x -4 x

(ii)

F :

- 3x

Factor

(b)

x3 + 6x2 + 9x + 2

x3 - 3x

over the reals.


48?

(c). What are the'zeros of each of the following?

(Consult_Exercises

174, No. 11.)


(i)

f: x

(ii)

F : x -4 x3 + bx

x3 - 3x
,

+ 9x + 2

Use the Factor Theorem to find a cubic equationwhose roots are

5.

and

3.

1,,

You are familiar with the fact that for the general quadratic, equation
2 '8X + bx + c = 0,

is

the sum of the roots is

and the product of the roots

Similar relationships exist between the roots and the coefficients

of polynomials of higher degree.

The following problems (Nos. 6, 7, and 8)1

are intended to illustrate:these xelationships for third-degree polynomials:

6.

Use the roots of the equAion given in Number 5 for each of,the
parts:

Find the sum of the roots.

(a)

of
(b)

x2

Comparethis result with the coefficient

obtained in Number 5.

Find the sum of all possj;b11 two-factor products of the roots.

That.is, find

(-2)(1) + (-2)(3) + (1)(3).

with the coefficient of

obtained in Number 5.

Find the product of the roots.

(c)

Compare this result'

COmpare this result with the con-

stant term obtainedin Number 5.


7.

If the roots of a 3rd-degree polynomial equation are

-2,

and

3,

find'
(a)

the sum of the roots,

(b)

tlietsum of all possible two-factor productp of the roots,

(c) , the product' of the roots.


t

'4 tdy Using the results of (a), (b), and (c), write a polynomial
..eqtiation of 3rd degree having the given roots.

.(e)

Check your result's by using the Fablor Theorem to obtain the equation.,
63

7:3

-o

V'

1-6'

8.

(a)

Using the Factor Theorem, write in expanded form a 3rd-degree polyr1, r2, and

nomialequation having the roots


(b)

r3.

From the result obtained in part (a), and from the fact that any
polynomial of 3rd degree can be written in the fOrm
.

(x

a3

2 2
1
0
+ -- x, + -- x + --)

find expressions for the coefficients

a
1

-a

,
'

9.

Find the polynomial function


and

10.

r2, and

-1'

a0

--

in terms

of the roots

and

r3.

of,degree three that has zeros

and satisfies the condition

-1, 1,

f(0) = 12.

There is a theorem known as Descartes' Mlle of Signs that states that the
number of positive roots of

f(x) = 0

tions in sign of the coefficients of

cannot exceed the number of varia-

A variation in sign occurs

f(x).

whenever the sign of a coefficient differs from the sign of the next
nonzero coefficient.

Thus_ x

- x

+ 2x + 5

has

variations in sign.
.

Since the roots of

fx) = 0,

f(-x) = 0

are the negatives of the roots of

the number of negative roots of

f(x) =,0

number of variations in sign of the coefficients of


f(x) '= x

4
-

.f(-x) = 7

x3 + 2x +,5

has at4most

+ x 3 . - 2x + 5

Nts

Cannot exceed the


f(-x).

Thus

negative roots, since

variations in sign.

Find the maximum number of pbsitive and negative roots of each of


the following equations.
(a)

x3

x2'- 1167

(b)

x7

x4 + 3 = 0

3x4

x2

24 = 0

2x - 3 = 0

(d)

x5 - l'= 0

(e)

x5 + 1 = 0

(f)

x5 =0

64

'7

4:

1-7

Rational Zeros

1-7.

If

is a polynomial

f(x)

whose coefficients

an, an,1, ...,:a0

rational zeros of

n-1
X
+
+ a x + a0, all of
n-1
1
are integers, then we may find all

anxn + a

by testing only a finite number of possibilities, as

indicated by the following theorem.

If the, polynomial
,

(1)

/ ;(' n

f(x)

+ a

.4 . a-

has integer coefficients

and

divisor of

n-1

\-as
+ a0
+ ... +-lx

f. has a

and' if

expressed'in lowest terms (that

are integers with no com* integer divisor

'greater than

an, an:1, ..., ao,

?ational zero F # 0, q > 0,


is,

n-1

1), then

p4

is a divisor of

and

,is. a

an.

We use the following argument to establish the 'theorem:.


.
..
.
..*
is a zero of f,
If
then f(), = O. By Equation (1)

f(2.)

. a (E)n + a
(E)n-1
n q
n-1 q

+ a (E) + a
1 q

= 0;4

or, when cleared'of fractions,


n

(2)

-+ a

n-

n
n-1-

Solving Equation (2) for

aoq

= .481 p

n-1

q +

+ alpq

qn

we obtain

where

divides
a0.

n-1

= O.

n-11

pn-lg

-n-2

n-1
q

,:nc1
-2
an -1P

a whole number,

8'1(1

r.
j

is an'integer.

Hence

of times.

We wish to show that p divides


4
To do this, we appeal to the Fundamental,TDeorem of Arithmetic,, that the
0

N,

factorization of positive integers is unique; namely, we note that since

and

and

have no common integer divisor greater than


4

a q

n-1

= -p[a pn-.1 + a
= pN,

n-1

65

1,

neither have

1-7
q

Hence, all the factors of

of

a0,

are factors of

and

is a factor

a0.
q

To prbVe that

divides

anPn = -11.a

(3)

Then we reason that since


a p

the number
than

and

n-1

n-1

and

in the form

n-2, 4-a

and

have no common divisor greater

Hence; all the factors of

pn.

is a factor of

+ a pq

22

divides the right-hand side of (3), it divides

Again, since

neither have

of an,

we write Equation

an.,

are factors

an.

The'foregoing result may be easier to remember if we state it in words:


If a fraction in lowest terms is a root of a polynomial equation with integer
o

coefficients, then the numerator of the fraction must divide the constant term
of the polynomial, and the denominator must divide the Coefficient of the
To keep things straight, we can always see how the

highest power of x.

theoremwoits for
b = 0, m / 0.

mx

4, only root is
divides

the numerator

divides

-b

bk1 while the denominator

m.

If the polynomial has, fractional coefficients, the theorem can be applied

after the polynomial has been multiplied by a non-zero integer 'to clear of
f(x) = 0

fractions,' because the roots of


(k / 0)

and the roots of

k['f(x)] = 0

are the same.


%.

0,

,Example 1-7a.

What are,the rational roots(of

1r
It is clear that

.1W

3x3

'

8x2 +,3x +,2 = 0?


If

is not a root.

is a rational root, in m

lowest terms, then


-p

diVides

2,

divides

3.

The possibilities are


,

t 2,

,s

(1411 1, 3,

so that
I.

q-

4.

2_

or, +2
_3.

We test these one'by one and find that the roots of the given equation are
'1, 2,

and

3
AO.

76_

1 -7

(Note that in the statement of Theorem A2 -2a, rye specified

q > 0,

,,..soP

the potsibilities for


1

-1

-,- -and
,' ..-...5

are all positive.

There Zs no point in testing both


0

.
)

...

:e ra ional roots4of
4

- ox-

,f(x)

3x
=

3x

+ 2x = 0.

- 8x- + 3x- +'2x

x(3x3

Now,

8x2 + 3x + 2).

f(x) = 0

if and only if either

P*41ft

x = 0

or

3x

(1+)

ox

+ 3x + 2

0.

By Example 1-7b,'the roots of Equation (4) are

1, 2,

and

Adding the

root: 0,

we see that the roots of

f(x) = 0

are

0, 1, 2, -

. We, can use our Rational Zero Theorem to establish a corollary for
integral ziotros.

44.

,r

If the polynomial
f(x) = x

+ a'

n-1

x +

has integer coefficientswith the constant term


with the coefficient of the highest power of
then the gnly possible rational zeros of
divide

0
a0

4 0,
0
equal to

are integers that

a0.

We establish this corollary with the follOwing short


.

oof.

Suppose

(in lowest terms),

7F0 = f(0) / 0,

divides- 1.
.

and

/ 0.

Therefore,

q > 6

is a'zero 'Of

By the Rational Zero Theorem,


q . 1,

an4

= p
q

67

le**

f.

Since

divides

a0

and

is an integer that divides

ao.

1-7

Find the rational zeros of

Example 1-7c.

f: x -)x 3 + 2x

- 9x - 18.

By our corollary, the possible rational zeros are integers that divide
- 18,

, +he ?Pros of

namely

- 3, -4,

and

are

3.

After we have found one zero of a polynomial function,

we can use

By this device,

;--a special device to make it easier to find further zeros.

we can cut down the number of possible zeros we have to test, and sometimes
4

we can even use it to help us find-certain irrational zeros.

We know from the Factor Theorem (Section 1-5) that


if and only tf there is a polYnomia.1

such that

f(x) = (x - a)q(x).

(5)

(x - a)q(x)

Since the product


q(x) = 0,

(x

(6)

can find one zero of

f,

consists of

x = a

(x

or

together

q(x) = 0)

f.

is one less than the degree of

Thus, if we

Equations (5) and ,(6) allow us to reducethe problem

of finding the zeros Of 'f


of lower degree.

f'

or

GI:

f(x) = 0)

Moreover, the degree of

x - a = 0

is zero if and only if either

it follows that the set of zeros of

with the set of zeros of

f,

is a zero of 'f

to that of fteding the zeros of a polynomial


q

Naturally we may repeat the process, with

if we are fortunate enough to find a zero of

apply the Fa ctor Theorem ,to

q,

say

b,

in place of
for then we may

and" write

q(x)

(x

b)r(x),

and
(x

q(x) = 0) = (x :x.= b

or

r(Y) = 0)

If we are successful in repeating this reduction until we have a quotient


which is either linear or quadratic, we can easily finish the job by solving
I

a linear or quadratic equation.

-1.

Example 1-7a.

Find all solutions of


2x

(7)

- 3x

tirect'calculation'show.s that
fore,

x - 1.

is a divisor of

2x

is a solution of Equation (7).

0.

- 12x + 13

.- 3x

Al2x + 13.

-3

-12

2.

71

Performing the divisioh,

13

-13

-1

There-

Thus
2x

- 3x

- 12X + 13 = (x - A
12x

and the solutions of Equation (7) are


2x
(1 +

By the quadratic formula,

- x - 131,

and the'solutions of

- x - 13 = O.

(1

- i5)

/1. 55)

f.15

and

areLthe

solutions of Zquation (7).

Example 1-7e.

Find all zeros of


f

x -'12x3

8x,2

- 21x

14.

This is the same function that,we considered earlier in Section 1-6,


Example 1-6a.

between

At that time we found that there are zeros between

and ,2,

and between

-2

and

and

1,

Thus, we know that there are

-1.

three real zeros, but we do not know whether they are rational or irrational,.
/

If all three are-irrational, the best, we can do is to find decimal approxi

tions. ,Butifat least,one zero is rational, then we can Obtain a function


reduced degree -- in this case a quadratic -- that will enable us to find the
exact values of the remaining zeros whether rational or irrational.

If the function has a rational zero, it will be of the form 2


theRational Zero Theorem of this section the possibilities for
t 2, t 7, t 14,

and for

are

l',' 2, "3, 4, 6, 12.

and by

are

! 1,

Thus, there appear to be-,t-

a good many values of 2 to test as possible. zeros of the given,function.


q

But

since we already know something about the location of the zeros, we need test

.only those'possible rational zeros p. between

0 and

and between

;2! and

-1, - until a zero is found.


.1

69

79

1,

between

and

2,

'f

1-7
Now the possible rational zeros between

and

are

b
1

1
2

By synthetic substitution, we find that


1

the zero lies

f(1) = -3,

values

1 1
,

and

and

f(-2 )

1.

and

f(0) = 14

Since

= 3.

Hence, we need riot test the

This is a,good example of how the Location

12

Theorem (Section 1-6p may save us unnecessary wor). -

Continuing, we knob that the only possible rational zero between


2
1

is

or

.2N

and

and we have found

Testing these, we find that fkt = 0,


3

12

the rational zero 7 .-.By the Factor Theorem,


_.

x -

is a divisor of

f(x),

3
2

and the quotient, obtained from the synthetic substitution of

is

q(x) = 12x2 - 21.


The ,zeros of

are the roots of


12x

which are

17 .
-7
and

1/7

- 21 = 0,
2

Thus, the zeros of tile given polynomial are

,
IT

VT
'

'

o.

78

80

Exercises

1-7

Find all rational zeros of the polynomial functions in- Firxercises 1

and find as many irrational zeros as you can.


(a)

1 .

'3x -

2x" -

3x

;1-4 2X3
2.

3.

4.

- 2x

(1)

x3 - 6x2 + llx

(b)

x14

(a)

x3 - 2x

(b)

x -3

x14

(a)

(b)

6x3 + llx 2 - 6x

+ 3x - 4
2

2x3 + 3x -="'4

-.2x3

-- x2

2x

-)2x4 - x3

2x

5. x -)12x3

- 40x2

19x

3x 3 - 10x2

+ 5x + 4

4x 3 - 10x

+ 5x + 6

8. x

x4 -12x

9. x

x4 - 8x2

,-,

+ 8x .4-12

16

- 5x + 5x 2 + 5x -.6

10.

11.

x -3 x5 + 3x4

x4

- 7x 2

+ x

+ 2,1

'7.

-6

5x

15x

+ /pc + 12

2
3
x -).3x 4 - es'
o x - 28x + 64x - 15

13.

Show algethaic'ally that the equation


if

is

a real

number such that

x +

1
-- =

has no real solution

In j < 2 .

Ni.

You are familiar with the fact that for the gdneral quadratic equation,

ax2 + bx + c = 0,
is
s

the sum of the roots is

and

the ,product of the roots

- .
a

Similar relationships exist between the roots and the coefficients of

polynomials of higher degree.

Thesfollowing problems are intended to illustrate

these relationships for third-degree polynomials.

71

81

. Z;

Approximating Zeros

- 1-8.

Methods for finding rational zeroTof, polynomial functftion:, are discussed


A simple method, known as the method of repeated bisection

in Section 1-7.
N.

canQbe used, along with theLocation rfleorem, '''c) approximate roots (either,

rational or iTational).

<

This Methodis easy to desbrzbe and i3 simple -bb


>"

The arithmetic can Lecome very complicated, however,

frogram on a computer.

v.;

04

(Another, more powerful, method is described

and the method is fairly slow.


in,Section 2-10.)

COnsider fhe polynomial function


At
-0

Since7'f(0):, -1 < 0

and that

x -)x

3>

f(1) =

and, 1,

namely

we know (by the Location Theorem)

0,

that there is at least one zero between


"

- 1.

and

1.

We take the average of

and find
/

r(7).= 7 + , - 1 > Q.
t

1
Th01,thereisazerobetweellOarld,.We
average again to obtain -74

v.;

,:'

Srnce

r....--7

f(;)

1 < 0,

we know that thereis a zero beteen 7 and


1

7r

Averaging these we'get

+ 2

3
-. 7

-2

and since
.1)

(4)'=
".

we have

1 >0,

and

0 ated a zero between

3
.

For convenience we now use

decimal: n ta ion and average again to obtain


0.25 + 0.375
2

Since

- 0.3125

we knoW :that there is a zero between

f(0.33Z) < 0

0.3125
C

41

/
72

00,

1.8

We could continue the process to show (Exercises 1-8, No. 1) tha,the


zero is between
the zero is

0.31

and

Having done.this we could be certain that

0.35.

correct-to one decimal place.

0.34.

Sorpbta_ining zeros of fu9ctions to some prescribed degree of accurac


the method.of repeated tisection is often .sad on high speed computers since
the process is easy to program,
,

Without -a computer we try to speed the process by shrewd guessing.

For

example, we might obser1Athat


X' 4 ,X - 1
.b

x >

iS DO.Fiti')V for

We could thentest

>

- 3x -1 >

0.333,

n.ive

3()- I > 0.

(7):)3

,0.332', -'8.331-,

0.330,

0.329, etc. until ,we

obtain a-negative value; and then average to obta.in


_ further accuracy.
7

Does

this really speedy the process?'

.IP

r
S

'

I.

73r
8:3
b

1-8

Exercises 1-8
3

+ 3x - 1

1.

Show that a zero of

2.

Extend the method of repeated bisection started in this section for the

x -)x

lies between

0.3

and

0.4:

function

to locate a zero of

to show that a zero of


3.

Find correct to the nearest


f

4._ (a)

x =ix

- ix -- 2x

x -)x

3...

2x

+ 3x - 1

between

0.31

0.5,

0.35;

0.32.

the real zero of

that lies between

and

correct to two decimal places, is

f,

Find, correct to the nearest


f

5.

x -gx

apd

0.5, the real zeros of

+ x - J.

(b)

Find the zeros correct to the nearest

(a)

Find a solution of

+ x = 3

0.1.

*.

p.
correct to one decimal place.

Cor Find this solution correct to two decimal places.


6.

Find the real cube root

of.

20

correct to two decimal places by solving


2

the equation

x' = 20.

1-9.

1-9.

Degree of Polynomial and Behavior of Graph


Suppose

is the polynomial function

0
,

x -;a

+ a x + a x2 +
1
2

+ anxn,

What kinds of information about the graph of


: expression?

an

0.

can we get easily from this

FOr example, note that


f(0) = ab

so that the constant tern a

0 is the y-intercept, that

0
(0,;0).

the y-axis at the point

This observation is, of course, quite,timple.


Shall show that the coefficient
graph of

at

(0,a0).

In the next chapter we

is' the slope of the tangent line to the

The othef coefficients of

ficance .as we try to determine the b.ehalhoi: of


rhe degree Or

the graph crosses

f will also be of signinear the point

(p,a0).

can also give us useful information.

Suppose we wish to knoow how many,times a lie4 given by

g(x) = mx + b

can intersect the grgloh of a polynomial function

f :x_)ax n

+a an

-1

a x

n-1

+ a x + a0.
0
1

".

...

Figure 1-9a

...

';;.

7 5

-111;?" '"4 f

6, on

"

1-9

That is, for how many values of

is it possible that

f(x) = g(x)?

other words, we are asking how many roots the equation

f(x) - g(x) = 0

have, or the maximum number.of zeros of the function

x -,anxn + a
too**

n-1

xn-1+

In
can

111

+ a x2 + (a - m)x + (a - b) .
2
1
0

We assume (without proof here):


if

is a polynomial function of degree

> 0,

then-

( 1 )

has at most
Since
zeros.

n 'real zeros.

is of the same degree as

14;

we know that

has at most n real

This means that a line can intersect a polynomial curve no more times

than the degree of the polynomial.


The x-axis is a very special case of a line given by
m

and -c

are both zero.

y = mx+ b,

where

Therefore, as a partibular consequence of (1) we

have:

if

f has degree

n >-0,

then the graph of

the x-axis no more than

cdn cross

times.

the expression for


of

If

also determines the behavior of

f(x)

far from the origin.

for va

For example, consider the function o.


x -4 1

is far fromIero (that is,

dominates the remain;ng terms.


for

- 3x
Ix'

+ 2x2 +

x3.

is large) then the cubic term

x3'

To show this we can rewrite the expression

f(x)
I

-/
as

+ x3 = x3 ( 1

- 3x +

c5,

4.

'XI _increaseS, the abs lute value of eachlpf

and

he terms

2
,

decreases so that

11

when

Ix!

L+1
x

is very large.

6'1 1/416

is close to

,
1-9

By_this kind of reasoning, J\ could show that for any polynomial function
f,

the term of highest degree will dominate all


This means that4the sigrrof

from zero.

term,of highest degree for

Ix'

other terms when

isfar

will agree with the sig

f(x)

large, and hence the graph of

of the

wil

above or below the x-axis according as the value of.this term is positive or
negative.

.,1
.

We combine this information with that previously garnered to sketch


the possible graph of

f :x -41 - 3x 4- 2x
f(0) = 1,

We know that

and that the araph of

particular the x-axis, at moss three times.


,

nates when

Ix'

+ x3.

\ ,
is large, so that forA

mustRte far above the x-axis and for


1
-.--

lies fa-belbw the x-axis.

x
,-

can cross any line, and in

Furthermore, the ter

,domi-.

far to the right the graph of

far to the
left, the graph of
.-

In partibu.1,ar, the graph of

x-axis to the left of the origin (since

f(x) < 0

for

mu ' cross the


far

ft and

Some candidates for the graph of t are sketched in Figure 1-?b.

if(0) > 0).

Further information is needed to show which graph mightbetan accurate picture of

f.

In the next chapter weistIll develOp thethods for determining the

behavior of graphs of polynomial functions (e.g., locating maximum and minimum


points).

For now we can eliminate five of the six possibilities pictured: We

--r

eliminate:(i) and (iv)

because to line shouldobe able to cross the graph


' more than three times;4

(!iii)

because

fl(0)

(ii) and (vi)

because

4k-11

must be positive:
must-be-gzeater tha

, 77
4
(.5

f(0).

//'

1-9

( i)

(7i)
.,.

FIgure 1-7c.
.

40,
Candid8tes for trkg grapti of x :41

-.

3x + 2x2 4: x3 .

78 ,

e.

9-

I
a

Exercises 1-9

1. Plotting as few points as possible try to sketch the graph.of


2

, 4x 3

- 12x

y = 3x. . +

2.

Plotting as few points as possible try to sketch the graph of


f

3.

+ 5.

"If

5x3

x -) x5 + "x4 -

and

g(x) = 0,- then

x2 +. 8x - 4.

are polynomial functions and


f

and

if and only if

f(x),= 0

are identical.polynbmial functions." -Refute

or defend this statemen


4.

Suppose that-there.are o ly a finite number of selected points shown for


a number of polynomial functions.
graph.

One could only gueat the complete

In each cage indicate the minimum degree that a polynomial func-

tion migirthave-ar:d still be satisfied by these points.


(a)

(f)
4!

"(g)

(b)

OM-

(c)
Is

.4
aA

(d).

(
It

.
I

79

-:8 9

1-9

5.

f -is a polynomialfun6tc
on of degree

Suppose that
-g

(a)

x -)f(ax + b),
Is

where

and

b,

.6

e constant,

n
a

and

O.

a polynomial function?Itg. If so what is its degree?

If

not why not ?


.
.

(b4 If

a = 1

If

b = 0

(c)

how ii=The graph of

related to the graph of

f?

how ere the, graphs related?


a.

(d)

Use parts (b) and (c) to indicate'the relationship.between th.

: .

graphs of f
g(x)
.

and

f(a(x -:- 12))


a

for general., a

and

g(X) = flax

and

Consider

b.

b).
4

tI

;1.

90

Chapter 2')

'

THE DERIVATIVE OF A POLYNOMIAIANCTION

Having discussed polynomial.functions in Chapter 1 we now turn to one


aspect of the calculus of polynomial functions.

The two basic ideas of the

elementary calculus are derivative and integral.

We can appreciate these

ideas intuitively and understaAd their usefulness- before we formulate them


precisely.

We begin with the idea of derivative.


, 5
..._.
If we select any point P on the graph of a polynomial. function and,1

draw a line through

with a ruler, it will be possible

direction of the ruler so that very close to

't

choose the

the line s ems to lie^altvpg

...

the graph.

When this is done,


.-_-

if we stay close enough to

P,

'.it will be impossible.to distinguish between the line and the

curve.

We may appropriately refer

to.the straight line which has this


propeKt.y as the best linear approxi-

P. The

mation of the graph It

straight line is also said to touch


or....12e-tangent to-the graph at

P.

In this chapter, we shall be concerte'dir

with the precise determination of the direction Of the tangent line at any
point of a polynomial graph.
te that our use of, the:sword "tangent" here is consistent with its

meanie

in the elementary geometry of circle

but it, is al'S7o more inclusive.

Graph of polynomial functions mdy lie en:tir

on one side or the other of

one of their tangents, as i circles do, but th y

ay also cross their tangents:

es

r
.

The derivative will help us to determine the direction of such tangents, and
,'

also the shape of the curve.


--

...

--7--- \

81

10

V t'
A

.4tfr'

.*

2-1

Once we know how to determine the tangent and the shape of the graph we
shall be in a position to find any points on a polynomial graph at which the
tangent line iS horizontal ,and the graph nearby.isentirely above or entirely
below the tangent.

Such points are called minimUm and maximum points, res-.

pectivelY.

max. point

min., point
P

The problem of finding the tangent to a polynomial graph at a point

and the shape of the.graph nearby is particularly simple if the point is on


the y-axis.

As we shall see, in this case the result can be written by

inspection.

At first we shall, therefore, confine ourselves to, this easy

special case, and later turn to the case in which the point is not on the

From these considerations we shall obtain a general' formula for the slope

of the,tangent to the graph of a polynomial function


(x,f(x)).

from

f,

at any point

The general result will be expressed as anew function, de'Aved


sometimes thoughtf of as.the slope function of

function wjiich we call the derivative of the function

f.

f.

It is` this slope

In the final sec-

tions of,this chapter we shall apply.these ideas as weCamine the behavior of


polynomial functions and 1; later chapters we shall see that the same basic
concepts can be used to discuss functions other'than polynomial functions.

12 -1.

The Tangent at the y-Intercept of a Gr4ph


I

. In this section we shall illustrate, the method of obtaining an equation


.

of the tangent to a polynomial graph at its point of intersectio


y-axis.

As indicated in

with the

he introductidh, the tangent we are seeking is

defined here to be the straight line most closely approximating the durve at

a. given point.

f
For a polynomial, t

method coral-es merely of omitting every tet1TWhose

degree is higher-than,on

*'

.02

'82
4

2-1

Example 2-1a

of

The tangent

y-axis 17 .P7(0,1),.

to

it

("4;

x - lx

4-

h:1,4

intt,r;;.-ct, the

tne ,:quation

y--14x
obtained ty -omitting the se' ond degree :ern::

4r

is easy to draw

from its equation.

Figure
-

of

is -Che gra

-41

4-

x -

y's-= 1 + x

is the.'graph of

4r
Moreover, since the omitted term

lies below T

except,

exc

is negative for- all values of

-1.x-

at

.)

P .

'

v.

I.--; ,.
"-.'itreIlse

2 .*.

Ellample 2-1

The graph` G

If we

p(O,2).

it the

,,

the tangent

T . through

x2

term and write

--) 2 t

'

y = 2

x1

'

'.;:i .;:,,

fie' equa4i.on of

wecip.:!,,,1"1,...

P. ' Inthis, case the tangent is, p'4

st1.4": y7,axis+cat :,

f..
to' the x-axis..
-

1
;,,,,

Since
7i

is positive for all

q_ .,
..

Because

except zero, all pints *,-,

T.

is the lowet point on

(See Figure 2 -lb.) .)

.,s.
-I

0.... -.,:f
4.

G,

'4

it is called the

..

9,

milimum. point.;

-,

.83

1:%:,,

,4\4:

,,,,74'

,.

of the .graph.

1:%ex-cept

,..

lie above the tangent_ line


1

x.

.0

-,

ti

Figure 2-1b.
G

The graph of

Ample2-1c.

7-4

is the graph of f : -42 + x2 ,


is the graph of y = 2

x3iintersects the y-axis

at Q,0). T,heequation.
y=x
of the tangent at

.is obtained 'Cy


omitting the x3 term. Since x3 is
positive for positive x and negative
for negative 5i,
is above T if
P

'x > 0 and below T if x < 0. (See


Figur. 2-1c.) he graph G therefore
rouses from one 'side of the tangent
,o the other: P is called a point of
_inflection of the graph G.

Figure 2 -le
T

1% the graph of f : x -4 x
is the graph erf y =, x

+ x5

pictures for Exainples 2 -la., by and ce seem to indicate that the procedure of omitting, -every tbrm whose degree is higher than one does indeed
,
prodre the iequatibn of ple tangent o a polynomial graph at Its y-intercept.
T explaip 4hy,
retuzn to Exe,mple 2-1a. We obtained the equation
The

y/= 1 + 5: of the tangtht to the graph of


c

I.
.1*

2-1

1)

x 7-4l+x-4x

at 4(0,1)

by omittif.g the term

-4x

We wish to justify this procedure by

showing that the.'1,,lhe obtained does represent the best linear approximation

to the graph.at the,poiit

This will entitle us to call

P.

equation of the tangent to,the,graph at

y = 1 + x

the

P. 7

From (1) we have

t
f(x) = 1 + x - 4x2,

which may be written a's


p

f(x) = 1 + (1 -16)x.

(2)

If

to

1.

making

is numerically small, the expression


In fact, we can make
x

1 - 4x

lie between
and

in parentheses is close

li-e as close to

as we please by

numerically small.

'Specifically, if we wish

-.01

1 - 4x

.99
.01,

and, 1.01,

1 -4x to be within

it will be sufficient to make

and therefore to make

lie between

f and hence to

of

.01

4x

-.0035

lie'beiqween

and

.0025.

'ThiAesult has a simple geometrical interpretation (see Figure 2-fa).


Let us consider,three lines

L, L1,

and

L2

through .P(0,1)

With slopes

1,41 + .01

and

1 -.01.

These lines have the equations


L

y = 1 + x

L .,y =

1/1101x.

L2; Y

1 4...99x

1'

L,

L
L2
L2

n .0025

.0025
A

iyare 2-1d
t

85

I
2=1'

.4,Pleir slopes are so nearly 'egual that the differences ca

Figure 2-ld only by distorting the scale.

e shown on

Let, AB' be the interval

-.6025 < x < .0025. .dn this interval AB, the graph of

-1.4s

between

and

L2

1': x -3 1 + (1 -

The numbers chosen were merely ilrustrative;

,0

They were
. degigned to give

a certain concreteness to the picture. We can make 1 .


1 + -c
and I - c for 'an arbitrarily small value of c,

between

-r
e

and

c
-,-;

We did not need to chodse

4x

..

Geometni-cally this means that if we keep values o

of

lies betwee

X -'1 + (1 - 4x)x:,

f):

merely by. choosing

clo e enoughto

zero the graph

lie between

c = .01.

"

'''.

4x)x

and, hgnpe,,in the 11itchedregion.

two lines.

y = 1 + (1 + c)x
L

y = 1 + (1

c)x

which differ in direction as little as we please..


which is always included between gucp. lines Li

L
Hence, tie see that

to f:x-4 1 +x- 4x

..1
1

at x=.0.'"

We can confine the-graph

and L2,is

y = 1 + x.

can indeed be regarded

The 0111p-straight line

of

gt

as

the best linear approximation

1 + x - 4x

of the hatched region in filsre 2-1d isy noting thet.

at the point

Hence, on the interval AB;

P.

to the right of

and between

and

to a smaller part

,d lies below. %except

lies between

to the left of SP.

L- and

L2

(See Figure

-9-

2-1

Exercises 2-1
1. For each of the following
.

(i)

(ii)

write the equation or the tangent tothe graph of the


function at the poiktk of intersection of the graph of
the function with the y-axis;
the ) angent line and sketch the shape of the graph

thaw

hear its y-intercept.


(a)

x>1 -

(b)

4 - x2

3x - 2x2

(c)

(f)

(g)

- x3 -.

(h)

1 + 2-x + x

(1)

x5

1-

x.

x3

"Po

(c1)

)1+x+

(e)

2.

(a)

3 + 2x +

x4

(j). x

x3

x4

For If x
+ x + x2 'show -that if -.01 < x < .01, then
.
1. -4- .99xe.:4(x) < 1 + 1.01x;
:

(b)' Strengthen ths, result of part (a) by showing that

(i). 1 + x <f(4) <1--+ 1.01x, for


(ii) 1 j- x < f(x) < 1 + 99x, for

x >.Q,
x

'

<f0.

Show the impl-oved results on a diagram.


(c)
,

Show that the results of pant (a) can be obtainers more simply by
noticinq,that except at=the 3t- intercept, the graph
f x
+ x + x2 must lig above the graph of y = 1 + x.
:

3.

In Example 2-lp -we could write f


(a)

as

Sa

(1 + x 2 ,)x. .

Show that

(i)

x < f(x) < 1.01x,

, (ii) Lax < f(x) < x,


(b)

+ x3

for 0 < x

.1

_.for 0 > x > -.1

Draw a figure to illustr;tte the geometrical meaning of the results


Ort

h,-( a) arid (b)..

87
A
_

th.,2 function

4%,*

'

(a)

At what point does the

)f the functt3n ,ro"."o the

axis?

f?-50'

show tat if

Ixr..< .01,

-x ) 2.99

3.01

and that

:etween

lie

fex)

3.01x

and

(c)

P-aw a figure to illustrate the' gsom,trical meaning.

noticing that the graph -of the

jtr-Agthen ;he re alt of -:,mte'r 4

function lies :Pio', the graph .?f the ,traig;nt.lin-

made in tne figure assopiat'el

c.,hat.dditional refinement

ith

N.mcer 4?.

Consider the function

f': x

x - 1 -

+ (-2

x < .01,, then the graph of

snow that if '0

lies between the

ltnes who,e equations'are


= -1 -=2.01x
and

= -1 - 1.99x.
/

(t)

/-

to show'the geometrical interpretation of this result.

D.aw a figuf

'''t

7,

Consider the function ',f.: x -)3 - 5x - 4x

(a)

F.:-:r

lies near the point

If it is desired that, near

the hra4ht lihes


value's may

'-'

determine the s/oltes of.the lines between_

-.02 < x < .02

which the graph'of


.(b)

-,

(0,3).

(0,3) ttie graph of

y = 3 - 4:998x

apd

lies between '

y = 3 - 5.002x, *what .I.

assume?

'I

8.

If.

is a polynomial function of degree 'higher than '2,

quadratic (or parabolic) approximation to

found by omitting every term whose degree

they bes

near its Y-intercept

is higher than two.

Thus, the

qua-di:atic approximaticrn to'

is

x, )1 - x + x2

-2x3,
-

x + x2.

(aT

On' the sam.. set Of axes draw the graphs

of

and its 'best linear

and bestlquadratic approximations near the y-intercept of


(b)

If

compute

**"

f.

f(x) - g(X).

-r

(c)

If

(d) 'As

g =,0.01, eompute

f(x) - g(x).

f(x) - g(x)

x approaches zero, what value,. if any, does

x2

approaoh?

9.

`-.

is found by o

roximation to a polynomial functi6n att,its y- intercept

The best cubic a

(a)

Determine the best, linear, quadratic, and cubic apPAqoximationeto


f

(1;)

t ng ,e,i)ery term whose degree is higher than three.

Graph

x
f

near

4;

2x

- x

near its y-intercept

e(0,2).

13(0,2), -making-A of the' informatibri you can

glean from its best linear, quadratic, and 'cubic approximations


near there.
(C.), Let

i(x)

graph of

value,

be the value of the best cubic approximation to the


near -its y-intereept.
f(x) - g(x)

if any, does,

As

approach?

x3

.89

99,

approaches zero, what

-3

Y.

2 -2

2-2.' The Behavior of the -Graph Near an Arbitrary POint

In Section 2-1 we restricted our attention to the behavior of the graph of

Noll We shall generkiza our-dis-

a polynomial funct,io;nnear its ynintercept.


.

mission to include,the behavidr If the

rgraph

of such a function near any point.

In Section 2-1 the behavior near the,point for which

the point for which ,x =,a,


'for

f(x)

say, can be determined if we have an ex&essidn


4

x - a.

in ascending powers.of

Totiegin we consider 4 specific function,

- 10x + 4x

Writing

f(x)

f.: x

at a

- 10X + 4x2

'

a = 1.

articular point, where

Example 2-2a.

The behavior near

in ascending powers of x.

a:10

from,the expression for

was determined

x = 0

Determine the behavior of the graph of the function


a

P(1,3).

near the point


powers of

le find

(x - 1)

A
4,

f(x) = 3 - 2(x1) + 4(x.- 1)2.i

(1)

.
,

shall see how to derive such'an expansion for

(Soon

f(x),. but for the

,g moment merely check that


*

3 - 2(x - 1) + 4(x =;1)2 = 3

+ 4x2 - 8x + 4 = 9,- 10x + 4x2 . f(x)

2x +

41

as desfred.)
A

translating the graph of the,function


right. .(See Section 1-3.)
X = 1

f' may be inteqleted as the result of

In this form (1) the firaph of

- 2x + 4x

one unit to

Hence, the behavior of the graph of

is identically the same as the behaVior of the graph of

near
near

x = 0.

0 .1
Figure 2-2a

ti

2-2
.

Since the'tangegt to
to

x2 >

Since

y = 3 2 2(x -

k3 1 ?x + 4x

except at the point_of tangency

(x - 1)2 > 0

- 1) + 4(x-- 1)2

the tangent

\:

the graph or

y = 3 - 2x,

In the same.manner, .since

x = Os .given by a = 3 - 2x,

fo1Call x/ 0,

above its tangent line

at

f at .x =1 is given by

x-i

for all

lies

Q(0,3).7

',theegraph of

lies above its tangenty = 3 - 2(x %- 1)

except at the point of tangenoY


o

The foregoing discussion a sumes that since or = 3.7 2x is:the equation

of the tangent to gat x


y = 3 - 2(x - 1),

then the equation of the translated line,

will reire ent the tangent to

at

x ='1.

lating the graph of

we can verify that.the line given by

is the tangent to the graph of

At ,x = 1,

'Without trans-

y = 3 + 2(x - i)

in filch the same way that wet/


0

,V'ried out the argument in Section 2-1 for tangents at the y-,intercept.
Writing (1) in a factored form
f(x) = 3 + [-2 + 4(x - 1)1(x

we note that if'x- is near enough to


small, the expifssion
words, for any

e,

however small,

that is ff

1,

(-2 + 4(x - 1))

1),
Ix

sufficiently

is arbitrarily close t.

-2.

In other

lies between

f(x)

3 + (-2 + e) (x - t)
and

3 +

provided that
3 - 2(x'- 1)

Iii(x %. 1)1 < e,

-2 - g)(x

1)

that,is, that

<

Ix

is the best linear approximation to

T'is thetangent to the graph

at the point

`Hence

near

(x)

x = 1 And

It should be noted

P 1,3)..

that we have followed the same procedures as before with

x - 1

in place of

x.

4
Thus, to describe the behavior of the gr ph of a function'
A

f.: x -4 13

+ blx- + b2?c2 +

only express

f(x)
c

bnxn

4 c

(x - a)

(x -

4: c

we need

tion (See...Exerciees 8 and 9, Se.ctio

,x = a

x= c.

is then :y = co + ci(x - a),


The best quadratic approxima-

2-1) is given by

x = a,

+ c (x -

'the equation of the tangent to

y = c

,s'

e po;nt where

in the form

The test linear approximation, to

near

+ c (x - a). + c
-'a)2,
1
2(x

d so Non.

IN

91'4.'

101.

2-2

(x

Now consider the problem of expanding a giVen function in power's of


+ c (x a) + c 2 (x - .
a) for some a. We want f(i)
.9

ri

Note that upon dividing ,f(x)' by (x - a) the q'uotient is

+ cn(x

40/

+,,cn(x -.8)n.-11 and the remainder is c0. Hence, to


- a) +
'find the first coefficient c0 of the desired expansion we divide f(x) by
- a) and record the remainder. Similarly, to find c1 we divide the
0
votient.

[c 1

+ c2(x

[el +
,

by

agaih.

a)

(x

c2(x

...

a) + c3(x - a)2

The remainder will tie

cn(x - lei)n-11
at\if

the (second) quotient k.41.11

[-c2 + c3(x -.a) + + c n(x - a)4 -2 1.

Continuing in this manner, we can find the coefficient


.
.
fx - a) in the, expansion of f(x).
,-

\ Is i

. for eacti power of

\L

For example, to expand f(x) 4 4 - 3x 1- 2x2 in powers of (x - 1)


dividei 'f(x) by (x - 1). By synthetic division we have \

2 -3,

we

4-

2 -1,

7:717

I.

indicating- that

,(2)

f(x) = 3 + (2x - 1](x - 1)%

Now we divid

the quotient (2x - i)

-I

(x - 1)

py

win:

'-1
t.

2
.

.which.taillsus',that (2x - 1) = 1 + 2.(x


k

1)'.

Stkbstituting in (2) we have'


-..:?

' or

.7

f(x) = 3 + (1. + 2(x - 1)](x - i)


-F. 2(x - 1)2.

f(x) = 3 + .1(x -

Note that the coefficients '3, 1, *Id

are precisely the remainders under

repeated division:by (x n 1) .
r(

1,0 2

92

2-2
o

/t

If

(.

f(x) were .15n exprgssion of higher degree, the process would ,be con-

tinued.

We simply divide each Successive quotient by

remaind4r, until

f(x)

'

(x - 1)

and record thq,

iscampletely expressed in- powers' of

(x - .1).
4"74.

Example 2 -2b,
f :

Determine the behavior of the grath of

2 :4- 3x + x2 - x3 'near the point at which -x = 2.


'

11'

-- -We need to expand

f(x)

in powers of

x - 2;

that is, tntfind the

, ..

coefficients in

f(x) ="c

As before, if, f(x)

x.- 2,

is.divided by

(x - 2)
.

ci. + c2(x - 2) + c3(x - 2)

the remainder is

further, division of

the final quotient

2)V+

+ c (x - 2) + c (x -

the quotient it

..

co

.-+ c
a2.-}

(x - 2),

x - 2

by

3.-

and

this quotient is divided,by

and the new quotient is

+ c (x - 2)

the remainder is

2
.

d5 (x - 2)3.

gives the remainder

A
.

c'
, 2

and

We proceed to carryout tIese divisions synthetically.

c3.

Dividing by

x - 2
-1

+1

+3

+2

-2

-2

+2

-1,

We obtain ;the first remainder

co = 4

and the quotient

= x2 - x + 1.

Dividing this quotient by

x - 2
-1

-1

12

+1

2 -6
-1

gives the remainder


this quotient by

the remainder

obtained:

-5

c1 = -5 . and the new quo'tient

x - 2,'

c2 = -5

the expansion of

-3

f(x)

co = 4, cl =

-x - 3.' Finally, diyiding

we ha'

and the quotient


in powers of
.62 = -5;

-1.

x - 2

The buccessive coefficients in


are the successive remainders

the final quotient. c3 =

can write
f(x)

= 4 - 5()c - 2) - 5(x - 2)2

93

03

1(x - 2)3.

Thus we

2,r2

Near the point where

we conclude that:,

x q 2,
s'

,,(1)

(2)

the value of the function is

the equation of the best linear approximatiori.to the graph


of
-5;

(3)

4;

is

y = 4 - 5(x - 2),

thus the direction (slope) is

and

the equation of the bes-d quadratic-(parabolic) approximation


to the graph of,

is

" .2

y = 4 - 5(x - 2) -'5(x - 2)

thus

the graph lies below the tangent on'bcith sides of the point
under'consideration.

4't

104

94.

2 -2

MO.

Exercises 2-2
1.

For each orthe' following express


(a)

(b)

'f(x)

in powers of the given facto'r.

f(x)

(x - 2).

5x:

f(x) = x3 -'7x2 + 3x + 4)
2.

?x - 2)

(4)

(c)

f(x) = 3x3 - 5x- .+ 2x + 1,

(d)

f(x) IL, x3 - 2x2 + x- 1,

(x + 1)
4

1-

2.

(x +.

For each of the following function's write the expansion of


powers of
of
(a)

(b)

f.

(a;f4a)):

x -4 3 + 4x 4- 2x2 + x3,.

x -4 3 + 2x3 + 4x2,-

a = 2.

a ; -3

di

(c)

in

and determine the equation of the tangent to the graph

x - a

at -11e point

f(x)

x -4 4x3

3x2 + 2x + 1,

a =

AIP

-14

t
(d)

,f

x -4

5x4

3x2
.

3.
.

(e)

(f)-

f :

.x -4

4x3. + x2 + 3x,

x -4 2x3 + x2

a = 3

- 16x - 24,

a .-- -2

For eacli 'of the following write the' equation of the tahgent at the"
t
/
nearby.
specifilnd point and sketch the shape of the
4
(a)

x -4 4 + 3x -

(b)

x -4 3C3-

(c)

x -4 3x4.- 4x3

.(d)

7x2 4x3

at

(2,:110)

tt 1

at

(3,-109

at

(1,-1)

at

(2,-1)

_60XF

t -4-20 -

40

- 5t + 9

3x2 - 12x

14, at

(e)

2x3

(f)

s -4

2s3,- 6s2 + 6s - 1

(a)

f(x)' = x3 - 3x

(1;1)
cs

4.

at

q.

(1,1)

fn terms of asbending powers of

h
(b).

Write inequalities to show the relative values of


(x - 2)2,

and

(X - '2)3: near the point where

instance consider' x = 1.9 or


y = f(1) = x3 - 3x

2.

(x 7 2) ,

x = 2.

(For

x = 2.1.)

write the value of _y

when

x = 2.

(c)

If

(d)

Write the equation of the best linear approximation to the graph


f

x -)y = x3 - 3x

near the point wh:te

4.

05

x = 2.

00

2-2

(e)

What is the direction (slope) of the (tangent to thy) graph of


x -)x 3 - 3x

near the point where

x = 2?

(x)

Write the equation of the best quadratic approximation to the graph


of f
x -4y = x 3 - 3x near the point where x = 2.

(g)

What is 'the coefficient of

represents the graph of

Near the point where

(h)

(i)

x = 2

-.3x

is the graph of

Compare the behavio?.of the graph of

+ x

and

-3x

near the

x -4x 3 - 3x,

in terms of -X-- a.

x - a.

x -*x3 - 3x

at its y-intercept.

Beginning withthe simple statement


x'

x -4x3 - 3x

with the behavior, of the graph of

x -42,+ 9x+ 6x

Again consider the fundtlon, f

near the point where

Why?

(b)

flexcl (concave) upward or downward?

point where. x = 2

(a)

in the parabola which best

x --*x

x = a + (x - a),

Write

x3 - 3x

express,

in powers.'Of

Make a table to indicate three successive synthetic divi;iolls of


6

x3

3x

and resulting-quotients by

x - a.

(c) 'Use your table from part (p) to write


(di)

x3

in powers of

3,t

Write inequalities to show thexelattonships between


/(x - a)2,

/and

4x3

(x - a)3

when

in .close to

(x - a

a.

(e)

If

(f)

What is the, linear function that best approximates the graph of


f.

at

- 3x,

find the value of

at

a.

a?

(g)' What is the dillectionlope) of the (tangent to the) graph of


)

(h)

- a.

near the .pint where


Foi what .values of

x .= a?

does a tangent to the 'graph of

ft have

zero elope?

46).

(i)

At what points is the tangent to Ae graph of

(J)

What is the quadratic function that best approximates the graph of '
f

(k)

near the point

hor,izontel.

(a,f(a))?

What is the Coefflicipnt of

in the best Paraboric representation

to the graph or-f near the point .(a,f(a))?

1.0 (i 96

2-2

(i)

For each of the points found in part (i) determine Whether the
(parabolic approximation to the) graph of

f is flexeA (concave)

downward on upward.
(m)

Decide which of the points found in part () is a relative maximum


and which is A relative minimum.

(n)

If the coefficient of the

the graph of

x2

in the parabola which best represents

near some point

is neither positive nor

(a,t(a))

negative, then the, graph is neither flexed, upward for downward at

that point.

(We refer to such a point as a point of ,,inflection.)

At what pant on the greagh of

x -)x 3

3x

does this

-phenomenon occur?
(o)

Use information acquired in other parts of this problem to quickly


sketchlthe'graph of

-;x3 -

I.

3x.

2-3

The 'Slope as Limit of Difference Quotients

2-3.

To find the equation of the tangent line to the graph of a polynomial

function

at the point

we expressed

(a,f(a))

in terms of powers of

1., Thus, we wrote

andalkhen omitted the terms of degree larger than

x - a,

the function

+ bnxn

x + b 2x2 +

as

-f-: x

+ c1(x - a) w`c'2 (x

+ c.riCx

to obtain the equation of the tangent line


As

y = c
4

+ c (x - a)
1

4"...

to the graph of

at the point

(a,f(a)).
)l''

We now.describe an alternative procedure.for


Tinding the slope
.

,.

Wa tangent line.,
.

..

Let

P(a,f(a)),

be a .point on the graph of


N.../

nearby point on the same graph.

and, let

See Figure 2-3EoWhere

P..

'"

of

',W.4.,

.
.'

Figure 2-3a

'103

98

Q(x,f(x))' *be a

is to the right of

2-3
i
:*

\The line that joins

ands, Q

f(X) -.f(a)
X . a
" .

'

has the slope

Consider what happens to this',"difference quotient" if we chobse


graph-closer' ynd closer to

P.'
,

......

. Figure 2 -3b shows intuitively-that the slope of therseeant


,

,approaching the slope of the tangent

PT.

PQ

is

'.4.

.
,

on the
A '

,Q

kk.

Figure'2-3b
-TO take a speeific,,example, let

and let P

be the pant
A

2
4x

(0,1).:zThen

?2)

=_X.
c

99 109
4

2-3
The slope of

PQ

is

x - 4X2

f(x)' - f(a)

4,

P x-

Since

0,

we aey divide and obtain the result


sloPe(PQ) = f(x)
f(a) - 1 - 4x.
x - a

x > 0

If

of

is greater than

PQ

To take
4'

PQ

the slope of

means to take ,x

',take 4x closer to zero; and


1 - 4x

differ from

x <0 the slope

and if

1,

1;

closer to

is. less than

closer to/zero, and henge to

1 - 4x . closer to

In fact, wecan mike.

1.

le by as small amount as we p/eake by choosing

lx)

imall enough.

In ft; if x.<

where

then

isa positive 5umber no matter how small,

r'..4,

1 - 4x >1 - vp = 1

:.,

e.

,Pthen

Siniilarly, if .x

1 - 4x < 1 - 4(- i)

.Since the slopt of

PQ

+ e.

1 -4x we_conclude

is

1 - e < slope(PQ) < 1 + e


,

provided that
.,'

e
-

<x

'

<

that is, if

e
11c I < V

'.

II

.......

For smaller and' smaller choices of e (> 0) the slope of the secant PQ
.
is thus brought arhitrarily close to 1. We have'learned to describe this by

saying that slope of' PQ approaches

ss x

approaches

In this example

Olt

we shall call the number


approaches

or as

the limit, of slope of PS as

In Figure 2-3b the line 1Yr

0.

slope is the tangent to the graph -at the point

approaches

P,

with this lithiting

P.

Gen4alizgliffrom this example, we introduce at definition.


The slope of the tangent to the graph af
is the limit of
f(x) - f(a)

x
.x

Thoaches

at the point

P(a,f(a))

.745',*-10410"16'

as

a.

100

110-

2-3
I

A ,convenient abbreviation for this phrase is


- f(a)

slim ,f(ix)
x - a
x
a

to be read "the limit, as

approaches,

of

a,1

f(x) - f(a)

divided .by

a."

We illustrate this 'definition- by using. the same function


,
f :x-41+x- 4X

but a different

'

pc:dr

on its graph.

,P(1',-2)

Then

f.

Let

be the ppint

(x,r(x)):

f(x) - (1)
- 1

.slope(P

(1 + x

4x2) - (-2)

x - 1
3 + x - Itx2
1

4x2

- ,x

x-

(x`- 1)(4x
= -(4x
Now

lim
x

-(11x

+ 3) = -7'

3)

[since

x A 1]

which is the elope of the tangent to the graph at

'

We, illustrate the use of our definition_ with two further examples.
I

ii

Example 2-5a.. Fird the elope of the tangent to the. graph

'';

of

2x - x3

at the point P(a,f(a)). The slope of the line through P(a,f(a))


'QCx,f(x)), x

a,

and

is given by the , diiference quotient

f(x)- - f(a)

(2x

is)

x - a

2(x - a) - (x3
ax - a

X= _a

- a3),
.

+ax +e ),

'

a3)

-.(2a -

x
6

101

- 1 1 it

'"

..r

/.

-4

a .

ti

2-3
As

'approaches

the difference quotient approaches

2 -

, 2

+ as + aa2)) = 2 - 3a

2
.

We conclude that
2,- 3a

= slope of tangent to the graph of

at

P(a,f(a)).

..,

'

Example La. Find the slope of the tangent to


f

at the point

1 - 2x + x

- 3x
a

P(a,f(a)).

The desi;ed slope Val be the limit of the difference quotient


f(x) - f(a)
x - a

Using the expression:for

f(x)

as

2
-

3x

,x -a
,74.

a.

Ire'have,

- 2x + x

f(a)

approaches

- a
= 2(x
x - a

- k1 - 2a + a21 x - a

:3E;14:)

14:

+
:

x - a

a )
3(x x - a

2
2
i,= 2 + (x + a) - 3(x3 + ax + a x + a3

As x, approaches

a,

x + a
and
4

73(x3 +

approaches

2a

2"

a.x + a3)

approaches

-3(a3 + aa
-3

+ a2a + a3) = -12a3

so that
i
0

tif(x) - f

x - a

approaches

+ 2a - 12a ,' the desired slope.

No

4. I

102

2-3

"
'f

Exerdises.,2.-3

1. For each o

rte the

the following functiona,, assuie, 'that


1

difference quotient

,t

'f(x) - f(a)
a.

in simplest form.

f:

(a),

-) x
7

(b) - f

(c) , f

(d)d

x -> x

For each of the functions in Number 1 evaluate the limit

2.

of the difference quotient

3.

approaches

..

Find,the slope otthe tangent to, the graph of each of the funCtions in

4.

Number 1 at the point

5.

(0,0).

-Write the equation of the tangent line to, the graph' of each of the P

ftinctiong in Number 1 at the point

'

(a,f(a)).

-a.

I.
.

......

or each of the following functions, assume that

x / a

and write the,

i ference quotient
r( x) - f (x)

1(a)

implest form.

Find the slope- of the. tangent to the graph of each of the 'functions in
Number 1:at the point _'(a,f(a)-)...

r(x).

4s

f : x -) mx + b-

k(b

f :

x =)Ax

A :

x -ritx

,c)

2
+ Bx 4- C
3

+ Bx

'

+ Cx + D

Or each of the functions in,Number 6 evaluate the limit as

7.

6 the difference quotient

8.

r(x).

ind the s1ope of the tangent to the graph

jle point

9. Mud
pdint

of

(it)

each function in

umber 6,

.(a,i(a)). c'

the slope of the tthge-i-rt te,the graph of


(a

apprOaches

x -)20x - 3x

at the

.
V.+

103

A.;
;4,-

2-3
IL;

10

Consider the function


(a)

Evaluate

lim

f :

f(z )
z

z -; x
o

approaches
(b)

Evaluate

x,

litn

x -) 1 -

f (x)
x
-

x3:
i.e., determine thel limit as

.:

of the difference quotient

i(i) !Ii..

f ( X )

z - x

f(x + h) - f(x)

h -) 0

kc)

What is the sl ope of the tangent to the graph


4

oftWat

the point

(x,f(x))?
11.

Consider the function


f

(a)' Find the limit as

x
x

1 + x - !ix

appropches

of the difference quotient

f(x) - f(a)
x - a
.$

(b), What is the slope of the tangent to the graph of

at the point

(f(a))?
(c)' Find the limit as h approatches zero OF

ence quotient
,
.

'f(X" +

(d) 'What is the slope of the tangent to the graph of

at the point

(x,f(x))?

1 ,.1,4b4

- f(x)

2-4

The Derivative

2-4.

Consider the function


ti

L 2

(11

x -,1 + x - 4X

..

nr

the previous section we showed that the tangent line to the 'graph of

(a,f(a))

The slope of the tangent to the graph of

has slope;' 1,- 8a.

(x,f(x)1

the point

isiven by

at

at

(See No. 10(c), Exerciadt 2-3.)

1 - 8x.

ti

The function/
,

x 1-41. - 8x

(2)'

is sometimes called the slope function for

value .1 - 8x

at a point

the graph of

known as the derivative

of

since its

The function (2) is more commonly

(x,f(x)).
f

: x -,1 + x - 4x

gives the slope of the tangent line to

(x,f(x))

at the pint

and usually denoted by

f'.

Thus the deriva-

tive 'of
2

1 + x - 4x

is the function
1 -

f :.x

f,(4.- at a point

The value
graph of

at

is the slope of the tangent to the

(x,f(x))

For brevity, we will often refer to this value

(x,t(x)).

as. simply the slope of the/graph of

ft(x)

at- (x,f(x)).

Our purpose in this section and the next is to Ob'taina formula for the

derivative
function

f'

(that is, the slope function) of an arbitrary polynomial 41

f. ,In the previous section we defined the slope of (the tangent

to) the graph of

at the point

(3)

(a,f(a))

f'(a) =

f(x) - f(a)

lim
x

to be

x - a
a

An alternate form will be more convenient here.

place of

in (3), and.substitute
f'(x) =

x'

lim

(x+) -2x

If we write

in place of

a,

(x + h)

in

(S) becomes

f(x + h) - f(x)
(x

h)

which .simplifies to'


1*

oo

The name is reserved for this very special function, in spite of the
fact that there are many, functions which could be, derived in other ways from.
.
a particular function under consideration.

('

2-1p

f(x

'fqX)

/-4

-f(x)

h)
h,

f(x + hh ). - f(x)

Note that

PQ,' where P
(x + h.

ca d

still be interpreted as the slope of a secant

..

f(x + h)),

and

(tft(x))

has coordinates

has coordinates

(See' Figure '2-4a .)

Figure 2-4a

TZ is

It is clear from Figure 2-4a that the slope of

h) " ffx1

f(x

h
As in the-last section we define the slope' of the tangent to,be the limit of
-

this differerice quotient as Q ,approaches

P,

that is, in terms of the new

s.
expression (4), as
and calling

f'

We are denoting this limit by

h approaches zero.

the derivativi of

f.

piyeAsthe_function_f
find

f'(x),

_x

3x2 -2x+ 1,

and the slope of (the tangent to) the graph of

use (4) to
f

at the point

(2,9).
J

f(x +

h) = 3(x +h)2 - 2(x + h) + 1


2
2
= 3x + 6:th + 3h - 2x - 2b + 1

f(x) = 3x? - 2x + 1

f(x + h) - f(x) = 6xh + 3h2 - 2h


f(x + h) - f(x)

lim
h
0

f(x + h) - f(x)
h

The slope of the tangent at


-1

(2,9)

f' (x)

6x + 3h - 2

6x - 2 =''x)'.'.
'f

Is

f' (2) = 6

2 - 2 = 12 - 2 = 10.

.
2.4
Example 2-4b.

Tofind the equation of the tangehtto the graph of

J
x -)x3

at

(1,1).
- x3

+
ft (x)\

h -*to

(x + h)3 = x3% 3x2h + 3xh2 +h3.


0

Hence
+ ;03. + x3 = 3x2h +' 3xh2 + h 3
,

(x +

x3

3x2- + 3xh + h2,

and

fqx) = 3x2
The required tangent hsis die equation

y = f(1) + f.(1)(X - 1)
= 1 + 3(x - 1)

or
y = 3x - 2.

This solutionhas the advantage that it enablei utto obtain tangents


at other points with little extra work.

the slope

f'(2) = 3 .22 -= 12.

.1

107

Thus at

(2,8),

the tangent has

2-14-

Exercises 2-4
Consider the function

1.

(fi)

1im

Find'

from part (a)...

f'(3)

Firid

- 1.

f(x + h) - f(g)

11, -.0

(b)'

-2

-4x

..

Find

(c)

as

f' (3)

lim
h -4 0

and also

f(3 + hh)

f(3)

using auccessively.

h = -.1; -.01, -.001.

..

Use the definition

.4;(x) =
to find

(b)

f(x) =.3x2 + 14

(c)

f(X).= 2x2 - x +

If

ax

lim
h -4 0

f(x + h) - f(x)
h

fpr each of the following:

f'(x)

f(x).= x2 - x + 1_

a)

where

+ bx + c

a, b,- and.- c

are constants, show 'that

2ax + b.

f': x
14.

h = .1, -..01, ..001,


.

3.

'

Construct a table of values for

(d)

h) - f(j)
h.

-1:

2.

f(3

v.,

Use, the definition

f'

(X) =

to fl.nd' --the

"f(x) = x3 + x

(b)

f(x) = x3; -3x

(c)

2
f(x) = 2x3.+ x

f(x)

'each`ok tkie following:

deiivitive

(a)

h)

li'm

41e

6x +

5.

If

f(x) = ax3 + bx 2 + cx

6.

If

f:

(a)' f'(a)
(b)

f1(0).

(c)

fl

(2) :

2x "-

xr2

d,

show that fqx) = 3ax

,evaluate

+ 2bx + c.

-'

2-4

7.

If

find the slope of the tangent'to the graph of

x -4 I + 2x - x ,

f :

at, each of the following points.

(0

(a,f(a))

(b),

(df(0))

(c)

(1,f(1))

(d)

(e) (710,ef-l0)).
.

If

(a)

x3 - 2x + 1,

find all

such that

f'(x) =

(x) =.22

(b)
tc.

9.

(c)

f'lx) = 0

(d)

ft(x) .F

Determine each of the following.

Oi

,6x

X,N404.`)4

(c)

(TheWrbol/"px,"

- a

in

(x + h)6

lim

:::

(10.)

x6

v.

What is the slope function of

I z.,-(c)

It

is merely another name for the quantity

x - a

(b)

x."

x5 lim z5z -- x
Z )X

(d)

10.

reid "delta )7."

often stands for "change in

lim

(b)

x3

(x + Ax),
DX

lim

.4

`.

What is the derivative of

3.
: x -4 X ?

4
: x -4x ?

What is the slope of (the tangent to) the graph of


at the point

x -4 x5

(x,f(x))?

"

11.

What

is

fr

if

Find
(a)

(b)

(c)" f

-4

ff(x)

(X +

if

x -4 x6?

1)2

x -4x(3x + 1)2
:

Ant

(x2 + 2x)(3x - 1)

(vria*'

J
0,7

The Oinph of

slope of th& graph of


and

C.

2-5

2-5.

Derivatives of General Polynomial Functions'

IneExercives 3 and 5, Section 2-44 you were asked to.show that if


2

4 bx + c

x -%ax

f': x -e2ax + b,
g

x -+ ax

+ bx

14

3ax

g': x

.,then

+ cx + d

+ 2bx + c

Such general,expressions allow us tb write derivatives of specific quadratic


and cubic functions by inspection.

For example, toobtain the derivative of

the function

'

x --t

5x3

7x2 - i1x + 13

".
.

we merely obserye that this,. is the above cubic polynomial .with


c:. -11

aAa' d = 13,

ti(x)

. -7,

so we know ihmediately that

9 3(5x2)

4:?(-7x). t (-11)i'=

15*2

14x - 11. .

These exa7led, and othet'exercises in SectiOri 2-4, suggest the following


general.expressions for derivatives of polynomial functions:
The monomial
sQ

f
(1)

x --t bxn

has the derivative


x -+ nbxn-1

f'

The-polynomial
f
ro

-4 170

b x + b x2
1
2

n-1

xr1-1 + bnxri

(2)

has the derivative


ft:

2b2x + 3b x

+ (n-1)b

n-1

3
;,4(

n-2

m-1

+ nb x
n

2'

ecause derivatives are calculated as 'limits otdifference quotients, tHe


ocess of obtaining the derivative of'a function is called differentiation.
Observe, that (2) states..that the'derivative of

general .nth

degree

polynomial can be obtained by differentiating each term in the sum, according

.44

J2 -5

to (1). We are claiming t 'hatthe derivative of such a sum of terms is simply


;the sum of the derivatives of the terms.
writing the expakions f6r
Formulas (1) and (2) can both be derived
f(x +
- .f(x) using the Binomial Theorem, and taking the lirt as h
by

approaches zero. We shall derive (1) 'this way. Fortunately, however; we can

deduce (2) from (1) wirout a long algebraic argument by first justifying the
claim we made above, namely,that the derivative of a sum of functions Is the
sum- of theii-' derivatives.
To show

(i),

let f'

bxn.

Then according to the Binoinial 'Theorem,

f(x + 11) = b(x +

brxn +flxn-lh +n(n - 1),xn-2h2+n(,n 2T) (,3 -2)

f(x + h) - f(x) = 'tinxte 1 h + n(n - 1)

(3)

+ hn1

and
- Pi

--

f(x +`4h -

1)(n)th -1 +ri(n
.

0-1

Note that every term in (3) except the first term, bnxn-10,, contarns II at
(
zero1 all the terms in (3)
pr
least once as a eactor. Hence, as h approaches
w,

of

.except

bnxn-1

..-

t;-

also approach zero. We conclude that .lim f(x ,;I- h)h - f(x)

h- > 0
,

and therefore, that the derivative"of f

unxn-1
u

is

x -+ bxn

f':.x

n-1

bnx

Now to see that (2) follows' from this rasa t we must fiiSt see why the
general property of derivatives mentioned above holds true . Suppose 's is
l
a function whose v alue s(x) eQuals fcx), 4- gx), wnere f anti' g , Eire'
functions of t x. (For exatnple, if ,f; x > 5x and :13.1 x -4 13x7, , then
s : x 5x + 13x 7.) Re may calculate the derivative.. of .s in general

5...

'directly frozethe definition, as follows:


,

lira

h) - s(x)

s(x

h -40
h
s (x + ,h) = (x + h)+ g(-X

s(x) = f(x) + g(x)

and

so
A

112

122

*40
o

...

P-,
s(x) = 1:(x + h) - f(x)+ g(x

h)c

g(x)

'-Therefore

ax

s(x + h) - z(x)

h) - f(x)

in the

lizni as

g(x + h)' - g(x)

h 7aproaches zero, it can be sh3041.4that (4) becomes

'

carthe-d.:-

The argument could easily be extended to sums, of three, four, or any

numbqr of functions. 'Hence, we'see that we can differentiate the general


polynomial function
5
n

b0 + b x + b x2 +

.f : x

+, briX

2-

term by term since it can be thought of as the sum of


From (1) we conclude that
f'(x)

(n d- 1)

functions.

A
1

+ -a; x

Translation and the .Derivative

3b x2

...

4; nb'xn-l'

Nowlet, us consider what happens to our differentiation formulas if we


.

replace x by

x 7 a.

To be concrete.we considet the function


x -42x

- 8x +

whose derivative is
f'

x -4

4x 4 8.

Let

be the function
2
g : x -42(x - a)- - 8(x - a)

a
that is,

+ 9,

40*

If a = -6,

g(x) = f(x, - a).

translating `the graph of

*t4

graph Of

six units to the left.

.313

.123

g. iS obtained by
(See Figure 4-5a.)

2-5

..
I,

.41
.

Y:

I,
,...,.4:

i
i

...

,,

St

Nlv

I
%

I.

%.-/
07-

""?

Figure 2-5a
-46

g : x

If

= 7,

becomes

X -4 2x

g : x

8x + 9
+

(2x + 6)2 =
2( X

of

obtained by translating the graph


Figure '2-5b.)

6) 4 9

8(-x - 7) + 9

7)2

4
4

and its graph is

seven units to the right.

(See

-4-

;
Figure 2-5b

f:x
g

: X -4

2x?

+9

3.

2(x -

7)2
1114

8(x - 7) +.9.,
,

,4044111
142-5

Now under a-translation a line is carried into a parallel line, and thence, the
.

slope of a tangent to a graph is unaffected by ranslation.


- a), /that is, the graph of 'g
1Pting the_graph of

(x,g(x))
gs"

point
g

f,

(x - a ,Ox - a)).

at

(10,3)

is exactly the same as the

,slope of the graph of f: x -'2x2 - 8x + 9

2-54 since

is the image of

Tu

g, at the point
at tile corresponding

For example, the slope of the graph of

7)2 - 8(x - 7) + 9

is the result of trans:

then the slope-Qt-tha-gi"aph of

is the same as the sinper-of thegiaih 9f

(See section 1-2.)

at

(3,3),

ander a translation

as shown in Figure

7 Units to the

right. -

a"
Figure 2-5c
...._

Since the slope of the,graph of a function is giv'en by its derivative, we have


concluded that
if

(5)

.X ,t4-f(x - a),

This conclusion en
rather easily.

then

gr() = f'(x - a),I

les us to differentiate, 'g

For the function


f

2x

o.
- 8x
o +

we find the derivative'

f': x -44x- 8
and replaci

by x - a

to obtain

4(x - a) - .8.

115

125

in the above example

2-5

Pt
we have

For example; if Aa = -6,

g : x --) 2(x' + 6)2 - 8(x + 6) + 9

and
.

'13': x --> 4(x + 6) - 8 . 4x + 16.

',.

11

.
.

We can also justify conclusion (5) algebraically using difference quo.

tient'g.

.,..

We know that
f'(x) = slim

(6)

Replacing

by

x - a

in (6) we obtain

g(x) = f(x - a)

- f(x -.a)

f(x - a +

lim

f'(x.- a) =

Nov/ if

f(x + h) - f(x)
h

we can rewrte,this as

f'(x- a) =

+ h)

limb
lim
h

g(x)

g'(x).

and the right-hand expression is simply the defiriition of

Hence,

fqx - a) = g'(x).
In general then,

The polyriial function

g:x4c0+c 1(x - a)

02
c 2(x

+ c

has the derivative


g': x
t

Example 2-5a,.

-find the derivative


(2j9)

a)2

+ 2c (x - a) + 3c (x
2
3

Given the, function


f'

+ nc (x -

con=1

x,

3x2 - 2x + 1,

and the slope of the graph of

use} (2) to

at the point

Using (2), we obtain


The slope-of the graph of

f': f -1(2)(3x) - (1)(2) = 6x - 2;


f

at

(2,9)

is

f'(2) = 10.

10.

.Example 2-5b.

5-

Given

. the tangent line to the graph of

Sind
of

is

f(1) = -7,

- 3x

+ x - 6,

find the equation of

at the point where

f.

the tangent passes throper

'

x = 1.

The derivative

(1,-7).

4.

A -45x4- 6x + 1 '

f'

so'that the slope of the tangent at


line'at. (1,-7)

(1,-7)

f1(1) s 0.

is

the tangent

Thes

is Horizontal and has the equation


Y = -7.

Note how74Ch easier it is.to find the equation of this tangent by using
the derivative formula (2) to, obtain its slope at

themethod of expreglang

Exam 1

y =

2-5c,

x + 1

-.4x3

f(x)

in powers of

x - 1

x = 2,

rather than using-,

as we did in Section 2-2.

Finethe equation of.the tangent to the graph of


(2,-45).

at the point

It is common to denote the expression for the derivative by

y',

so that

(2) gives

yt = -12x

's

- 7

the'slope of-the tangent at the point

(2'-45)

we, replace

To find

(x,y).

This is the slope of the taftgent to the graph at any point

by

to

obtain
-12

- 7 = -55.

The equationof.the tangent line is


y = -45 - 55(x

Example 2-5d.

For f

x --)x

2).

- 3x

and

gi(1)
4

'We have
:

X -4 3x2

so that
.

x -4

- 2)2 - 6(x - 2)

12 7

x -if(x

find

and hence

gl(1) = 3(1

2)2

- 6(1 -

J'

Example 2 ..21.

Suppose

t.ngent to the graph of

f : x -t(x + 1)10.

We could use the binomial theorem to expand


differentiate.

(x + 1),1

However, it is easier.to use (7) which gives


,

tfe

10(x + 1)9

so that
1

the equation of the

at the point 4 ,102hid


1024).

P(1) =

10 x 29 = 5120.

Hence, the desired tangent has the equation


y = 1024 + 5120(x - 1).

1161

28

ands then

2-5

Exercises

1 (a)

Find

f'

if

(b)

Find

g'

if

(c)

Compare.

2 (a)

(b)

f'

x2

.4-

2:2

2)0+ 1-, using (2)-.

x (x + 1)2 using

(7).

and

Find the deriVaiive of each of the following functions.


(1)

(ii)

g 1: x

2(X + 6)2

" 8(x 4-,6) + 9

(iii)

g2: x -) 2(x . 7)2

- 8(x. -; 7) + 9

: x -) 2x

- 8x + 9

Find the slope of the tangent to the graph of each of the functiOns
in part (a) at the point: indicated:

(c)

(i)

f 'It

(ii)

g1

at

(-3,g1(-3));

(iii)

g2

at

(10, g2(10)).

(3,f(3));

Show that the tangentssto the graphs of

f, g1, and g2

at the

points indicated in part (b) are parallel lines.


(d)

Indicate the function obtained by shifting the graph of each of the


functions in, part (a) as prescribed:

the graph of

the graph of

g1

the graph of

g2 nine units to the left.

two units to'the left;


four units to the right;

(a)) Find the derivatives of each of the following functions.


(i)

F : 2c-) x3 -43x

(ii)

(iii)
(b)

x -) (x

2)2 +

-p2) + 2

1)3 - 3(x + 1)2 + 2

(0)

(ii)

ft(0)

(iii)

gt (0)

What is the equation of the tangent to 'the graph of' each of the

functions

F, f,

and

at the y-axis?

11129
4

.9(x

Evaluate:
(i)

(c)

+ 6(x

x -4 (x - 2)'

2=5'''

and

F, f,

(d)

Compare the functions

(a)

7nd the derivative of

g.

F.

(0- Determine

if

f'

x )x

+ 6x

+ 12x + 8.

f : x > (x + 2) 3.

(x + 2 + h)3 -.(x + 2)3

lam

Evaluate

(c)

. h )0
,4

(d)

Evaluate

F*(1).

(e)

Evaluate

ft(-1).

(f)

Evaluate

lim

14,

(x +

2)3 -

(-1 + 2)3
x + 1

x,--> -1

Consider the. function

5.

1)10.,

(x +

(a)

Find

(b)

Evaluate *f(0)

(c)

What is' the equatioh of the tangent to the graph of

ft.

tt(0).

and

at the

y-axis?
ft(-1).

and

(d)

Evaluate

(e)

Find the equation of the tangent tothe'graph of

,where

f(-1)

at the point

x = -1.
and '.ft(-2)._

(f)

Evaluate

(g)

Write the equation of the tangent to the graph of

f( -2)

at the point

x = -2.

where

6. %Consider the function

x ;(x - 2)15.'
1

f.

(a)

Find the derivative of

(b)

Evaluate' ft(1), ft(2),'

(c)

Find the equation of'the tangent to the graph of

and

f!,t(3).

(4 ,32768).

at.the point

.D

t
f

: x )3(x + 2)2.

(a)

Find

(b)

Whatis the derivative of

(c)

Compare

(d)

For

(e)

Determine the derivative of

f'

if

f with g,
:

x >3x2 .7+ 12x + 12?

with

and ,f*

x 70(3x + 6)2 find

g'.

F'.

G : x.>9(x + 2)2.
,

P.
.

(f)

Compare F wi

G,

and

F'

with G'.

'11

120

130

8.

Find

f,1-

cif
2

(b)

X -)

x3

"
"nt",

(Note:

from 1

+.57..

x9

x4 + x

-2

x ) 1- +

3* x 4
+
x7

31.)

(c)

-x

+.

read

8
x

is the product of all the integers

"11, faCtorial,"

through the positive integer


= 1 .2 .3

...

n.

.(n - 1) .n.,

11Z.

For example,
9.

(a)

Find two points where the slope of the grAeph of-

k
(b)

10.

5: = 1 .2 .3 .4 .5 = 120.)

x -,2x3

9x2 - 60x + 5

What dOes the graph of

is zero.

look like at these two points?

Consider the functions

-3

x -)x

3x

+ 1

x.

2
6

(a)

Find the. associated slope futIctio(!

(b)

Evaluate

(c)

It each case write an equation of the ]tine tangent to the graph of

and

fl(1)

f'

and

g2.

e(1).

the function at the point where

11.

x7-

x = 1.

(d)

What is the relationshipof these tangent lines to one'ano;theri.

(a).

Using

7
x = (x - a) + a

and the Binomial Theorem express

x'

-.)

in.

bt

x - a.

pok.;ers of

(b)

Using part (a) determine

(c)

Evaluate

lira

(d)

Determine

(10

x 7 - a7
x - a

for

'

x / a.

x7 - a
X - a

7
xfbr

& / 0..

(See No. 9, Exercises 2-4.)

2=5
(e)

As

increases by an amount

(x + 6x)7
'"change in

&c,

the change in

x7.

This quantity is often labeled

y."

Determine

is

representing.

"4,,"

OP.

ti

r
s.

(This limit is, of course, the derivative of


symbolized by

y'

or

y = xer ,

'

and is often

3Y. .1

dx

,)

)
FI

122

182

2-6

Applichtions of the Derivative to Qraphing

2-6.

*The derivative, f'

of a polynomial function-- f

,
very "Afful for

f.- In particular, the sign of

, obtaining information'about the graph of


f' (x)

will enable us to determine exactly the int,ervari- over which the graph

of

is rising

or

falling and to locate precisely the high and low points

44

of 'the zraph.

To be specific, consider the functionf

. .

2x

-12x-i. 2.

3x.

Its derivativeig' given by


2

f': x 7)6x .- 6x - 12.


The value
point

f'(x)

(x,f(x)).

can be interpreted as the. sioPe

of

the graph o.f

In factored forin
f' (x) = 6(x -1- 1)(x - 2),

'4
from which it fpllows that
ft(x).>` 0

for

x < -1,

f'(x) = 0

for

x =

p(x)

ft (xi
and

0 for
= 0

f'(x) > 0

-1 < x < 2,

for:"

x = 2

for

x > 2.

See Figure 2-6a for the zraph of .f :

2x

3x

- 12x + 2,

together with

these facts.

,
123

133

2 -6

i',(x)<o-4.4.1(x)

f'(x) > 0

fl(L1)

p
.

fr

(2 -18)

Figure 2 -6a
0

,,

The graph, of
.

Note that as

3
f : x -42x

- 3x

..

- 12x + 2.

....

increases, the graph of

rises over the intervals 4n

f'(x) >0 and falls over the-Intervals in Which

which

f1(x) < 0.

Tat; is
41.

as we might expect from our experiencesAfith pOsitive and negative slopes or


lines.

We now show, for any function

f,

increases through

a.

is rising as

that if

124

134

f'(a) > 0,

the graph of: f

Figure 2-6b
:By definition,
f'(a)

lim
h -4 0

Hence, if

f'(a) > 0,

f(a + h) - ffa)
h

then
f(a h)* - f(a)-

life

> O.

-40

This limit can not be greater than zero unless


* h) - f(a)
h

(1)

for 6.bfficienty small values of

Suppose we take

enough so that (1) holds. *Then multiplying by

positive and small

ip (1) we have

f(sy- h) - f(fa) >0


or
(2)

f(a +

> f(a),.

This inequality (2) hays simply that the graph of


right of

a.

rises immeditptely to the

(Sea Figure'2-6c.)'

,.

125

135

2-6

1A

a + h

Figure 2-6c
f(a + h) > f(a)

for

h sufficiently small and positive'.

Similarly,. if we take

negative in (1), then o4m1tiplication by

h, (1)

becomes""
f(a + h)

Or

- f(a),< p

f(a + h) < f(a),

which says that the graph of

falls away immediately to the left of


L

(See Figure 2-6d0*

a.

iy

f
co,

f(a)

f(a + h)

a + h q.k a
..-

Figure 2-68

f(a + h) < f(a)

for

-h

sufficiently small'and native.

126

1 .8 6

2-6

I-

We 'have shown the following.'

If

then

fl(a),..> 0,

f(x) :increases at the. point

and the graph of

(a,f(a)),

increases through the value

rises as

x e a.

NI'S A completely analogous argument.may be carried out for the case in which
ft(a) < 0.

We merely state the result.


If

then

ti(a) < 0,

'

f(x)

decreases at the point

(a,f(a)),

and the graph of ,f

increases through the value

falls as

x = a.

A simple but impbYtant corollary of (3) '4nd (4) is the following.


The graph of

(5)

only if

is horizontal at the point

(a,f(a))

f'(a) =10.

For if the graph is horizontal at

(a,f(a)1,

ing noX decreasing at

1'1(0:

x = a,

qc

then

f(x)

is neither- increas-

can neither be greater than zero,

according to (3), nor less than zero, according to (4).

Hence,

f'(a),

must

be equal to .zero.
f : x -)2x 3 ,- 3x

Returning.to the function

- 12x + 2

we note that we

could ha4e predicted its intervals of increase and decrease without having
seen the graph in Figure 2-6;.

the to

of the poZt

Furthermore, we could have pinpointed exactly

(-1,9)

where the graph of this particular function

ceases'to rise and,begins to fall (as


c lled a relati

,poin

ban all the,values of

similarly calle
The point
interval about

is increasing).

maximum of the'function

ow
f

nearby.

The point

a, relative minimum.
(a,f(a))
a

maximum of

is a.relativekInfaximum if and only if


,

(i)
.(ii).
,

(iii)

f'(a) .= 0

i.,(x) > 0

for

x < a

fl(x) < 0

ftir

x > a, and close to

and 'close to

..

12

is

f..4Xi es for x < a, and falls for x > a.


.

(iV(a))1

(2,-18)

'

at that

f, if and only if in some

Ixpother,,w,ords,
e)Ipoint

./

is ellela&A,

the graph 'of

Such a point is

because the value of

,01F14.r;4..

a.

The point
val about

(a,f(a))

is a relative minimum if aid only if ,in some inter-

the graph of

falls for

x < a

and rises for

x > a.

In

other words,
oint
the point

(a,f(a))

is a relative minimum if and only if

(i)

f' (a') = 0

(ii)=

f'(x) < 0

for' x < a

and close, to

(iii)

f'(x) > 0

for

4 > a

and close to

a.
.

A final word about notation:

in discussing the intervals over which a

function is increasing or decreasing, it As convenient to .use the symbol


[a,b]

to represent the closed interval from

points

and

That is,l[a,b]

V.

[a,b]

(a,b),

whglech excludes the endpoints

derivative

suchthat. a < x < b.

of

f'

x -42x

- 3x2

less than zero on tie open'interval


[ -1,2]

and

such.that

a <x < b.

from the open interval, denoted


That is, the interval

b.

(a,b)

Note, for example, that the

- 12x + 2
(=1,2)

including the end-

b,

is the set of all

Often it is necessary to distinguish

is the set of all

to

in Figure 2-6a,is strictly

while in the closed, interval

this is ndt so':

Example 2-6a.

Determine the relative maximum and minimum as well as


intervals of increase and decrease for -f i x 41-+ x - x2 - x 3 .
We have
,

"r,

j,ft

x -4 1 - 2x - 3x

= -(3x - 1) (x +.14)

The graph of

1
x = -5

has a horizontal tangent when

and when

x = -1.

Examination of the signs of these factors,leads to the conclusions:


0
Qr.

Thus,

_044 f' (x) < 0

if

x < -1

f'(x) >'O

if

-1 < x < 1

f'(x) < 0

if

x >

decreases if

x < -1

or if ,x >

and increase's when

between -1

and

1
.

In particular

f(-1) = 0

is a'r'Aative minimum

and

f(4
3

.3-2-

is a relative maximum.

27,

128

1 32,-

is

,
2-6

This information ehables us to give a quick sketch of the graph of


shown.in Figure 2-6e.

f,

Of course, further accuracy'is obtained by plotting

`,more taints, but Such quick sketches are often all we need to have.

falling
27),

J, 31,

a relative maximum'

rising

(-1,0)

-2

a relative
minimum

4,

falling

Figure 2-6e
400

x - x2

X -)1

Example 2-6b. *The zeros of the derivative do not always lead to relative
maxima and minima of a function. 'Consider the function:.

x3.

3x

Its derivative is
f t

whigh has the zero

x = 0

of multiplicity

has a horizontal tAitent at

(0,0)

2.

The graph of

In this ease

s'

ft(x)
f

if

f,

therefore,

but thispointisnot a high point or a


crosses its tangent

Such a point is called ajpoint of inflection:

(0,0).

that is,

2,

In fact, the graph of

'low point on the graph of ,f.


at

x < 0

or if

x > 0;

is increasing on either side of the origin.

1.29

(See Figure 2 -6f.,)

2-6

r.

=x3

increasing-

graph crosses its


tangent (which is
the x-axis) at
this point

increasin

Figure'2-6f
This

example serves to remind us that to determine the relative maxima


f

and Minima, we locate the critical points (that is,-the zeros of

test the sign of f*

ft) and

on each side of a critical point to determine if that

point is a relative maximum, minimum, or point of inflection.

Example 2-6c.

Graph

+ 4x - 13x2

18x3

9x4

first by

plotting points from the table below Figure, 2-6g,andconnecting them with a
smooth curve, then by finding the critical points and intervals of increase
and decrease.

1,

X30140

4.1

(.

(3,-.304)

Figure 2-6g
'Plotting points for
f : x -) 44 + 4x - 13x
.

from the table

'
"f(x)

-2

'-1

-304

0
44

18xr -

9x

44

-304

We have plotted st,few points and connected them with a smooth curve in

Figure 2-6g (using a compressed vertical scale for negative values of

f).

The graph suggests the possibility of a relative maximum point between X0,44)

and

The derivative of,

(1,44).

f'

is

x -a4 - 26x +54x2

We suspect that f'(x) will be zero somewhex'e in the interval


1
a
indeed gives
Testing x =

0 < x < 1.

= 0'

so that

factor of ft(
.

f'

Upon, factoring

2
- 2 -3ox
A

131

x -

+ 36x - 8).,

141

we obtain

2-6
We can factor further to obtain
f

*<.,

-36(x - P(x - i)(x -"f)

Note that there are three critical poiRts between

(0,44)1i and .(1,441.

The product of three factors, will be positive if and only if all three factors
are positive, or exactly one is positive.

fl(x) > 0

Therefore,
1

<

if

f'(x) < 0

if

f'(x) >0 if
f'

< x < 1

(x) <

<x <

x >

if

2
3

We conclude that
f

is increasing for

, t

x <

and for

< x <

while
f

is decreasing for

< x <

and for

x >

2
3

In summary we can say that the graph of


(1
3 '

then falls to the point

i(1))

(1,r(2)),

(i, 4).

and falls beyond


f(3)) = 44

'

and

9i(1) = 44

i is

f orises until it reaches

rises again to

(2

f(2))

In particular we know that


2

t,

f(-3 ) = 44 7 are relative maxima, while


Y

a relative minimum.

We conclude that Figure 2-6g correctly indicates the increase for


x.< 0

and, the decrease for a > J.,

0 <'x < 1.

but is incorrect for the interval

A more accurate representation of

in Figure 2-6h.

on this interval is sketched

132

,1 42 i/

2-6
4.

I.

3
FiguYe 2-6h

f:

x "4 44 + 4x - 13x2

+ 18x3.

- 9x4 for 0 < x < 1.

tr

133/

2-6

Exercises 21,6
1.

Make a careful sketch on the interval

graph of the fuection'f

[0,1]

x -41 + x - x

2,

(i.e., 0 < x < 1).

- x

of the

given in Example 2-6a.

Does the graph confirm the conclusions of the text?


2.

For each of the following functions locate and characterize all extrema
(maxima, and minima).

On what intervals is the function increasing?

decreasing?

3.

(q

(b)

4.

4oc3

Prove that, for


is, let

8x2

14:x4

f : x 14x2

01,

is an increasing function.

(That

xl > x2 > 0 and show that

xk > x22 .)

Employing information gathered by procedures suggested in the text,


sketch the graph of each or the following polynomial functions over the
interval indicated using convenient scales.
(a). f : x -4x3 - 3x + 1,

-2 < x < 2

f.a x -4x3 + 3x + 1,

-1 < x < 1

(b)

(c)

5.

(a)

'f

: x -4x4 - 4x3

8x2 + 6:,

-2 < x'.< 5

Describe the behavior of the graph of


(Maxima?

(b)

Minima?

-12x

Intervals of increase, de

Sketch the graph'of

f : x -42x3

3x.2

- 3x

on

[ -1,27.

eRse

op J-1,

6.

(a)

Describe the behavior of the graph,of- f

(b)

Sketch the graph of

x -4-35E4 f 8x3

on

.O 3
+ ox

on J-1,31.

[ -1,3].

p.

7.

Determine the maximum ialuk of the function

f : x
k=

8.

8
2
x

.
le.

What is'the greatest possible number of points where the tangent to the
graph of a quadratic function

x -4Ax

+ tx + C

may be horizontal?

2-6
9.

Consider the function


f
(a)

-(b)

Find

x -4 Ax

+ Bx

f'.

What is the maximum number of zeros thet

can have?

How many relative extrema (maxima and minima) can

(d)

If the graph of

If

f have?

has a relative maximuth point, must it have a,

Explain or give. examples.


X

(e)

f'(xl) = fqx2) = 0,

determine

x2
1
2

Consider the functions

and
.t(a)
(b)

11.

f'

(c)

relative minimum point?

10.

A / 0.

+ Cx + D,

x -4(x + 1)3(x t 2)

x -4(x + 1)2(4x + 7)'

How are the functions related?


Sketch the graphs of

Suppose that

on-the same set of axes.

are zeros of

and. x

and

x -4Ax

+ Bx + C,

tts

A > 0. r

Show that

has a minimum at
x

x1 ,

x -0

12.

Determine the relative maximum and minimum points of the graph of ,


f

N/43x

- 12x

12x

- 4.

2 -7

2-7.

Optimization Problems

The post office limits-the size of a parcel post'package by requiring


that its length plus, its girth, may not exceed

72

inches.

Is there a package, subject to such a restriction, which has greater volume:


than all other such packages?

If so, what are its dimensions?

In this form,'the Problem is hard to hpdle.

Suppose we simplify it try

asking ifthere is a (rectangular) package with a squai'e'cross-section which


has maximum volume subject to postal regulations.' If we let

x -represent &-,

the width in inches of the square cross-section,,then the girth of the package is hoi: inches and its length is it most 72 - 4x iriches, accor4-ing40
the post office.
4

4'

,e

V.;

14 4
t,

2 -7

Hence, the volUme in cubic inches of the package is at most

V =x

'(72:-.4x).

This f4rmulg. defines a polynomial function


(1)

V : x -472x2 -'4x3,

and we wish to, find its Maximum value.

do not restrict
as

if

x:

Note that

x < 0, V(x)

x.'becOmes negatively infinite.

[0,18]

has no maximum if .we

Hqwever, this causes no difficulty because

we are Interested only in values of 'x

thiinterval

V(x)

is positive and gets arbitrarily large

between

and

because only in

18,

are all the dimensions of the parcel post package sen-

sible positive lengths.


Hence, our idealized model of the probletris:

Finthe maximum value of the function


for

in

V : x -$72x2 - 4x3

[0,18].

The desired maximum may occur at one of the end points of the closed interval
01
1.0,1uJ,

fr

if for example, the sxaph_of

V looked 'like one of the curves shown

in Figure 2-7a.

18

Max. in

x = 0.

Max. in

18

[0,18]

at

x = 18.

Figure 2-7a-

,--

'However, if the maximum value

V(x)

occurs between the enaioints. of the

interval

[0,18],

.,J

i.e., for some

the remarks in Section

x = a

-..6 Show that

in the open interval

V'(a) = 0.

be higher points immediately to the right 6, 'a,

If

V'(a) > 0,

and if

would be higher points4iffiediately to.thp lerLof _a.

(0;18),

then

there

ould

V'(a) < 0, th

2-7.
Thus the maximum volume for our parcel post Package occurs where
x = 0

or where

V'(x) = 0,

or

x = 18.
or -x = i2.\'

V'(x) = 12x(12 - x) =0 then ,x = 0

If

The possible maximum volumes aie therefore


V(0) = 74292 - 4(0) 3 = 0
V(12) = 72(12)2 r 4(12)3 = 3,456

'

V(18) = 72(18)2 - 4(18)3-= 0.


Clearly,

is the largest of these and-by the above ymarks.it must be

V(12)

the relati e maximum of

on the,ifterval

[0,181.' Hence, the dimensim ns

of the most) vOluminous parCel of this sort acceptable tro the post office are

width = ?C inches = 12 inches,

height = x inched = 12 inches,


length = 72 - 4x inches = 24 inches,
and its volume is

3,456

cubic inches'.

The folj.owing examples do not begin to indicate the wide range of

possible applications of the methO used above.

v.

.4

Example 2-7a.

A man proposes'to make an open box by cutting a square

from each corner of a piece of cardboard


ups the sides.

12

inches square and then, turning

Find the dimenSions of each,square he must cut'in order to

obtain a box with maximiim volume.

Let the side of the squarjbe cut out be


box will, be

11"-

12 - 2x

the volume V

inches.

The base of the

inches on each aide and the depth will be

in cubic inches will be

V = (12 - 2x)(12 - 2x)(x)


= 144x - 48x

+ 4x

Figure 2-7b'

138

'148

inches.

2 -7

We suppose that

0 < x < 6,

for otherwise

V will be negative.

Our problem

is to maximize

subject to the condition

x -4144x - 48x

0 < x < 6.

+ 4x

The derivative is
TMe

ft

-tgo the ,zeros of

x -4144 - 96x + 12x2 = 12(6 - x)(2 - x)


are

f'

We know that this maximum must occur at one,

2, 6.

of the points
c = 0,

= 2,

or

c = 6.

We find that

f(0) ..4f(6) = 0

lnrgedt iialue of

and

so that

f(2) '> 6,

on the'Uterval

'f

0 < x < 6.

With a

from each corner, the box will have dimensions

8 x 8 x 2.

f(2) = 8 x 8 x 2 = 128,

128

Example 2-7b.

the maximum volume is

must be the

f(2)
2

inch square_cut
Since

cubic inches.

We wish to plant one square and one circular flower bed,

surrounding them with *15

yards of fencing.

What shodld be. the dimensions

of the two' fences so as to contain flower beds of greatest posbible area?

Let ,s be the side of the square bed and


bed.

the radius of the circular

Denott the suns of the areas of th,p two beds by, A.

(2)

A = s

Then

2
+ mir

and

454 2vr = 15.

( 3 )

'
S

Figure 2-7c
Solving (3) for

and substitUting this into (2) gives, our. area in terms of

the circular radius

r:

A
sit

/12_121EN2.1. gr2
/

= (14q)r2 -

r +
Akt

139,

149

2-7

We can suppose that

< r<
3
_

the endpoints of this interval corresponding to the respective situations of


Thus we seek to maximize the function

no circular bed and no square bed.


6'

2
f

over the interval

+ /4vN 2

Iv

---E---ir

o <.r < 12
2v

The derivative is.

15v

+ 4v)r

'

2.

Solving

r +

gives

f'(

15

r -

+8

Hence, ,f must hive its maximum value on the interval

1.2

0 < f <

amt one

of the points
1

0, c -

We examine the values of

or

+5

2o

for each of these values of

277

f(0)

c:

14.06

e.

f(2g9 8) 7- 7.88.

1)

*.

and
4
f(224t)

sz

We see that the maximum' value of


2
f

...;(v

+ /4v)r2

lin

15

subject to the restrict Ion

r <

2v

is attained at the right endpoint -g

Our conclusion is that the problei has no solution in the terms posed; a
square and a round flower bed together will never encompass as great an area

as a single round bed Whose perimeter equals the total length available,
,

S
1401

150

t;

2-7
44,

Similar problems requiring the minimization of a functiog over an inter,

val often occur, and can be solved similarly.

If

is a polynomial function

defined on some interval


x =' a, x = b,

the minimum value

',[a,b1,

or where, f,' (x) = 0.

To ,justify this claim gyre need only ob-

serve that if the minimum occurs at'f x = c, a < c < b,

there are lo wer points immediately tp theleft of

then if
and if

c,

there are lower points immediately to the right of

Example 2-7c.

occurs either'at

f(x)

f'(c) > 0
f'(c) < 0

c.

What shouldbeithe dimensions Of theqrlower beds in

Example 2-7b so that the leastpossible area is encompassed?


f

Following the analysis of Example 2-7b, we wish to minimize the function


,.

v2

+ '4% N r2

(-4----ir

over the interval


where

f'(r) = 0.

221.

r +

16-

:'.--

[0 4].

f(r)

is minimum either at

r = 0,r ,.--.

12

or

an

In Example 2-7b we fOtand that

if and only if

15
+ 8

Hence, the possible minimum values for


.

f'(r) = 0

are

f(0) v. 14.06

15
f (2v + 8 )1.- 7.88
f(12i)

We-see that the minimum value or

= 17.90
.

occurs when

15

Hence, the

combination square and circular garden surrounded by


has the least area when the circle hits radius

the square has side

s =

15 yards of fencing
5
r =
= 1.05 yards and
2v + 8.

.(15 - 2vr) 74 2.10

yards.

I
Example 2-7d.
the point

Find the point on the graph of y=irk 2

that is nearest

A(3,0).

Recall that the distance between

(xl,y1)

2,

0,2

- x2)

(x2a2)

and

(Y1

Y2)

is

*a.

2-7

A(3,0t

The distance from

to a point

P(x,x, )

on the.graph of

y = x

is

thus given by

AP = J(x - 3)2
Our prdblem is to choose

(x2)2.

is least.

so that this distance AP

(x x

minimum dibtance
.

A(3,0)

Figure 2-7d

This expression for the distance AP

is not a polynomial so our tjchni-

Note, however,, that

ques cannot be directly applied.

(AP)2 = (X - 3)24 (x2)2

=9 - 6x
'Which is a polynomial expression.

x4

Furthermore, if

least, then AP will also be least.


f

4,x2

is such that

Thus wq,,aeed only choose

9 - 6x

x2

(AP)2

is

x so that

x4

has its least,value.

The derivative of f

is
C

ft

The factor

4x

t4 -6 +

4.. 4x + 6

2x7-13

= (x - 1)(4x2 +14x + 6).

is always positive, since we can complete the

,square to 'obtain

4x2 + 4x+ 6 = 4x2,4- 4x 421 + 6 4

.tr

Nx +

152';

+ 5 > 5.

2-7

(x) < 0 if x < 1 and f ' ( x) > 0 if x > 1. Therefore, by the


(x)
remarks in Section 2-6 weJAnow that, f has a relative minimum at x = 1.
Hence, the point (1,1) is the point bf the'graph y = x.
which is sest
to A(3,0).
Thus,

Exercises LI

ti

1. Consider the funct).on

+ 4x3 - 12x2 :1- 5. ,

Determine the behavior and sketch the graph of f.


2. Find the pxtrema (maxima and pinima) of the funcL tion
if :07x 44x5 - 5x4 - 40x3 + 100

14e

'

on the 'interval -3 < x < 4. Sketh the graph of f .


.

3.:. A ball is throwen upward so that its height


above the earth, where

s =96t

seconds

later is

S'

feet

16t2.

What is the maximum height the ball will reach?


t
,
4. Shdw that of 'all the, rectangles having a given
'perimeter _p,
, .
square alas the largest area.
,
V.
5. Sketch the graph of the function

f
over the inte'r`val
(0,181,
.
,

L.

x 4--Ix

-4-n72x
.

indicatingy/extrema.

t.

...

6: A Tectangutlarbox with square base and open top is to be made from a..0 ft. square piece' of cardboard. What is the maximum volume of such
.
'a box?
,

7.
,

tie

"

field is. to be , adjacent to a river and is to have fencing


..
on three sides, the 'V.de'on the river reqUiring nO' fencing. If 100
yards of fencing is-available, find thelimerisions of the field with
)
largest area'.
,
4
A

rectangular

8. The sum of two positive. numbers is


.

N. -1.......termine the'
. numbers so_ that

thert, product of one and the square of the other will be a

..

.7

maximum.

9. A wire24 inches long- is cut in two; and then one part is bent into .thex shape of a circle acrd the other, into. the,shape of ,s3 square. aw
should it be cut if the sum of the area is to be a minimum?
1143

...

-.

'153

2-7

Given the requirements of Number 9, determine how the wire should be cut

10.

if the

surN

the areas is to be a maximum.

A fpur-ft. wire is to be cut into two pieces:

one piece to become the

*Perimeter of a square, the other the circlamference of a circle: Determine


hot it should4be cut to enclose.
minimum area;

(a)

ti

114

.maximum area:

(b)

feet which

12. 'Determine the dimensions of the rectangle with perimeter' 72


will enclose-the maximum area.

eight. of the right circular cylinder of greatest

Determinetheradius and

13.

ight circular cone with radius

'volume that can be inscrib


height

and

h.

A man has

14.

yds. of fencing which he is going to use

600

enclose a

reanguler fieldand then subdivide the field into twodots with a

3,1

What are the dimensions of such a fieldiif

fence parallel to one side.

theencloseddarea is to be a maximum?

An open box is-to be made by cutting ouysquares fromthe corners of a

rectangular piece of cardboard anetheeturning up the s

pipe of cardboard is

12"

what are the dim hsions of the bpx

24",

by

If the

of largastvolume madeein this way?


,

16.

A rectangle has two of its vertices on the x-axis and the other two above,

ithe.axiston the graph of the parabola,

x2.

What are the

sions of such a rectangle if its area is to be a max.imum?

'

./e

17.

A stone wall

1Q0

Pail or all of it is to

yards long'stands on a ranch.


,4

'

be used in forming a rectangular corral,, using ay Viditional

260

yards

Find the maximum area which can

of fencing for the other threepsides


ge so enclosed.

Find the point on the graph,of the equation

18.

to the point

(2,1).

y2 . 4x

Find the dimensions of the right circular cyiinder of maximum volume

19.

'

inscYibdd-in a sphere of-radius


A

.a).

which is nearest

10

inches.

a,.

What number most exceeds its square?

154

.4,44

I
/'
...

21.

Suppose that the babe of the parcel post package mentioned in the text
is taken to be stilare.
1

Ad

Find.the package of this shape which has maximum volume subject to the
postal restriction,that the sum of its length and-girth may not exceed
72 inches.
,

22.

A rectangle has two of its vertices on the x-axis and the other two
above the axis,on the parabola

y = 6 - x

2
.

What are the dimensions{ of

such a rectangle if its area is to be a maximum?


23. A rectangular sheet of galvanized metal' ip bent to form the sides and
bottom of a trough so,that the cross section has this
t,

If the metal is

14

shape: U

inches wide iloweep must the trough be to carry

the most water'?

24. A rectangular sheet of iplvanized metal is to be made into a trough by.


i
-nnd ing it SQ that the cross section has a ' 1
shape.
If `the metal

-is

10

L___1

inches wide, how deep must the trough be,to carry the most water? 0
'44144

25.

Pro/0 that with a fixed perimeter

P the rectangle which has, a maximum

area is a square.
26: Determine' the area of the largest rectangle that
can be inscribed fn the;
,region bounded by the graphs of y 2 = 8x and x =_4.
27.

Show-that there is=no

to the point 11,01i that

on-the ellipse given by


1

(3

4y

closer

'28., Find the alti,tudelif the

one ofMaximum volume that can be inscribed

,.,

a ..sphere,Of r,idiuS" r.

145

1,5 5
7'

in t

(
29.

A rectangular pasture,with lie side bounded by a straight river, is


fenced on the remaining three sides.

If the length of the fence is

400 yards, find the dimensions of the pasture with maximum area.
30.

A farmer plans to enclose two chicken yards, next to his barn with fencing,
Find

Ihs shown.
(a)

the maximum area he can


enclose with

feet of

120

Barn

fence;
(b)

the maximum area he can

Chicken Yards

enclose if the dividing


fence is parallel to the
barn.

In the follqwing problems (Nos. 31-35) meaningful replacements for the


variables are obviously restrieted to'positive integers, but we must consider
the functions to be continuous in order to apply the techniquts of this
chapter.
31.

10,000

A printer will print


-thousand.

'labels at

a base price of, $1.50 per

For a larger order the base price on the entire lot is

decreased by

cents for each thousand in.excess of

10,000: For how

many labels will the printerts,gross income bele maximum?


32.

100

A manufacturer can now ship 'a,mgo of


per ton.

tons at a profit of

He estimates that by waiting he oan add

20

$5,00

tonsper week to

the shipment, but that the profit on all that' he ships will be reduced
250, per ton per week.
33;

A peach orchard now has


400

peaches per tree.

How lOng will it be to his advantage to wait?


30

trees per acre, and the average yield is

For each additional tree planted per acre, the

average yield is reduced by approximately

10

peaches.

How many trees


e.

lier,acre wile give'the largest crop of,peeches?


34.

A potato grower wishes to ship as early as possible in the season in


ordeg to sell-at tbe best price.

tons at a prof

of

If he ships-July 1st, he can ship ,6

$2.00 per ton.

By waiting hesestimates he can add

tons per week to his shipment but that the profit will be reduced by

dollar per ton per week., When should he ship for a maximum profit?

I
f

11.65 (3'

2-7
35 A real estate office handles 80

apartment units.

unit is

$6o.00 per month, all units are occupied.

creased

$2.00

When the rent of each


If the rent is in-

repairs and maintenance).

$6.00 worth of service a Month (i.e.,


What rent should be charged in order,to obtain

the most profit?

A right triangle with hypotenuse

is rotated about one of its legs..

Find the maximum irblume of the right circular cone produced.


37.

Determine the dimensions of the rectangle with greatest area which can
be inscribed in a circle of radius R.
P

38.

a man-ti, on the average one further unit remains unoccupied.

Each occupied unit requires

36.

71

Determine the dimensions of the rectangle with greatest perimeter which


can be inscribed in a circle of radius

R.

>

2.47 157

44,

2-8
2-8.

Rate of Change: Velocity and kccelaration


The derivative

ft

of a polynomial,function

function whose value at


at the point

is the slope of the tangent line to the graph of

(a,f(a)).

In many physical situations thejwillie

velocity.

f 'has been defined as the

can also, be interpreted as

f'(a)

Let us loolOat an example.

Suppoie a solid ball is droPpAd froM a. 2000

foot tower.

its distance (in feet) from the top of the tower at time
it.is released.
t

Experimentation has shown that

Let

denote

(in seconds) after

is approximately related to

1Dy the equation


s =

Thus, we sometimes say that the fallen distance (s


time (t
function

seconds).
f

More precisely, the equation

feet)
e 2

s = 16t

isa function of
specifies the

t -)16t2.
2000 ft.

= 400

1500

position at

:V= 5

s = 1600

500

position at

t = 10

Figure 2-8a

We wish to formulate a suitable concept of velocity, so that we can


answer questions such as:

5
val

seconds?
t

How fast is the, ball falling after it has fallen

To do this we first define,the_average velocity in the time inter-

< t < t
-. 2
1 -

as the ratio

2-8
f(t

2
t

- f(t l)

-2

feet/second.

This is just the ratio of the distarice traveled in the time interval to.the

length of the interval.


For example, in the time interval

[4.5,5]

(i.e.,- 4.5. t < 5)

the

ratio is
.,.

16
,

(5)2 - 16 (14:.)
5
4.5

400

324

0.5

152.
.

Therefore, the average velocity between


In the time interval

0525.2]

16..5.2)

and

4.5

seconds is

152

ft./sec.

the ratio is

- 16 (5)

4 2.64
0.2

5.2. 5

'0_13.2.
,-....,

Whence the average velocity between

and

5.2

eeonds is

163.2

ft. /sec.

''''

,....

suppose that
the time interval'

li

..

1...... ...... .._...........

is a small positive quantity

The average velocity in

5 < t. < 5 + h is then

ti
f(5

f(5)

h)

ftilsec

This is just our old friend, the difference oott t used in approximating the
-derivative.

We know that.asthe time interval becomes shorter, ,h Approaches

,zero and the ratio expresing average velocity approaches, f'(5),


soff' the.ddrivative-of the distance function

at

='5.

the value

We therefore, adopt

the, following definition:

The (instantaneous) velocity of a body whose position

after
In our example,
".

It

seconds is

seconds, s given by

f'

160

t --)32t, and

f(t),

f'(5) = 160,

is

ft(t).

so that the velocity after

ft./sec.

In summary, the function


f

t a s = 16t2

describes thepostficin of the ball at time


f': t

describes its veldaty aetime

t,

= 32f

t,

lilt 59

while

2-8

Velocity is a measure of how the position of a moving body changes over


time.

It is often' characterised as a rate of change of nosition.,with respect

to time.

The acceleration ofa moving body is also a rate of change -- it

,,

measures how the velocity is changing over time.


1 Is
'

Acceleration
flpo-

We,haVe.seen that velocity, like distance, 'can:be exl?ressed;as


tion of time.

As st,solid ball falls from a tower, for example, *ts 'velocity

(in ft./sec.) after

seconds is given by

\t

y = 32t.

This formula "Specifies a velocity function


g

32t,

and Allows us to determine how"the velocity of the falling body'i3 changing


The rate of change of velocity with respect to time is called

over time.

acceleration.
4

Just asoie4efined.average velocity over a time interval, we can define


ft

average acceleration over a time 'interval

40change in velocity to the length of time

as the ratio of the

l' 2

t2 - tl:'
I

g(t2) - g(ti)

<'
,

In the time interval J4.5,51

- t

the ratio is

32(5) - e(4.5)
5 -

1111gt

.5

32(0.5)'

32.

0.5

Hence, the average acceleration of a falling body between

4.

an
,

seconds of lewd time is 32,^ft /sec. per sdcond.


In a short time interval

,5/

the expression for the'averages:

[t,t + hi,

acceleration is

1(t

h) - -g(t).

ft

h
..,,

As the time interval becomes shorter, ;11

l
,,
44.

approaches

and,thd ratlp

expresaing average acceleration -approaches

gt(t),,

the derivatle of the

.
-:.veloci.:ty, function. 4As before, we therefore define-:
I
.

-e.

' The (instantaneous) acceleration of


ajfte6r

.4

t" secondS,is given by


_

.:,

g(t),

body whose velocity,


is

g'(t).

2-8
In bur example,

and

g': .t -4'32,

indicating that the accelera-

tion of a falling body in the absence of air resistance is a constant appraxi.


mately equal, to

32

ft./sec. each second.

In Chapter 7 we shall see - that constant acceleration of an .object

(e.g., a = 32)
A

-(e.g.,

guarantees that its velocity is a-lir-lb:1r function of time,

32t)

and that the Aistarice At travels is a quadratic function of

time, (e.g., s = 16t2).


ligte

Example 2-8a.. gh""ii-ti..ietre velocity of the ball dropped from the

top of

the: 2000 foot buildpag,akt-the time it strikes the ground.?

The distancefunction
f

16t2

andthe baIl.iskdropped from a,height of 2000

feet.

Setting

16t2 =: 2000,

.411

we see that the ball strikes the ground when

i5555
=

)t

-----55.

Since the velocity function is


ft

we find

V(515) 2,357.8.

3?t,

Therefore the (impact) velocity after 5/5 dec.

is approximately
0

357.8

Example 2-8b.

ft/sec.

A car is being driven at the rate of

60 jmi. /hr.

(88 ft./set) when the brakes are uniformly applied until the car comes to a
complete stop.

Suppose that the function


2
f

describes the distance

t -4

k t - 7.-

10)%

s = f.(t.)- in feet traveled in,

seconds after 'the

brakes are applied.


(i)

How many feet doei the car move before it stops?

(ii) .Show that the acceleration is negative and oonstant.

no.

The velocity function is


-6 4 88 - -E-88 t

the distance traveled in

f(5) = 220

t = 5. -Since

gives

Solving

0.

and the car will stop at the point where the velocity is

ft(t) = 0

seconds is

220

feet:
.

The derivative of the velocity fUnction

t -488 -

88

(renamed for

convenience) is the acceleration' function


-

g': t

Therefore,' the acceleration after

88

seconds is

88
-

feet per second per

second, which is indeed constant negative acceleration (deceleration). , This


refle,pts the physical fact that the car is slowing doyen due to the applica' tion of the brakes.

The,fact that the acceleratiori isaonstant is based op

the assuAption that braking is uniform.

tO

(rue to brake fade, the pressure'

applied to the brakes must increase to maintaina-uniform deceleration.)


It is often useful in'other situations to interpret the value of the
Given a polynomial function

derivative as a rate of change.

the average rate of change of

[a,b)

in the interval

x -4f(x),

is defined to be

f(b) -, f(a)
,b - a.

Since the.limit of this as the length of the interval approaches zero


apprOaccies

of f

as tti

is

a)

(b

f'(a), At is appropriate to refer to the deriVative

(instantaneous) rate of change of

This is consistent with our interpretation of

at the point

f1'

x = a.

ft(a) ...as the slope of the

for the slope of a line does measure its rate of

tangent line at

x = a,

rise (or fall).

The tangent line is the line of "best fit" near

(a,f(a))

and hence its slope (rate of change) gives a measure of the rate of change

of f

at that points

ry

152

162

2-8
Example 2-8c.

The volume pi' a sphere isa function of its radius.

the rate Of change of the volum4' with respect to the radius.

rite of change when the radius is


Letting

Find

What is this

inches?

'V' denote the, volume (in cubic

hes)

and

the radius (in'


,

"inches),,we have

V=

In Other words, if we let


is given by

V,

V = f(r)

when

r =

0.

denote ehe function

r >(li n)(3r2)

The rate of change of the volume


2

vr 3 .

(in cubic inches).

ft
Ltvr

4
3
r -4- vr

The,derivative of

'

the volume
f

is

4nr2.

V, with respect to the radius 'r

is thus

or

; the volume is changing at the rate

C!
146 ire/ unit changy in radius.

, Speed

We have defined velocity as the (instantane4ps) rate of change: of the

position of an object with respect to time.

This definition implies that

'velocity hes direction, because it involves not only how fast the object is
moving, but, also from where to where.

For motion along a straight line thgn,

the velocity of a moving object is a signed quantity:

A,

it is positive if the

motion,is in the direction we define to be positive, and it is negative if the


O

motion is in the opposite direction.

What we intuitively think of as the

speed of the object is independent,of direction.

We define speed to be the


-

absolute value of velocity.

The following familiar example (Exercises 1-3,

To. 12) will make the distinction clear.

Example 2-8d.

Suppose that a rock is thcdwn straight up from the ground

with .an initial speed'of

64

ft./sec.

Contrary to our analysis of the ball

4,

dropped from the

2000

foot tower, let us take "up" as the positive direction

and "down" as the negative.direction.

Hence, the inAial yelocity.oi' our rock

.
'

is

+64

ft./sec.

..

g
.

We know 'that if the rock were not a cted upon,4 gravity the pAitfon- of

the ro* after t

seconds would be given by the function


f,

64t.

AZ'
a

153

63

.:

J.

2-a

But the force of gravity acts on the rock, just'as on the ball dropped from
2

the tower, adding a downward (in this case, negatiyj component,

the position of the rock at time

to

Hence, the function

t.

Ot2

t -464t

-16t

describes the position of the rock at time

and

t,

f': t -> 64 - 32t


i

gives the velocity-of: the rock after

for

t < 2; f'(t) > 0,

for

t = 2, f'(t) = 0,

and for

t > 2, flt) < 0,

indicating that between

and

We note that,

seconds.

seconds the rock isrising (moving in the

chosen positive direction) while after

seconds have elapsed the rock is

falling (moving, in the negative direction).

At

t = 1.5

seconds the velocity of the rock is


Jr.

f'(1.5) ft./sec. = 12 ft./seg.

At

t = 2..5

seconds, the velocity is

al/

f "(2.5) ft./sec. = -12 ft. /sec.

We conclude that the speed of the rock is the same at


namely' 12. ft./sec.

and t = 2.5,

t = 1.5

Similarly, the initial speed and the final (impact)

speed of the rock are the same since the inittal-ve1ocitrls--164 ftisec.

(up)and the impact velocity is

-64

ft./sec. (down).

Renaming the velocity function


g r t -* 64 - 32t,

we see that the acceleration of the rock is given by


111

gl: t -t-32,

indicating that gravity accelerates the rock at the rate of

32

ft. /sec.

eachsecondlOk downward, of course.

Note that this negative acceleration

detreases the rock's velocity constantln.on the 'way up the;ZeC-i-easq in


VIP
velocity-amounts to a slowing of the speed of the rock, while on the way
.,

down the "decrease" in velocity (becoming more negative) creates an increase


-

. -ipthe-speed of the rock.

...

.,

:,,

.
4

,=

tP'

)
I.

II

158

4.

.'

%a

2=8

-4
Exercises 2-8
Determine the rate of change of the area of a circle with respelIK ,
.,
to its radius r. Compare your result with the formula for the
circiiittfriide

C 4.0f a circle in terms of radius

r.

-.
.

(b)

Wbat, is the rate of change

Cot are this resul\with the formu a for the

'r?

.4

from the time


?

L'4'

't =.8,

r-

L' -

z.,.11.1-

iSf a sphere 'in terms of radius. r.

2 A certain motion is described


3-

with respect
;6

to its radius
surface area

'

`the volume of a spher

t = 5 *until the time

by the equation

r
.

s = f(t) r 2t3 -- -39t2


(a)

We subinit tliat

f(i) = 2t 3

f the distance

"hen the velocity

is given by

39t2 + 25 t - 535,

=6t2

at time

v .at 04the

252t - 535.

Is

- 78t + 2

Explain why :this is, tr ue.

(-b) 'Sketch the graph of th'e ;unction

t -4 s = 2t

on the interval
(c)

ft

on the same interval

4C

-39t

5 < t <

Sketch the graph of

&

+ 252t '- 535

v = 6t2 - 78t + 252


p5,81.

DetArmine the zeros of

ft.
M

14

(e) ' When does the ptxticle whose motion Is being describedcome to rest

--

for an instant,as it shifts direction?


,(f)

Cy

Whenis the particle the ,greatest distance from its starting point
,on

[5471?

,
(g).,..1.114at is the greatest distance Of the patticle from its starting

point on
(h)

(5,71'1.

When is the next time on

[5,8]

when the distance of the particle

from its starting point is as great as its greatest distance on

N
,

(5171

4
155

65

4's
s

2-8
When is the particle the greatest distance from its starting point

et)

[5,8]?

on

When is the speed of the particle the greatest on

(j)
3.

of the particle whose motion is

'What is the acceleration at time

(6,7]?

considered in Number 2?

4. 'Hecall Example 2-8b


(a)

For a cat traveling at

mi./hr., how'many seconds are required

60

after the brakes are applied (and held) before the car comes.to a
.

(b)

complete stop?

mi./hr., go after the brakes.are

60

How far will a,car traveling at


applied?

(c)

"

t = 0

Suppose at time
60

velocity

the brakes are applied on 11 bar moving at

mph and kept on until the car is broaght to a stop

producing.a constant deceleration (negative acceleration) of

1
(a)

Given an approximation for

ft. /sect.

wi4 continue to travel only

100

Ustig your approximation for

,to ensure that the car

f . after the brakes-are applied.

a.)

from part (c), determine the dis-

tarice the car would require to stop if it were traveling at

30

mph.

5 Let us assume that a pellet is projected straight up and after awhile,


comes straight down via the same vertical path to the place on the'ground

After

from whiCh it.was launched.

aPoveth

ground.

followin

table.

seconds the pellet is' ;s 'feet

are given in

(t,$)

Some of the ordered fairs

the
.../.4

144

256

336

384

400

384

336

10

We shall intentionally avoid certain physicefl considerations such as air


resistance.

Moreover, we shall deal with simpll numbers rather than

qtantities measured to some/prescribed degree of accuracy which,mfght


.

arise from the data of an actual projectile prOletlin engineering.


''(a)

Interpolate from'tqe data given to detefmine the height of the


projectile efter eight and nine seconds respectively.
using syminetry as your guide.)

(Guess,

Does extrapolatiO;1 to find values

of

s 'for,

or

t =11

make sense_on physical ground's?

After how many seconds does the projectile appear to have reached
its makimum helig? What seems to be the maximum height?,

156

2-8
(b)

appea

Doe's

to be a,function of

so, discuss the domain

t?
,

,physical' considerations into account.

and range, takin


(c)

If we were to plot a graph of

s = f(t),

is it, plausible on physical' grounds to restrict our. graph to

(11)

the first quadrant?


Does 'the dat4 suggest that the scale on the s-axis (vertical)

(2)

should be the same as the scale on the t-axis (horizontal)?


-

(d)

Keeping in min
(t,$)

your responses to part (c), plot the ordered pairs

from the table.

Connect the points with a smooth curvy..

On

What is the name of the function suggested by the graph?

physical grounds is it feasible 'that there would be a real value of


s

forfevery reaAnumber assigned to

0 < t < 10?

over the interval

Were we probably gistified in connecting the pointsR

(e)

Assuming that the equation

s = f(t) = At

+Bt + C was used to

develop the entries im our table, find- values for constants


B,
(f)

and

A,

C.

Sketch the graph given by the equation

s = 160t - 16ta

over the

interVa41-. 0 < t < 10., Using a more carefully plotted graph of the
above set, connect the point whei-e
t = 2

with a chord.

the,slope of the curve at


'(g)

If the units of

with the point where

t = 1

and

Estimate

t = 2.

preTeet and the units of

are the units of slope?


ratio of units?

t = 1

What is the slope of this chord?

are seconds, "Ale't

What word is .commonly associated with this

What would you guess are the physical interprets-,

tionsof positive, zero,_ and negative values of this ratio?


(h)

Draw the graph of

v = 160 - 32t

Compare the values

of v

for

over the interval .0 < t

t = 1

and

t = 2

your estimates for .the ,slopes of the graphs of

respectively with

s = 160t - 16t2 .in.

pgA (f).

X767
4

2:8

ci)

Average the values of

and

for

an

i.= 2

compare this

average with the slope, of the chord connecting the points where

t = 1
(4).

t = 2

and

in part (f).
are

If. the units of .v

ft./sec. and the units of

are seconds,
What

v = 160 - 32t?

what are the units of 'the slope of the line

word from physics is commonly associated with this ratio of unite?

Does the minus signalong.with the particular numerical valUe of


this slope have any special connotation from your experience?
6.

Aoprojectile is fired straight up and after awhile comes straight down


via the same vertical path to the place on the ground from which it was
2

launched.

t- seconds the projectile is,

After

s = 160t - 16t

ofet

above the ground.


.

(a)

After how many seconds,4oes the projectile strike the ground?

(b)

What is the velocity of the projectile after

(c)

What is the initial velocity?

seconds?

(d) What is the tmpact,velocity?


(e)

How high is the projectile after

1- .seconds?

(f)

How high is the'projectile after

(g)

After how many seconds does the'projectile reach its maximum height?

(h)

How high does the projectile 'go?

(i)

How far has the projectile traveled after6 seconds?

7.

seponds?

A ball is thrown upward from the ground so that after t,seconds its
height 's

feet igiven bythefuriction


.

-4

f
(a)

, 2

t -4 s = 96t - lot

The path of the ball,is straight up and straight down.


.

of the functiOn

What is the

f?

What is the velocity function?

(b)

What is the derivative of

(c)

How high is the ball after,on& second?.'

(d)

How high is the ball after

(e)`

How far has the ball traveled after

(f)

What is the initial velocity of the ball?

f?

5,

seconds?

168

seconds?

4,

2 -8

(g)

How long is the .tell in the air ?'

(h)

What is the impact velocity whep the ball strikes tha ground?'

(i)

What is theconstant acceleratiOn acting upon the ball?


Give a distance function

where

t;-'

is the number,of

t :,seconds, appropriate for the situa-

feet above the ground after

tion if the ball were thrown stVaight upward with an initial melocity of

96

ft./sec. from a to9er

Give a'distance function

(k)

200

where

t *4s,

G :

feet,:above the ground after

ft. high.
s

is the number of

;seconds appropriate for the situa-

tionif the ball were thrown st&ight downward with an- initial
96

velocity of

ft,/sec. from a tower .200

Give a distance function q'

(2)

where

feet high.
s

the number of

feet displacement from the top of the tower after t,rseconds if


the ball is simply, dropped from the-top of a tower.
,8.

The velocity Of an object, whose location on a straight line at timer

t= t

is given by

is the limit of the ratio

f(t),
.

f(t) - f(to)

t - tO
as

approaches, to.

This limit is the value of the derivat/v

Experimentally it has been established_ that the distan

t = to.
in time

by

freely falling body is proportional to

fore it can be represented by the function


0

positive constant.

ft

,covered

2
,

wher

at

ft

Show that the velocity of a ireely felling

is a

ody is

directly proportional to the t' me.


Suppose a projectile is ejected
second, at a point

P which is

eet per

th initial velocf.4roof

20

feet above the ground

friction and assume that the projectile moves up and down

eglect

a straight

pf

line.

Let

attains

f(t)

denote,the height (above

seconds after ejection.

P) in feet tha

Note that if gravitar onal attraction

were not acting on the projectile, it would continue to m


p constant velocity, traveling a distance-..of

thatits: height at-time

would be given by

feet e

0
f(t) = vo .

e upward with
h second, so

We knowthat.

o slow doWn anti

qstvelocity is zero and then trayel_baek-to r-the earth.

On the basis of

.
0,

the force of gravity acting on thietprojectile causes it

I
4

the projectile

159

169

2-8

'1"
g

f(t) = v0 t- at2 ,

physical experiments the formula'

where

represents

the acceleration of gravitOs used to repreSent the height (above P)`of


the projectile as long as it is aloft. Note that f(t) = 0 when t = 0
2vo
,This means that the projectile-returns to the initial
and when t =
.

2v

20 foot level after

--- seconds.
g

(a)

and
(b)

t = to

Find the velocity of the projectile at

(in terms of

g).

Sketch the

s vs. t

graphs on the same set of

v vs. t

and the

$.

axes.
(c)

v0) the time required for the Velocity to

Compute (in terms of


drop to zero.

(d)

What is the velocity on return to the)nitial 20 foot level?

(e)

Assume that the projectile returns to earth at a point


below the initial take off point'-1).

30

feet

What is the velocity at

impact?
10.

V2

ft./sec., it will reach a maximum height of

11.
*

V0

Show that if a ball is thrown upward with an initial velocity of

''Elsie Can throw a ball -148

go when it leaves her hand?

hand is .4

feet.,

How fast does it

feet straight up.

(Assume that when the ball_ is released her

ft. above the grounds)

12.

A ball is thrown upward with an initial speed of

64

ft./sec.

Simultaneously a ball is dropped from rest at a height of

128,

ft.

When} does impact'occur and how fadis each ball going dt the time of
impact?
13.

Determine the average velocity of a car for a total trip if,it averages

60 miles per hour going and. 30


14.

miles per hour retjUtning.

Find the velocity of an object whose location along a straight line is-deacribed.by.the.equation

s = 128t - 16t2.

Sketch the graphs of

s vs. t

on the isme set of axes.

and

v vs.

(a)

During what time interval or intervals is t400object moving toward

the location
(b)

s = 0?

What are the values of

and

160

170

when

is a maximum?

2-8

41015:

A ball is, thrown upward with a velocity of

height

in feet after

feet per second.

32

Its

seconds is described by the, equation

s = 32f - 16t2.
(a)

What is the velocity of the ball when its height first reaches.
12

When it again reaches

feet?

12

feet?

(b) 'How high does it go, and how long after being thtlOwn does th,e ba

reach its highest pos.glon?


16.

An object is projected up a smooth inclined plane in a straight line.


Its distance

in feet from the.starting point after

described by the equation

s ='64t - 8t

2
.

seconds'is

After the object reaches,its

highest point it slides back along its original path to the starting
s =

point according to the equation

- t

Here

tance of the object from the highest point and

to

is the dis-

is the time it takes

the object toreach the highest poidt.

f
(a)

Determtnejhow long it takes for the object to make the up and down
trip.

(b)

Sketch the

s vs. t

of coordinates.

graph for the up and down motion using one set

Do the same for the

71.

v vs. t

graph'.

2-9
The Second Derivative

2-9.

In the preceeding section, we found that the flunction expressing the

derivative of its

acceleration of an object with respect to time is

velocity function, which is in turn the derivative of the function


cribing the object's position at time

t. Hence,

can be obtained by differentiating' f

twice.

derivative of

f" and

e acceleration function

For any function, f

the

is called the second derivatiVe of _f....--The-second derive-

f'

tive,.denoted by

des-

f",

gives us valuable information about the graphs of both

f.

.Consider the function


f

x -' x3 - 3x.

Its first derivative is


x

-4 3x

- 3,

and its second derivative is


f":

and

describes the slope of the tangent to the graph of

Since

f"

'since

f"(x) -7, 6x < 0,

for

Now to say that

increases.

x < 0,
f',

then

f'

is decreasing as

the derivative of

an interval, is the same as saying that the slope of


that interval.

x < 0

f,

is decreasing over

is decreasing over

.T

From the

sketch of Figure 2 -9a we

T
2

see that if,the


slopes of
I

successive tangents to the

graph of

are decreasing,

the graphza
f
4WA,

ii "bending

Hence, we are led to

down."

believe that the'graph of the

function
3

2",

x -)x

3.

Figure ?-9a

- 3x

is similarly "bending dowh" as

x < 0

increases, since

f"(x) < 0

.474a,

43.4o,

72

for

x < 0.

fr

2-9

In like manner, since

given by
for

fqx),

f"(x) = 6x > 0

for

the slope of

x > 0,

f ,f

is increasing

Figure 2-9b indicates

x > 0.

that in such a situation the graph


of

his "bending up."

Putting

these two pieces of information


about the graph of

: x -) x

- 3x

together with what the first derive.tive,

f',

tells-us about it, we

obtain the following graph of

f.

Figure 2-9c
The 'second derivative,

f"(x) = 6x,

is bending downward, and for

x = 0, f"(x) = 0

x > 0

f':"x -) 3x

x < 0:,

the graph is ben

and the graph is apparently 'Changing

The first derivative,


e

indicates that for

the graph of

upward.

At

Aa

m one to the other.

- 3 .r. 3(x , 1)(x + 1),

indicates that the

slope of the graph is zero at

we conclude that

x = 1

and from the way the graph 14,bending

has a relative maximum at

x..,= -1

and a relative
-

minimum at

x = +1.
/

163

173

Zi4

a-9
A more rigorous interpretation of "bending - upward" or "downward" uses the

f.. We say that a func-

notion of the convexity and concavity of the function


tion is convex in the interval

[a,b]

if its graph inthis interval lieS.,


*

above each of its tangents in the interval.


which is convex in the'interval shown.
concave in an interval
in that interval.

Figure 2-9b shows a function

Similarly a function is said to be

if its graph lies below each of its tangents

[a,b]

See Figure 2-9a for an example of a concave function.


s'

Some texts use concave upward in place of convex and concave downward in
place of concave.

The ideas of convexity and concavity are said to describe

the flexure (bending) of curves.


e intuitive remarks motivating the use of

x3
(1)

f"

in graphing

- 3x lead to the characterizations that:


a function
f"(x) > 0

la,b1

is convex in the interval

for all

between

and

if and only if,

b (a < x < b).

and
(2)

a function
f"(x) < 0

is concave in the interval

for ail

between

and

[a,b]

if and only if

b (a < x < b),

An important consequence of these two characterizations is that if the


grgph of a polynomial
p(c,f(c)),

then

crosses its tangent at the'point of tangency

f"(c) = 0.

nt,

Figure 2-9d

The graph of f

crosses its tangent at

P.

The reader familiar with convex regions from g geometry course will
observe"that we are defining a function to be convex over the interval in
which the region above the graph of f is convex.
4,

le

2-9

'For if the graph of

crosses

at

(c,f(c))

x = c

then7'near

Figure 2-9d),'the graph must lie.above the tangent on one side of


below the tangent on the other side of
'one -side of

and (x) < 0

c.

e(x) > 0

Hence,

oh the other.

Thus,

and

for

f"(c) = 0

(see-

since a

polyAomial function must pass through zero as it passes from positive to


negative.,

If the graph of 'f, crosses its tangentatpoint of


tangency
of

tHen P

P,

is called a 'point ok

k.

The above argument shows that 'if _P(c,f(c))


of

f,

then

f"(c) = 0.

The converse of this may be false, however.

f"(c) = 0 at points
of

f.

is a point of inflection

(c,f(c))

Consider the graph of

It is quite possible that


P

where the tangent does- not cross the graph


f

14,

at the origip.

(See EXehises

'2-9, No. 2.)

To summarize, it, is instructive to view the graphs of

together to see how the zeros of 0 and

f"

f, fel, and

affect the graph of

f.

f"

To.

show the relationships most vividly, we illustrate the graphs in Figure 2-9e
without y-axes.

4E
Alk

165

2-9

'

maximum

1
1

1
1

point of
i flectilrom

f"(#

6x

31

Figure 2-9e

3.661:7G

>

2 -9

Note that a relatie maximum occurs at

e(c)

< 0,

(c,f(c))

since the tangent to the graph of

zontal and the graph must be concave.

when

f'(c) = 0

and

when

at

ft(c) =$0
x

ik

add

must be.hori-_

Similarly, a relative miamulioccurs


Alt

since the tangent muse-behorizontallas

f"(c) > 0,

before, but here the gra'ph' must be convex.

s
.

A point of inflection occurs'at

the other side of

When

(c,f(c))

is positive immediately to one side of

x = c

f"(c) = 0 \and cf"(x)

and negative immediately to

x = c.

Example 2r9a.

Determine the intervals over which the function


f

, 2x

+ 1

.rooffodi,

is increasing, decreasi g, convex, and concve, and

ate all relative

maxima and minima, and hll points ofinflectione


Intervals on which

is ivreasing and decreasing can be determined by

the sign of the derivative

f' :

4x3 - 4x = 4x(X-+ 1)(x - 1)

'The critical points (points where


and

f'(x) Cs 0) occur where

x = -1, x = 0,

x = 1.

10 <

f'(x) < 0

for

f'(x) > 0

for

-1 < x < 0

ft(x) < 0

for

0 < x < 1

-1
(3

f'(x) > 0. for

From these signs we conclude that-as


X < -1,

rises between

`finally rises again for

-1

and

0,

1 < xt

increases the graph of

falls again between

f4lfalls for

and

and

1,

x > 1.
-

Intervals, of convexity and concavity can be determined by the sign of


the second derivative

f": x -,12x2 - 4 = 4(3x2 = 1) =


4

f"(x) = 0

if and only if

/T=

x = I

if

(xlq - 1)

if

(x-.1-3- +

and
0

(x13- + 1)

and

- 4)(xl3- + 1)
IF

or

x = -

11-71-

f"(x) < 0

if and: only.

have opposite signs, which happens if and only

(x13- - 1) <00,

that is, if and only if

,167 177-

1
-

< x < 1 .

r.ft '

-.

a2 S

st),

..

2-9
0

f"(x) > 0

.10

Similarly,

- 1)

if and only.if' (x,/

same sign, which happens'if and pul?.y if

and (x5 + 1)

have the

-,..

,.

x,q+ 1 < 0

x)q - 1,> 0,

or

i.e., it and only if

<', ,,-

1
mor
x >

We conclude,that

is concave

4
over the interval

-- < x < --,,

and convex for all

x < 1 -1- or all

'...

N.,

4.

x. > -2-

The graph of

)q

x =

since

- (I.

3.

r"

thus has points of inflection at

41/..

= 0

and

..
is convex on one-side of each of these

Together with the information that the

point and concave on the other side.

Wirst derivative of

x = =.1. and

is zero at

x = ,:1, 0, and :1,

these intervals of

_convexity and concavity show that

and

Finally,he graph of

hese relative minimum at

x = -1,

f.

hes.a relative maximum at

x =

6,

hag a relative n*1imum at

x =

1.

has poi is of inflection at

x = -

-1-

and

x =

4,T

-11,T

f(! --

= 0 and f is convex on one side of each of these points


o
IT
and concave on theolher,side.

,since

In Figure 2-9f we sketch the graph of

using the foregoing informa-

tion.

Figure 2-9f
f : x'73 x.

2
- 2x + 1

0
168
.

17 8

Exercises 2 -9
1.

Determine the second derivative of the function


f

2.

t -,2t

- 39t

+,252t --535.

Characterize the origi&for each of.the following functions (by determining whether it is a relative max. or min., or poirii of inflection):

3.

(a)

f : x

(b)

-c4

4):3

Consider the function


f

x -)4x5 + 5x4

20x3

50x .- 40x.

(a)" Find
(b)

fl(x), and

e(x).

Characterize each of the points

f(-1))

tnd (2, f(2))

as

maximum oAminimuni.

,Consider the function


(a)

Determine fl(x).

(b)

Determine f"(x).

,f

)(

2x 3

-4

+ x

on

[-2,2).

(0' Evaluate- f'( 1).


(d)

EValuatete(-1),

(e)

Describe the behaviok of'f on

[-2,2) -(by determining Sub-

intervals of increase, dedresse, convexity, and concavity, and by.

locating-relative maxima and minima andpointq of inflection, iftr


any of these occur).
(f)
5.

Sketch the graph of

on

[-2,2).

Determine the relative maximum point'and relative minimum point of the


graph -of

f : x -)(-x

2) (x - 2)

6.

Sketch the graph of

f : x

+ 5x4

Ifrx

20x3

50x2 - 40x.

(See NO. 3.)


7.

Sketch the graph'of


extrema (adxiMa an

- 3x2

- 12x + 2,-

indicating relative

minima) and points of inflection.

169
4_1 -

4'

2-9

8.

The.point

(1,1)

functions:

lies on the graph of each of the following

For which is this point (Wa relative maximum, (ii) a.

relative minimum, (iii) a point of inflection, (iv) none of these?


x

(a)

(b) x

9,

6x2 + 6x - 1

- 4)e,+,5

2x3
3

( c)

x -4,2x

( d )

x -*2x3

+ 12x - 10

- 3x

3x2 - 12x + 14

Consider the function

+ x 3 - 2x

- 3x

.over the interval'

-2e< x < 2...


(a)

At what points is a tangent to the graph of

(b)

What are the relative minimum points ?,

(c)

What is the minimum value of

f?

(d)

What is the maximum value of

f?

(e)

Sketch the graph of f..

10. Claspify each of the points

of x 3 - 6x2 + 9x

- 4

(1,0))

(2,12),

and

horizontal?

(3,-4)

on the graph

as a relative maximum, a relative minimum, a

point of inflection, or not' bf these.


11.

The figure at the left stows four


polynomial graphs and their common
tangent

yi'=3

at '(0,3).

Match each graph

(A, B, C, D)

with one of the following equations.

3-

(a)

Y =

(b)

Y = 3

(c) ,y = 3

,(d)

= 3'

:,4 + x2

x
-,-;

%..

'4- X3

(e)

r= 3 4-

x3

(f)

=3+

x.

(g)

= 3 4-

(h)

y . 3 +

+ x

+ x3

I-

'110111011

2-9
12.

15.

Classify(see'No. 10) the


(a)

x -,(x - 2)2

(b)

x -.)(2 - x),3

(c)

x -)(x

Find an

(2,0)

for each of the following functions.

2)4

classify each critical point (local extremum or point of

inflection)' for each of the following functions.


(a)

x -)2x

(b)

x --)x

(c)

-2x

(d), x

/14.

12x - 7

+ 3x2

- 12x + 160
3

+ 3x2

+ 12x + 7

1)2(x + 2)

Consider the function, f


(al

9x2 + 24X - 18.

Determine

(b) Loca4 the ralatiVe maximum and minimum points of


(c)
.(d)

(e)

Determine

f.

f".

What is ae point of Inflection of the graph of.f?


What is the slope of the tangent to the graph of

(0 Determine

f'(3 + k)

and

f'(3

k).

(g)

Sketch the graph of

(h)

Discusp the symmetry of the graph of

at

(3,0)?

4.

f.

f.
=

15. SIT that;the graph of a cubic function must have a point'of inflection.

16. (a)

Sketch the graph of


f

x5 + x

5x

- x.

+ 8x - 4.

Respond to each of the following by inspec tion of your graph for 'part (a).
(b)

What are the zeros of

f?

at the points where

(c) ACe9cribe-the flexure of the graph of

f(4 . 0.
(d)

Describe the flexure of the graph of


,4.2(x) = 0

and

at points for Which

f(x) > -5.


2

mine those points on :He graph of

17.

X5) _ 2x
rs-

at which

e flexure is neither upwarld nor downward.

,
I

'A
1

2-9

18.

Characterize the point

-,x4

on the graph of

2x3

7x2 + 10x 4210.

(-1,-11)_ on the graph of

and

Characterize the points ,(0,0)

19.

+ 3x5 + 10x.
,

20.

: x -,(x + 1)2(x - 2)

Consider the functions

and

g : x -,3(x + /)(x - 1) .

and

gr'_

(a)

What is the relationship between

(b)

Characterize each of the following points on the graphs of

and

g.
41.

(i)

C-1;0)

(ii)

(0,f(0))
(0,g(0)),

(c)

(iv)

(1,g(1))

(v)

(1,f(1))

(vi)

(z,f(z))

t4

Consider the function

21.

and

f : x

Sketch the graphs of

(a) Locate the zeros of


(b)

on the same set'of axes.

- 3x + 2:

f..

Locatthe relative maximum, relative minimum and point of inflectibn.

(c)

22.

Sketch the graph.

Consider the function

'3

f : x -, x

- 3x

2
4..,

(a
/,

Locate the zeros,of

f.

y kb i

Locate the relative maximum, reiative minimum and point of inflection.

i (c)

Sketch the graph.

Show that the grail

23.

of

Bi + C,

f :

A F/ 0,---has no point of

inflection..
1

Find an equation pf /the tangent to the

24.

at its point of inflection.

graphf
-o f :

x3

+ 3x2 - 4x -'3

';

'?
4

15.

2-10

2-10.

Newton's Method
In Section 1-8 the method, of repeated bisection was presented as a means

for approximating zeros of a polynomial function.

In this section

present

another method, known as Newton's method for approximating such zeros.

This

method makes use of the derivative and is more efficient than repeated bisec-

./

tion.

Newton's method proceeds as follows.

Suppose''.f

function and we wish to approximate the real zero


graph of

r.

is the given polynomial

By inspection of the_

synthetic substitution, repeated bisection, or some other device,,

f,

we obtain a first-approximation of

r.

Let us call this first approximation

4
1

Figure 2-10a
:

If the graph of

looks like thatMlown in Figure 2-10a, we should:

expect that the .tangent line at


point `x"

2'

which 4s

oser to

will intersect the x-axis at a

(xl,f(x,))
r

than i

'

The tangdnt line at

x
1.

has the e uation

(x1,f(x1))

y = f(x1-..Y + fl(x1)(X - xi):

that is,

This crosses the, x -axis at .(40);


4:

1'

0 = f(x1) +ft(x1)(x2 -x1).


,

itstuningthatmo/ 0,

we cansolve for -x2,

obtaining the formula

'

f?x

-,(4

i
\.

1
'

..1

X2,7 xl
..=1

SI

173 183

i
lh

2-10
x2 . instealdof

We can now repeat this process, using

to obtain the

1-

new approximation
f(x2)

.x3

TTTT2Y

x2

(See Figure 2-10b.)

in place of

Repeating &pill, using x3

we obtain the fourth approxima-

tion
f(x I-

x3

x4

.
.
.,..

process; havin 'arrived at

'Thus, equation (1) it the basis for an iterative

define i.new appr&imation

xn,

!the approximation

I'

(2)

n+1

=x 01

Example 2-10a.

by

f(x )

m)

ft(x

it

n+1

n)

.
r

For the polynomial function'

.
.

.,4
.

----f

x --.0 x3 --F.

2:+

x - 2

..

the value of the real zero which lies be ween


4.

141

.1

-;,-.

eLmate/

and

1.

.1;

/
A method o f CallilaIlOn which consists of the repetition (iteration)
of a basic process, especT7lly useful for writing a program for a compulkg
machine%;
...(

...

174
i

I,

/V

1,84

2-710'
Since

f(0) < 0

real zero between

and

f(1)

and

1.

we know that-there is

Further calculation shows t

f(0.8) < 0

and

as our initial estimate.

We have

ft

and

Le'

f(0.8) = -0.048.
2

f(0.9) > 0,

so that the de4red zero lies'between 0.8


.7'

at least) one

3x

+ 2x + 1,

us take

= 0.8

Since

"

we have

f'(0.8) =
whence formula (1) gives the second estimate '
)

-...

= 0.8

olts

= 0 81.

% .47?7

...

Now we calculate to obtain


f(0.81):= -.002459
and

ft(6.81) = 4-.5883.

We use (1), with

replaced by

to obtain the thii.d estimate

x2,

x3 =-0.81 -

o.81o5..

Correct to two decimal places the zero of

Example 2-10b.

f, is

0.81.

Use Newton's Method_to estimate

313.

Since .q i's a 'lot of thp equation x3 = 3,


function

must be a zero of the

f : x

Since, f(it)' and

approximation.

3.

f(2) 'lave apposite signs we take

The derivative of

xl = 1.5

as our first'

is

No-

3x

f"

2
f
A

*so that (1) gives


f(xl)

(i.5)3

x2)%x1

1.5
,3(1.5)2

f'FT,c3.,,,,-,

t 1.444 z 1.44.
Using :1.44

as our second approximationrrwe obtain


A

,k

/75

--1

A11

2-10

'4(1'44) tt 1.442.
f't1.44)

- 1.44.
..-

.Correct to two places

1/

1.44'

f4
1.

Exercises 2-10

t,

Use Example 2-10a to respond to each of the following.


a
(a)

What is the slope of 'theftngent to the graph of


x -.x3 t'x2 +

(b)

at the point

(0.8, -0.048)?

Write the equation of the tangent to the giaph of -f.-at t1

point

(0.18, -0.048).
(c)

The line tangent to the graph of

f ,at the.,point

(0.8, -0.048)

intersects the x-axis at a point close to the place where the graph
f

drosses the x-axis.

,What is the x-intercept of this tangent

line?
2.

Use Example 2-10b to -respond to each of the following.

(a) -Find the slope of the tangent to the graph Of


at the point

x -.x3

0.375).

( -.5,

(b). What is the. equation of the tangent to the graph of

at

(1;5, 0.375)?
,

(c)

Find the value of

at which the tangent of part (b) intersects

the x-axis.
3.

(a)

What is the positivezero of the' function

(b)

Show that a zero of f lies between

(c)

Use Newton's me

od to approximate

Consider tkii\function
(a)

f : x

and

If

to three deci

such that,

0 < r < 1.
ti

lb)

Find

(c)

Evaluate f(0)

(d)

To two 15140,mt..ajpaces apprOximate

fl.
4

and

fl(0).
h:

x
2

if

xi = 0
ni

(e)

Use your estimate from- part; (A)


a

and.

es.

x3 - 12x t 1.

Show that there is at least one. real number


and

2.

x' - 2?

o dhow that
4).040512

.e

f ( x2) =

.L5

9,308.

- ;

'f(xl)

9.xl

f(r) =.0

2-.10
(f)

Use Newton's Method and the results of parts (a) through (e) to
compute the zero of

.5.

between

and

to three decimal places.

Calculate to two decimal places the zero of


f a x ) x 3 - 3x
which is' between

6.

and

+ 2.

3. -.

Find an approximate solution of


2
x-

+ 3x = 7

correct to twOrdeAmal places.


7.

Suppose
(a)

is a polynomial function and

If the derivative

f'(x)

f(r) =

5, < r <I5'.

changes sign over the interval*.-N,b1,

it is possible that Newton's

method will fail to generate cloier

and closer approximations to

r.

Sketch a picturesshoWink such a


I

situation.

,
(b)

(c)

If

f "(x)

changes,sign over the interval

[a,b],

then even if

f'(x)

does not change sign it is possible for Newton's method to

fail.

Sket

In vid

a picture showing such a situation.

(a) and(b), what precautions should you take in applyink

Newton's method?

il

SS
I

'

'14

''*7

4,

2.al.

Higher Derivatives and Notation

We have denoted the derivative of the function


There are other notations in common use.

by the symbol

In graphing

f'.

we often write

f,

so it Is natural to write

y = f(x),
I

for the value of

f'

at

x.

is

Andther alternative symbol for

This notationallows us to abbreviate such statements as

f : x -4ax2 + bx + c

if

f'(x) = tax + b,

then
'^by'writing
2

D(ax

+ bx + c) = 2ax + b.

The symbol
R

sLY

dx
bj

which was introduced by Leibniz (1646 - 1716) to represent

14(x),,is suggested

10.have defined

by the'difference quotient used to calculate it.


4

'

If we replace

f'(x) =

by the symbol

lim
0
h

f(x + h) - f(x)

(read "-delta x ") to j1i icate a differ-

"Ax"

ence in x-coordinates, the difference quotient becomes

f(x + 6x) - f(x)


6.x

Th

ressio

(x + 64 - f(4 ..stands for the corresponding difference in

coordinates, so we write

+ 6x) - f(x)
6x

bx

This prompts the hotation

for the value of the derivative if

Ax

= LI.L4

dx

f(x).

The-sym

dv

/-

is not a ratipl

*r.

it stands for the limit of a ratio.

It is a tribute to the genius of Leibniz,

tigtha.4-9'a notation which, as we shall ste, anticipates some

c:Iyeyer,

732

properti6 of, derivatives which permit us to handle their valdes ps though


they were common fractions.
Corm,sponding

the foregOingtsymbols ?Or the first derivative, we have

the following symbols for the value of the second derivative!


2

f"(x)or", D2f(x),
dx
a2

The Leibniz notation

is again suggested by the difference quOtient and

`-'=-,Y=
.

dx

the A symbo\for "difference "


ALL.

f'(x + bx) - fqx),

"'dx)

bx
.
.

..

d(c(ii)

To symbolize the limit as & -)0,

we write
,

d(dy)

or

dx

dx dx

.1

..

..

*-

in theshort-

,
,

2,

dx

We have seen how to diferentiate'any polynoMial function of the form


..,,

x+ a2x + .
.

x -)a

+ a

. + a ijY''-'

F.'

'
.

Since the.second derivative of f

is still a poLynomiallwe may compute its,

.J.,

'derivative and call,it6,..the__...........third derivative of

f,_ denoting its value by any

..

of the symbols.

(.

-,:

f '"(x),

t"', D3f(x) ..,.,,,,l..


t --.-

.,

de
11

..
'

'

'4

t'.rivatf4e.bf-f
'r.''

14(.

Similarly, we could find the fourth

*.

4N

'

. y ditfer,entiatinf.he

;;,....

third derivative, and so on, to fifth, sixth, .014,,tir

higher derivatives.

.1.r1W
Geometrically, we have seen that
of the tangent to the graph'of

,f

f'(x)

_.

can "qep. f rpreted as the slope

at the point

(x,'1"

derivative cab.be'interpreted as an indicator of, the vQ


High

and that' the second

mature of the grtiph.


t..4:

derivatives do not have such vivid gponietrii.intedriretations:Tor


h

but 4hey do have. important algebraic relationships tethl coeffictents of the


term

of

f.

'

I
lc

179`,89

,11.1,

2-11

To be concrete let us look at a general third degree polynomial function


.

x+a2x 2.+a3x

f:x-4a.0

(1)

3:

We observed in4the first chapter that


t

f(0)-7 ao-.

If we obtain the derivative


`ft

x -4a

+ 2a x + 3a x
2

we observe that
ft(0) =.a1.
We. differentiate

to obtain the second derivative

ft

f"

2a.
,

+.2

3a .x;

and observe that


f"(0) = 2a2.

0.

If we difTereptiate

we obtain the third derivative

f",

`k
a

f"7 x -4 2 -.,3a

In this ease

fm(0) = 2
We summarize:.

for the cubic polynomial function


I

f :x-4a
th

3a3.

+ alx + a2x2 + a3x

coefficients'are related to the values of

tiv s

ft,,f",

and

and its'succe sive deriva-

by the-formuIab:

ft"- ateo-x = 0,

0
1

= f(0)

= ft(0)

a, = 7 f"(0)
a3

r75- ft"(0).

180

1,9
a

2-11

Now we express

in terms bf powers of

x - a:
4

(2)

k 7./20 + bi(x - a) + b2(x - a)2 + b (x - a)3.


3

Such an expresioll for


2-2.

can be Pound by synthetic division as in Section

We can sholfthat the coefficients

b0, b

b2,

and

1,

'b

are given by

-b

= f(ael.

bjc- ft(a)
(3)

f"(a)

alb

V
To show that

= f(a),

--1

fom(a).

2x 3

we let

x = a

in the expression for

f(x)

to

obtain

f(a) =b0 + bi(a- a) + b2(a

(a - a)3 =.bo

a)2 4-1)
3

The remaining results (3) are almost as easy.

We'differentiate/f

to obtain

the derii/ative
A
,

a'

2b 2(x - a). + 3))

a). 2 ,

'kx

'

whence
ft(a) = b
'Differentiating -f'

+ 2b (a - a) + 3b (a - a)

= b

1,

we obtain

ftt
Therefore,

.3b (x

4have

,r(p) =

f'

"

(a) .

Another differentiation
't.

gives'

f'": x --)

3b3
r

so tht
fni(a) = 2 .3b
7

3-

int(

and hence 'b


3

1
wt 1

'2-11

Using the foregoing proces for a fourth degr e polynomial function


.111.

x -' b

+ b

- a) + b (x

a)3 + b (x -

b (x

we could obtain
10
b
b

0
1

(4)
'2
b
3

b
4

whereftf"
f*.

of

= fea)

= ft(a)
Vi(a)

2 ,
=

1
2 X 3

fm(a)

" 1

2 >-

is the fourth derivative of

fins (a)

f;

tihglt is,

ft"t is.the d#rivative

It is common to write
;

(4)
,

,'
for ihe fourth deritrative of

f;

rather then f"


similarly we use the notation

...,
.

for the fifth derivative, the sixth derAVative41


,
etc.
'

*ea'

..

....*

k!= 1 x,2 x 3 x'4 x ... x k'


;-)

with the convention that

0! = 1.

i
0

N
..-

Our results can be generalized:


x

(5)

(6)

A polynomial functidri can be

bo + b1(x - a) + b2(x - a)2 +

where.

and

(6 )

It is also Common

to use the factorial notation

(5)

.10

!..

'

1 f(k)(a),

k!

2-11's
/

Example
x + 1.

EXpress

f ;

in terms of powers of

1 .

- (-1)7

-2 + 28x3,

f'

f"

x --)84x2,

f:(4):

x = -1;

We have the successive derivatives:

fen

Since

3 - 2x + 7x

168x,

X -4 168.

we need to find the values of these functions when

we have

f(-1)

fl(-1)

12;

b0 =

-30;

b1 =

84;

b2 =

f"(-1)'f

0!
1

1,

X,(-30)

-30

x 84...42

ft"(-1) = - 168; b3 =

J4)(-1) = 168; b4 =

X 12 x- 12 ,

x (-168) ._-28
1.-

x (168)

= 7

Thus, we can write


,

;re-t 12 - '30(x + 1) + 42(x + 1)

cs,

- 28(x + 1)3 +-Y6c-r 1)4.

The same result can, of course, be obtained by synthetic division. .(See


Exercises 2-11, No 6.)
,0000!%.

ti

183

193

2-11
Cs.

Ecercises 2-1a:
fi
1.

hai/e?.

2.

degree,pdlynomial functibn'

n-th

How many nonzero derivatives can an

If we write a fifth degree polynomial function in the form


f

x b0

then

+ b (x- - a) + b (x-- a)2 + b bc6- a)3-+ b 4 (x - 04


1
3

= k

'b

What is the value-of

f(5)(a)

a)5,
5

k?

5
3.
e,

and again consider the nine-.

We.rel:leat Tertsof Number 11 of, Exercises


tion

We submit a table to show three successive synl

f : x -+X 3 , 3x.

.and resulting quot ients by,

f(x)

thetic Divisions of

73

'2

1-

-1

l'

1.
8
9

(a)

Determine

g(x)" and

f(2)

if

f(x) = (x,
(b)

Determine

p(x) and

g(2)

g(x) = (x
(c)

Determine

p(2)

and

q(x)

2)g(x) + f(2).
if

2)p(x) +.g(2):
if

p(x) = (x - 2)q(x) + p(2).

(d)
(e)

What is

q(2)?

Show that, for all

x,

we can write

f(x)'= (x - 2)((x - 2)((x - 2)q(2) + p(2)] + g(2)) + f(2).


(f)

Using the results of parts (a) through (e) of this problem determine
A, B,.C,

and

if, for all

x,
1

f(x)= x3 - 3x = A(x - 2)3 t B(x - 2)2 + C(x - 2) + D. .


(g)

Find the first, second and third derivatives of

-x3 73x.

'1

(h)

TI,vluate

(1)

Evaluate

f(2), f'.(2), f"(2), :and


'f(2)

0'.

(.3)

f'(2)

'

f"1(2).
ftst(2)

f"(2)

and

.3!

'

Compare the results, cir- parti (f) and (i)

a
.

14.,,'Consider the functions


4
G.

.- 3x

x3

42

-42 + 9(x - 2)

x -42

-,-.9tx

x -4

+ 9(x

- 2)

+ 6(x

+ 6(x

2)2/

2)2

+ 1:(x

2)3.

(a)

Find the value of each of these functions when

(b)

What, quadratic function best ,repreents the cubic function


G

(c)

x3

3x.

near the 'point where

What fiction as)

x = 2?

).

as the best linear approximation to

near

=2?.

the point where


"(d)

x = 2.1.

What function se4yes as the best quadratic appr.oximation to

near the' point : where

(e)

x = -1?

What .function serves as the best quadratic approximation to

near the point where x= a?

5. F.'ind

the first four derivatives of each of the following functions

(O.

(C)

x.

xx
1

,5

3
#

7.!

x4

X2

4,x

14

9
xx 11

+ 9!

x6

11?

x10
.e

6.

.(a)

Compile a table similar' to Number 3 toindicate four successive'


4
synthetic divisions -of 7x - 2x + 3 by x + 1.
00-

Us-e. the table of part a) to. write-

Of' powers'of

+ 1.

47x4

- 2x

+3

in terms

Compare your result with the result of

Example 2=11ab.

.185

195

w.

.4
,
Write the functions which give the best lineal:, quadratic and cubic
k
7
apprdximatiohs to the'graph of f : x -Ox - 2x + 3 near the point

(c)
.

.,

(-,1,12).

(d)

us

7.

..

At the,Point

(-1,12). is-th6 graph, of

the graph.of

Is

rising or falling?

f/tpted upward or downward near the point .(,-1,1_2)?

Show that the third derivative of the function


2
f

x -->ax

-'
/

p 0

+ bx t c,

is the. zero function,.


r

'
.90

tb

196

is

Chapter 3
4

CIRCULAR FUNCTIONS

Unlike the polynomial functions we have considered in the first two


41apters certain functions have the property that their function values repdat
themselves in the same order at regular intervals over the domain.

having this property are called periodic.

Functions

Included in this important class

are the circular (,trigonomerc) functions.


0

The simplest'periodia motion is that of.a wheel rotating on its axle.

Each complete turn of the 4heel brings it back tp the position it held at the
beginning. After a point of the,,wheel traverses a certain distance in its
path about the axle, it returns to its initial position and retraces its
course again.

The di

ante traversed by the point to a complete cycle o1' lts

motion is aga' a period, a period measured in units of length instead of

unity of time.,

If it should hipPenthat equal lengths are traversed in equal

times, the Fotidn becomes periodic in time as well and the wheel can be used
as a a1 ock .4
The model of a wheel rotating provides a basis for our definitions of
the sine and cosine functions, whose values are defined as the second and
first coordinates', respectively, of points on a circle of radlus one.

'

These.

definitions are compatible with those of ratios of sides of right triangles.


By defining the sine and cosine, functions 1.n terms of a unit circle, their
periodicity is immediately evident'.

Furthermore, we can use the geometric

properties of circlesto obtain the properties and graphs of these circular.

functions (Sections 3-1, 3-2, 3-3).


These definitions and'results are applied to uniform circular motions
.

('such as rotating wheels) in Section 334.d The basic addition formulas are
e

deriv4 in'Section 3-5, again by making use of the geometry of circles.


.

These

'Rre
applied in the next section to the study 6f pure_weves, the simplest type
.
..

vof periodic motion, 'While the final section points toward some of the ways.

114044 lle. lrcular functions can be used to analyze more general periodic
phenomena*.

se.

3:87
.

tg

197

Sine and Cosine Functions

3-1.

We assume that.you have0some familiarity with the sine and cosine functions, so that much of the material in this chapter is review.

You may have

previously defined-

ese functions in terms of ratios of sides 9y right tri-p

angles.

nstead, to define the sine and cosine as functions of arc

We prefe

length on a circle.

The usual angular definitions in terms, of degree measure-

.ment can be obtained frot our dvfinitiou by a suitable change of scale.


.definitions, in terms of the circle, have two great atyantages:

. Our

first, we

can easily read off many properties of sine and cosine from pr,,pperties, of the
circle;

second and more important, our choice of scale will simplify our

dtZferentiatiOn formulas.

For convenience of definition we use the Aide With center at the


1;, the unit circle whose equation is

origin and radius

2
u-

+.17

m
The circumference, of the unit circle is

= 1..
21t

units.

For any real number

We measure 'x

units around this circle heginnif at the point


(1,0). If x
4
is positiv4 we meaeure i-na counterclockwise direction and if x is negative
we measure in a clockwise direction.
coordinatts

nate of P

(u,v)

We obtain in this way a 'paint

on the circle given by

is called the cosine of

'is called the sine of

x.

u, +

= I.

P with

The first coordi-

while the second coordinate of P

(See Figure 3-1a,and 3 -lb.)

14

Note that x > 0' and we


measure in a counterclock:Wise direction, obtaining
with
= cos 8,
P(11,v)
v = ain 3

ii

ylgore 3-la

14.

40 3'

3-1'

Note that x.< b, arid we '


measure in a clockwise
'direction, .obtaining P(u,v)
with u = cos(rl), v = sin(-1).

Figure 3-lb
Two functions, cosine and sine (abbreviated

cos sand

sin), are defined

as follows:,

cos

: x

sin

= cos x = the first coordinate (abscissa) of

(i)

The values of

= sin x = the second coordinate (ordinate) of


cos

sin

and

P.

are easily obtained in certain cases:


-

,For example, referring tQ Figure 3 -lc, we see that'since

P' is the point

(1.,0),,, we have, by definition

cos 0 = 1

and .sin 0 = O.

Since the unit circle has-circuldiference

2v

vf4110

units we can measure

2v

units

around (in either'direction) to obtain again the point P. of Figure 3-1.


Thus
cos 2v =' cos(-2v) ='1.
sin 2,

= sin( -2n)' = 0.

'74.!

This point
has coordinates
'(l4)
(1,0)

To meahlre 2v units
around fran- P, returns
us to P.

Figure 3-1c

189

199

3-1.

To traverse one-fourth of the way around, the unit circle is to thrive


,

thro'gn

2y
2

x = - we'have

units. -Thus-if

'

with coordinates

(0,1),

at

so

a
Ise

cos

x =

if

- then we get

with coordinates- (0,-1),

P2

cosy(*

(See Figure371d.)

sin

= 0, sin(-

sO that

T -1.

A Figure 3-1d
d

Further calculations are indibated at the end of this sectibn a d in


the exercisei.

0 .
1

The sine and dosine are often defined in terms of ratios of sides of
right triangles.

In Figure 3 -le, the sine and cosine of angle' AOB

defined by

AB

sin LAOB 2

opposite
hypotenuse

cos LAOB'-

ad acen
_,OB'
OA
hypoten se

OA

(2)

hypotenuse
opposite

adjaceht

.4,

Figure 3-le

are

3-1
44.

To relate these definitions to our,YiarlierIones, we can place the

u and

axed as shown in Figure 3-1f, letting ,.x

denote the.distance along the circle

trom:,R(1,O)' to

P' are

P.

The coordinates of

(cos x, sin x).

Figure 3-1f
Whether

OA > OP

(as shown)or

OA < or,

ccsr-x = OQ a

we(have, by similar triangles that

.0B

cos aoB

and

AB
OA
.

Ms the angle

AOB

LAOB.
f

corresponds to an arc of length

and

cos x

and

are respectively

sin x

cos LAOB

and

sin LAOB.

The right,tNengle definitIonS are somewhat restrictive as the angle

AOB must always be between the zero angle and aright anglerthat-is, the
arc length -x- must be between

and

such restriction and ertle us ta define


number

x.

Our definitions (1) involve no


in x

and

cos x

for itU real

Thus (1) gives us anaaturain extension of the definitions (2).


t

1..

3-1

"N.

Angular Measure
It is also common practice to measure angles in degrees`.;

is established by dividing the circle into


angle

Degree measure

equal units, measuring an

360

AOB 'f/ the number of units of arc it includes. For example, if

includes

1
;.

LAOB

of the circumference wt would say that the angle measures

X 3600 = 60

We can also measure angles by grs length.

Note that LAOB determines


the arc length x units
(on the unit circle).
(1,0)

ti

Figure 3-1g
In Figure 3-1g angle
.

+ v

= 1.

cuts off an arc of length

We say that

a measures

units.

x -on the circle given by


Thit type of measure is

Called radian measure, its unit being called .Qkradian.

illustrate an angle of

.1

Figure 3-1h we

radian.

).

No 'measures

radian.',-

oMINowli

Figure 3-1h
192

202

A moment's'''tchaug

measure.

tndi-cates the relationship between radian and degree

Cleeti.ly, if the radian measure of an angle is doubled, the degree

measure also.doubles.

The same result is Qbviously true for halving, tripling,

etc:

In general, we have that the degree measure M

,y directly croporticV1 to the radian measure

of an angle a is

Thus.

M = kx
."'

where

Since M= 360

is constant:

when

x = 2n

we have

;360 = k(2n)
l70

ao we must have k = -77 .

Thus

It

We thus

180x

radials corresponds to

(3)

degrees.

cc that degree'meksureis obtained from radian measure by changinE

scale.
.

We note the following cqnsequences.of (3).


'

(4)

'

180 ,

--- a 57.296

radian corresponds to

lib
degrees
r"

degree corresponds 'Co I7 = 0.01745

11

radians.

In working with radian Mea,Sure,,it is customary simply to- write, for

example,

S.

When we mean 3- ,radians.

use the degree symbol, such As

Example 3We see 't

11

corresponds to

we measure

radians.

745 X'990

90 0, 45 0, etc.

sin 00.

Evaluate

t ,990

11

With degree musipc we shall always

,units around the unit circlq in

a counterclockwise direction.

= (27t) '47

If we

3a
2

we indicate two times arbund the circle

urn, suggesting arritpl at

plus a

the point
Thus

(0,-1).' (See Figure 3-10

sin 9900

'.

ni

3-1
-

Example 3 -lb.

If

n'

is any multiple of

determine 'COS x

and

sin x.

3- '

Solhion.
to an angle, of

OQP

The arclength 60 .

7 th

of a circle and hence corresponds


:.

At Figure 3-ij shows we thus know that_ angles

are equal and hence that

- and

has length

OR

PR

ioQ

and

herd length.

p (_._

it

p (.1

1 2'

-1,0)

- 12
(

2' 2

Figuge 3-1k

Figure 3-1J

s'fb.?

Thus

P (1,0)

has coordinates's,'

cos

Measurements of

2n

TY

3n

.a

1.

4n as

and An

,;

f 7 f QJ f

give the respective points

P2

P3, P4,

of Figure 3-1k., The coordinates of thqse points are easily.found using


6.
the same techniques as above. This gives us enough information to find nos x

13

5)

**and sinx. for

n.

x any integer multiple of

find

for example, if we wish to

cos(-111),

we obsel..ve that

.,.

.-

-1n
-

57(

3'

We measu
point

P2.

in 4,pinckiwise d4eCtion first' 5y

then, - units to obtain the


3

Thus

cos(

1.6fn

-)

1
,

the first coordinate (abscissa) of

2.

--Throughout our'discussion we make use of the facts that.

f. ( 9

sin(x - 2n} = sin 'x


cos. (x - 2n) = cos x

The general'form is stated

(5) inSectiod 3-2.

298 4
J
I;

3-1
A

Exercises 3-la
1.

Change the following radian measure to degree measure.


(a)

8g

2.1-r

(g)

18g

(b)

(c)

(h)

-27t

(a);
-

(e)

4
2v

(r)

'5.;1-JEA-Aela.following degree measure to radian measure.


(a)

270-

'(b)

.30o

(g)

810-.

(h) :190

(C)

135 f

(a)

(i)

18

la

{-3)

o.4

60:

195

(k)

1620

(f)

-105

(2)

180

Express the following radian measure, in terms of the smallest positive


angle:'
(a)

What'is the sula of the measures of'the angles of aAriangle?,'

of a'rectangle?
0

(b)

Given:'4'a polygon of

of the interior angles?


(c)

What is the sum of the measure

of the exterior angles?

The smaller of the two angleS between the hands of a clOck at:
11:30

(d)

sides.

has a-measure of

Over which part of a radiate. does the mi

in. 15 'minutes?

in

25

to hand of a clock move

minutes?'
.e

(e)

itoW many radians does the minute-hand

sweep out in 1

2 ,

in

hrs.

50. Min.?
e

195'

205
it

.,

hours?

3-1

44

Give the coordinates of the point on_the'unit circle corresponding to


$

3000

'(c)

.(a)

A
(b)

12000

15g

(d)

Express each of the following angles in terms of a positive angle

5.
.

(a)

0 and

between

radians.

2g

16

13 g
4

(b)

7 g

11

17

(d)

3n

(--,-/.

6.

'W4te two equivalent expressions for each of the following angles in

o4

terms of
(i)

(iii)

n (2g)' +

integer,

lal < 2g

n (g)

+a

integer;

kd <

n (2)

+a

integer,

jai <
2

28g
241v.

'5

In
,
7..

1..

(a)

2gg

12
lN

:Extending the information readily

available from the

30-60-90P

tri6ngle in Figure 3-,13, find


cOs x

and

sin x

for

x,

by drawing a

multiple of
2

unit clrcle similar to the one

to the Fight and labelling the


a,

-coordinates
For

(cos x, sin x).

91, Q2,

9i2

(similar

to,ngure 3- lk)."

O.

'eco

(b)

Which of these points duplicate multiples of

IL

in Figure 3-1k?

3
(c)

.,

Whiich of the points

multiplies of

Qi, Q2,

9.

.44

196

.2013

411.2

have coordinates for thee

3-1

Since

(d)

is midway between

and

you can deduce

conjecture whether or not

cis T by averaging the value of

cos ;

and

cos

y
?

' Can you make a general statement about reading off values of such
-A
functions fromthe drawing
it

91

it

cos

41 &)S
3

4 -,
ta)

Using re'ationships betWeen the


sides of a
find

45-45-90

cos x

sin x

and

a multiple of

1.d

triangle,
for

x,-

by drawing

a unit circle similar to the one


to the right and labelling the

cc.

(cos x, sin x)

coordinates k

,Ki, K2,

for

eee, K8.

(b)'

'ro4 this circle read off,the following values:


(,1)

sin 4:1

(2)

cos

1-1-

(3). sin 41
JAI. _cos

(5)

(6)

cos 135

(7)

sin 315

(8)

-cos (-225) .

72

sin ( - 21f-).

' (9) -sin (-135)%.


(10)

it 0 7

cos (3 .360-+ 45)

./

3-1
a

Using the coordinates of the point


indicated on the unit circle to t
right answer the following:.
(a)

Find the value of


sin

&51

sin

sin

cos

COS

COS

v
;

14-

v.
.

Wat is the relationship


between
.and

sin
ir

sin -5

sin

v
IT

cos

V
.

and

v
cos ;.1?

and

cos

&/-..

In this first qpadrant, what is the sign of the sine? of the cosine?
(b).

Find the value of


of

sin

5v

of

sin iT,,

an1_ d_of

sin 7 ;

of

cos

and,of

cos'--

5n

cos 7-6-

2v

3n

,-3n
7

2v
3

In this second quadrant, what is the sign of the sine?

of the

cosine?-

Find the value of

sin

7v

of

0.

5v

sin --17
4

andof

and of. cosIT

sin T ;

of,:

cos

of

4v

cos

In this third quadrant,, what is_thesighiofthe sing?


cosine?
(d)

ofithe

Find the valde of


of

11v
sin -g- ,

cos

11v
,

of

sin

of

cos

7v
,,,

7n/

5v

and of

sin

and of

cos --i-..

t. 57(
',

In this fourth'quadrant, what is the sign of the sine?, of the


cosine?

198-

203

3-1

(e)

(i)

In which quadrants Wsi/le positive?

(ii)

In which quadrants is cosine positive?

(iii)

In which quadrant is,shie persitive and co'ine negative?

sine n gative?

sine nega.tive. and cbtine positive? ,bot

cos ne negative?

positive?

both

negative?
10.

(a)

What are the coordinates of

P,

indicated onthe,cii:cle to the


right, if the circle has a radius

of

(b)

l?' 2?

R?

1:47::T?

What are the coordinates'pf' T,1

indicated on the circle to the


right, if the arc measure is
and the radiug is

2?

7?

11. 'Given a circle of radius 1:

An arc which measures


length

radian has

radians

radian has

1;

.an 9, which measures


has length

x?

Given a circle of:Adius


An arc which measures
length

R;

an arc which measures x, radians f


has length
(a)
T

xR.

Shona by similar triangles that the length of the arc is proportional


C

to the measure of tp,earp-,.,and the constant of proportionality is

the radius, or

s = Rx

9V 9

3-1

Hi- The minute hand of a clock is


far does its tip travel,in ''1.5

(c)"A Circle has a radius of


measures

60 ?

inches long:

inches.

15

Approximately ho\

minutes?

How long is the arc which

72 0?

(d)1 What, is the radius of the

circle to the right if the

AB

measure of

is

and the length of


A
(1)

AB

is

in.

in.

(ii)
3

(iii)

10 in.

(iv)

(v)
(e)

3x .in.

What is"the radius of a circle if the measure of


the length

(f)

in.

of AI

x?

is

If an arc of length

2x?

12.

is

x,

lOx?

has a measure of

of an arc of a semi-circle?

Al,

14 ,

what is the le gth

of one-third of the circumference

From geometry we know that.ip any circle the areas, of tw

sectors 0

a,

circle are proportional to the measures of their arcs; for example:

lea 'sector AOB


Area sector AOQ

x
'

AOB =

i.e., -Area sector

2x

R x

Ir

Li
C

2002:14),

..
, r

3-1

(a)

This formula can be re-stated in this way:


In a given circle the area of a sector
AOB = constant x arc measure

AB

the constant of proportionality,

.1

or

A = kx,

k,

is

where

that'is,

2
ac

What is this constant of proportionality for a circle, of radius

2? 4?
(b)

1?

Ih a circle of radius

a,

if a given arc has axeasure

m,

what is

the area of the sector?


(c), In a given circle, how is the area of a sector 'affected if the arc
'

(d)

measure is doubled?

halved?

tripled?

In the beginning of this problefwe stated that the area df sector

AOB

is proportional to the arc measure.

AOB

sector

Obviously, the area of

is also proportional to the arc.length.

What is this

constant of proportionality?
(e)

Wha4.is.the area of a sector of a circle of radius


the'arc of the sector is

(f)

(0

inches long?

tripled?

Illustrate (6) geometrically; that is, show

sin (x -2n) = sin x


cos (x (b)

inches if

How is the area of the sector affected-if the arc length is doubled?
halved?

13.

12

18

= cosix.

ShoL that
sin (x . +"2n

= Sin x

cos (x + 2n 7r) = cos x

where

an integer.

20r-

-4-

4.1

The Use,S Tables


In a separate bookletewe give tables of approXimate values of

sin x' for decimal values A

and

up to

,x

1.60

cos x

which.is slightly more

than

(The number

x,

of'course, represents the measure of

on the unit mircle, iie., radian measure.)

arc length

More complete tables: tables in

terms of degree measure and tables for converting from radian to degree mea.

aUre are also foundvin the Booklet of Tables.

t,
The ,following,examplesindicate some of the ways of using these tables:

Example" 3 -lc.

Find, sin .73

and

cos .73.

We simplyrel;d from the tables the values


sin, .73

.6669

cos .73 0 .7452.

Exarti le 3-1d.

sin 6.97

Find

cos 6.97.

and

e our tables do not include

6,97,

we do knew that
t

sin x.=sin(x. - 2v)

and

cos x = cos(x - 2v).


.

Using

6.28,

'11e. have

sin 6.97 p sin .69 0 .6365.


and

cos .69 0 .7712.

cos 6.97

Exam

e 3-1e.

Find

41

sin
,

Using

v ss 3.142

we have
v

ei

0 .524.

The tables t Lve


.tz

sin .52:0 .4969


sin .53 0 .5055
,,Interpolatin , we obtain

sin .524'0 .4969 + 10 (.5055 - .4969).

202

.2 i 2 d;

3-1

.5003.

sin

Therefore,

Oficourse, we can observe'that


'1i (radians) corresponds to

and.read the result

30 0

tin i

_1
- 2.

from Figure 3 -12.

.3
'4

Example 3 -1-f.

Find

x.--r..1<-1 that

0 < x <

and

sin x

.885o.

From the, tables we see that

sin 1.08 ;4

.8820

and

sin 1:09 Ps .8866.

Interpolating we get
x

Example 3-1E.

Find

1.08 +

30 ,
(.01)

1.0865

sin 2:

Referring toFigure 3 -lm we see that


1

14,7!.

sin `2 = sin (It - 2)


\I\
t

q sin (3.14 - 2)

= sin 1.14
.9086

'Figyre 3 -lm

203
.

3-1

sin 100.

Find

sin 100 is,located

.To find where the

on the unit crcle,we ask how Many times


2v

divides

7=
"

guess that

2v p_6.21g3

Since

100.
100

In, fact'

lo.

16 x 6.283 m 100.528,.

16 x 2v

100 = l6 x 2v - .53.
P

-that point

so that

We show (Figure 371n)

sin
100'

units around the

100

is

we

Ptcos 100,sin :00)

unit circle (counterclockwise) from ,(1f0)


or

.53

units short of

16

-653.

revolutions.

The tables give


sin.:53'm :5055
'Figure 3-2n
4

in 100 m -sin .53 m

Find

Example 3-1i.

Since there are

plus

.2.p

We have

in one revolution,we write

Five and one-half counterclockwise revolutions


,.
gives a point on the unit circle
20o - into the third quadrant..

2000 = 5 x.360 +
0

360

cos 2000. -

;-

x 360 + 20.

o.

cos 2000 = cos 200,= cos(180 + 20) =,-cos 200.

We use the ;-'

table of No. 5 of Exercises 371b to find that

-cos 20 m -.914C).

radian tables we first note that '20, corresponds to

To use our

.35. (approximately)

s6 that

cos 2000

-0.9394.

.,,

o-

o.

204

21,4
e

'3-1
Exercises 3-lb
For problems 14 2, 3, 4

Find sin x

1.

(a)

0.73

(b)

-5.17

when

is equal to

x
41

(c)

1.55

(d)

6.97

2y .1 6.28)

(Hint:

and, .

(a)

sin x

0.i098

(c)

sin x

(b)

cos x a 0.9131

(d)

cbs'x v. 0.5403

Using

3.

cos x

x ,when 8 < x <

Find

2.

and

use Table 3 in the Booklet of Tables

0.6518

314, .approximate the following, interpolating where

necessary.

11

(a)

s;in

-37

(c)

sirf 11.5'

(b)

cos :IT
5

(d)

cos 417

Find' x

where

0.< x-< 1.57


4.

(a)

sin x = 0.2231

--,-(c)

sin x = .8714

Cb)

cos x = 0.7135

(d)

cos x =..1759
.

,-:.

5.

Bela/ is a table giying values of

in degrees, .Sin'x

and

dos x

sin x

Pi

and

90

cos x

for angles between

read from the top and left, sin x and


45

and

cos x

are read froi the bottom and right.

hen x
0

and

is given
459,

are

for angles between

For example,

20 = cos 70p = 0.342.

4
4444016,

2sinc'x

0
COS' x

00.

50
10
150

20

Yr

0.000

1.000

0.087

0996

0.174-

0.985

0.259.

0.966

0.342

0.940
":"

25.

0.423

300

0.500

35

0.574

40

0.643-

0.766

45

0.707

0.707

''01.90e

o.866

vt
Using.4the table above find the valiteof the following:

6.

(a)

sin 75

(c)

sin .!;80?

(b)

cos 140

(d)

cos( -460)"

Using the table itTNumbCr 5, find two values for

in degrees

0 < x < 360.


(a)

lin x = 0.574

(b)

cos x = 0.643.

sin x = -0.8194

(0I

(d)\ cos x =*-0.087.


1

A/

\
\

4
.4

a
11.

2'06

ro

VI*

!%
3-2

'3-2.

Properties of the Circular Functions


We have defined the circular functions,

arc length along the unit circle

+ v

cosiie

= 1.

by-measuring

In this section we der've

perties.

v
....,

The valu 11111cos x

sine,

Many properties of these two

functions are'easily derived 'from this &efinition.

a few of these

and

and

P(cos x, sin x)

sin x

were defined as the coordinptes (cos x,


sin.x).:0 a point
u

,%2',.
-1-1,

. 1

on 'the circle

such as in Figure 3-2a.

Therefore, the coordina,tes.of

must
u

satisfy this equation, that is:

(1)

.cos

+ sin. x =1

4
This identity will often be useful.
We have followed the usual convention
'of writing
(cos x)

2 '
,

cos
in

Figure 5-2S

x
x

rathen.then
rather than

(sin x)2.

Since a square is never negative it f011ows thatCOS*.2AX

< cos2

x,

NI- sin

and

sin

x < cos

x + sin

x.

Combining j.hese with (1.) givesfthe two inequalities

. cos

x < 1

and

sin2 x < 1,
Ns

which can be rewritten as

-1 < cos x <1 and

(2)'

-1 < sinex < 1.

Another_covequence of (1) which will be -useful in our approximation


disddsslons in the next chapter is the inequality
1%40.

(3)

0 < 1 - cos x <


4

2c2

x2
2

-e

3-2
o

*4,

To establish (3) we uses the familia_r distance, formgla to get (in Figure

3-2b) the distance from 15 to Q:


'

sin2

- cos /C)

P(cos x, sin .x)

As.

....

Figore 3-2b
'''.3."`

...(

.;

s,rjee,the short e4distance Letween two


.
Thus
points is measure -along the straight i&ne joining tl4kit.4
A..
This distance can li-t exceed

lx10,

'

Scluating

/(1 1 cos x)

'

- sin2 x < lx)


cos x),

and th en multiplying out

/- t

4.

gives;

2
2
2 cos x 4- cos x 4- sin x < x2,

and .hLnce (1) gives:


.

-" 2 cos'
' x < x2 .
2%-

4111* 0.41400

Di

0< 1

2
2, we get 1 - cos x < 7 ., jioting tkt cos x <
xviding

by

cos x, we complete the proof of


.

(3)

..

X2

0 < 1 -- CO s X <
-7 2

2082

s8

and hence

--,

auk
...

3-2
Periodicity and.Related Results
t.

Thee are several formulas which


relate the values

sin x

at different points.

and

'

cos x

For example, if

we traverse the unit ..circle #2g,

units,

we.arrive at our initial position, since


the circle

+ v

= 1

(See Figure 3-2c)

2g.

has circumference
Thus we have

sin (x + 2g) = sin x


'(4)

cos (x + 21) = cos

Figure 3-2c
-Functions which repeat their values at equal intervals are said to be
periodic.

In general, if there is a number, a > 0

stichthat
o

f(x + a) = f(x)c for all


then we say that
and_ cos

f' is periodic with period

are periodic with period


sin (x + 4rt

'21c.

a.

x?

Thus the functions

sin

As consequences of (4) we have

= sin ((x + 2g) + 2g)

sin (x + 2g)

=sin x
and

.
sin (x - 2g) = sin ((x - 2g) +'2d)
= sin x
.In-fect, for.-any'integer

nl we can make the general statements

4:7
sin (x + 2ng) = sin x

(5)
cos (x + 2nd) = cos x.

Other useful formulas can be "read qff" from the properties of the unit
circle given by

+ v

,
,

= 1.

For example, the points'

are symmetric with respect to the u-axis.

Apnd *(u,-v)

Consequently, we'see (Figure 3-2d)

that

cos (-x) = cos x


sin .(.J1) = -sfn x

209

(u,v)

3-2

has coordinates
(cos X, sin x)

Q has coordinates
(cos(-x), sin(-x))

Figure 3-2d
Using the unit circle we can also derive the two familiar formulas:
,g

sin(- - x) = cos' x
2

(7):

cos(1 - x) = sin x
2

It Figure 3-2e triangle


congruent to tr angle
Then

P(u,v) an

related so that

OQS."

Q(u1,v1)

U = v1

and

OPR

is

(Why?)

are

y= 91

It follows that
cos x = Oi = QS 7sin (i - x)
and

sin x = PR = OS = cos (i - x)s.

'

'210

229

Ji

3-2
The useful formula

cos x = sin (x +

.
Q

dan also be derived,by geometric arguments using the unit circle.

Here we

,derive it Using (6) and (7), as'foll9ws

cos x = cos(-x) = sin

(- ;-x)]

= sin (x +

We have given but a sample of the relationships which can be derived' from
the unit circle.

Other such results will be derived in the exercises and, as

we naed them, in SeCtions 3-5and

ti

221
211

3-2

, Exercises 3-2
1.

Using

f(x + 2nsc) = f(x)',.

f(31)

(a)

and

x cos x,

.1

find

f(?-567-1)

(e)

f(171c).

(f)

f(_ 121.c.)

2; If
3.

find the values of t in Exercise 1 above.

xj

'f

For what values of

(if'any) will

(a)

sin x = cos x?

(b)

sin x = -cos x?

(c)

sin x = sin (-x).?

(a)

cos x = cos (-x)?

0,,m401101(

USe the fact that

Hint:

(cos x,. sin x)

repi.esents a point on the

unit circle.
4.

(a)

of

cos e

are reciproc als

csc e

sin e, 'respectively, show that the expression

and

sec e
sec e - csc e
(b)

sec 6 and

Using only the definition that

is identically equal to

sin

sin e - cos e.'

Adding to the definitions of part "(a) the definitions that

tan e

A#
is

(1)

sin e
cos 9

and .cot e

tan e + sec e
sin e c6t.8

can be changed

1 + sin e
cos

show that
to

tan 9.
4

show that the expression


to

(ii)

is the reciprocal of

1 + cot e
csc e

1 + tan e
sec e

and

can both be changed

sin e + cos e ; and

show that

sin e csc e,

are all equal to

cos 8 sec e,

and

tan 9 cot e

1.

212

222
a

3'-2

5.

(a)

Formula (1)

sin

e + cos

e = 1- can be used to an advantage in

changing the form. of many trigonometric expressions. As_ing -this

relationship,
show thdt

cos2 e

sin2

can be written as

2 cos

- 1

or _1- - 2 sing 0;
(ii)

show that

(iii)

shOW that'

2 csc

tan e-+ cot e

-can be expressed as- csc e


cos 0
-

1 - cos e

can be expressed as

1 +- cos e

and

0;

(iv)

show thEZt'Qth

cot

csc e

and

cos 61

are both equivalent to

sin
(b)

is simple to prove

sec 8 - cos e

sin e cot e = cos e

and

cos e tan e = sin e.

(Why?). With these relationships, those of Exercise '(b)(iii), and


others developed earlier, prove the following:

6.

(a)

(i)

(1 -;lin2 0)seC2 e = 1

(ii)

(1 - cos2 e)csc2 e = 1

(iii)

coe e (1.= cost e) = cos2 e

(iv)

eq2 e (1

cos2 e) = tang e

Starting
with the relationship
.
cally that

(b)

1 + tan

e = sec

sin

e + cos

2
e.

By considering iAT and.


the unit circle to the right,
prove geometrically that
1 + tang e = sec

DW

213

223

e = 1,

prove analyti-

3-2

(c)

analytically that
(d)

1 + cot

proVe

sing e + cos2 e = 1,
2

Startitlg with the' relationship,

e = csc

e.

Adding the relationships posed as problems in parts a, b, c

to the

earlier ones discussed.


show that

(ii)

establish that
sec

(iii)

show that

identically
ct

1;

+ csc

is equivalent to

in two ways; and

sin

e kl
(1 + cot

9) + cos

Okl + tan

2
e)

is

2.,

sin(x + h)

- sin x < PQ.

From this restlt prove that


Isii(x + h)

Using the figure to the right


prove

(b)

'tan e
cot e

sec

e csc 2 0

always
(a)

sec e
cos 0

(1)

- sin(x) <1h1 .

.Again, using the, figure to the

right prire that


'cos( x + h)

P(cos x, sin(x)

cos( x) I < Ihl

and

are

Q(cos(x +7), sin(x + y))

indicated on the drawing to the

right.
(a)

By the use of similar tri %ngles,

read off the coordinates of

Q;

i.e., prove

cos(x + y) = -cos x,

sin(x + y) =
(b)

and

x.

Similarly, prove
cos(y - x) = -cos x,
sinfy

x) = sin x.

214

224

and

3-2
9.

(a)

Using the figure to the right


read off the coordinates of

to show that

Q'

cos(x

= -cos x,

sin(x
(b)

y) = -(sin

and

x.

Use formula (6) to extend the


results of (a) 'to show that
(i)

cos(x - y) = cos(y - x) = -cos x,

and
(ii)
10.

sin(x - 1)-'= - sin(x - x). = -sin x.

Read off the coordinates of

to show

cos(x + 71).= -sin x,


2

sin(x +

,cos x.

f4

11.

Using the relationships (6)


(7)

(a)

prove

(b)

prove
(i)

- x) = cos x,

cos(-x) = cos x, sin(-x) = -sin x;

cos(

- x) = sin x,

cos(x+ 2) = -sin x;

cos(x + y) = -cos

and
(ti) sin(x + y) = -sin x;
(c)

prove
(i)

cos(x

411

y) = cosky,- x)

-cos x

(ii) sin(5V y).. -sin(x - x)


)

P3-5

225

-sin x

and.

----1- 2.---The-1 equality

phis- /section.

0 < 1 - cos x

< 7- was established as formula

By numerical substitution of various values of

(3) in
x,

let

us n w investigate this relationship

Using the table which gives the cosine of angles expressed in radian
measure, complete the following table.

X
1 - cos x

cos x

(in radians)

0.1

0.15

0,36
0.5

'

0.6

Q.7

0.8

0:9
1.0

1,42

1.5
2

6
i

(b) ,From the completed table, conjecture for which values of

this

inequality is most usefUl.


13.

We know that the functions

cos x -and

Find the period of the functions


(a)

sin 2x

(12)

sin --x
2

(c)

cos 4x

(d)

cos 2.x

Show thqt the functions


than

2v.

2v.

x -4

14.

x )sinx have period

sine

and

cosine

216

226

have no positive period less

Ars,
3-3

3-3.

Graphs of the -Circu

r Functions

The sine and cosine functions have been defined in terms of


arc length
on the unit circle give
functions, the graph

by- ua,+ v
f these f

1.

As was the case for polynomial

tions prOvide another geometric device

for understanding their behavior.

At,this point for nonpolynomial functions

our primry procedure for graphing is the plotting of points.

Fortunately
we can make use of the results of the previous section to simplify pur procedures.

We first plot some points for


(1)

.y = sin x,

Tale" 3-3a lists some values of

0 < x < A.

'sin x

which were obtained in .the previous


These points are plotted in Figure 3-3a.

section.

Table 3-3a
,Values of

y = sin x

y = sin x

ir

1,

2
A
--ig-

= '71

If

74.,

3:

rt

2A

.17

3g

'V

2A 3A

-3

2.1!:
b

87

2..

,e

I--

17

..11

.71

"7

5A )*'::

1.

*-6-

Figure 3-53a.

Values of yt sin X) 4
plotted from Table 3-3a.

0
If we connect these points
with a smooth curve we obtain the graph shown in
Figure 3-3b.

A more com-

'plete, picture can be obi

tained.using more points


but this will suffice for
rt

Figure 3-3b

o'ir present purl:loses.

217
t

)9

.3-3
Now we can make use of the properties obtained in the previous section
0 < x < yr.-The, identity

to extend our graph beyond the interval


-

sin( -x),= -sin x

11

.(2)

',

_,-

is' the
tells us,thatthe graph is symmetric with respect 1.o_the origin;_that
(Such a functionAl also
if it contains --(x,y).
graph contains (-x,-y)
,

called an odd functtOn.

Later we shaIl show how to-approximate the sine

function by a polynomial function with o ly odd degree terms.)


t enables us to obtain Figure-3-3c froM'Fi

Equation

re 3-3b.

-n

We call this, one cycle of the


sine function.

Figure 3-3c.

sin x, -n < x < n

Next we use the identity


sin(x + 2ng) = sin x,

(3)

which holds for all integers


graph shown in Figure 3-3d.

and all real numbers

x,

The identity states algebraically thatthe.graph

of,thesine function coincides.with itself under a tranalition-Of

nits'

2ny
.

(to the right if

7'

to obtain the

is A negative integer and to the left if


..,

tive integer).
't.

Figure 3-3d.

218.

y = sin x

228

n lea posi-

{Arr.^ 4,

3-3

The gra'ph of the cosine function can be obtain


.
.

n.a ..-aimilar. manner, ,

for we knowthat

(4)

----44-........=..=-J--- -,0,,,,.,,
.

cos x = sin (x + 2) .

Thus we can picture a shift of the graph of thei;ine funcion

x
2

units to

ta graph of the cosine".functiolf.'

In Figure 3-3e we indicate this relationship by superimposing on the same


axes the graphs ig the sine and cosine functions.

cos
sin'

Figure 3-3e

Translation and Stretching


We havebbserved that the Cosine .functioh can be obtairied from the sine
A

function by translation;' This process generalizes.

The graph of

-y = sin (x + C)

where C
of

is a constant is easily obtained by,suitsbly tratslating: *le graph'

y = sin x

33d)

ICI

on the x-axis.

units to

We can think of shifting the graph (in Figure

the right or left according

_is negative or posi-

tive. For example, in Figure 3-3f we show the graphs of


y = sin -(x -

under the .graph of

-.tained from the graph of

y =

sin

y = sin

x,

3,-=

in (x + 2) *and

to show how .each can be ob.

by an appropriate translation.

,.
r

219

.2 6

= qin(x + 2)

.....

.0"
....

Figure 3-3f
In Figure 3-3g we picture the graphs of

y = sin x,

y = 2 sin x,

and

y = sin 2x.

y =sin x
y = 2 sin x
y = sin 2x
"o

er-,---

Figure 3-3g

We can describe the graph of

y = 2

An

x ,ad being obtained from the graph of j

y = sin x by "stretching" each ordinate by


graph of

y = sin 2x

factor of

2,

and similarly; the

being obtained by.'"shrinking" each abscissa.

M3-3
The graph of
y = A sin (Bx + C),

called the general sinusoidal curve, can be obtained by combining translation


and scale change.

FQr example, to graph

(5)

y = 3 sin (2x + i)

we observe that
sin (2x + ir) = sin (2(x + V)
1

so,that the graph of (5) can be obtained from


graph

units to the left. The graph of

frorn'that of

y = sin x

5;16= 3 sin 2x

y = 3 sin 2x

by shifting the

can be obtained

by "stretching" each ordinate,by a factor of

"shrinking"4each abscissa by a factor of

2.

(See Figure 3-3114)

y = sin x
y = 3 sin 2x
1
y = 3 sin (2k + .11)
2

Figure 3-3h

ti

221

231.

.........

and

3-3
-

Exercises 2:1
For egth of the following, sketch the graphs bf the three equations on

one set of axes over the interval(s) indicated.


1.

2.

3.

4.

5.

'7.

0 < x

y = 2 cos x

(0)

y = 3 cos x,

(c)

y= 2

(a)

y = cos 2x

0 < x < 27'

(b)

y = cos 3x

O < x < 2v

(c)

Y = cos

(a)

y = cos (x

(b)

Y = cos ()

< x < 3v

(c)

y = cos (x +

-7( < x < 37T

(a)

y = -cos x

O < x < 2v

(b)

y = -2 q.b,rx

O < x < 2v

(c)

y = -cos 2x

O < x < 27(

-(a)

y = -sin 2x

O <x<2n

0< X < 2v

COS X

0 < x < 2v

O < x < 47(

(b)

y = -2 sin 4x

(c)

y =-

sin

< x < 37(

< x < 2v

O < x < 2v

ye= -cos (x -

O < x <.47(

Y'= sin (x + 7t)

O<x

Y = 6os (x + 2)

O < x <_47(

(a)

y - 1 = cos x

0.< x < 47(

(b)

y + 2

=sing

-0 < x < 47(

(c)

y +

1t'in I

2x

< 476.

.0 < x < 47(

CJ

132
'V

2v

(a)

0,

3-3

8.

,P
(a)'

y = Isinxl

(b)

y o=

(c)

y = iisin

(a)

0 < x < 2vO < x <.2g

ilsin 2x1
4

5 xl

'

.9.

O <x <2n

= -Icos xl

O < x < 2v,

.
'(b)

y = Isin (x - i)I

0 <.x
_ < 2v

10.

(c)

y = !sin (x - 2)1

(a)

y = sin

(b)

y = cos

2
2

(c)

y = sin

IcOs xl

0 < x < 2v

0 < i < 2v

x + cos

0 < x < 2v
2

0 < x; < 2v

.s.

.
I

CO.

..;;:'

.J

tf

41,

A-

-'233

3-4

Uniform Circular Motion

3-4.

units pr second.
After one second

seconds

after

point

(r,0).

2s

omits away from

and similarly

P0;

st units

which is

P(u,v),

In Figure 3-4a we show a point

after

units from its starting

is

will be an arc-distance

P.

represent the initial position of


(r1,40)
0
will be an arc-distance s .units away from P0'

We let

will be an arc-distance

seconds I)

moves at the constant speed of

in the uv-plane, and suppose that

around a circle of radius

Let us now consider the motion if a point

(measured clockwise if ,st > 0) around the circle given by

from P

+ v

= r

2
.

Figure 3-4a

We wish to describe the coordinates


the sine and cosine functions.
x

x
u

2
+ v = 1

of

(u,v)

Since we defined the functions

r < 1.)

We can express. these coordinates in terms,of


st

and

some constant factor

x
'k?

directly proportional to

if

crosses

instead of

x.

What

is doubled, tripled; halved, or multiplied by

What is
t.

OP

The line

Pi(cos x, sin x).

the unit circle at the point

happens to

r > 1,

(While we illustrate the case where

in Figure 3-4a.

and

x -*sin x

in terms of a unit circle, we also drat the circle given by

our reasoning will also hold for the case where

in terms of values of

when

st = 2Ar? We know that

It follows that
224

23i

x.

st

is

is directly proportional to

3-4:
t;

that is, if

is a constant,

the cirdle given by


,st = 2nt

u2

x = 2o

we have

2
v

= r

When

x = kt.

then

has completely traversed


it follows that when

st = 2or;

(since the unit circle has circumference

units).

2o

Tthes.we have
,2or.

'2y = kk---),
s

from which it follows that k =

Alternatively we could reason that

st

2o

we get

x = kt,

Since

kt
2n

st.

2or,

k =

whence we arrive at the same result:

To summarize we can say that the coordinates

(u,v)

at any time

seconds are given by


u = r cos (2-)t
`r

'and
s.

v = r sin ()t.

k = r

The constant of proportionality


is called the angular velocity of

the measure of

w =

In

t, seconds

moves an arc-distance of

It is called angular velocity because

P.

POOP

central angle

is commonly. denoted by w and

(Figure 3-4a) may be written as

OP rotates through an angle measure of


st

units.

If we let w =

wt

as

we can write

u = r cos wt'
v = r sin wt.

When wt = 2y,
the point from Po

P will again be in the position


back into

Po

Po.

again is called a cycle.

This motion of
The time inter-

val during .which a cycle occurs is called theperiod; in this case, the
perivd is

21

The number of cycles which occur during a fixed unit of

time is called the frequency.

We give a commonplace' example of frequency

when we refer to the alternating current in our homes as "60-cycle", an


abbreViation-for "60' cycles per second." .

225

3-4

Consider the motion of a point

Example 3-4a.
.radius
3

Suppose that

in the uv-plane.

units per second.

per second; the coordinates of

w'=

Since

u = 2'cos (i)t
2g

TT

the period is

2g

the angular velocity

P(u,v)

and

around a circle of

moves at the consta t speed of

units

are given by
'v = 2 sin (i)t;

and the frequency is

.4;7

To visualize the behavior of the point


the motion of the point
P

in a different way, consider

which is the projection of


Q

moves around the unit circle,

on the v-axis:

As

moves up and down along a fixed diameter

of the circle, and a pencil attached to

will trace this diameter4repeatedly

If, however, the strip of

-- assuming that the paper is fixed in position.

paper is drawn from right to left at a constant speed, then the feribil will
trace a curve, something like Figure.3-4b.

4-

Figure 3-4b.

Wave MotiOn

An examination of this figure,w111 show why motion of this type is called


wave motion.

We note,,that the displacement

position is functionally related to the time


tion

such that

we may have either

y = f(t),
y = cos wt

of

y
t,

from its central

that is, there is a func-

By suitably"locating the origin of; the ty-plane,


or

y = sin at;

thus either of these equa-

tions may be looked upon as desctlfoing a pure wave or, as it is sometimes


called, a simple harmonic motion.

The surface of h body of water displays a

wave Motion when it is disturbed. -Anothet familiar example is furnished by


the electromagnetic waves used in radio, television, and radar, and modern
physics has even detected wave-like behavior of the electrons of the atom.
222

1:

A I.

3-4

One of the most interesting applications of the circular functions is to


the theory of sound (acoustics).

A sound wave is produced by t rapid alterna-

tion of pressure in some medium.

A pure Musical tone is produced by any

pressure wave which can be described by a circular function of time, say:


(2)

p = A sin at

where

is the pressure at time

positive.

andthe constants

The equdtiot (2) for the acoustical pressure,

and

A
p,

are

!o

is exactly in

the form of one of the equations of (1) even thoUgh no circular motion is
involved; all that ocwirs is a fluctuation of the pressure at a given point
df space.*

Here the numbers

The positive Number.......Ae

and w have Asect musical significance.

is called the amplitude of the wave; ,it is the peak

pressure and its square is a measure of the loudness. 'The number


portional to the frequency and is .a measure of pitch; the larger

a)
a)

is pro-,i,
the more

shrill the tone.

The effectiveness of the application of circular functionqlto the theory


of sound stems from the principle of superposition.
individually produce acoustical pressures
produce the pressure

the sum

+,p

pl + p2.

If

pl

has the same period.

pi

and

If two instruments
p

then together they

and p2 have a common period then


This is the root of the principle of

harmony; if two instruments are tuned to the same note,,thOy"will produce no


4o
strange new note when played together.

Let us suppose, for example, that two pute tones are produced with
individual pressure waves of the same frequency, say
u = A cos mt
v = B sin at
,where

and w are positive.

A, B

According to the principle of superposi, r

tion, the net pressure is


p = A cos at + B sin mt.

What does the graph of this equation look like?

We shall aqpwer this question

by reducing the problem to two simpler problems, that is, of kraphing (3) and
(4) above.

For each

t,

the value of, p

isobtained.from the individual

graphs, ,since

The acoustical pressure Is defined as the difference betwetn the gas


pressure in. the wave and the pressure of the gas if it is left unOisturbed.

227

37

3-4

u + v.

To' illustrate these ideas with specific numerical values in place of


and

w;

A, B

let

A = 3,

B = 41 vw =n.

Then we wish to graph


p = 3 cos,nt + 4 sin .nt.

(5)

Equations (3) and (4) become


(6)

u = 3 cos itt,

(7)

v = 4 sin nt.

By drawing the graphs of (6) (Figure 3-4c) and (7) (Figure 3-4d) on the same
set of axes, and by adding the corresponding ordinates of these graphs at
each value of

t,

notice that certain points on the graph of


nates.

You will

we obtain the graph of (5) shown in Figure 3-4e.

are labeled with their coordi-

These are points which are either easy to find, or which have some

special interest.

'

;,.
0

2g
-2"

Figure 3-4c.

311

Figure 3-4d.

Graph of

211'

Graph of '

v = 4 sin.gt.

u = 3 cos itt.

228

238

3-4

The points

(0,3), (0.5,4), (1,-3), (1.5,-4)

find since they are the points where either


and

(0.29, 5 )

(1.29,-5)

u = 0

and

or

are the first zeros of ,p.

and zeros of

The points

v.-2.0.

are important bqcause they represent the first

maximum and minimum points on the graph of tp,


(1.79 ,0)

are easy to

'(2,3)

(0.79, 0)

To find the maximum and minimum points

ipmplves:the use of, tables and herice we shall put off a dis-

cussion of this matter until Section 3-6, although a careful graphing should
produce fairly good approximations to them.
V

,/
p,u,v
6-

(0.29, 5)

5-

(9.514)

4
ee

and

0.3)

(2,

%
%

3(

%
%

1-

%
%

\'
i

%
%

i\

0.791

,
,
,

.r%

(1,-3)

/2

il
/

\.

-3-

(1.79,0)

/A

1%

/11

-45,-4)
- 5(1.29,--5)
s. .

-6-

Tr

Figure 3-4e.

211

The sum-of two pure waves of equal period.

Dashed curve:
u
Full curve: p =
are not the same
in order to shoW

= 3 cos yt. Dotted curve: v = 4 sin yt.


3 cos Itt + 4 sin yt;
0 < t < 2.
(The scales

on.the two axes; this distortion is intipduced


the detaili more clearly.).
'

229

239

3-4

1.

Exerciges 3-4

Sketch graphs A' each of the following curves over one complete cycle;,
. and state what the period is, and what the range is, if you can.
(a)

y = 2 sin 3t `

(b) 7 = -3 sin 2t

2.

(c)

y = 4 cos

(d)

y = 3 cos (-x)

(e)

y = 2 sin x - cos x

(a)

Find the length of thdarc traversed when co = 3

(b)

(1)

t = 4

(ii)

tom= 2

Fora given w

t = to

(iv)

in a circle, how is the arc length affected


tripled?

Find the length of the arc traversed when w = 3g


ir
(i)

r = 5

(ii)

r =

if

t = 6

if the radius is doubled?

IIN

rs= 3,

3.

(iii)

r = 10,.

(iv)

r = R

if

t = 3

- *

/
(d)

For a given w and a givemtime, how is the arc length affected


if the radius is halved?

(e)

doubled?

,...

Find the length of the arc traversed under


.
o

(ii)

(iii-)
..

--

5-

(iv)

if

W ''.ia'doubled?

J.)

).

quadrdpied?
. _..

-i

-Foe the following, sketch and identify the


(i)- period.

:.

('il)* location of maximum point(s) and


(iii)

minimum points. in this interval

(a)

y = -

(b)

y = 2 cos

221 sin 2x

(c)

0 < x < 2v.

y = 2Icos

.0

42-.

If theotime is given and the circle fixed how is the length of the
arc affected if

cIS = g

(f)

r = 10:

0
.

(i)

4
-; 230-,

240-

,),

3-4
4.

(a)

On one set of axes, using the same scale over the interval

0 < x < 2v.

(b)

Sketch the graphs of

(i)

y = sin X

(ii)

y = cos x

(i)

Using the sketches and the scale in part (a), sketch on the
same graph

(ii)

sin x + cos x.

From the graph of

y = sin x + coS'x,

conjecture the period,

and the maximum and minimum point(s).


(c)

(d)

(i)

Sketch y = I cos(x

using same scale as (b)(i).

(ii)

Sketch y =

using same scale as (b)(i).

sin(x

Compare the,graphs of (b) and (c).


Have you any conjectures?

tt,

to.

231

241

3-,5

3-5

The Addition Formulas

In Section 3-4 we added the corresponding ordinates of the graphs of'


t -) 3 cos it

and

t.-4 sin it

at each valve of

to obtain the graph

of
f

over the interval

0 < t < 2.

We could have obtained the graph of

easily if we had been able to express


f

nt

3 cos lit! 4, sin

t -) p

more

in the form

t -)p = A sin (lit + a) .

In this section we shall derive formulas Which will enable us to show that,
for all real values of

t,

3 cos it + 4 sin it = A sin (lit + a),


where.

A = 5,

cos a = 11, and

sin a =

The formulas that we shall derive will also hepo us to discuss tangent
lines to the graphs of 4trcular functions and areas beneath them.
We begin by deriving the basic formula
cos (a - 0) = cos a cos 0 + 'sin a sin 0. .

Q(cos a,sin a

cos 0,sin

o
"
V

AEI

4
To begin our deriva-

(We,illustrate the case for whibhO < 0 <

tion we refer to Figure 3 -5a.

P to

"

_ _

Yoll may have derived this formUla En an earlier course.

The distal-de from

is

c-

p.
232

2.4

3-5

PQ = Acos, 0 - cos a)2 + (sin 0 - sin a)2

(2)

444

We now use the principle that arplengthon a circle"depends only upon the
unit of measure and not on the choice of axes.

axes (Figure,3-5b), we see that


the coordinates

If we choose the

now has coordinates

(cos(a - 0) ,sin(a - 0))

(1,0)

and

u'

and

v'

has

Q(cos

0) sin(a

- 0))

P(1,0)

Figure -3-5b

In Figure 3-5b the distance from P

to

is

PQ = 141 -4os(a

(3)

05)2 + (o - sin(a -,,0))2

-,

We,equate this with (2) and square both sidesto obtain


(Fos 0 - cos a)'4- (sin 0 - 0110)2 = (1.- coS(a- 0) 12 + to - sin

0)12:

Expanding and regrouping, we get,on the left


!'

11

(cos2 0 + siri2 0)

cos2 a + sin2 a) '7, 2(cos 0 cos a + sin 0 sin a)

and oh the right


V._11.11,1

',1,4-JcOs26Z -

sin (a -

233.

243

2 cos(a

- 0) .

3-5

Sines, for all real

x,

sin

x + cos

we have

x = '1,

1 + 1 - 2(c6s a cos 0 + sin p sin a) = 1 + 1 - 2 cos(a- - 0).

Therefore, we conclude that (1) holds; i.e.,


cos(a - 0) = cos a cos 0 + sin 0 sin a.

While we could use a similar argument to derive the formula for


we elect to use (1).

cos(a + 0),

Replacing

by

-$

in (1) we have

cos(a + 0,)= cos(a - (-0)

*.t

= cos a cos (-0) + sin a sin(-0).


.

...

cos(-0) = cos 0

Sine

and, sin(-0) = -sin 0 owe have


....

..

( 4)

cos(a + 0) = cos a cos 0 - sin a sin 0.

Earlier we showed that, for all real


sin x =, cos ( - x)

(5)

and

x,

cos x = sin

- x).

We cap use (1) and (5) to obtain


sin(a + 0) = cos

- (a + 0)]

cos

a)

*CP

13]

St,

= cQs

- a)co

0 + sin (2 - a) s in' 0
4

ox
)

sin(a + 0) = sin a cos 0 + cos a si

(6)

,Replacing

by

-0- in (6) we. get

sin(a - 0) = sin a cos 0

cos a sin O%
A

The following examplleshoW some of the many fordulas which can be


deril/ed froth the foregoing addition (sum and difference) formulas.

Show thatfor all real

,ExamPie 3-5a.

cos

(8)

We use (4) with a = 0 = x

2 .

x,

1 + cos 2x
2

to obtain'l

2
2
cos 2x = cob x cos x - sin x sin x = cos ,x - sin x.
1

Since

cos

x + sin

2
x '= r,'

can r,Oritee this as

11

3-5

cos 2x = cos
cos2 x

Solving for

x -

cos2 x) =2 cos2 x - 1.

we ,get (18).

Show that for ali real

Example 3-5b.

sin (x- +

(9)

= sin x + cos x

We use (6) with

a=

x,

to obtain

0 =

sin ,(x +

+ sin x cos,

) = cos x sin

r.

Since
x
n
cos '' = sin -.=

we get

cos, x t sin x

sin (x +

sin x + cos x =

Thereforej

).

A slight generalization of this

processkyill be used in Section13-6 to rewrite. (from Section 3-4)


3 c s Itt +,4 sin nt
.

as

5 din Ott + a),

Example 3-5c.

(11)

cos

11.

sin

5'

a=

Show that for all,geal numbers

a, b,

and

sin ax cos bx = ifsin ((a + b)i) +,sin'((a - b)x)].

(10)

We let

'where

= ax

and

f5 = bx

to obtain

sin((a + b)x) = sin (ax +1'bx) = cos ax- sin bx + sin ax cos bx.
go

Formula (7) gives:


(12}

sin((a - b)x) 4 sin (ax - bx) so sin ax cos bx,- cos ax sin bx.

Adding (11) and (12) we get


sel((a + b)x)

'Dividing by

cos((a + b)x) = 2 sin ax cos bx.

we obtain (10):

235

: 245

fi
3 -5

Exercises 2:21.

Show that for all real

sjn

(a)

sin 2x- = 2

(b)

cos 2x = cos

x cos x
x - sin

= 2 cost x - 1

= 1 - 2 sin

(c)

2.

si,2.
n x

Sketch
(a)

1. - cos 2x
2

and show that

(0 < x < 2n)

if

cos x- + sin x =

sin (x

+ Ii) =

1ff cos (x -

*(b)

c os

if cos (x + p = -4

cos x - sin x =

4.

sin(x

li)
-741)

-I)

= ,r2 s in (x + 341)

3.

Using formulas (1), (14),


(a)

tan

.
(b)

(c)

tan

(a

r3)

(a

4.-43'

tan

14.

tan

a
2

=-

1 L tan

+ tan 0

tan (3

a
tan2 a

2 tan
1 -

(d)

and Exercise 1 show that

tan a - tan 13
1 +'tan a tan (3

tan 2a -

(6), (7)

1 + co

1 -'cos
sin a

sin a
1 + cos

Use the law of cosines to derive formula


Show that for all numbers

a, b,

(3).

and 'x
.01

(a)

sin ax sin bx =

c
2-1 cos(a,-

b)x - cos(a + b)x]


1.

1.

(b)

cos ax cos bx = --LcOs(a 2 b)x + cos(a


2

236 6

2 4:
,tt

4:.

b)x)

fi
it

"J

6.

Using any of the formulas developed in this chapter, find:


sin

(b)

;COS

(Hint:

12.

12

,0

21'
1

fl

lig

..

cos .15-

(d)

Tr.

ir

(a)

..

.
7.

Using Iiriy of the formulas delloped in this chapter, show that for all

values where the functiondare defined the following are identitiesj


cos4 e - sin4 e = cos 2e

(a)

tan e + sin e
2 tan e

tos2 31 e
2

(b)

1
1
%2
(sin -a
+ cos -a
2
,

(c)

1 + sin a =

(d)

(sin e' +, cos e)

(e)

sin 2e _

2 tan e
+ tan

+ cos e
sin e

()
8.

= 1 + sin 2e

sin e
1.+ cos e

sin

You derived the formula:

cos 2x = 2 cost x
Solve this for

(a)

cos

function of

1.

thus expressing

cost x

as a linear

2x.

Consider

(b)

cos4 x

as

(cost x)2

and. by the-iame methods as used


---, ,

in (a) show that


4

cos . x = 5.0 +

x,,..derive the formula for

x:

.
1-

(''''''

..-,_._..,

Thin
(..

,..

cos 2x + cos' 4x) .

cos 2x = 1 - 2 sin

9. "-Using the forthulEi

sin

._

1,

,.

or

x = 513 - 4 cos12x-+ cos.4x).

f,

237

4.7

T
3-5
10.

"
that is, they are true for all

Show that the f llowing are identities:


values for whi
(a)

sin 2e co

(b)

sin(x -

(c)

'(d)

cos 2e sin e = sin' e

cos z + sin(y - z) cos x = sin(x - .z) cos y


1-kcos
,

sin .3x sin 2x =

cos e

.7.

("0sin3 6
(f)

the functions are defined.

sin 13 tan 2e

d<

x - cos 5x)
cos 36
cos 20
3e)

114(3isin

144

sin x + sin 2x + sin 3x = sin 2x (2 cos x + 1)


+ tank \ 2

1_

1 + sin 2x
1 - sin 2x

Aft

.4

238

48

Pure Waves

3-6.

We promised thatformulas derived in the last section would enable us to


write

.P = 3 cos nt + 4 sin nt

(1)

in the form
(2)

p.= A sin (at + a).

We apply the formula'(6,of Section 3-5) for the sine of the sum of two
numbers to (2) to obtain

A sin (wt +

= A sin at cos a + A cos at sin a


= A sih-a cos wt + A cos a sin wt.
.

NoW if this ins to be the same as (1). we must choose w = n,


A s trr a = 3
4

A cos a = 4. .
.a

To find

we .take ifie sum of the squares irk' (3) to obtain


2

L2

A2 sin a + AA2A. costa = 3, +, 4.


A
-2

2,

N.

A ksin a + cos

a) = 25,

.
A2%., 25.
.

Thus we can choose A= 5


air; 6 =

(14)

aqd then choose- a^ so that


and

cos

a=

"A

From tables we get a g .643.


..

We have..seown that, for all-'real Nglues of

we can write

t,
b

(5).

fr

o'

3 cos nt + 4 siti nt = 5 sin(nt + a)

...
r

.r

.:

...

where a .1:

.6143.

:.'

,.

We can use the same procedure which we have followed for' our articular
.

numerical example to express any equation of the type


.

(6)

y =,B cos wt +46 sin at

in the foria
(7)

ti

,y = A sin(wt.-.)

By using the such formula for

sin(wt + a)

A sin v = B,
r
4.4

we obtain the two equation

A cos a = c,
,

?39.,

r
:4

"

d.

which can be solved by putting

(8)
d

A'=

choosing a

so that

(9)

cos a.= -

sin a =
-A

Neer a such that

We can choose exactly one

We know from (8)

0 < a < 2y.

that
2_

'=

(1A-3)

t.

that

2
P

+,v

= 1.

on the circle given


0 < a < 2y

There is then a unique a on the inl-erval

a units around the unit circle from

is

13 %

(- ,-)
A A

by

with coordinates

so there is a unique point

(1,0);

so

that is,, so that

(9) hOTds.

Consider thefunCtion.
(10)

y = A sin (wt + a),


1

where

A > 0

and

a <7a < 2n.

The graph of

is called a pure wave (or

'

sine outve)." We call


of the wave.

The amplitude

MiniOnvaque.of
(9r minima)

(11)

the amplitude;

2y

the phase; and

to,

is the maximum vallie of

and

the period,

-A

is the

The period is the distance between successive maxima

f.

We can rewrite the eglAtioli_of (10), as

y = A sin (w(t

4.
C.

Fr,om this,we see, that the graph-9f (10) can be obtained from, the gr&ph'Of
(12)

by shifting the y-axis

y= A sin cot.

.w

unitsto the right.

,
1.rated in Figure 3-6a for the graph of

'his ibformation is illus-

la = 5 sin4(Itt + 0, where 4a

.643.

sc

4
.

a
1,*

24.0;

r
12 ti

maximum

the period is` 2.

Maximum

44

minimum

Figure 3-6a

y = 5 sin (at + a), a


E'

.643

We have seen that ingeneral thgre Ts a better way to sketch the graph

y,= B cos &V+ C sing

than to_add the ordinates .octhe.graphs of

4
.

y = B cos wt

and

y = C sin at.

To discuss a fUnctiOn

.defined by
4

4 aV4

y = B cos wt + C sin ut
expediently aid to be-able to graph it quickly, we can write
y = A sin (at + a).

If we write the function in this form we can tell by inspection the period
(-1),

the amplitude

(IA1),

the maximum and minimumvalues

(tIAI).

Since

A sin (wt + a) = A sin w(t + 2),


,a)

we can obtain the graph by shifting the, graph of

y = A sin wt

_units to the

.eft or right according,as

a
w

is,positive or negati.
ve

22 1,-,1
41-,

.44

y = B'cos wt + C sin wt

-ft'wou10 be just as convenient to express

the form y = A cos (wt

0).

in

We leave this for the exercises.

We wish to discuss and sketch the graph of the function

Example 3-6a.

given by the equation

y = 2 sin It

(13)

- 3

cos*2

t .

We want to write (13) in the form


y = A sin (it + a).
Our addition formula enables us to write
A sin (it ;I- a) = A sin

For all real values of

a.2

cos a+ A cos

t sin a.

we require that

A sin 3 t cos

it

3
+ A cos t sin a = 2 sin 3

- 3 cot

3t
2

Therefore, we must have

A cos a = 2' and

A sin a,= -3.

Following our earlier procedure we write


2
12 ,...(cos

whence we get
that the point

A =

A7.

a + sin

a) = (2)

+ (-3)

Referring to the unit circlp of Figure 3-6b, we see

.=

lies in the fourth quadrant.

il3

a
ID

242

252

4'

4.

We/now find

Figure 3-6b

a'

our.talaes we get

at

so that

7,1

.98.

cos a/ =

2
'---

and

sin a/ =

From

wehave

Since a = 21t - a'

_a s", 6,28 - .43 =

.831.

5.30.

- _14

Therefore, we can write (13) In the Convenient form


(14)

Y4,71.447 sin it

a),

where a *60.

att.

By insp ction we can'tel that the period is

241.

the amplitude
3.
2

is

1/1.5,

the phape is

5.30.

A sketch appears in Figure 3-6c.

"3

214'3

9. 5 3

Figure 3-6c

Graph of

y = 2 sin 3. t
2

- 3 cos 21t =

sin (1t + a),

where a L. 5.30.

'nercises
1.

'

Sketch each orf-the following graphs-over aeleagt-twOeits Periods.*

Showtheaoplifude And period_of each.:_

= 2 cos 3t

(a)

,y

(b)

y =2 cos (4)

(c)

y = 3 cos (-2t)

(d)

y = -2 sin

(e)

y = -2 sin (2t + v)

(i-)

(f) 'y = 1 cos (3f +

244
t,

2 4

3-6
4,

2.

Without computing the value of

a,

find the amplitude and the period of


.

eadh.
(a)

(b)

) ---

y =

si:;*\3t

'

+ cos 3t

i,:

,,,

y = -2 cos itt + sin\yt


I

(c)

y = 2 sin

(d)

y = 8 cos

t - 2 cos lt
2

+ 6 sin ICI
3

(e)

y = 6 sin

(f)

y =

2t

4'1',

- 2

cos

cos

2.1iL

- 4 sin

57rt

Express each of the following equations in the form of. y = A sin (x + a),
(where

4.

0 <'a.< 2n):

(a)

y = sin x + 1/

(b)

y = -sin x + cos x

(c)

y =

(d)

y = sin x---cos x

sin x - cos x

Express each equation in Number 3 in the form


(where
(1)

cos x

0 < a < 2y),

y = A cos (x + a),

by two methods:

Use the formula for the cosine of the sum of twoangles;.i.e.;


cos (0 + cp).

(ii): Convert the answers of Number 3to A cos (x + a)


trigonometric identities, suchas
sin (e + 2n) = sin 0,
(a)

sin 4:p = cos (-

cos (-e) = cos 0,

Using the addition formulas, show that

by the use of
- (p),

etc.

y = 914T sin Tit- 3-16 cos ni.'

may be put into the. form of any one..of the following '(0 < a < 2y)
H

(i)

y = A sin (at + 0)

(ii)

y = A sin (wt - a)

(.ii) y = A cos (wt - a)


(iv)
(b)

y = A cos (wt + a)

r.

By the use of trigometric identities show that the four expressions


of part (a) are equivalent.

Sketch the graph indicating the period

:r the amplitdde

(using the form y = A sin (wt + a))1 indicating the phase.,


...

A,' and

3-6

,6.

Express each bf the following equations in the form


A

for some appropriate real numbers


(a)

and

y = A pbs (nt - a)

a. -

y = 4 sin nt - 3 Cos nt

(b)'y = -4 sin /ft + 3 cos nt,

7.

3 cos nt

( c )

y = -4 sin /ft

(d)

y = 3 sin Tut

4 cos rut

(e)

y = 3.sin /ft

Ycos nt,

and a
and

the coefficients

can be d---rmtwi...frO

sin at

show on a diagram how

a,

Without actually computing thagalue of

and

of

cos at

if each of the fo lowing expressions of the form

B cos at + C sin at

is made eq

A cos (at - a).

'1 to

Compute

a,

and find the maximum and minimum values of each expression, and its
period.

Give .reasons for yotir ansWers.

in 2t + 4 cos 2t

(a)

'3

(b)

211in 3t - 3 cos 3t.

(c)

7sin

ce

+ cos (2)

Verify that the superposition of any two pure waves


and

B cos (at - p)

A cog (at - a)

is a pure-wave of the same frequency, that is,'

that there exist rehl values

and

such that

A piwt (at - a) + Bcos (at - p) = C cos (at - y).


9.

Show that any wave of the form


y p B cos (gt - a),

(g # 0),

can be written in the' form


$.

y = A cos (at - a)

where A

is noprnegative, 'd),, positive-and

0 < a < 2n.'

246

25G
1

3-7

3-7.

Analysis of General Waves.

Period
2v

We.have seen that the superposition of two waves, each with period

--

s1ich as-0those given by

y = B cos at

and,

y ='C sin at

gives a pure wave


y = A sin (at + 0)
AKOh

of the same period.

Now.we direct our'attention to the superposition of two

waves with different periods.

Suppose, for example, we had to deal with


11,

(1)
.

The period of

sin 3x

is

sin'3x - 3 cos 2)1.

av

--

the period of

cos 2x

is

Point a simple observation is, helpful -- namelypif


then

2a,

a, 148, etc.,are also periods of

f.

?1(

-- = v.
2

At this

is...a period of

For example,^V4 know that

sin 3(x + 21) = sin (3x + 2v) = stn 3x


3

two

and hence it follows that'

sin 3(x + 2(4)) = sin 3(x +

I)

= sin [3(x + 21) + 2v]


3

2v%

= sin 3(x + --) = sin 3x.


-

In general, we must have4for each'integer


.
foe'

,27(

sin 3(x + n( --)) = sin 3x


and
cos 2(x + rig) = cos 2k:

In particular, we have:

sin 3(x + 2v) = sink


cos 2.(x + 2v) = cos 2x

so that

2v

is a period for both

defined by (1) has period


of the respective periods

2v.
2v

sin -3x

The number
and

and
2v

v.

247r,

z. 5'T

cod 2x?-

Thus the function

is the least common multiple

*-

ow.

3-7

- In general, suppose that


=4_

y = A sin (at + a) + P si

(2)

bt +. (3).
9

If

,a = b,

we.can proceed as in the prey us section to express

pure Wave With

a
period

. If

a /.-

then

v,-

is no longer a pure wave.


Suppose

as a

may still be periodic but


;

27t

but th t

,2t

and

have a commoh%multiple, that is,

'4 a

there are positive integers

and

(3)

We can then choose


holds.

m and

such that

2n

= n

b
2n

so that they have no common factors and (3)

In this case, the relation (2) is periodic with period m

is exactly the situation in (1) where a = 0,


we can choose

M = 3,

n = 2

The period of (1) is then

is not periodic.

a = 3,

2n
.

This

0 = - 2, ,b = 2

and

so that
2rt

tic

m a

n 17

2n

(2n)

- en,

Of courser it may be that


case (2)

2n
a

2n

and

have no common multiple, in which

For example, the function

y = sin nx + cos x
notp ;periodic.

This is/difficult to proVe and ita proof is omitted.

Theeriodicity of (1) is thus easy to determine.

There is little elise

About all we can do to simplify matters

we can conclude ih general-about (1).


to.sketch separately the graphs.of

v = 3 cos 2x
4
The result is shown by the three curves in Figure 3-7a,
u = 2 sin 3x,

and

y = u - v.

As
2 5,, 3;

* 3 -7

Al

y, u,v

I
1

v =3
I

u =2
'At
1

2 'TT ?(

v=-3

t%

Figure 3-7a
u = 2 sip 3x,,,,y = 3 cos 2x

y = u - v = 2 sin3x -

-cos 2x,

0 < x < 271.

The-superposition of sine and cosine waves,of different, periods can

produce quite complicated curves.

In fact, with only slight restrictions,

any periodic function can be approximated arbitrarily closely as a sum of a


finite number of sines and cosines. The subject of harmonic analysis of
tt

Fourier series is concerned with approximatingperiodfc functions in this


way.
,

The principal theorem, first stated by Fourier, is that a function

of period

cap be approximated arbitrarily closely by sines and cosines

for each ofNwhich some multiple of the fundamental period is

f(X)

AO + (Al cos

211x

+ (A2 cos
a

`oi
o

47tx

+ B
2

sin ax)
a

..

Spicifically,

2eX '

+ Bi

(1)

+ (A

a.

44g+

cos

2n7tx

2nxx.

n a
e

+ B

cos *--).

4.

.--1

'and the more terms we use, the better is our approximation.

259
21'

As an example, consider the function depicted in Figure 3-7b.


by

-n < x <11

.efined on the interval

function

This

0, ife = -n
=

For all other values.of

we define

if i= 0

0,
1,

-n < x < 0

if

1:00

f(x)

(2)

if 0 < x < v.
.r
f(x) .by the, periodicity condition.

f(x + 2n) = f(x):

This function has a particularly simple approximation as a series of the


t.

form (1), namely,


x

(3)

11'

sin 3x

sin 5x

sin(2n - 1)x\.
'
2n - 1

0
qi

Figure

3-7b

Graph of periodic function.


.

0,

x =

if

1,.if 0 < x < v


x -4f(Z) =

0,
-1,,

Fourier series.

f(x +

x = 0

if

= f(xr..

if -v < x r 0

4 s n x

iinja

sin 5x

....4..,.sin(2n

1)xN.

2n - 1

As an exercise, you may graph the successi'e approxiMations to

ttaking.one, then two,

y = f(x).
Q.

250

260,
-

f(x)

by

en thine terms of the serieb,epd iet-how the succes-

sive.graphs approach the graph of

"

.S:1

-,
.

...

---NN,

Exercises 3-7
,

:-

1.

Without setcfling,
find the periods,of the functions defined by the
.
following expressions
.

:.(a) 2 sin x + cps"gic

'

ti

.
-

(b)

(c)

cos i -,sin

-sit; 71- gx -

cos ItX 1r

.
j cos 12x,- 2 sin* 3 x
1."1

(e)

...(f)

1 - 2 sing x + 2 sin
Icos xl + sin

cos

x
2

rr
(g)

(h)

41simp4

11cos 4v xl:

2.

Sketch the graph of


y.=

sin x + sin 2x

by Urst-ketching y1 = 2 sin' x,4


axes for the interval
'3.

Sketch graphs, /or


(a)

(b)
.

1x1 < v,

sin x
n

on the same set of

for each of the following curves.

y = - sin x
n

then y2,= sin 2x

0 < x < 2y.

sin 3xN
+

4 sin x
= ...(_-_
__ +
v
1
.

-,

'

sin '3x..4. stn 5x)


.

3. ,

5
6

'4.

(a)

Find 4e periods of each of the successive terms of the series (13);


namely,

sin ..x

0$

sin 3x

,.

(b)

sin 5x
5

1.

'

What terms of the general series (1) are missing?


properties-'of the fl.petion

From the symmetry

defined by (2) can ypu see a reason

fdr the absence of'certain terms?

40
I

.251

261.

.
.

3-7
5.

The symbol

[x]

the real number'


(a)

stands fOr the greateStinteger less than or equal to


x.

Graph the greatest integefunction

x -,[x].

(This function is sometimes called the "integer part" function.)


.(b)

Graph the periodic function

x --4x - [x].

Indicate its period and its maximum and minimum values.


(This function is known as the "fractiobal part" functiOn.)
(c)

Graph the periodic function

(i)

sents the distance from

where tx)

x -4(x),

to the nearest integer.

Indicate_

values of the function.

Sketch the graphs of

(ii)

x -4(2x)

and

On each graph indicate period, maxima, Oidlminima.


(iii); What is the period of

x -,(nx)?

What are its ...maximum and minimum values?


.

6.

Graphtthe periodic functions defined below in the interval,


[Note:

See No. 5(b) for definition of l

Ca)

f(x) = [sin tocl

(b)

f(x) = [cos tx]

0 < t < 2.

Best integer function.]

(c), f(x) = [2x] - -2[x]


7.

The function

defineXi by the equation below is periodic.


1, x

rational

0, x

irrationala

Why ? -

f(x) =

4'

'r

AK

Chapter 4
14

DERIVATIVES OF CIRCULAR FUNCTIONS

In Chapter 2, we discovered that the,derivative of a polynomial function


is another polynomial function (of one lower degree), which can be obtained
.algebraically using the idea'of limit,

We can show that the derivative of a

Icirculaifunction is another circular function;

Using simple geometric

arguments we shall show that the derivative of the sine function As the cosine
function and the derivative of the cosine function is the negative of/the sine
function.

The first section of this chapter indicates how to obtain these

results for the particular cases at

x = 0,

using the same wedge method to

find eqUations 9f tangent lines employed initially for polynomial graphs.


These resultsoare interpreted in terms of limits of difference quotients
in Section 4-2.
terms of motion.

Later

hey are extended to pure waves and interpreted in

In Section 4-, the idea of approximating circular functions

hoy polynomials is introduced,

The resulting approximation formulas are useful

for constructing tables and finding limits.

4-1.

The Tangenc'at the y-Intercept


As was our approacil with the polynomial functions,, we begin our discus-

sion of the calculus of the circular functions by considering the behavior of


.$' their graphs near the

First we want to find the best straight line

appro;imation to the.graph of y = cos x


cos 0 = 1,

at the point'where

we are talking about the point

X0,1).

x = O.

Since

We conjecture (from
*,,

Figure 4-1a) that

(1)

the line given by


graph of

y = cos x

y = 1 , is- tangent to the

at the point

253

g63

(0,1).

Figure 4-1a.

Since the curve is symmetric with respect to the y-axis it is sufficient


. to consider positive values of x. We know, for .C.< x < 721
that
.

cos x < i

(2)

and hence the curve lies below the line given by y = 1 .in the interval
0 < x'<

2.

We now wish to show that near the y-axis the curve lies above

'the line given by

where

EX

Y =

(3)
E

is some positive number.

(See Figure 4-1b.)

I,
y = 1
I

Figure 4-lb
Obr

Y = cos x

0. and near zero the graph of


lies inside'the shaded region.

*111-

254

264

4-1

To establish this we use the inequality

(4)

1 - cos x <

which we established in Section 3-2.

than

For

x2

2
the quantity

x / 0,

is larger

so we have

1 - cos x < x2,

if

if

x / 0;

0.

This can be rewritten as

(5)

cos x > 1 - x

2
,

that is,

cos x > 1 + (-x)x,

It 0 < x < e,

if

x / O.

then we have
+ (-x)x > 1 - ex;*

whence we, have


(6)

cos x > 1 - ex

for

0 < x < e.

In summary, we know that-Tor any positive number

the graph of the

osi,ne function liesinside the wedge of Figure 4-lb on the interval.


0

where

is the smaller of

and

Since we can make

as small as we please, we conclude that the line given by

y = 1

is indeed

the tangent to the graph of y = cos x

at- (0,1)-.

established.

is the best straight linelapprdxima-

The line given by

y = 1

Conjecture (1)-is

tion to the cosine curve at the y-axis.


Having established conjecture (1), we now consider-4ie-problem of finding
the equation of the tangent line to the graph of
point of intersection with.the y-axis.
the line,given by

graph of

at

(0,0),

its

We conjecture' (from Figure 4-1c) that

.y = x

y = sin x

y = sin x

is tangent tq the

at the point

(0,0).

..

,
255

265,

.10

sin

Figure 4-1c,

First we shall consider only points to the right of


lar we shall restrict our attention to the interval
lar

To establish our conjecture for

(0,0);

0 < x <

in particu-

in the first quadrant we shall need

two inequalities for the sine function:


x(1 - x2) < sin x < x.

We canderive'these inequalities using our unit circle definition of sine.


Since, for the moment, we are only concerned with the interval- 0 < x <
we picture only, the'first quadrant.of the unit circle in Figure 471d.

..
D(1

sin xc
cos x

sin x
cos x

cos X

IA

. Figure 4-1d

Part of the Unit Circle


256

?66

The line trom

to

is perpendicular to the line from

hence is shorterthan the arc


BA

is

x,

sin x

AB.

to A

and

If the measure of the length of the arc

then,the measures of the lengths of segments

BC

and

OC -are--

respectively; therefore

and . cos x,

(8)

sin x < x.

This means that, in Figure 4-1c, the graph of the sine function lies below

the line given by

y = x

to the right of

as we have indicated.

(0,0),

Our second inequality uses an area argument.

again, weychoose
dicular to
radius

OA..

on the line through

sin x
be ----- .
cos x

and

The first coordinate Of D


sin x
OB
has slope ----- ,
cos x

so that

must be

Sidce

OA).

Referring to .Figure

EA -is perpen-

(the measure of

the Second coordinate of

Thus the region enclosed by triangle

-1d

'

D must

OAD _has area.

xl

(9)

2
"cos xl.
2(1)/sin

24010(AD)

The area oc_ztp circular sector


the area of sector

OAB

x;

that is

is'constant.

To deter-

OAB

is proportional to

mx,

where

is given by

mine

m,

.we note that if

x = T.

then the sector

OAB

is.one-eighth of

the area Of the unit circle; thus

i
1
.2.
8(71(1) )

jso that

m =

area of the sector

Thei.efore, the area of sector

OAB

OAB

is

is less than the area of triangle

aince. the region of 'sector

OAB

2 x.
Ot

x < 1 sin x

2 cos x

The cosine function is positive for


(10)

0 < x <

so we have

sin x > x cos x.

We know .front (5) that if

then

cos x > 1 -'x

2
.

We use this'in (10) to obtain the inequality

(u)

tihst

is within the triangular region ,ODk), the

area of the' sector is less than the area of the triangle; that
1

Since the.

sin x > x(1 - x2),

for

257

2d7'

0 < x < 12 1-

With-inequality (11) we can show that


the tangent to the graph of
value' e,

= sin x

y = x
(0,0).

at

is indeed the equation of


Suppose, for some positive

we have the line whose equation is


y = (1 - e)x.

If

< E

and

0 < x < 2 ,

1 - x

then we have

> 1

E.

Therefore, we get

sin.x > (1 - x2)x > (1 - e)x-.

In summary, the graph bf the sine function, to the right of


---J

when

close as:me please to zero (that

between the lines-given by y = x

y(1 - e)x.

and

and

(0,0)

0 < x <lp

and as'_

must lie

(See Figure 4-1e.)

The graph
lies in here.

s.,

2.*

X"
Figure 4 -le

The case when

x < 0

is now easily handled since the graph of the sine

.function is symmetric with respect to the origin; that is,

so,thet when re- < x < 0


given by

y = x

and

(x, sin x)

any points

y = (1 - e)x.

must lie between the lines

(See Figure 4-1f.)

258

268

sin(-x) = -sin x,

4-1

Figure 4-1f
We have now ettablished both conjectures:
(10: the.line given by

to'the graph of
(7)

the line given by


to the graph of

is the best straight line approximation

y = cos x at
y = x

(0,1);

is the best straight line approximation

y = sin Z

'296.9

0).

Exercises
1.

,a,
(a)
f

y = cos x

Write the equation of the tangent to the griph of

at

(0,1).
(b)

What is the slope of thetangent to the graph of

y = cds x

at

(0,1)?
(c),

2.

(a)

cos (0 + h) - cos (0),


h

lim
h -40

Determine

'Write the equation of the tangent to the graph of y = sin x

at

(0,0).
(b)

What is the slope of the tangent to the graph of

y = sin x

at

(0,0?

(b)
3.

Use the results of

aim

(a)

h -40

sin r(0 + h)

lim
h -40

Determine

and

cos h - 1
h

- sin (0)

to determine

sin h

(a)

To the right is a portion of

the graph of

and to the right of


P

near

y = cos x

x = 0.

is the point on the curve


x = 0;- Qi(i = 1, 2,

where
6)

are points on the,

curve where
.2,

x = .5,

.4,

.3,

11, -ftir--FffrTRe slope

'
of !!Q1' PC42;
the table provided.

PQ6 U

(b) lFind the equations (in the


form

y = b + mx,

is y-intercept and

the

slope) of each of the lines


determined,by

Had

Cos

Sin

.5

.87758-

.47943

.4

.92.106

.38942

95534

29552

..2

.98007

09983

.999995

.ol000
A

whard b

PQ (i = 1, 2,

.01

4).

,.
4-1
5.

(a)

To the right is'a portion of


the graph of

sin x,

and to the right of

near

x = 0.

is the point on the curve

.P

where

x = 0;

= 1, 2,

are points on the

..., 6)

curve where

x = .5, .4, .3,

.01.

of

Fi9dthe slope

PQ1, PQ2,

PQ6.

Use

the table provided in Number 1.


4

(b)

Find the equations 'of each of the lines determined by

PQi (i = 1, 2, ..., 6).


Assuming the following relationships established in this section
P

1 - dos x <

()

cos x#>'1 - x2

(8)

(10)

sin x < x

,.

x dos x,< sin x


...

show that the following inequalities hold:


(a)

xcos x < x
1 - cos x
x
<
x
2

(b)

1 - Cos

(ii)

>

x > 0

x <0

xi <,s

(iii)

(c)

(d)

sin x > x -

sin x
-

x3

7-

o<x<

< I - cos )5.

ti

.1

261

271
1

lx1 <

4-1
4t
7.

Given (from Section 3-2)


2

0 < 1 - cos

h < 7

and (from Section 4-1)..,,

h cos h < sin h < h,

for

0 < h <

show that

1 -

8.

<

sin h
h

for 0< Ihl <2.

Using the premises of Number 7, show that


- cos hi
h

for

h / 0.

'

9.

Use the results of Numbers 7 and 8 to estimate


for

10.

h = 0.01

and

sin h

and

1 - c os h
r
h

h = -0.001.

Use the results of Numbers 7 and 8 to determine


sin h'

(a).

h --,0

(b)

lim

cos h

h --)0

II

27142

4-2

ik
4-2.

'1

The Derivative as the Limit of a Difference Quotient

The deriyatives of the sine and cosine fundtions are, res pectiv4l1, the
iCe

cosine and the negative of the sine functions.

These geweral results Trft

obtained by first finding the equation bf tangent lines to the graphs'of the
sine and cOsihe".functions and then discovering the slope functions (derivetives).

We shall, however, obtain these derivatives directly

askmit;'of

certain difference quotients.

First we shall show that for the sine function


f

x -) sin

we can obtain the derivative


f': x -*cos x

by considering the definition of


(1)

f'(x) = the limit of

f'

given by
sin x

sin (3. + h)
h

as

approaches zero.

The geometric interpretation is that for some fixed Point


graph of

f,

(x,f(x))

the difference quotient


sin (x + h) - sin x
h
AA

is the slope of the line passing throughhe points


and

(x + h1, sin (x + h))

(See Figure 4-2a, where

(x

is shown negative and rather large.

slope = doses
tangent

line

s
slope

sin; (x + h) - sin x
iv)

x + h, sin (x + h))

Figure IP-2a

j63
10

2 73

on the

y;

0:
1+.2

We wish to show that


J

r-.4.
.

.i.

(2)

sin kx

-..,

h)

- sin x

h -)0

- cos x.

iThis c n. be shown:13y arguing (as we did in Section 4-1 for,x = 0) that the

graph of the sine function lies inside any given welge about the tangent when
is small enough.

ilif
.

We shall use a more direct method An which we employ

the results of Section 4-1.


Section 3-5.

We begin by usAng the addition.formula (6)


(6) se

Wa caniwrite

sin (x +'h) - sin x

Il

sin xicoS h + cos x sin h - sin x

,cos h - 1%
sih
n h
) + cos x
h

= sin.xk

7,
.

We must now show that as

approaches zero, the first term approaches

zero and the second term approaches

cos X.

In other words we moist show that

lim
h--) 0

..

cos h - 1.0
h

In Section 4-1 we showed

'a 'nod

lim

Lii....12,

h -i0 4,

7 1.1
t

'the horizontll line given. by

the best straight line approximation to the graph of

y.= cds x

y =i1
at

is

(0,1).

Since the.slope of the best straight line appro44mation (horizontal tangent)

to the graph of

y = cos x

is zero, we have

cos (0 + hr- cos (0)

lim

x = 0

of

h -) 0

Therefore,

. lim

cos h - 1

h -40

- o.

In Section 4-1 we also showed that the line given by


straight line approximation to the graph of

y ='sin x

at

y = x

is thg best

(9,0). ,Since the

c,

1 ke of

the tangent to the graph of


lim
h --) 0,w

y = sin x

at

sin (0 + h) - sin 0

x'= 0

1.

is one, We have

.44-2

Now we can conclude that

sin (k + h) - sin x

lim

(2)

cos x;

h -) 0

We have establiihed that

(3)

D(sin x) = cos'x.
The derivative of the cosine function at any point

(x ,cos x)

can be

obtained by calculating the limit of the difference quotient


co (x + h) - cos x
h
as

We leave it to the exercises (EXercises 4-2, No. 1)

approaches zero.

to sivi; that

D(co; x) = -sin x.

...,0000m0.000

Example 4-2a.

sin h

Find the limit of

as

approaches zero; that


*

11, evaluate

sin 2h

lim

h 40

We can argue that since

sin h

for small

a 1

1h1,

.
it follows that
J

sin 2h
2h

ts 1

if

-is small.

12h

We can write
sin 2h
h

- 2

sin 2h

2h

g.
4

If

Ihl

is so small that

7iiil small, then

Ighl

sin 2h
a 2.
h
\

Alternatively, and more directly, we can say


I-

-sin 2h
.

- 2

2(1)

= 2.
s

sin 2h

h t40

h -to 0

lira

/
Example 4-2b.

Find the equation of the tangent line (best- straight line

approximation) to the graph of the sine function at the point

It

'when

(a , sin a),

7t

7(

a =

We can write

so that

x =.a + (x - a).'

(6)

addition formula

of Sect

sin x =sin [a

(x - a)].

Using

n 3-5 we get

sin x = sin g Cos_f_x -J.0) + eos a sin (x- - a).

If we let x

approach

cos (x - a)

and

by 1,

sin (x - a)

and

fore, as we let
approximition t'o.
,

siiit (x

then

- a)

by

x..approach
sin x

x -

,approaches zero.

We can replace

by their best linear approximations:, cos (x


- a
a,

(7)

(from (1) and

..f Section 4-1).

a)

There-

t14 expression which gives the best linear

is

sin a(1) + cos'a(x - a);

that is, the equation"' of'the tangent line

at

the

point

(a , sin a)

is

y = sin a + cos a(x C a).

When

When

the equation of the tange'nt to the sine curve is

F.

7,t71

e,

the equation of. th

tangent is

y=1;
that is, -the tangent to the sine curve at

a =

It

,1)

the equation of the tangent is

is horizontal,

When

C'
1

266 \

`7;

4-2

It

57t

16

3:k

y = 1

sif

7 + 7`x
I

y =

5y.

1/

Td_7,--

...

the tangent,at
the tangent at

the tangent' at

Figure 4,2a

Tangent lines to the graih of

y = sin.x.

Note that the slopes of-these three tangent lines are

o,

and

-AO
.-.-

15
2

respectively.

As we should expect, these values of the derivative,

could have been obtained by direet substitution in (3).

p;

262 7

Exercises 4-2
1.

(a)

Use Addition formula (4) of Section 3-5 to show that the difference
quotient

cos (x + h) - cos x
doo.,

can be written in the form


(cos h - 1.
cos x k
)

2.

sin x(2411).

(b)

Show that

(a)

Assume (from Section 3-5 ) that

D(Cos x) = -sin x.,

Sin (a + 0) = sin a cos 0 + cos a sin 0


sin (a - 0) = sin a cos 0 - cos a sin
TAgt ,ei.j;. 0 = x + h

and a

0 = x.

e?

Show that

sin (x + h) - sin x = 2 cos (x

3.

f3

(b)

Use part (a) to show that

(a)

Given (from Section 3 -5):

Isin (13).

D(sin x) = cos x. L-

cos (a + 0) = cos a cos 0 - sin a sin 0

0.

cos (a - 0) = cos a. cos 0 + sin a gin 0;


0
4

Prove;cos(x+0-cosx=-2sin,kx+-h9sin
2.
.

(b)

Use part (a) to show that

I
1
.s/

From the inequality

D(cos x) = -sin x.
c

4.

sin (.01)

1,. estimate

.011

1(b)

1;

From the inequality

cos (-.00 1)

0005,

cos. (-.001).
2

5.

.(a)

(b)'

Using the i

quality

(i)

sin ( 1)

(ii)

sin ( 01)

(iv)

1 -

cos (.1)
)1.

estimate

'Op sin (.001)

"Using the i equality/

Isi n h
< 777- < 1,

).
...

'sin (.0001)

2
h

--f- 5 cos h <1,


(iii)

cep i.01)

(1-v)

estimate

cos (.001)

cos G0001)

28
(

278

sin (.01).

estimatq

6.

hi

11

(a)
a)

Using

ILL'

<

11 -

the value of

h t 0,

and the results of Number 5(b) find

for

(i).

h = .1;

(ii)

h = .01;

h = -.1

h = 7.01

(iii)

h = .001;

(iv)

h = .0001;

h = -.0Q1 4.
"h = -.0001

eS

2
(b)

1 - 11 <J-222-1 < 1,

By the use of the inequality

sih h
--,E-- and

the difference between

as the values assigned to

becomes smaller and smaller

decrease.

!hi

Show this for the following values of


(i)

h = .1;

(ii)

h = .01;

h = -.1
h.= -.t/
)

-7. The limit of

sin h

h:

(iii)

h = .001;

(iv)

h = .0601;

h = ;.001
h = -.0921

,,

--rs- as

t symbolically as

illustrate that

lim
h -)0

't

This may be stated

tends to zero is one.-

sih
n h

lim

LE 2I

x -4 0

li

.Find the

1.

by two methods:
.

(a)

by u se of the inequality

(b)

by direct applicatioh

8.

1 -

sin h

h-- < lo

sin 11

lim

- 1

h -) 0

/)
.

When appeOpriate use


sin h

lim

h 40

1 - cos h _

lim

and

- 1

h 40

'
0

in evaluating the following:


sin x
3x

lim

(a)

(b) -'lim
x -)0

lim
e

lim

sin 2t

lim edot 2e
' sin h ,-

sin

lim
(f)

t- 02t2 +t
e

lim

(i)

'x
h

tan 2x
lim
sin x
X -4 0

sing x

Jam 1,
(d)

1 - cos, e

lim

X -4 0

7x

sin 3x

cos e

F
e

cos 7X
cos 3X

tiM

lib

(2)

x 40

269

279

sin 5x - Isin
xl

.44

r--

4-2
9.

Evaluate

(a)

lim
x

(b)

sin x - sin a
x - a

Usk the result of part (a) to show that

lim sxX_1.
0

10.

Find the slope of the line tangent _to the graph of

(a)

y = sin x

at

the point 'where.,


(1)

X =

(ii)

(iii)

x =

IT

3
3v

(iv)

X = V

(y)

X =

(vi)

x = 0

graph of

Find the slope of the line tangent to the

(a)
a)

at

each of the points in part (S).


11.

y( = sin x

Write an equation) of the line tangent to the graph of

(b)

y = COS X

at

the point where


x =

(i)

(iii)

(iv)

x =0

21

(ii)* x

Write ari equation of the line tangent to the graph of_ y,=- cos x

(b)
,

12.

at each of the points in part (a).

does the,graphof

.For what values of x

(a)

y ='sin x

have a
ft

tange t line?

.fiorizonta

4 '',S1

%
'r

For what

(b)

alueso

does the graph of

y = co: x

have a

horizontal tange t line?1,1


4'13.

(a)

For w
.tange

by

line given by

or a

y = x'

y = x?

sin x

have the

p iqa lel to the line given

x dOes the graph of

.(a, For what values of

( )

does the graph of

Answer the above question for the graphlf

/b)

14.

t values` of

Answer the abo

qubstion/for 'the:graph of
I

4
t

y = bin x

have the

or a:line parallel to the line given

y = -x

tangent' line given by


.S
by y = -x?

y = cos ,x;

270

280

y = cos x.

,
15.

(a)

For what value of

16.

or parallel to that line?

Answer the above question for the graph of

(a)

If

f : x -+ sin x,

f!(-200% - i)
(b)

If

g : x -)cos x,

(a)

If

(b)

If

f'

then

Find

x -, cos x.

and

x -o sin x,

: x -o -sin x.

g'

e( og -

x 11n x

S x.

tt

f' (640n),

0(600n),

iL).

'f' = h

show that

x -o cos x

Find

and

h' = -f.
g' = j,

f'..7 h,

show that

h = -j'.

h' = j / and
(a)

,then

y =

f' (60n - 4) .

g'(-200n - g.),

18.

y = sin xs"have the

2y = x

(b)

17.

does the graph of

tangent line whose equation Is

4.2

In the interval
what values of
function,

< 2n

lx1

for

does the

I
x

x -0 cos x,
-2n

increase?

-n

For what values

2n

-1

does the function decrease?


(b)

Sketch f

and

(on

f'

different graph's but uqing

the same scale) to illustrate

x
-y

-27t

2:t

your answer.

-1

19, Without fihding'its'derivative, determine, for. lx1


values of

the function f : x -*cos 2x

TO?kAshich

< v.

increases and for which

, -

.'g43.

this funttiOn"decrea es by the'fol owing two protOOliresv


t

(a)

Extend th

(b)

Inspect the graph of the function

solution of part ( ) of Number 1 to the pres nt problem.

-0 cos 2x.

e4
20.

In the interval
erval
values of

determine 'by the following two procedures the

ILI

for which the function

x -0cos(2x

increases and

for which it decreases.

,,

(a)

Extend tile' solution of part (a) of Number

(b)

Snapect the graph of, the function,


1

.
I

I
,

x -* cos(2x +

n%

2
0

4
\ 271

'r281.

21.

For what Value:, of

over the 'I terval

0 < x < 2y

tion and the cosine function both ncrease?

do the'sine func-

For what values of

do

both functions decrease?


22.

For chat veil/es-of


.

over the interval

functions al? increase?

d < x < 2y

do the following

all decrease?

(a)' x -*sin ac2

23.

0,

(b)

(c)

x -) sin 2x

How large must the constant

be for the following functions

increasing?

(a)

f : x

ax - sin x

(b) 'f : x - ax + cos x

24.

(a) demonstrate the concavity (downward flexure) of the graph of

x -sin x

at

x = .4

by showing that at

x = .3

the curve lies below the tangent to the curve at


(b)

and at

x = .4.

Demonstrate the convexity (upward flexure)Is. of the graph of


x -*sin x
x = -.5

at

x = -.4

by showing that at

x = -.3

and at

the'curve lies above.the tangent to the curve at x

0.

'282.
j

+VP

4-3
4-3.

Linear Substitution
Now we" urn to the problem of determining the derivative of a function

specified by an equation of the type


= k sin (ax + b).
Such*E1 function is obtained from

is,

= sin x

is replaced by the linear expression

To detetmine4 DEk sin (ax + b)),

by linear substitution; that

ax + b,

and

by

we try to obtain the limit, as

approaches zero, of to difference quotient


k sin [a(x + h) + .1)3 - k sin (ax +io)
A

which can be written as


k sin [(ax + b) + ah) - k sin (ax + b)
h

Using addition formula (6) from Section 3-5, we can express the n

r of

the difference quotient as,k sin {ax + b) cos (ah) + k cos (ax + b) sin (ah) - k sin (ax + b).

',7actoring, we obtain for the numerator


a

k sin (ax + b)[cos (ah) - 1) + k cos '(ax + b) sin (ah).

The difference quotient can be expressed in the more useful form


k sin-(aZ

0(cos Tah) - 11

o[sin (ah)1.

J +-k cos (ax

As we prepare to take the limit as- h

a proaches zero wm make one further

change (suggested by Example 4-2a and Ex rcises 4-2, No. 8) n the form of
t/

e difference quotient:
l

rcos

14,.

Ell -.1

ak sin (ax + bfl.

ah "

,
.,-0(sin (ah)1
l + ak cos (ax +

As

approaches .zero,

Section 4-1) that


,

ah

(ah)

approaches zero; ancrwe know (from (1) and (7).of

cos (al - 1
approaches zero and

(ah

sin (ah)
(ah) 411

approaches one.
.i

We conclude that

1D(k sin ax +

b) = ak sin(ax+ b)1 Lim cos( .-ah

ah).

h
in(ax+ b)

0 + Eik

sirx(ah)

+ alc co-s ax + 1))

cos(Eix;+b)

n's

(, ah)

1
*go

= ak cos(ax+b) .

./

213 2 8 3

eq.

4-3
6

We can express Our result in the form


f

if

x )ic sin .(ax + b),

then

14)

ak cos .(a)t + b).

k cos (ax +

3.<

Similarly we 'could show that


if

(2)

cly

Using

sin (ax +

then ,gt :

and (2), we can obtain the slope of a tangent to any sinusoidal

curve directly.

Eiample 4-3a.,

If

find

3 sin 2x,

=6

x ) 3(2 cos 2x)

ft

f' (n)

From- (1) we have

cos-2x

so that
f' (n)

Example 14-3b.

cos 2n

= 6.

Find the'equation of the tangent line to the graphof


(7-1- ,

at tti9 point'

.ty = cos (x +

=6

2).

Using

to denote the value of

y'

the derivative, we use (2) to obtain


6

alt"
.

(37i- ,2)

At the point

-sin

(fik +

the slope of the tangent to the graph is

= -sin i = -

I-3-

The equation of the tangent is

L,

II

'

.e>

Id'
Ir.

I.

We can use the - graphs of

y =, cos (x.4

geometrical interpretation of this result.

Example .4-3c.

(3)

.)

and

y = cos.

to give one

(See-s-Exerciees 4-35

NO. 8.)

Use (1) and the facts (from Section 3-2) ;that

cos x = sin

- x)

and

sin.,x = cos (ir- 7 x)

to obtain the derivative of the cosine -function.

/?.

We ha e from (3)
cos:

'x

We use (1) with

k = 1, a = -1,

cos':

and

h = i to obtain the derivatilie

(-1) cos (i. - x) = -sin x.


4104

Exarftple 4-3d.

ind the derivative of

.1

The function

x -4 sin x

sometimes written
o-

is the functim wholue at


x

we carp express

x -) sin

is the siile of

degrees.

Since

..

2y
x
300

x,

radians,

as

Thus, formula (1) gives

icos

fit x

x).

The primary justification for defining

sin x

in terms-o'fferc length

rather than degree measure, is that we then obtit the-simple formul


....

4:..

n x = ,cos x

4 .6.%'1

If we had used,degree
4

formula

'asure then we would htive had the 14s satisfactory

.
,

...

r
silof

Scos4

l): =
'

.." \

...-,.,..

.. .:',
,

.o.
.

11

Cif:

10*

4-3
Exercises 1111
Using formulas (1) and (2), find the following.
(a)

D[sin (3x -

(b)

D[-2 cos,

(a)

Dr5 -cos (i

(e)

D[l sin (-x

;I)).

D[-Ain (

(c)

2.

)]
3

) ]:

+ 721)]

D[-n cos (2n - 2x)]


I

(f)

For the following functions determine


(a), the slope. of the graph at 'the. point indicated for each:
.

(i)

77(

x -4 cos (-x + 71-)

x = - -- ,

3
(ii)

x -4 -sin (2x -

(iii)

f :

(iv)

3 cos
1

x = g ,

X=

+ 2x),

sin (

3n
2

x.
-

x = 0;

)
72-

the equation of the line tangent to the graph of each function in

(b)

part (a) at the point indicated.


s

\....

.
.6

If two functions

there is a ,constant

..,

.
,

and

are directly proportional; we mean that

such that

g(x) ...cf(;)
1

(a)

Ijow are their derivatives related?

(b)

Illustrate your answer to part (a) when

'

f(x) =

sin

2x,

and

g(xi = -2 cos (2x - 7;).

4.

(a)

(i)

For what value(s)

of x

0 < x < 2n

over the intP'rval

tie graph of the function


x -*sin x + cos x

have a horizontal tangent

linefr

OA), ;Give the equation(s) of the tangent lineks).


(b) =Answer the- questions, of.' part (a) for the function

g: x
;

,7

+ 'a.. cos

(c)." Answer-the questions or


off' (a) for tile function
4

Aih 4 ;

X -4 3 sin (2x 4- T) ,,4

3 ccs (2x +

4
1

2+6

does

-4-3

5.

Show that if
f : x -"k cos (ax + b)

then
f'

x --ka sin (ax + b),

Employ the method of linear substitution used for the function

lix,-ksin (ax

.+ b)

at the beginging of this section.


,

6.

In Example 4-3c we used the facts that


cos x = sin (1 - x)

and

to obtain the derivative of

sin x = cos (i
x.

x)

Uge the same method to fihd the

derivative of
k cos (ax + b),

given that

D[k sin (ax

b)) = ka cos (ax + b)


o

_
.

(a)

(b)

Find the,limit of 'sing

(a)

On one set of
and

1:))

Find the derivative of


'f

8.

as a -'approaches zero..

coOrdinatir

x -)cos x.

axes sketch the graphs of

x -)cos,x

x -)cos (x + r

Show that th
(11- 'f0))

tangent line to,the graph of

is parallel to the tangent line to

at the poi

(12

at the point
he graph of

g(21-),
12-r,

9.. Use the difference quotient1definition of derivative as it applies to

the sine and cosine functions tOevaluate each of the following


sin (x + 11), - sin x

(a)

lim
h -)0

(b)

li

cos x

cos

lice

cos (3,

(c)

+ h) 1 cos 3x

h
W."

witot
I

.5

4-3
-3

sin

4,4-

lim

* (d)

+ h) + 3 sin

2
(n

sin

lim

nl

x + n

h -4 0

(e)

+ k)
k

sin (i)

COS

h) 4. cos

lim

(f)

2h

k
o4
6

Sketch the graph, of each of the following on the--

10.

val' [

-n,n].

y = siryx

(a)

y = sin (x -

'(b)

y = -sin x

y = sin
-

Y = cos x
Leos (x -

(g)

11.

- x)

sin

(e)

4-x)

(h)

y = cos (JX)

(i)

y = cos (

Find

..-."

for each of the following.

dx

y = sin x
(b)

y = sin (x

(c)

y = -sin x
4

(d)

(e)

(9

(-x)

y
=

- x)

sih,

Olt

y = cos x
.1A

(.,g)

(h)

= cos (x
y = cos
-"y = cos

(-xt
(12-1

..'*

II`

721)

- x)

278,

'

24-4. 4Velacity and Acceleration

In Section 3r4 we discussedkotion in a circle with radius


,
FigUre 4-48.)

If lobint

is moving'

(See

1.
;

Jr;

countercloCkwise _starting at

A(1,0)

he

and if the aA- is,

't = '0

covered at the uniform

of

unit pAsecond-,. the coordinate6 Of

at time t

are

x
(1,6)

0 cot Q *

A ( -1 , 0 )

X = cos t
y = sin t.

fir

'Let us draw a line thTugh

11

perpendicular t,o thel5c-axis.

Q be the foot of the perpendicular.


Figure 4-)la

We shall study the motion of theipoint


as

When

increllvs.

B on the y-axis and':

0 to

moves from

back to O.

t =

has reached

A' (-1 ,0)

At' t = 2n,

ands forth on tile rx-a3ti s

P ,hasqenhed .

(a* quarter circumference'',

As

0.

to t =

t = n

Er'om

2
As we see,

is back at A.

increases from

moves

to

n,

from

AT

oscilla/es back

This motion repeats b..t time intervals of

Let us find the velocity, of

point

at any time

at time

2n.

To-do this

we take the derivative of the functi n


f

(1)-

t -) cos

which- represents the poaritift of the point

derivative of

(2)

"4

During thi first halrrrevolution, -(0 <, t


.which shosfs that
and . t =-Tr.

At

As 4e know, the

t.
t.

is

t.
'n),

the velocity is negative

Is moving to, the left.". The velocity is

t = 5-

then

Q = 0,

f' (i) = -1.

' absolute value of the Velocity, has its maximum value when. t =
During the qecond half - revolution
fg h t

(31) = 1
2

at

The sperm, which

(n < t < 2y),

the

is moving td.the

is the maximum velocity and speed during th'; time iner-

The motion relleatt6.* after' t

ao,

2789

A.

4-4

at an3/ time

-What is the acceleration,of' Q

The acceleration is the

t?

,derivative of the, velocity 'ft, that is, the second ilorivative

f.

f"

In this case, sitce


to

f': t

(2)

(3)

bl

t -4-cos t.

'

The velocity decreases when the acceleration is negative and increases

decrease of

is negative but

f"

the.fi.rst quarter-revdlution,

During

Let us lee how this works out.

wnen the acceleration is positive.

is also negative.

f'

means,an increase in its absolute value, that is, an increase

f'

in:Speed.

Between

f' 'is'negative and

Hence,

Between

t = *

is positive;

f"

ancrthe speed decreases from

to

f'

is

positive.
-1'-to

increases from

10.

sin t

t = 21

and

is negative .and ;sin t

cos t

t =

and

t = -2

and .cos t' are both negative so

'

that both
t
0 to 1.

f'

and

are positive.

f"

Both velocity and speed increase from


-

Finally., for . /1_ < t < 2y,

f"

is negative and

f'

is positive K, that

Q slows down again to 0.

xample 4-4a.

Show thp't

In what quadrants is

f"

(a} positive?

'

(b)

negative?

.
f"(t1 = -cod t

Since

f"(t)0= -f(t).

f(t

and.

cos t.

fi(t) = -f(t)

,
(more briefly,

fu

_f).

Hence, .f"(t)p'.is positive when` f(t) = x


quadrants II and III;

f"

is negative when x

is negative, that.is, in
is positive; that is, in iva-

drants I and IV.


vcf

280

290

r".
.

Example 4-41o.

t =

wheri

To find the position, velocity and acselera

pn of

Wes have

3
)
.

cos

f'(3) = -sin

% f

Henee, after

lc

(n)

- --". -0.86'
2

-.cos -5

seconds, the point

is at xf

90-

and is moving to the_

,/".

left at

--

distance units per second.

The
h acceleratlon is

.1 .

distance
2
units per second each second, which means that the velocity 1.-s decreasing at
.
c
1
the date of
unit per second per second.
(The speed is increasing at this
2
2

...

rate.)
4%.

We cin'generalize oar discussion 'by considering uniform motion on a circle


of radius r Where the point P moves the

diatine4

in ,to seconds.

4 -.4b.)

Let

AOP in

'radians.

(See Figure

be th4,Meisure of angle

Then
kt'
r

e =

As in Section 3-4 we let

so that

ci)

1;'

The coordinates of

are

x = r- cos at
Y = r sin wt.

Figure 4-4b

We 'let
46

(.4)

r cos wt.

6
o

Q' is the projection of

-If

P.

..1-_

tion of

on the x-axis, then the velocity and accelerao

are given by the functions

..()

1:.

and

f': to-4 -wr sig,at

t..

(6)
4--

..

f": t ) -2r cos wt.

281

29I

c_

r"

4'4
.'

Now

respectively,

o?cillates between'' A (r,O)

and

A' (-r,0)*

The time

requtred, for One os,cillation back and forth.is obtained by. .....
using the circum.
T\\.:the 'period
this
We
call
ths
time
ference 2nr Ids the arc, so that kt'=

.,

'4-of the motion. ..Then

r
2nr"

2n

%ff
.

Note that
f" =

(7)

f(x) = r,cos tut


4 r

which is called ,r; differential equation.. We say that

(I

solution'apf this equation.

moves on a circle of radits! 10 feet a the


,r
Find the position of its projection Q and the
rate of 2 feet per second.
1
k , 2
.tince w'=
velocity and acaeleration of 'Q after -5. seCondi
5 1.
10
.r.
A point

Example 4 -4c.

is

- \,

t -410 cos

f'

'and

10

.i,

t"

5'
=:72

k-

.,

sin

For

t = 5

cos t

we'have
f(5) = 10 cos

1:=
-1.68

fe(5) = -2 sin 1

0 (5) =

cos 1...;z

since

`cos 1 tt 0.54

and

sin 3.

When

t = 5

seconds, cQ
1.68

moving to the left a


in g at

0.22

ft./

0.84.

ft./sec. approximately.

each second so 'that

O. .It'is

feet to' the right of

5.4

is about

Eli speeding
.

up ,at this rate..

t:

.,
r,

..

"Ow

fr kr

282

909

The velocity is decrees=

-.

4-4

Let

P move around the cir.C1e-, of radius, 'r

time let 'us study the motion of the projection

as in Figure 4-4b but this


R" of

P' on the y-axis

AI

(Figure 4-40.

The point

is at,

y = r sin 8 = r sin'cut..
.If

is the functiOn

g.

(8)

t -)r sin wt,

then

4'

rm cop

(9)

and
2

(10)

=no

g11-: t

sin cot.
v

The point
.

starts at

= 0
A

at the point

and. oscillates up

*.11

and down on the y-axis between


and . B' .

'

It completes an oscillation
2n

after the time --

V.gure 4-4c

Note that

"

g"(t)

?ej

-tc2 g(t)

or more briefly

..c02

g"

We have discussed 'd-the motion of ecil of the pointi

in Figure

However, we may, interpret

and

g' (t)

ft(t)

Q and

indicated

'11

\,

4Q\

%%ZI

as the horizontal and

..

, '

vertical components of the 'motion of

itself.

-i.,

Notice that
Ilkkt

f' (t)/ = -closet),


'

IF

;
A

r
r

Figure 4-4d
04.

283

293

14-

4( ,,
.

g'00 = co f(t)

,,'

-,4
,

4-4
.4.
-

The poSition and velocityof

are giAn'by.

at time4 t

= r cos wtft(t1 = -ru) sin cut;

and of

R
o g(t) = r sirr cat

'

The actua

Sngle

PQ

velocity 'of

0'0

rco cos wt.

using the Pythagorean Theorem in tri7

is feund

. .

liff'(t)12.4-,[gt(t))2.

42[g(t)I2t+ 2[f(t)
cbi g(t) )2 -4-(f(t))2
cur.

Since
k

au =.k; ' so that as we iight expect,

a.) =

at "which

,t

e arc is changing with time.

that "the dire c ion of

motion Of

is the constant rate

You can easily convince yourself

is perpendicular to thi radius

OP,

so.that the

s in the direction of the tanvpt, as,it must be.

If we work with the horizontal and vetica

compoli

of acceleration
accelerpton

of P

in' the s Me way,.we find that

'

=If"(t))2
= 14-032i(t)]2 +

)2

g(t1)2

irf.t) 12 4= [g(t)?
car'
and that the dire

ion of the acceleration is toward

0.

f
.P284'

2'

-Exercises 4,4
e

1.

Suppbse

move's around the circle at the rate of

Ii

. It

arc Onite. per


-

second- 'Find-%le horizontal and vertical position, velocity, and


" ....6.

acpeiti-ation when the following values 'are &reit tM Adius and time.
,

r = 1

(a)

.
.

',

..

'

(I) i

t = 0

(ii)

t = 1.

iii)

t ,-- 1.,

iN0

t = 2

% )

'

(t),

r = 2

,,

(c)

010%.

(i)'

t =0

(ii)

t ='rl

(iii)

t = 2

(iv)'

t = 4

6 ,
't

t =0

(iii)

(iv)

t = 3

Iting the result's of NuMber 1 compare the 'position, the horizontal

and the vertical Alocity, and the horizontal and the vertical acceleration of the following if
(a)

(b)

r =2 and

(C)' IL= 6
3.

t =

anis

f.N.-1;.

and

ana

Using the results dumber 1 compare the horizontal 'and vertical


position, the horizontal and the vertical velocity, and the vertical
acceleration when

t = 0,

and again when point. P

has traversed 'the

entire circle far,


(a)

r'

(b) TS irk.=
2.

at

t = 0

(i.e.,' at

t = 0

and at 't = 2)

and at' t = 4)
,

{61
,

4.

r = 6.

(i.e.o;

at

t = 0' and at

t =.12)

Show that the square rOptIof the ,sum of the 'squares. of the acceleration

components is the product of the speed and the angular velocity;


4

prove that -Ar(t)12-+ Eg"()12 = s- a),


<'..the angular velocity.

where
.t

.e
I

'

28

35
rC's

sr is the speed And 'co

11,

1,
--.

Suppose a massis atached to'a spring as

5.

wall

spring

mass

surface

-"'

GD

4
.

' When ti-ip spring is unstretched,!the Pointer i,s at

Suppose that wee'movethe mass

on the scale.

unitsto,the right and release it.

The mass wil4 oscillate back and forth,

..

s.

It aan be shown that the spring

acts fn such away thai the position of the pointer is given by

Pit

f(t) = A cos

where

and

is a certain number and

are doniants that measure,


.

the cart.

the stiffness of the spring and the mass


5

(a). What is 'the value of "A?


C

./.

14hatlis the'periad.6i the motion?

(b)

how that

(c)

f'01 = 0;

that is, the initial velocity of the sring

is zero.
(d)

Show thj

f"-(0)_

(e)

Show that

1r! ...,Interpret the negative sign.

is a-solution to the differential equation

_AO How far left doeb the - pointer move?

(g)

Show-that at title

t = li

ms

11 the'pointbr crosses the point


In what

What,are its velocity and acceleration at this time?

f" =

direction is the,mass moving at this time?i

(h)
,

at time does the pointer cross 0

and acceleration at this time.


moving; at thia time?.

i
1

At

again?

Find its velocity

In, what direction is theIlass

''..)

4'

,
6. Draw a figure to justify the,atatement that in the case of 'uniform motion
4.0.

in a circle, the acceleratton is toward the c enter.

the horiiontal,and vertical acaeleration;of


.
.

28t

296

P.)4

(the'? arrows to show


,

II.

4-5

%*

4-5. .Higher Derivatives and Approximations


,

,..

When we,differentiated a'polynomial function (Chapter 2). we got another

polynomial function.

We discovered that if we took the first, second -and-

'higher derivatives of-a-zpolynomial function, we eventually gota deHvative


with value zero for 611 values of

Th6. will not be the case for a circu-

x.

lar function, since the derivative Of'a circular function is a circular function.

We shall see the

if we take the derivative of a circular funCtion, and


.

then the, derivative of


*

he derivative, etc., we will ?Soon obtain the very func70


.

_/:

tic)* with which we begA.

''

0,

For example, if we begin with the function


f

sin,x

we get:

tlie first derivative


.

at -4 cos x

the third derivative

Of

x -' -cos x
,

and

derivative x -4 -sin x

l t)pe `second

tin fourth derivatTrve

x -4 sin x.

As was the case for polynomial filnctions we can write ...Q9 = f', D 2 f = f",
f4,
fiv,,D5f
fv,
=
=
=
etc.
The Roman superscript notation becomes

..

cumbersome for high orders and it becomes more convenient to use-Hindu-Arab i'4
13

,.

order

, -

derivative of

as'

itself: ,f) = f.)

f,
e

...

,,

If we take successive derivatives`-of the sine function we get


:

;,',

ft

X --4 cos x
1

f"

(1)

x -4 -sin x

fm: x -0 -cos x

(4)
f_

x .4

sin x.

%.,

Since the fourth derivative is the original function wecan see the pattern
,

4?

In Leibnizian notation we write 1

Dny =

(-1:) y = --2.
dx
n
dx

This can be further,atbreivated:

we write

.4

' .

287

(13)

- numerals parenthetieally as D f = f
. ,Thus the nth derivative* of f
a
Ls written n-DfFf (n) ., (It is also a, usIful convention to define the zero-

"

(n)
',

when' we mean

n'''

D y.

',-

4
4-5

f.= ff(4) = f

f=

0
---

(8Y

f(5) 2 p(9)

-(1a)

f(10)

f(6)

f (7) = f (11) =

flit

etc,

4
This result is sdnetimes summarized by saying that each function
here

y = sin x, cos x, -sin x

or

-cos x

is a solution of the differential '.

equation
.

(0

y/(4)

where

(4)

= Y,

represents the value of the fourth derivative of

We know from our polynomial discussion that tale process of differentiation


lowers the degree of the pqlynomiar function., .Thus; if

x --vy

is a poly-

nomial functiOn of degree

then

(4.1)
n

e--

(3)

is a solution to the equation

f(x)

Y.

,= O.

The first derivative can be interpreted as the slope of,a tangent line

or as velocity; the secondderiyative can be thought of as the rate of Olehge)


of,the slope function or as abcele'ration.

While physical and geoilittrical

interpretations of higher derivatives-are more difficult to contemplate, higher


- :derivatives ere useful in approximation discuisions.

Approximations
We want to find a polynomial,function which approximates the sine faction.

To do tiis we turn to the problem of finding a polynomial function


to

whose firstand higher derivatives "fit" the sine function near zero.
precisely for each positive
polynomial function

'//integer

More

we shall show that there is a unique

11

which satisfies the conditions:

(a). ,the degree of .1) < n

(4)

(b)

*
,,,,..

p(0) = 0 = sin.0

-tc)" ,the values of the first

are the same for

derivatives, of

x = 0.

and.the sine function

To construct such a polynomial function


cients:

a0, a

a2,
l',

'

p(x) = a

(5)
.

',.

we need to find the coeffi-

for

+ a v+ a x- + ..; + a
1,

x!t.

Since conditions (4b) and/(4c) determine the-values

288

,298
ti

p(0),, 15'(0); p" ((S),,

only shold

. we need

p(n) (0),

that these values 'determine the coefficients of

P(x)

-in

(5). ",
First try .to' approximate the sine functioK by. a first degree polynomial
function;.

i.e

we suppose that

n = 1

p(x)

so that
+ a x
1
.

-7 a
0.

and

p' (x) = al

Therefore, we have

%ow

(6)

and

p(0) =a0

al.

pt (0) 7

1
,For f: x ---) sin x,

f(x).= sin x

and

cos x

f 1(x)

we hate

ft(0)-=

f(0) = 0,

(7)

.
If tthe polynbbial function" p

require that

is to satisfy the conditions of (b) we

,,,

'

p(0) = f(0)

and

fig).

p'(o)

Combining these requirements with (6) and (7), we conclude that


a0

and

al

Thus there is exactly one polynomial function


= f(0)

of degree

and ,p' (0) = fq0); and that functiop is given by

Consider now the case when

= 2 ' so that

13. has the form

p(x) = a0 + a x + a2x

,.

2
.

In this case, we have

and

pi (x) = al + 2a'2x

ph..

(x)

= 2a2

,.

,p,0

that

'p(0)

(8)

Sitice

ri < 1, such that

x -+ sin x,

ao,

p'(0) = a-1 ,

p "('9)

wec have (from (1))

289

299.

2a

2 -N

p(x) = x.

3
.

f(x) = sin x,

f"(x)

fl(x) = cos, x,

so that.

fq0)= 0,

.k)
If p
.

f"(0) =

f'().= r,

is -to satisfy the conditions of -(4) with 411 = 2

(10'

p? (0) = ft (0)

f(0),

p( 0)

we require that

= f" (

p" (

10) .with (8) and (9), we conclude thti

Combillingthe reqUArements of
=

Et

The cases
p(x) = x.

n = 1

0,

= 1, .and 2a

result in the same polynomial

n = 2

,and

'

2
p

given by

This result is not surprising since the Second derivative of.the.,

sine function is zero when

turn to the case for'which

Now

x = 0.

p. is'a polynomial function of degree

i.e., we suppose that

.p(x)

= 8

+.8 X
1

8 x

n = 3;

given.by

+ 8 xJ.

Differentiating, we get
.pt(x).= a

p"Tx) = 2a

et(x) =

and

For

x = 0

(11)

+ 2a2x + 3a x2,

+ 6a x.
3 4

6a3,'

we' hale

p'(0) = a

p(0). = a0,

p"(0) 7 2a2,

1,

Using (1) for

x = 0

"1(0) = 6a.

and

w,p obtain

'.

f'(0) = 1,

.,f(0) = 0,

"k

and 44"(0) f -1..

fill)) 's 0,

Comparing these values with (11) under the requirements of (4)., we get

1 .
a

= 0,

= 1 2a 2

Therefore, the only polynomial

and

= 0,

= -1.

6a
3

of degree

or-less thich satisfies the

conditions of (4) is given by


,

,
3

p(k) = x

AS we attempt to express a result for arbitrary


1
notation
.

'

,kt = 1

'

3 ...k

We can show thai if'


.

290.

'3 0 0
V

and

Of = 1.

we shall use factorial


1

'4-5
.

A
p(x) = a

+ alx + A2x

n
+- a nx

then' we have -

=(09a0

p(0) =

p'(0) = al = (11)al

Z4b

,
-p"(0) r 2a2

(29a2

p"(0) = (2

3)d3 = (3)

(4)(

0) = (2 .3 ,4)a: = (40a4
"

= (2 .3

p(n) (0)

f.: x0--,sin x;

If

= (n! a

then, 'from (la), we can write


%k %

da0 46f10) = f

rt)a

1 =)ft OCI,

_i

(4)

N
(0) = f (8) (0/
= ;.
.

(5)

1
(0/
= f (9)
'

(01

,..;

(6)
'0 = f(p) = f
(01 = ji())(0) '. ..
-.,
,.....1..!

0-1 = f'''.(0)
..--,

To satisfy

fkm-479Noryfk,1

4) we must have
.,

(0) =

v(0) =' f(0);.k,pq0)


= fl(0),''13"(0) = f"(0),
,

Nowwe observe that

(13)

-all the coefficients Ofterdts in

of evenIdegree are

0..

The ocid.degree coefficients are


al = 1,

(14)

a3 =,-

1
,

For exaniple, the polynomial

a, =

'

p ,of degree

10

or less which satisfies'

the conditions (4) iS given by


.

p(x) =

= x

tx3
X5
A. "5(9
- -37 +51- - 71,4-t

x3: x5

x7
'x9
120:: 5E76 + 362,880

In summary, for each positive integer


function

of degree not exceeding

n
la,

to

there is precisely one poly4omaal


such that

andits f

st

n.

4-5

derivatives agree

n. derive-

with the sine function and its first

x.=

tives.

The polyp

ials

are known as the Taylor approximations at

to tie sine function.


.

A similar process will yield the Taylor approximatons to the cosine


.

-1

factiod:

'

it

x
1

1,

-2:'

x4

41'

2:

you

In Exercises 4-5, Number 8

Approimation Error

e asked to obtain these.approximations.

.
.

4
-

Now we wa nt to determine the accuracy of successive Taylor approximations.

The.approximations,

y = x,

y 7 x-

x3
,

y = x -

x3

x5
+ 5T ,

y = x

Y = x
.

Qt

to' y = sin x
.

x5

x7
,

and

x3
,3

x9

x3-;1

,t5

x7

-r

7
5.

-7 + -T
9.

3.

are graphed in Figure 4-5a.

7.

,,

Note that as the degreeOncrease0 s,

the Taylpr approximations become better in the sense that subsequent approxi,

mationsdmprove the "fit" near zero and also give better approxictions
further away from zero.

0 10

292

302

Jr
)

Y1

-pi

y., = x

x5

x3

x9

x
x7

.
/
1

--

' ;4

II.

/
I

y = sin x
S .... ...."

t.

3.

\\

'

\\

\
X

3
X

Y
Y2 'x

3t

\\ ,
\

.\

X3

X5

X7

'51.

7!

+ --- -

--

Figure 4-5a

Wye shall prove in :chapter 7 that these approximations are alternately to

large and

In fact, for ,x > 0,

small.

ti

-ix.> sin x,

-.

'

x3"

x - -- < sin x,

3.
.

x -

x3

x5

> sin x,

Hence, the error made in using any oneof:these approximations is easy tcs
estifate.

For example, if we use the,approiimation


.

x,-

'sin x r.. x , 7.,

In practice
5
since there is no need to compute values of sin x to /.

the 'result is too small Py an amount which is less than

x -is.less than
X

>

1,

In this case, each of the successive terms

4.

.. .
e?

.*+.
.

x5

x3

' k

'

3'.

t
x

is smaller than the preceding one..Morreover, since r <


'

we can approximate

sin x

7 when

large

as closely ds we please by choosidg n


4

enough!

x < 1,

'

-k

;Similarly, it will be shown that the approXimations


,

..
'

1,

'It

to

cos x

x > 0,

are alternately too large and too ,small so that, for


. .1 > cos x,

2:

< cos x,
4

FF+ u>,cos x,

1 -

mo9
.

fir

,Henge, for example, the error in using the approxigtion


0 .and

1 -

2
2:

,
is between

x4

Example

4ta.

Use Taylor approximatiohs to estimate

sin 0.5.

Let us begin with the approximation


sin x 2 x

which gives

x3

.0.125'

sin 0.5 z 0.5.

The result is too large by an amount less than

0.020833...

If wd use the approximation :sin x z x

x3

"\;

we obtain

sin 0.5 2 0.5 - 0.020833:.. = 0.479166",

small by an amount less than


cludp that to

x5"

0.03i2 5

Tr. =

120
decimal place accuracy

2914N

0.00026.

This estimate is too


We can therefore con-

in 0.5 Z_10.479.

4-5
2

Example

4-5b

Use the approXimq ion

,x

for

cos x. to estimate

2'.

cos 0.2.

Also estimate the error pommitted in using this approximation.

(0.2)2'

cos 4.2 %. 1 .

,27
o.o4

7 0.

The error is less than

0. 0016

0.000067:

It is often uleful to irite our results in thrms of a remainder

'

R.

Pot

example,,

3',
sin x =yx

lo

where

and

0 < R<
2

co s x =

'

31-- +

4
< RI < rx

where

4.

We can Use,the Taylor'Etpproximations to determine certain limits.. Let

us beginwith some familiar ones.

Example

4-5c.

To find

lie
x-->0

We can write

sin
x

and

1 - cos x.

lip

x.

x-0

where 0 < R <

sin x = x - R,

3t

Asin x,-

1 -

"x

0 <

Since

R'

x2
---

2:

we conclude that

approaches zero and

approaches '1.
Similarly;
.

cos x = 1 -

+ R',

<
2

Hence,

1 '- cos

x
x= Tr

4
r

29g 0.5

0,

R'

Since

1 - cos x
x

-and

2i

and

R'

x
1

both approach zero

-x,

lim
x
0

lim

To find

Example 4-5d.

x
.

1 - cos x
sin x
0%
x

cos x = 1

Since ,

1 - cos x
- 0.
x

(0 < R' < E-)

+ IV

' .

'.

(0 < R <

sin x = x 7 R,

and

,
x2
-- - R'

x -

sin x

R'

1 - '

The numerator approaches

the required limit is zero.

This prodess can be abbreviated by:recognizing that

1 -

cos x

2-

and

sin x = x.
ti

Therefore,
_,,X2\

which approaches

as

x2

2/, _
x

approaches

0.

re

296

30

1.
(

-.

1 - cos x
'sin x

and the denominator apprdaches

2'

Hence,

4-5
This example may be done in still another way.
1 - cos x = 1-- cos x

simx

1 - cos x

x
sin x

approaches

approaches
sin x
We can multiply these two limits to obtain the required limit.

FVom the previous example


1.

We may write

and

It can be shown that the Taylor approximations are the best polynomial
approximations near zero.

For example, if

f : x -'sin x

an

then
f(x)

p(x)1',.

as

approaches

and

0;

approaches

is the only pblynomial of degree

with this property.

or less

x3

it

4-5
I
Exercises 4

(b)

(i)'

(ii)

f(1)(x)*

(x)

(ii,i)

.(iv)

-) Cos

Given:, g
(i)

determine

x,

8(31)(x)
g

-2.

(.1)

determine

-' sin x,

(a)"

(42).
(

x)

fr,
f

(35). xi'

(18) (x)

g', g", gm,


g

(20)

thin find

),

g(14),

then find

(x)

g(101)(x)
'

deterdine the value of

x,

(i)

(iii)

fm(ii)

(iv)

(ii)
(b)

Given:'

(i)

3.

(a)

(0

given:

(10)

5g.

(-6-)

gt(i)

(iii) 011(1)

8"Pir-)

(iv)

f : x

A sin ax,

(ii)

f"(X)

(iv)

x-4B

cos bx,

(4)- 5g

(--(7)

find
(iii)

'Given:

(4)

find

x -) cos x,

f'(x)

find

81(;)..(ii)

(a)

(i)

-"

r(i)

Given:

(iv)

g"(x)
f

x ->3 sin nx,

(it)
g

,
(*)

find

(i)

ft(x)

(iii)

fftl(x)

(ii)

e(x)

(iv)

fiv(x)

g*(x)

(iii)

gm(x)

g"(x)

(iv)

giv(x)

Giventz .g f

2 cos 2

(i)

298

1047

4-5 .
5.

If

-) 3 sin (i +

f(0)

(al

ft(x)

(b)
(c)

(d)

'

(e)

6.

If

find

(i)

f"(x)

(i)

fm(x)\

(i)

f(4)(x)

g-

4a)

-2 cos (2x +

'find

g(Q)ti

().) 7,(,i)

g'(x)

(:k1(1)
e

811-(:)

(d)

ii). g"'(x)

(e)

(1.-)

04)

g'(v)
(ii)

g"(-

(ii)

g'"(0)

12

8()*

(x),
-

7.

Find a formula for the nth derivative Of the' sine function.

8.

Show that the Taylor approximations to

. x --t

x2

2!

2!

cos x

are given: b3

+-7

and, in general, by

1
where

is even,and

9.

(a)

9,

k xn
(4'1)

n!,

k =

$N

.1

For Numbers

10, and 12, use the Taylor approximation for


using

Calculate iin(0.2)

= 4.

sine

aria...t

s-.

1.
1.

(b)

`.

Estimate the error.


.

10.

(a)

Calculate pos(0.2)

(b)

Estimate the error.

using

n
.

r
4

k-

.299

309

cosine.

;,

11. (a) "Calculate

cos

using'

n = 2

(i)

'(iii)

r-

=4
=6

In each case estimate the error, ascertaining the number of places


.

of accuracy in the approximation of

cos

1
a .

Show that the sine function is not equal to a polynomial function onlpany
.

interval; i.e., given


x

a < b

and any polynomial

p,

such that

there is a number
.,

11

a < x"< b
sin x = p(x),

Suppose

(Hint:

and

sin x #

a < x < IS

p(x).,
and differentiate 'several

times .)

Find the limits of the following expressions as


(sin

approaches

0.

(a)
X

sin x

2
-

x6,

(b).

4x2 .sin. 4x
(c)-7

te

x2 =

{Let,

4x2

and find the limit as

),0) /

6
3x - sin 3x

(Let

(Let ....3x .= t)

6x
2

1 - cos .x

x
1

x3

Ps,/

(f)

".

300

I .

14.

Use Taylor approximations for 'sin x

cos x

and

to evaluate each of the

following limits (if the lithit exists).


NN

---

sin'x
cos x

''''''-'

lim

(a)

x -4 0

lim

(b)

7-

cos x - sin
x

x -4 0

P
V

lim

(c)

4.4p+

4x )0+"

(Note:
15.

means that

approaches zero from the right only.)


C_

Find the limit, of each of the folloying'expressions'as

approaches

0.

sue_
x

(a).

x_

sin x
- cos x

(b)

nt
x

i
sin
x

t-

(c)

cos X)2

(1

16.

Show that the Taylor approximations to


replacing

17.

by

2x

inthe approx4tions fox

Taylor approximations at
(a)

xl )sin 2x

x.

n,

there is a unique poly -

such that

deg p

(i)

a /,0

Show thpt for each positive integer

nomia4

can be obtained by

p(a) '= sin

(iii)

the first

values at

x = a

derivatives of

have the same respective

as the first

derivatives pf

x 1-4sin x.

These polynomials are called the Taylor approximations of

the sine function at x=ja.


(b)

Show that for

(c)

g'n a

p(x)

'

si n a

+ (cos Fa)(x -,a)

2!

(x, -.a
4

Show that, in general,

'

.p(x) = f(a) + fT(a)(x-.a) +

f"(a)

* f (n) (a)

2
(x - a). +

(x.. a)n.

it

d) Derive a formula corfesponding to that 4n part (b) for .the cosine


Taylor approximationS.

501

A
J

''

-1

3 1 I
t

5-1
41,

Chapter 5.

EXPONENTIAL AND RELATED FUNCTIONS


Ak.

the preceding four chapters we studied polynomial and circular functions, the former being defined algebraically, the latter in terms o& arc
lelith bn the unit circle.

In this and the net chapter we take up the study

of exponential functions'and several otherfunctioni which can be described


in terms of exponentials.
x

In brder to define an exponential-function, such as

x
,.,

we must showhpw.iptational powers are defined.

Then we must obtain

the properties of these functions.


.

In order to provide a basis for our'dvelopment of exponential furictions


and their inverses, the logarithm .functions, we beginAhis chaptei- with

Peview of thailaws of exponents..To assist us Pin our discUssion`of exponential functions, we Show in Section 5-2 how.the function x -uN2x _serves as a

model for growth.

The laws'of rational exponents and the fact that

increases as

(rational) increases are established in Section-5-.

method fon defining irrational powlSof

2x

is indicated in Section 5-4,

4 where,it is noted that laws of exponents for arbit*ry real numbers hold and
that the graph'of the resulting function x -42x is rising and has no gaps.
These facts are used in Section 5-5 to obtain-the definitions and propertiet
of the general exponential function.

The final two'sectiOns use the inverse

eoncept to define and analyze the logarithm functions.

-1.

'

Exponents '

You are doubtless familiar with the,behavior of exponents.


.

Since, how-

ever, we shall need to use them extensively, it seems wise to put our know-,
ledge of theM in order.'
Le

s consider the sequence of.numbers

2,*2 X 2, 2 X 2 X 2, 2 x2,x 2 x 2,

...

which we abbreviate as
(1)

,using thel'exponents1 1

1
,

3
,,o2s

.4

, 2

3, 4, to indicate the number of equal factors.

I.

3,01

Can we give meaning to


0
2

-1

-2
,

2/2, 22/3, t

or to

equal factors or

-2

Certainly we cannot talk sensibly about

equal

4fdtors0

If

to the exponent

look at the list (1) we note that aftiltion,of '1

results in a multiplication IL

24

Thus

2.

this principle by saying that subtraction of


1

result which is

23+1 = 2 23 We can restate

from the exponent gives a

Thus,

the otiginal one.

23 = 1(24),.2.2 = 1(23),
2

from the exponent

in

We can get to

represent?

20

What number

20,

by subtracting

Therefore, if we are-to'maintain the principle,

we should have
2

]..;

.=

-t

from the exponent, We get

Subtitcting another /
2

-1

0
1
= -fke

Continuing in this fashion we find successively


2

-2

=3

2 ,

-4

1
=
=

g1

= 7,

'

and generally'
2

-n

1
=
1

2n

,for every positive integer


To interpret

1/2
2

n.
.

we shall assume that equal increasmin the exponent

correspond to equal ratios of the numbers.

%w7
With integer expoffents this prin-

ciple takes the form that each incr.ease'of

in the exponent cotrespondi to

a ratio o,'

What ratio corresponds to an increase{ of

2.

this unknown ratio

r.

Then
.

304jQ

(.1

1
?

Let A call

5-1
.
t

t t

21/2 ='

+1/2

2i1

and

r 2 = r 1 = r
r

21/2 +1 /2-

/But

1/2-_ r2
r

916

21
61.

Hence:
4
r.,,

and

1.-

....
-

.,

'>

'

= /E.
.

We reject the possibility

which would-not t nicely do the

r = -1/

(See Figpre 5-1a.)

grapild

Po.

4.4

FigUre 5-1a
Powers of

2
,

similar argument shows

that to maintain regulartty we should take


21/3 = 3AT

22/3 .

andso on.

3.

Generalizing, we are led to define

2=
p /4

12 /75

where-,p

and

are positive integers.

1%0
We can reverse our principle and say that every time we st.lberact a given
amount from the exponent we divide theaiiimbelby a fixed amount.
clude that

.3953
t

i4

Then we con-

5-1

-1/2

1
=

-2/2

, 1

ig

JE

-3/2

'

1*

: 27-..:::--

'2

end so on.
.

,,

`--

ac

..,

With these interpretations of negative integral exponents and functional


exponents,, we see that if

and

are_.any
two tational numbers then
--a
g

..

2r

'

2s = 2

r+s

'

and

(2r)s

More geprally, if ,a

2rs.

4
is any positive number and if 'r

rational numbers then it turns out that


'

(2),

eras = a

r+s

and

(3),

c(ar)s.

arS.

These equations express the familiar laws of exponents.

4,0''s

are two

Elercises,5-1
Write each of the following as a positive power of one number

1.

(a)

x5

x.?

(a3/5)5/3

(b)

'

(g)4 ;TLIU

10-7

.169

62/3

(c)

a3 /5.

a5/3'

(h)

3 .2

1/2
32

32

2 24
(d)

2 (73-).

(11

-- o
1/2

ti

22

i(i)

(e)

1.1

814/3'

23/4

r.

(J) '25 2755 (125 )-2/3

Find the value of

2.

if:

14m-r _442/3;

= (23)2;

(a)

8rn

(b)

8m .=

(c)

2(45) = 16m;

(a)

(24)5 = 16m;

2(32);
/

if)

571

= 0.2;

(g)

(pm

(h)

17m = 1

I;

O
.

Eva

'3.

100018-2/3)

ex.

3(9)-3/2.

-)4.- Area ge the following in order of magpitude:


22/3,

5.

(1)=riti;

..

Show that if

x =

22 ..7

then 7x = 4

2-3,

(2 -2/9)9

10,--7,-

.,.

.
.

(45/2)0-1),

6.

Carry out an argument, like that in the text 0 show that


.0.-

31/4

(b)

4-3 =

41/31.

43
(c)

2:1/2

500"

307

31G'

'41

-2

The Exponential Function, Growth and Decay

5-2.

In this and the next three sections we shall be concerned with assigning,
a meaning to

where

a > 0

and,

is an arbitrary real number.

This

will lead us to the general exponential function


x

'where

is a constant

The number

kax

is called the base of the exponential

function.
./

An exponential function can serve as acidealized mathematical molel for


growth and decay.

Suppose a biologist grows a colopy

Let us consider growth.


/

He wishes, to study, how the number, of bacteria

of a certain kind of bacteria.

Under favorable circumstances he find:, that so long as

changes with, time.

the food holds out, the time required for the number of bacteria to double
His

does. not seem to depend on the time at, -which he starts the experiment.

hypothesis is that .the time required for the bacteria to double does not

depend upon the time when the initial count is made. 'This is one instance
of a general growth principle which 'is important in social, phytic0s, and.
biological science.
(

N0

To be concrete, let us .suppose that on a given day there are

bacteria

presentsand that the number of baoteria doubles every day,' Then there will be
2N0

,twice

present one day later.


2N

the number

22Nd after three days twice

or

.After another day the number of bacteria will be


0

23..2,:.
After
v
of bacteria present will be given by.tne';qs,tat'ion

N(n)

2 N.. ;or

days

ti
0

N(n) = No2

(2)

__where

n
,

is a positive integer.

If we asamme that the number of bacteria increases steadily throughout


/

any given day,,ye might want to determine how many bacteria are present
after the s

rt.o; how many were present

1
2 .f

II*

1-f

day

days before th'e initial &mint

was made.

To answer these questions we must generalize equation (2) to


N(r) = NO2r,

where

pity take any rational-values, positive or negative.

expressions as
to

and
J5.

271

If such

are to have meaning, we must generalize further

308

3j7

-r

5-2

where

is an arbitrary real number.

Let As suppose that (3) serves as a model for the growth of the bacteria
colony.

Can we 44duce from this growth equation that the time required for

the bacteria to double doesn't depend on the time. the initial count is made?

Since thiswas thgAnitial_hypothesia, it should be true if (3) is to serve


as a model.

SinCeye_pan be certain of the meaning of -2

only, when

is

rational '(from SeZtion,5-1)we begin with that case to check the'hypothesis.'


Suppose we take a count
take another count

days after the 'start of the experiment and then

days later:

x and

Here

numbers, positive, negative, or zero.

may be any "rational

From,(3) we have

N(x + t) = NO2x4t.
-

m-Seet-i-on-5-1,-_tlie_exponent la

'r

,(2r +s

= 2 2

for- -r

and

rational)

gives

N(x + t) = N0 2x

Replacing

by

N 02

N(x)

(4)
414

we obtain
N(x + t) = *2tN(x).

In words, if 1N(x)

number N(x + t)
The factor

isrthe bacterial count after .x

of bacteria after x + t
does not depend upon

depends only upon

t,

x,

days

is

days, then the

2s times as great.,

the time of tfie,first count; it

the time interval between counts.

For example, suppose that there are one million baiteria ruSent initi,a1g4r;
i.e.,

N0'= 10

6
.

Then the number of bacteria one-half datlater is given by


N(2),=

106(21/2)

=10

Aftegone,and one-half days the number of bacteria in the colody is


;

N(p.... 106(23/2)

r= 106 (21,2)
r-%
= 200 6Y2)

= 2 N(t)

309

318
d

1'

5-2

If we use (4) we obtain the same result:


N(i) t N(i" + 1) =2N(2).
ea:

Assuming that the conditions of growth were the.same prior to the initial
count, we Stu:mid expect, that the number of bacteria one day before the' initial

cdunfis taken would be


,t!(=1), = 1.06 ,

-1

= 500,000.r

An exponential function.with any positive base can serve as a griowt,IN


model.

If,we-replacie the assumption that the number of bacteria dolpot each

day by the assumption that the number changes by a factor of

a_

each day, we

obtain the model


N(x) = Noax.

Si ce the laws of exponents hold for rational exponents and an arbitrary


base w

obtain the relation

N(x + t) = at N(x).

We see that it is also true in the general case that the growth factor
depends only-upon the length of the time interval

observation time

(x).

3100

(t)

(at)

and not on the first \,

4,

LI

Exercises

In the first four exercises consider the equation

N(n) = 106(2n),,

where N,(n)

represents the number of bacteria present at the end of

1. ) Plot the points for which, n =,0, 1, 2, 3, 4


smooth curve.

days.,

and connect them yith,a'

(The unit chosen for the vertical axis may be one million.)
7-*

2.

The bacteri'a count at theend of


as the count

3.

n + 5

day's is how many times as great

days after the beginning of the experiment?* -

n + 2

Onp week after the inftial count was made the number of bacteria present

was h& many tunes as great

the number present three days before the

experiment began?
4.

bacteria present after

If there are

days, after how many days

;.00

weri there k present?


5.

200,000

Suppose that in a new experiment there are


at the end'of I4ree days and

Compute:

days.
(a)

the number present at the end of

(b)

the number present at'he end Of 1-

days;
1

days;

(c)

bacteria pFesent

1,600,000 present et the end of 47

the number of days at the end of which there are

806,000

bacteria present.

issume that the number of baCteria present at the IINginning of

Hint:

the. experiment is

and that at the end of

a N0

24

6.

hours the count is

The number of bacteria in a certain culture is obseryed to double every


day.

If t

reewere

105' present at the first count, the number of

bacteria

fl.t after

days is given by

N(t) = 105
days?

(a)

How many Were there after

(b)

HOw many were there one day before the count?

2 -days?

2t.

two days before the

count?
.

(d)

How many were there one-half day before the count?


aftpr the count?

ti

What is the ratio of

,311

320

N(2)

to

one-half day
N(- 2)?

5-2

7:

Suppose `N(t).= Noa


(a)

is 'the number of bacteria present at time

If this formu a is to represent growth,- can


.

la with

Use the/for

a =

1
,

a = 1, . and

be-smeller tlian

a =

t.

1?

, to ;stetch graphs- - -'

illustrati g your answer.


(b)

Calcula
41,

N(t'+ 1)

NM'

77.7,

N(t + 2)
N(t + 3)
N(t + /) ' __N-(t f 2)

(c)

Ie general whbi is the value of

(d)

Suppose that

MItr + n 4,1) 9
N(t +..n)
,

N(1) = 105

and

N(2) = 106.

Find _N0

'

.....

and

a,;/

radioactive substance (such as radium) decays so that the amount

8.

/present,

N(t)A

.,--

at time

is satisfactorily given by the same I

formula as growth:
N(t) = N at,
0

where- a

is a positive real numberr-=-

(a)

What is the amount present at :time

t = 0?

(b)

If this is to represent Aecay, can

(c)

Show that the ratio .of the amount present at time

t + 1

amount present at time

and .is smaller

than

exceed

doesn't' depend /upon

1?

to the

1.

No
lttt

If .N(t) = N0( --)

find the value of

for' which 'N(t) =

JO

312

0 2
.7-

...
_

'5-3
1

More About Rational Exponents


sr

Before attempting to define irrational exponents we shall further examine


the'function
r -4 a

where

a > 0

and

a = 2;

r' is a rational number.

r ) 2r,
r

and

To be concrete let us suppose

we consider the ?unction

( 1)

For

rational.

r,

rational, the laws of exponents are

2r2s

+s
(2)

0-4_6

2rs

(2r)s

We can show that the function (1) is increasing; that is,


2

< 2

if

and

are rational and

r <-'s.

We firsts observe that

if

(3)

a > 1,

where

then

> 1, a 3 > 1,

4.., a

> 1,

is a positive integer.

.r:

Similarly we'note that


(4)

a = 1;

if

/'

then -an = 1,

and

0 < a < 1,

if

(5)

Now we assert that


m/n
2
were equal to

which is false.

m/n

1,

> 1

then

for any positive integers

If, on the other hand,

neither'less'than nor equal to


r

and

m and

n.

'If

wouldleaa to the result an` =.2m = 1,'


then 0) vwould
2m/fl

would lead to the result 'Ian = 2m < 1,

Now let

0 < an < 1.

were less than

it must be greater than

1,

1, .then,...(5..)

which is also false'.

Since

be any.twp rational numbers such that

s -tis.a positive rational number

2m/fl

is

1.

r < s.

In conclusion we have

Then

d
/

2s

-r

. 2

m/n

2r(2s-r)
2s > 2r.

32.3

3.22

> 1

2r

r -)2.,, where

Is rational,

unc ion.

is aryfcreasing

... /

'

If

is a large lager then certainly

Ailk

is very large.

Estimates

'

can be obtained by usIng the binomial theorem to,expand

2n

of the size of

n >

(1 t f)n. .For example, we have, for'

(14- 1)n = In An In -1 .1 + n(n

12 +

1) In -2

All of these terms are positive, so if we cmit any terms we can-only decrease
ms we
In partibUlar, omitting all but the first and second
the size.
obtain

>

2n = (1 + 1)

6,(6)

+ n.

Consequently, if-we go far-enough to the right the graph of


x

y = 2x,

(7)

rational

must lie above any given horizontal line.

An6ther consequence of (6) is the

fact that the, negative x-axis is an asymptote fot the graph of (7); that is,

the graph of (7) approaches the x-axis for

k negative and

Ix!

large..
,

show this we can take reciprocals 1(6) to obtain

:(8)
Since

-n

r and since
= -2

.0

1 + n

2**

approaches zero as

Inl

increases, then

1 + n

y,= 2x

1-:.g.

approaches zero as

becomes negatively infinite.

\In summary, the graph of


!x - axis (when

x < 0

iises rapidly for

and

!xi

y = 2x,

rational, increases from near-the

is large), crossing the y-axis at

(0,1),,

and

x > 0.

.;---.tr.

314

323

*.

5-i

.'

r,

4
-

Table 5-3. . Values of

) 2-r

.001

1.000 693 4

.005

1.003 471 7

.01

1.006 955 6

..02

0.999 307 1

0.99

54o 2'

0.993 092 5az


0.9vo 23

)1

1:01 3 96

.03-

1.021:01 -

0.979 42

.64

1.V 11

0.972 66

..05

1.106

.10
o
.15

26

0.965 94

1.071 77

,0.933 03

1.109 57

0.901 25

1.148'70

0.870 55

to

-"J

.20

.25 .

1:189 21

0.840 9b

.30

1.231 14

0.812 25

.35

1.274 56

0.784 58

.40

1.31951

0.757 86

:45

1.366 04

0.732 04

1.414. 21

...0.707 11

.50
.55

1.464 08.

0.683 02

.6o

1.515 72

0.659 75

.65

1.569 17

0.637 28

1.624 5o

0.615 57

1.681 79...$

0.594 6o'

.70
.75

.1

.8o

1.741 lo

_85 -

1.862 5o

0.554 78

.9o-

1.866 07

0.535 89

.95

1:931 87

0.517 63

a.

1.00

Rational Values of

2-.00o 00

0.574 35

-6.560 00

2r

Wile 5-3 gives rational powers of


use the entries to three place accuracy.
used to find

'

2.

Ordinarily it is sufficient.to

The laws of exponents (2) can be

2r for values not,listed in the table.

"a.

>

3!S2 4

5-3

DiT2212 5-38. Find

21'68.
de

We note that
2(1 +

21.68
=

1.

1..1

Find

Example 5-3h.
--

2r

0.65

000,

4 .03)

20.03

20.65

21

2(1.569)(1.0021)

0.57

write
+.0.63

2-0.31 =

2...(2o.6o oro.03)

11213.6o)

0.03)

'U*516)(1.021)
2

EITIP12, 5-3" Find

Note that.

22,

0.774

43'21.

so that

"43.21

22x3.21

(22)3.21

26.42

26

20.02>

21374o

64(1.320(10314) = 85.663.

Later we shall be able to use Table 5-3 to maculate


expressions as '3r.

general positive

a,

8r, 16r and such

(In Section 5-5 we shall show how to define


in terms of powers of

2.)

ax, tfor
.

316

325
4.

.
fi'

5-3

Exercises 2:2
1.

(a)

.(b)

2.

5 4
2-/

Calculate

by using the data in Table_ 5-3;

by noting that

2514 = 2

21/4

= 2 .

liding.the data in Table 5-3?' calculate


(a)
('0

21.15

.y.65

(c) 20.58
(d)
3.

4.

2-Q.72

With the aid. of Table 5-3, compute


0. 84

(a)

(b)

0.25-0:83

Extend Table 5-3 by completing the following table.

Table 22 (extended)

Values of

2r

2r.i
'

-3.6
-3.2
-2.8
-2.4

r..

-2.0
-1.6
-1.2

ft

..

1.4

'1.8
2.2

'

'

?6

3.0
e

Ak

Plot the points

(x

2x)

for the rational values of

shown in

1%.,._
e

Eibie 5-3 and Table 5-3 extended (Number 4).


6.

(a)

For what positive values of the constant


f

increasing?

decreasing?

r -tar

constant?

317

32G

is the function

5-3
(b)

For whht positive valus of

is the function

0r

increasing?
(c)

z.

we.

constant?

decreasing?

For what positil7e values of

is the function,

-r 2
f

increasing
(d)

If

decreasing? constant?
for what values of

+3

r '-

.ShowAhat if

n > 2

is the function

(2b + 3)r

then
2

>

MO.

nkn - 1)
2

I'

(b)

constant?

decreasing2

increasing?
(a)

(a)

'

Use (a) to show that


100
2

99

100 > 2 r

210,000
116,000 >

(ii)

(e)

As

9, 99
2

becomes large does

2"
n

become large?

Justify your'answer.

ti

3.

318

327-

5-4

onents

Arbitrary Real

ctions, we dealt with the properties of

In the preceding
for

We want to give meaning to these expressionS-if

rational.

irrational.

ax

and

is

'For example, we want to assign meaning to

n,

3
x

To be specific, we want theexpression


irrational values of

x;

to be defined in a natural way for

that is, we need to extend the function x -)a

that its domain is the set of all real numbers


suppose

2x

a = 2.

x.

To be concrete let us again

In the next section we shall show how to define

general positive

a,

in terms of powers of

(X

for a large number of rational values

ax,

for

2.

Of course, the meaning of "in a natural way" is ambiguous.


however, what we wish to do is clear.

so

Geometricqlly,

After plotting the points

2X)
then we just connect these points

x,

with a smooth curve and oetain the desired graph of `


x
x --) 2

Then, for example,

2 2

is calculated by measuring the second coordinate of

t11, point on this graph whose first coordinate is

(See Figure 5-48.)

(2,4)

The second coordinate of


this- point should be.,
2

(0,1)

if

Figure.5-4a.

Finding
"holes" in

by filling in the
x 742x,

319

rationat.

'

s/

This graphic prooess js not quite satisfactory es it doesn't lead to pre.

cision as to the meaning of

We shall outline an approximation procedure

which will enable us to define 2if

and in general

2x

for irrational

x.

The function so obtained is increasing everywhere; that is


)

if

u < v,

then

'2u <*iv.-

In fact, there is only one increasing function which has the values
all rational numbers

r.

u+r '= 2 u 2 v

hold for,all real numbers

and

and

for

(2u)v,= 2uv

(rational or irrational).

The graph of

x -2x. has no "gaps"; that is..


for any positive number

(3)..

For this function the laws of exponents

(2)

this function

such that

y,

there is an

2x = y.

. In the sequel lie shall assume that indeed

is so defined that,

(2) and '(3) are true and examine the consequences of these
assumptions.

Now we turn to outlining the process used to define

If we wish the '.

.increasing property (1) to hold, then for all rational numbers


such that,

and

s,

r<

we must-have
8,
(5)

...

2F < 2

< 2s.

Obviously this places a severe restriction on the value we assip-to 2


and,,.
as,we shall see", determines it completely.
The ordinary decimal approximations...,
,

to V give us_a handy .collection of values.f16r


.

4'.

,-

and

s;

we know that.

'

,
1

)..4

1.401,
.

fit

'

< 1f < 1.5


.

<12- < 1.42

1.414: < 1 <'1.415


,1,414'2

_1.411121

< Vg,< 1.4143

< If

<1.41422

and so qn.' The iriequa ities (4) and (5) then show that

2'7 must sattsfyethe

'4
320
-

5-4

1.4.

< 2

21..41

< 3

1.414

< 2

21 4142

If
if

< 2

if

< 2

1.42

1.415

21.4143

21rf

1.41421

1.5

< 2

if ) 1.41422
2
< 2

<

and so on.

We replace the rational powers of

appearing in the last set of

inequalities by appropriate deciffial approximations and arrive at the following


.

estimates for

if

1.4

2.639 <
2.657 < 2

2.664.< 2
2.665 < 2
and sa'nn.
2

if

< 2

1.41

< 2

1.414

If

< 2
< 2

<

1.4142

< 2

,/,"f

< 2

1.5

1.42
1.415
1.4143'

< 2.829
< 2.676
< 2.667
< 2.666

Thus;' if (1) is to hold, we know that, to

decimal places,

ff

= 2.665

The pinching down process that we use to estimate

if

is indicated

in Figure 5-4b.
V

1.4142

1.41

1.41

1.415

21
.

4 2

1.42

21
4,

.2.6

2:7

2.8

Figure 5-4b
,,

Pinching down on
TI:0 generalize to any real number

rl, r2, r3,

Yn,

decreaSing sequence

than X

x,

2'6.

we choose,any increasing sequence

ofrationalnumbers all less than


si, s2, 53," ..., so,

such that the difference

r l'i

can be made arbitrarily small.


,.,,
',

3?).
'i-t

and any

..., of rational number's all'greater

330

,
r1

1.We:compute the sequence of numbers


s

sequence of numbers
,

intervals

< y

1
,

3
,

r2

r,

and the

, 2 J, ..., 2 n,

,-2

and then look at the

..., 2 n,

sn
.

s3

2 q-

pinching down on
2

x
r

crowding in on x.

41,

fi

rn
r

s2 sl
n

Figure 5-4c
Pinching down on' 2x.

It ip a, prdpertY,of'the,teal. numher.qstem tIlat as- x


s

confined by

to successively smal4r intervals, the corresponding intervals

and

on the y-a is pinch down to a uniquely determined number, which we shall


define'as

he number

2x.

The number obtained is independent of the particular


.

choice

n"

of

the sequences

rn, ...

ri, r2, r3,

;..

3223

and

si, s2, s3,

r'113 "-*
1

SHIM

... ....

OS.

.....
..............
::: ::::::::::::

SS

MSS

......

SS

OS

S OY

113

sass
.....

11

1.1

Is

r1

-...

T:fin:2112

SEIM

.... ....

::: :
MEWL.

" " .......

31U.

..........................

....

WRENN

111

COO

NA

.....

..
.......... . :

'11::..... SS .........

:
I
I

WIN

e:

........

1:11:313:13
L..13:1111'11..

....... elas

..............
Wan ...............
%SS

.........

SW=

...........
....................

e.

3111:1-'13111111:11:11

................

/Il :

i..

....... ............

....................

00

...

00

.....................

Of ......

..... ........

.......................................
..........................................

II ............................. g

........

.........................................

.......... :
RS

........................................

II

......
1/8

5-4
4 Exercises 2:4
1.. Use the graph of
2

2
2

x --)2x

t5) estimate the value of:

1.15

2.65
0.58

2-*72
2.

Compare your results in Number 1 with your answers to Number 2 in

Exercises 5-3
3.

se the graph of

(a) 2
(b)

x -)2

to estimate the value of:

17 f

2g

(c) -.2-g/4
4.

Is there any value of

for which

= 0?

Gie reasons for your

answer.
5.

Use the graph of

x -)2x

r.
to estimate the vilify of

if:

(a)

(b)

23c

= 6
'II.

(d)

2x = 3

(e)

0:7'

,4

!I

= 0.4

(c)

iim

--'

= 3.8

= 2.7

*a

6rL

Fl
324

5-5

5-5 Powers of the Bape

as Powers of

We have concentrated on the function


f

x -)2x.

You are now familiar'With its graph andhave worked with a table of its Values.
We shall now study the function
f

where

x -)a

is any positive real number.

from1scratch because we scan express

Fortunately we do not have to start


as a power of

as lie proceed to

2,

show:
o

The graph of
right.

Also,

x -)2

f(x) = 2x'

lies above the x-axis and rises from left to

becomes arbitrarily large for

sufAficiently ,far

to the right on the real number line, and arbitrarily close to zero for all
..

sufficiently far.to the left on the real line.

The graph has no gaps.

sequently, if we proceed rrom left to right along the graph,


steadily in such a way that any given positive.number
once and. only once.

a,

increases

will be encountered

That is, there must.be one and only one value-of

for which

a=

(1)

(See Figure 5-5a) and therefore

may be expressed as a power of

a
Figure 5-5a

Graph of x

Con-

showing that

325

334

= a.

2.

x,

say

**)

5-5
We can find the value of a

by means of the graph (Figure 5-4d) or

Table 5-3.

Example 5-5a. Find the value of a. for which


We loo

1.11 = 2a.

1.11

or

in the second column and read backward to find the


0.15
corresponding value of a in the first column. Thus, .1144 = 2
(approximately).
J

Exam le
Welhave

3.25

Example

-7c.

Find the'value of a for wh'ich' 2a = 6.


x -.2

The result is

6.

we look for the abscissa corresponding to the


2.6

(approximately).

./if we use Table 5-3 to express


,

6 = '22(1.5).

2a.

21.70.

21 (20.70)

2(1.625)

On the graph of
ordinate

3:251in.the form

Express

'

as a power of

2j

4- first write

Interpolating in Table 5-3 between the entries for

and 0.60 we-obtain

20.58 m 1.50.

Therefore a Z 2.58,'

by itterpolation.

The expression. a

for

o= 2

!fence,

irrational and

x = 0.55

2 . 0.5 8).
A 2.58
0..50) ,S4'2 k2
4 d

2.

a / 2

has not yet been

defined.

We Could follow the Kocedure of Section 5-4 to assign meaning to

ax when

is irrational.

the function

x -,ax

Since we can write

a = 2, we can simply define

by
x

{2)

ax

where

a =115.

The laws of exponents will holt


)

and the graph of

ax

axa"and

( x)Y = PxY

will have no g p

These dre consequences of (2),

as is. the fact that


If

13)

a >1,

thee X-*ax

To prove (3), for ezamplei note that if.

a <'0

(for 'if

'2a1 < 2a7;

then

that is,

2a < 1).

Thus tf

a > 1
x < y

y.

E9. < a

is increasing.

32d3

= 2'

then ax <ay

where

a > 0,

so that

(5-5
The graph

a;

scale by,,the factor

where

a'= 2a.

then we just "shrink" the

= 2,

end

is obtained from the graph of

a = 2a,

then

For example, if

a = 4

scale by a factor of

Roy changing

so that

0 < a.< 1
x
In this case the graph ,of x -*a

will be negative.

is obtained by changirig scale in

x -,2x

If

2.

and reflecting the g'faph in the

vertical axis. /Three cases are illustrated, in Figure ,-;b.

These considera-

tions will be useful in our subsequent discussion (Chapter 6) 9f tangent, lines

to graphs of exponential functions.

(i)x

x -)2x

Figure 5-5b
,

The following examples illustrate the use of formula ,(2) Table 5-3 in
0

calculating

ax.

'

'

4 4 4.

i,

.
(0
.

Example 5-5d. Express


value of
t

To

3.7

as epower of

0.7

2,

and find the approXimete

'

;:d

the value of 30T ''W first eZpi.ess i 3 as a power of 2. 'Thus,


1, 0.58
1.58
3 = 2 (1.5) =.2 (2
)
-',2
(approximately). :(Verify :tili's from Figure
in#

'

-\.

5-4d.).

'30.7

Now

(21.58) 0.7

*106

21.11

a
.

m 2

(1+0.10+0.01)

= 2 (2

0.10

m'n1.072)(1.007) = 2.159.

)(2

0 01

,'

327

3 3.6

." p

5 -5

Example 5-5e.

Calculate the value of

(6.276)0.4.

We note that
//

6.276 = 4(1.569)

22(1.569) z 22

(Verify this from_FigureZ4d.)


(6.276)'-!1

22.65
=

Hence,

2(2.65)(0.4)

(22.65)0,.4

(20.65)

(1+0.05+0.Q1)

= 2

21.06

20.05

2(1.035)(1.007) = 2.084

20.01
2

(approximately).

Exercises LI
1.

Express

2.

Write

"A

3.4 in the form 2a

of

2.64

in the form

2.64 = 2a

and then find the approximate value

(2.64).3.1
.

3.

Find the approximate value of (6.27)0.6.

4.

Find the approximate value of

(5.2)2.6.

5.

ShoW that if

6.

By finding a suitable value of d,

0,< a <%.,4

and

v > u,

then

express each of the following


i

function's in the form of


X-) 2
(a)

x -o 4x

(b)

(c)

x -3 (5.736)'

(d)

x -3(0.420)x

aV < au.

QX
.

(3.66)x

328

337

5-5
oV.

7.

_ are positive and different from

a ,and

Suppose

Consider the

1.

two functions:
x

and

ax

bx

'

which cpn be respectively written as


x
(a)

If

(b)

Su pose

a < b,

_,2ax

and

what is the relationship between


close to

b.

Is

close to

and

(3.

Illustrate your

(3?

ans er by completing the following table:


For these values
of b

8.

2.0/.44

2-.3
---.1

ax

ax = 2

Given:

ax

(D)

(b)

If

(c)

ShoW that

Igt

is increasing/ what is the 'sign of

a?

is decreaping, what is the sign of

a?

is independent of

f(xf,(+x)1)

x.

.)

If

(1)

10t

what ban we conclude(about the a*vtiuotientt

> 0,

'

(ii)
(d)

'

9.

(a)

If

a > 2,

If

what can we ;c
z,

q ,< 0,

Where, does

a ,,

(2

he graph_of

e.

,44-1

lude' about the above quotient?

xs:,

:show that

.for

> 0.

Jai .> ,Q)

`*

.cross the y- axis?

Does

,..,

your answer depend upon


,, ,

(b)

%;

c ',t'

J. '

i-

."

Find the' point(s) of inte_, ectTon,,if `any, of the two grapl2s.; ,


.

(c)

a?

.7

W,

.'..

--)

aZ and

.x
x -, 2(2a) .
.

Find the pctint(s) of' intersection,

I.

if any, of the

.!

,,

f ,,4 %,
.

.,..

two graphs:

es.

r.
,..

x -> ax' and

is a real number greater than zero).

(where
St

x -,b(ba)X

u.
.

(d) ,Find the paint(s) of intersection, if any, of tlft twographs:..


x -i4x
(where

and

and

x -> b

n.

,x

Om)

are real numbers, greater than zero),.

I.:

-:'

.'-

5-6

The Logarithm (Ease_21

5-6.

4 Since the exponential function


every horizontal line

y = b

where

x -02

is,increasiAg, its graph crosses

b > 0.

Therefore, the inverse function


.

is increasing, its gr .K h has no gaps, it becomes arbitrarily large as

be-

comes large and it comes arbitrarily close to the x-axis for


lx1

large.

In particular, if

x > 0,

negative And

there is exactly one real number

such that

...

JO%

2y =x.
This 'number

by

is called the "logarithm of

to the base 2 ".and is denoted

Thus the function

log2 x.

log2 x

log2 : x

x > 0

is defined-only fkz.

With base

2.

and is the inverse of the exponential function

These two functions are related by


log2(c) = d if and only if

(1)

2d = c.

'

In terms of graphs, this tells us that


If '(c,d)

lies on the graph of

x -4kog

then

(2)

(d,c)

lies on the graph of

and conversely.

X' --'2x

As was tha case for other functions and their inverses that we have
studied, the graph of
x -32x

x --olog2 x

over the line given by

can be obtained by folding the graph of

y = x.

(See figure 5-6a.)

139

-.

5-6

Calculations involving

log2

can be carried out using the relation,(1).

For example, since


' 23 = 8,

we know thAX

log2 8 = 3;

and since

-6
2

1
"we have .log2(.alo= -6.

In fact, any table of valuet..of the exponential.functton .6 the base

also give.4tUes of

log2.

will

For example, Table' 5-3 gives


2

.20

= 1.14870

so that
log

1.14870 2:

A number of useful properties, of

log2

can be derived from propertiea

of the exponential function by using the relation (1).

Some of these are

ti

-331

'140

.--"

(a)

(3)

log2 1 = 0

(b)

1og, 2 = 1

(c)

log2

((i)

log2 x > 0

(e)

log2 x < 0 'if

is an increasing function

x > 1

if

0 < x < 1.

The properties can be observed in Figure 5-6a.


example, let us prove.(c).

Y1 = log

Suppose

They can also be proved.

x0 < xl. 'Put

y0 = log2 x0

For

and

so that

If yo

were not*fess than

= 2Y

yl

and

= 2 1.

we would have

increasing function, the condition

y0 > yi

2y0 > 2

y1;

y0 > yl.

Since

is an

implies that

.
,)

.,

so we are
contradicts the assumption that x < x
0
1
This
log
is an
log2'x0
<
log2
xl.
This
proves
that
forced to conclude that
that in,

5:

0 ?

x1.

increasing function.
\,..

The laws of exponents


x+y

x y
= 2 2

and

(255Y = 2xY

give rise to the following logarithm laws:


(a)

log2 xy = log2 x + log2 y

(b)'

log2 xY = y log2 x.

(4)

For example, to prove the first of the;e (4a) we let


c = log2 y

b = log2 x,

a = log2 xy,

(5)

4g,

so that
2c.

2b,

Observe that

xy

is then also given by

so

= 2b

2c =

2b +c

so that
0}

b+c

= 2

332

3 41

5-6

Wp.conclude that a = b + c;

that is (from (5)),

log2,xy = log2 x + log2 y.

Formuld,,(4b) is left to Exercises 5-6, Numbel 3.

These formulas can be

used with Table'5-3 to calculate logarithms of numbers not appearing in the


table.

'

Example

Find

log

3,25.

Upon looking at the second column of Table 5-3 we see tOlt

3!25

doesn't.

4T

.appear.

Note however that

3.25 =2(1.625)
so that
log2 3,?5 = log'114,+ 16g2 1.625.

'
Reading from t.he second .column to the first column in Table 5-3 we obtain
log2 2 = f,

log2'1.625 r. 0.70,

so that
log2 3.25 z 1.70.

. The next example shows how inequalities for the logarithm to the base

are obtained from inequalities for the exponential.

EXample.5-6b). Show that if n

is a positive integer then

lbg2 n < n.

-)

Since.

and log2

is an increasing function we must trete

log2 2n >.log2 n.
Formula (14b)

gives
-,

log2 2n = n log

Since

log 2 = 1
2

we must have

n > log2 n.

-tr

--333342

5-6 ,
Example 2-6c.
The function
1

can write
f :

Sketch the graph of


f

is defined for

log2(1 - 2x). = log2 [-2(x

LI

x -elog2 1-2(x - VJ

in four steps.

: x -elog94 - 2x

1 - 2x >0;
'''

)J.

can be obtained from the graph of

'

by

We

g :-x -elog2 x

we begin. with the graph of

Second, replace
Third,

The graph of

(See Figure 5,-76b.)

First,

x < 5

1.,
-

thatj s, for

2x...

fold the gfgph over the line given by

x = 0.

(the y-axis)

Fourth; shift the graph one-half unit to the right.

2/

1 og2

4
log2 (-px)

T6g21- 2(x -22:-)

es

Figure 5-6b

5-6

2.6

Exercises
1.

Prove that for any real number


/1%

log2(X)

that

log2(x

0,

-x1-)

= 0,

and hence.

= -log2 x.

Prove that for any real nurnber'xi > 0,

x2 > 0,

x
1
log2 (x2) = log2, x1 - log2 x2.

.log2xY = y log2 x

38""ffProwe foimula 14(b):

4.

.(a)

For what values of

is

log2 x

.(-10,)

For what values of

is

y=

log2

(iv)

(ii)

y=

log2 (-x)

(v)

y=

log2 (x

5.

1)

less than

less than

(i)

(iii)

114,

(vi)

0?

greater than

0?

= log2 (1

y=

1og2(gx

= log2

greater than

0?

0?

x)

- 3)

(3 -

Graph the following functions on one set of axes over the interval

-5 < x-< 5.

r
ea

X -4

6.

lipg

X -4 log2 (1 - x)

x -4 log2 (-x)

x -4 log2 (x + -1)

x,-+Aog2 (x

x -4

- 1)

log2g-(1 + x)1

On one set of axes sketch the graph of each of the following functions

< x < 8.

over the interval


x

log2 x

log2 (2x)

-4 log2 (-2i)
log
x -0

(31)

lg2 (-

On one set of axes sketch the,graph of each of the following ftmcti6ns

<x< 8

over the interval


(i)k:

(b)

x -4 log2 2x

-4 log2 (2x - 2)

(c)

x -4 log2 (2x

(d)

x -4 log2

335

34

(7

- 7)
- 2x)

-6'

Write each logarithmic statement in exponential form:


110i2

(b)

3 log2 x = 5k

(d)

;2 log23 = -t

Write each exponential statement in base

(a)

2x,=

(b)

4k =

logarithmic form:,

2.

(c)

(2)-3 = mn

{,d,)

= xy

Evaluate each of the following:

10.

(a)

log2 2

(e)

log2 1

(b)

log2 4

(e)

log2

(c)

log2 8

(8)

log2

(d)

log

(fi)4

log2 g

1
T;

1
7-

4
1

16

Using'the results (and extensions) of NlaMber'10 abover bat without the

11.

use of tables, locate the-values


of eachlbf. the following between conA
,

.-

stcutive integers (e'g.,

log2 11 < 4,

(e) 1082 S.

t( i)i

since

23,< 11 524).

log2 31
.

ti

;'(c)' log2-21+ log2 14= 4

(a)

t:.

(b)

log2,5,

(j)'

idg2 34-

log2 6.

(k)

log250 ..

(.0,

log2 99

'

e)
(d)

ldg2

gip

(m)

(f)

(n)' log2(i)'.;

(g) log2 13
(h)

12.

r;

.(eY 1-82 2

log2 10

log

18
2

(o)

log2 (4)

(p)

log2 (0.18)

In evaluating the following, first estimate your answer from your


solutions to Number 11, then make a closer'estimate from the use of

Table 5-3.
(a)

lOg2

(b)

log2, 5

(c)

/
log2 131,

336.

3'4 5)

5 -6

13.

iUsing the results of Numbers 10 and 12, estimate


(e.g.,

=3

log2 72 = 1og2(23)(32)

2+

loge

3z

2 log2

6.28)

each of

the following.
,(a)

(b)

(f)

log2 12
2

-(c)

log2

24

'.(d)

log2.

(e)

14.

1282

(b)

1282 54

(h)

log2 36

log2 27

which makes the following assertions true.

Log2 x = 0
log2 x =

(c) slog2 x = -1
15.

..(g)

log2 52

Eirid a 41ue for


(a,)

log2 169

Show that, if

log2 x

(e)

log2 x

ti >. 1,

1
2
1

log2 x =

then
log2 n(n

(Hint:

(a)

Use the fact. that . 2

1) < n + 1.-

( n(n - 1)

>

;s

'

4
set

3,37

:3 4

5-7

o.
'5-7.

Logarithms (General Base)


a >'0

If

is defined analogously to

loge,

denoted by.'

a 1 1 .then the logarithm to the base la;

and

log2.

Thus

log

ay

= x.

is defined onl

for

x > 0 and is given by


loge x = y

(1)

if

The techniques and ideas of the previous section extend easili to this more

For wthple, the graph of

general case.

loga x

x.-ax over the line given by

is obtained by folding the graph of

y = x.

We summarize other easily obtained properties:

(2)

(a)

loge 1 = 0

(b)

log 'a = 1
a

if

(c)

then

0 < a < 1

is an increasing function and if

a > 1, loge

is decreasing.

loge

The laws of exponents


,JVC-Py

axay,

(ax)y

axy%

give the correaponlifig logariithra'formulas


..-

log

.(a)

my = log

(3)

log

a id'

(b)
_

x + log

x ).0

and

y,> 0

x,

= y, log

for , x > 0.

a,

Logarithms tp the

for

y,;

are very useful in-calculatfons, due t8Fthe,

base 10

act that our number systeb is the decimal sys1ei.

Logarithms to the ,base

are of increasing importance due to the use of the binary/' system in computors

and in information theory.

In the next chapter, our discussions of tangent

lines. will establish the importance of the

base e,

where

is the symbol./.

for the (irrational) number, which correct to two places, is giyen'by

2.72.

To-conirert.'from one base to another the following formula is usefUi.


a, b

and

are each positive and unequal to

).

-logc b

r.

(4)

logab

log
e-

To prove this wellet.

logc; b

so that

338
S

CI A '7
CY 4k

1 --then

If

5-7
a

Now take the logarithm to the base

= b.

of each side.

This gives

loge ax = loge b.
Formula 3(b) gives
loge ax = x loge a
ea that

x loge a.= loge

Unce loge a / 0

we can divide by'it to obtainv(4)-

Another formula of interest is


4

If

and

are both positive and not Olual to


loge' b =

logb a

then

The proof of this is left to Exerqises 5-7, Number 18.'

Exercises

1. Write each eiptession in simpler f


loga3

log,5

(a)

(d)

(e)

16

(f)

32

7..",

2 logn3
(b)

(c)

log 2

1/2 loga3

log54.

2, What is She value of


r,

3.

If

ath. =

(a2)m,

4. Prove that for

hence

5.

Prove that

'
s,

x -if

52-= 4x?

what is the value of. ml


any real.numbpr > 0,,

loga(x

= 0,, and

-loge x.

log

(N
x
2

= log x - lo r wa-x
a I
2'

'339

5-7

Write this equation,in exponential fdrm.

loge a = 1.

6.

Show that

7.

Express in exponential form

8.

(a)

log10 35 = y

(b)

log2 25 = x

(c)

2 log10 5 = x

Given. log10 2 = 0.3010

find
128%

logio.5,

9.

logla(-5).

Express each of the following in logarithmic form.


.

(a)

3 17

(b)

10-2 = 0.01

(c)

274/3 = 81

(d)

'0.043/2 = o.o08

= 5

(e)

log6 (x + 9) + log6 x r: 2

in terms o,

Express each of the, following` g

r, s, "p,nd

t,

ig
A

r = log10 2, s

(a)

logia 4

(b)

to o

(c)

logio4

(d)

loglo

lOgio 10

= log10 5.
(e)

(f)

1og16 2.5

-kJ

(g) jgto 9 13"


.

(h)

log10 8 3,6535

.11

ov

14

-51-7

12:- Write the_ following logarithms as numbers.

log10 1000

(f)

(b)

log0.01 0.001

(g). log2

23

(c), tog3 (k)

(h)

logio Ar)

log4

(i)

log81 27

(,j)

log2

(d)

32

(e)

13.

logo.

(a)

,(0.0001)

In each case determine the value of

log 5

log '5
(a)

10810

x.

log 3c

- 2 log10 (x - 1) = login 3

log 5
x
14.

=5

Solve_ the following equations.


(a)
(b)

-(c)
,

.1.5'.

10810 x = 0
10810 X*+ 1 = 0

10810 x = 1
t'.

For what value(s) of

(0)

log

(e)

logio, ?c + 3 = 0

(f)

log10 (2x - 1) + 2 = 0

10

(x - 2)

=3

does it hold that

(a) :loge x= 0
(b)

logx x = 1

(c)

logic
= c

3.0gx 2x =-2

16.

17.

(a)

Show that if

(b)

Show that

(a)

How are the graphs of


(Hint:

(b)

18.

Use

then

0, < a < 1,

log

then

is an increas.ing function.

x -4 logs X,

x -'loge x

and

is a decreasing function.
logb x

related?

(10.)

At what point does the graph of x

Prove formula

(5)

of the.text:

not equal to

1,

then

if

a > 1,

log

if a

and- b

cross the x-axis?'

are both positive and

1-

logb a ,

t
341

'o/

,A1-1
,'1

Appendix 1.

FUNCTIONS AND THEIR REPRESENTATIONS

A1-1.

Functions

The precise definition of function can be formulated in many ways;


as asset of ordered pairs (usually, ordered pairs Of numbers), as an pasociation or correspondence between two sets, etc.

But no matter what definition

wg choose, for a function, three things are required:

a set called its domain,

a set called its range, and a way of selecting a member of the range for each
member of the domain.

EXampleA1-1a. The multiplication of integers by

defines a function.

The domain of this function is the set of all integers; the range of the
function is the set of all even integers.

We choose to define a function as an association between elements of two


sets; thus the function of Example Al-la associates with 'each integer its
double.

If with each element Of a set


element of a set

from A

to

andthe set

B.
C

B,

there is associated exactly one

then this-association is called a function,

The set

is-called-the-domain -of the-funcLijn,

of all members -of

assigned to members of A

by the function is called the range of the function.

In what folloWs we-shall be exclusively concerned with functions whose


domains are subsets of real numbers and whose ranges are also subsets of,real
numbers. More complicated functions (like 'vector valued functions;) may be

built fromthese.
c.:

The range

may be the whole set

B,

in which case the function is

called an onto function, or it may be a proper subset of


we generally take for
=

usually specified before its range is considered.

343

351
4

B.

Inany case,

the whole set of reals, because a function is

.7"

A1-1

It is common practice to repres ent a function by the 1


letters such as

g,

f,

associates with

x.

AUSTotion

at

f,

or simply

x,"

"f

takes

"f

function

into

relation between

and

x,'

takes

takes

or

9,

defined as follows:

Thus, if

takes

is an element.of the

associates

with

3.

Coltcisely

represents any number'in the domain of

f,

lat

lik

x:

Is merely a symbolic description'Ofthe

it

x;

into

into

x.

x. ")' An aiics.,C,cs

f(x }.

In general, if

f(3) = 9,

then

of

This notation tells us nothing about the

each number of the domain into its square.


domain, then

is an

C9nsider a function

Example Al-lb.

"the value of the

f(x)

or "f

with

(other

denotes the element of

f(x)

(Read for

f(x)

f(x)").

or the element

tten

'at

used to suggest the association of

(read

used). If x

etc , will also be

0,

h,

element of the domain of a function


the range which

-7-

F,

er, f

f(x) = x

or

2
.

,---The-function is-not adequately defined uttil we spedify its domain.


domain iS the set of all integers

If t

(..., -2, -1, 0, 1, 2, ...), then

the range is a subset of nonnegative integers,

(0, 1, )4, 9, 16, ...).

If

-we choose the set of all real numbers as domain, then a different function-is
defined, even thipigh the rule of,associationis the same; in this case the

range of the function is the set of, nonnegative real numbers.

1
Observe that a function from A
reverse association from B
and

Example Al-lb, f(3) = 9


would assign both

and!,-3

while the reverse association

f(-3) = 9,
to

is a one -way association;, the

is not necessarily a PUndtion. , In t

to

to

9,

violating the definition of a function.

It is often useful to think of a function as a mapping, and We say that a

function maps each eldm4ntof its domain upon.one and only one element of its
In t1is vein,

range.
f(x)

x---(x)

is called the image of

can-be read, "f

under the mapping, and

maps
x

upon, f(x)";

is called a

preimage of f(x).. This notion is illustrated in Figure Al-la, where


elements of the domain

and range

are represented by points au ,the

4,

mapping is suggested by Arrows from the, points of the domain to corresponding


points of the range.
8

r.
341;

'52

'

Al-1

A.
A
f

Figure Al-le
Note that 'each element. of the domain is mapped into a unique element of the

range; i.e., each arrow starts from a different point in the domain.

This is
the requirement of our definition, that with each element of the domain there
is associated exactly one element Of the l'ange4

definition of function contains the rather vague phrase, "there


is associated."

The manner of association must be specified whenever wewe

'dealing with a particular function

-In this cour

be defined by a formula giving its value:


g(x) = x2 + 3x +

7.

for

Other ways of definin:

, a.function will generally

xample,

f(x) = 3x

5;

function include verbal de-

scription, graph, and table.

The notation

f(x)

is particul

ly convenient when we refer to values

-t

of a function; i.e., elements Ln the range of the fdnction:

We illustrate

(this in the next example.

'Example A$-le. ,Consider the function


f

x--- -0-3x

2
-

who,se domain is the set of all real numbers.

;.(x)

f(-2)

a + 1/17

=,3(-2)

is a real nUmber, then

Then

5,

= 3x2

5 =

f(0) = 3(0)2

and if

5 = -5,

f(a + 4.) ,= 3(a +

A-N 2

5:

I
,r

345

353

Al -1--

We note, since
3x2

25

> -5,

less than

y be any nonnegative real number, that;

and hence

e range of

is the set of ail real numbers not

-5.

a
As mentiviii earlier, a function is not completely defined unless the
lh

domain is specified -., If no'Other'information is given, it is a convenient

practice, especially when defiling with aNfunction defined by a formula, to

assume that the domain includes all real numbers for which the formula
describes a real number:
function

real numbers except


g(x) =

-1x2,

domain is
-2

(x.

to

Fo4 example, if a domain is not specified for the

2x

f :

ti

then the domain is assumed to be the set of all


ll

-= 9

and

-3.

Similarly, if

is a function such theK

we assume, in the absence of any other information, that the


-2 < x < 2);

that is, the set:of all real numbers

from

inclusive.

We note here that two functions

and

they have the same domain and f(x) = g(x)

are identical if and only if

for each x

in their domain.

The graph of a function IS-perhaps its most intuitively illuminating


-representation; it conveys importarit information about the function at a
glance.

The graph of

is in the domain of

is the set of all those points


and

y = f(x).

Example Al-ld.' The graph of the function

tr

the semicircle shop in Pigure Al-lb.*

(x,y)

for which .x

'

x --spy =

47-7x7

is

The graph gives us a clear picture

of what the function is doing to'the elements of its domain, and we can,
moreover, usually infer from the graph any limitations on the domait'and range.,
Thus, it is easily determined from Figure Al-lb that the domain of

is the

sdt of all
that

such that

0 < y < 5.

-5 < x.<:'5

and the range is the set of all

such

These sets are represented by the heavy segments on the

x- and 1-axis, respectively.

* In this figure a complete graph is displayed. The graph in Figure Al -lc,


as well as most of the graphs in the text, are necessarily incomplete.

Al-1
j

Figure Al-lb

We remind you of the fact that not every curve is the'graph of a fundtiOn.

In particular, oar definition requires that a function map each element of


its domaincnto only one Element of its range.

In terni's of points of-a graph,.

this means that the graph of a function does not contain the points( (xl,y1)
and '(x ,y ) if y
y
1
1
2' i.e., two p nts having the same abscissa bat .,
diffeient ordinates. This is the a' s fJer the "vertical line test":
if in
xy-plane 'we imagine _41 possible lines which are parallel to the ylvis,
'and if any of,,thete lineS'cuts the graph in more than onepoint, then the

graph represents a relation which is not a function.

Conversely, if every

line parallel to the y-agis intersects a,graph in at'most.one point, then the
graph is thet,of a function.
,

t -

Example Al-le.
radiuS

+ y2 = 25, whOse graph-is a circle with

and center at the origin, does not define'a function.

5'

interval

The equation x

-5<_x<: 5, every vallie of x

an

(3-4)

On ike open

is associeifeci with `two different

Wiluestd". y,--'cbht43.ry to the definition.Of functibn:

Specifically,

,(3,4>,

..,

are two points of the circle; they determine.a line parnllel"t6

1#.

the y-axis and intersecting the 'circle intwo points, thus illustrating that
t'
he circle is the graph.oft.
the
latioh that is not a
1

separate the circle into two


X

.)

We can, howevt,, '..-

- .. 4 5.--

1- circles- -the graphs of


the functions
.,,

(Example Al-ld)

and

tc

127577

....

'",)

'-'i

o
'

^.

ft.

'

-1,

Throughout this diACussion we have used dthe letters

'

,,.

mi

'

x and 'y.

Specifically, if

c'

f. ls therfunction
.

.,

,,-,.

..

a'

',_represent elements of sets.

*
,..

-- = f(),
: x..?

'

,...k..

14

.----.4!

r
e

TT

.347
,

3 5 Lit

4_

'74

od ,

'

A1-1
.

then

.represents an element (unspecified, in the domain of

represents the corresponding element.in the range of f.

and

f,

In many textbooks

are called variables, and since a particular value of

range depends upon a particular choice of


the independent variable and

in the domain,

in the

is called

the dependent variable. The functional


OP
relationship is then described by saying that "y is a function of x."
y

For the most part this language is not used in this textbook.
. r

We =elude this section with a summary of several different special


functions; you 'are undoubtedly acquainted with some of them.

The Constant Function.


function

is an arbitrary real number, then the

If

which associates with every real number

is called a constant function.

f :

the value

Morse generhlly, spy function

whose range contains exactly one number is a constant function.


a constant function, say
to and

: x--..c. for ali real x,

b,

The graph of

is a line parallel

units fro:1i the x-axis.

icl

The Identity Function.


each number

in A,

...

function whose domain is

...

Let, A- be the set of all real numbers; 'With:

associate the number a.


..
.

and whose range is

A,

This association definei a


A
A, namely

More generally[.for /any domain such a function is called the identity function.

If the domain is the set of all real.numbers, then the graph of


line with equation

is the.

y = x.

\.

_ TT
,

The Absolute Value Function.

With each real number the absolute value

*
-function associates-its absolute value- :,.

._..

'1
.

x for.

f'

x > 0

...

....

ili

f 4 x---....1x1 .

-x

for

X <'0;

Alternative definitions:
x

f
f

lxi = max (x,-x);


I

X --. Ixi =

x..

3118

5.13

A'

A1-1
.

The-graph o

is shown in Figure Al-lc,

it is the union df two rayS____/-

issuing f4qm the origin.

Figure Al-lc
The Integer Part FunCtion.

the sum of an integer

Eve

al number .x

and a real

can be represented as

such that

110-

x = n +

and

r < 1.

1
,

For example,

5.38 = 5

+.38,

3 =.3

O.

-. +
We call nthe integer part of

.6

and denote it by

it follows

[x] = n ;

Ix] < z <[x] + 1.' Tliwe see that to each real number

that

corresponds a unique integer part

[x],

there.

and this correspondence defines the

integer dirt function


f

0$

/t

*
Sometimes called the greatest integer function.
I

"/
349

3 57 v

ff`

A1-1

A graph of this function is shown in Figure Al-ld; it is called a step graph;


i.e., the graph%of a step function.

Figure Al-ld
The Signum Function.

With each pos.'iice real number associate the number


40#0.

+1,

wiith zero associate_the number

associate the number


symbolized by

-1.

sgn x.

0,

and with each negative real number

These associations define the signum function,

Thus

1, x

0.0

S gfl X

0,

1 0,

x = 0 ,

-1

x <0.

We leave it as ah exercise for you to sketch-the"'graph of this function.

EVen and Oda FuncfrOns.


"whenever it contains

x.

Let

f be a function whose domain ccbfitains

The function

f -is said to be even if,

f(-x) = f(x).

-x

For example,, the runction f with values f(x) = x


ip
2
%
( -x)2 = x
for all x. Geometrically the graph of an even

even since

function Ls, symmetriC with respect to the y-axis.

The function
functiOn

said to be odd if

with values .f(x) =X3

f(,

-r -f(x).,- For example, the

is odd. since

(;x)3 t4. -x3

for' all x.

Geometrically the graph. of an odd function is symmetric with respect to the


'origin.

350
es

35

.4

t4

Periodic Functions. 'Certain functions have the property that their


function values reptit themselves.in.the,same order at regular/ intervals over
the domiin (FiKure_Al-le).
'14

Figure Al-le

Functions having this property are called periodic; included in this important
class are thg circular (trigonometric) functions, to be discussed in Chapter 2
3.

and

A function

is periodic and has period

for all 4 in the domain of f,

x + p

p # 0,

p,

if and only if,

is also in the domain and

f(x + p) = f(x).

(1)

From the definition we note that each successive addition or subtraction


O

of

I;

bripgs usback to

f(x)

For example,

again.

f(x + 2p) = f

(cxi p) + 4

= f(x

p)

= f(x),

and
f(x - p) = f ((x - p) +
= f(x).

Ingeneral, we'infer that any multiple of a period of

is also a per

that is,

for :ny integer, n.

'f(X + np)c:_f(x)

For a constant function

f
it is obvious that' f

111'

is periodic with any period


f(x + p) = c = f(x),
_.

351
0

359

'1D,

since

A141
,

'It can be. shown that for nonconstant periodic functions (continuous at

one point at least) there is a least positiye value of


true.

for which (1) is

This is called the fUndamental period, or simply the period, of'such

a function.

44'

Example Al-lf.

x-----wx - (4

x = n + r
*
,part,
then
If

where

real, is a periodic function:


*
is the integer part..of x and_x1 its fractional

o
n

..-;t

f(x) = f(n + r)
= (n + r) - [n

r]

=n +'r - n
= r,
and

f(x +I) = f(n + 1 + r)


.= (n-+ 1 + r) - [n + 1 + r]
qv

= n + 1 + r - (n + 1)
= r.

Thus, as was asserted,

i6 periodic and its period is

1,

gas shown in its

graph (Figure Al-if):


9

-- Figure Ai-lf rf

the fractional part of


We note That since' f(x) ='r,
is sometimes called the fractional part function.
.k

x,

this 'function

352

36.0
.t,

A1-1
Exercises Al-].
1.

'Below, are given examples of associations between elements of two sets.

Lecid/w4ther each example may properly represent a function.


requires you to specify the domain and range for each function.

This also
Note that

no particular variable has to be the domain variable, and also that some
of the relations may give rise to several functions.
(a)

Assign to each nonnegative integer

(b) .Assign to each real number

the number

the number

7.

the real number

r.

2n - 5.

ti
(c)' Assign to the number
(d)

10

Assign tO-each pair of distinct points in the plane the distance


between them.'

(e)

Y = -3

(f)

x = 4

(g)

x + y =4""2

(h)

y = 2x2 + 3

(i)

y2 - 4 = x

(for all

x)

(for all

and

z)

y"< 2x - 1

(k)

f(x)
,x2

1/217

y2

I
16-

-2.

3.

Sketch the graphs of equationsl(e) - (i) of Number 1.

A function

is comp/4tely defined by the table:


x

f(x)

4'.

'

-311

13

(a)

Describe th

(b)

Write an equation with suitably restricted domain that defines

If

domain and range of

x2 + 3x

4,

_f.

'find

f(I)

(a)

f(0)

(b)

f(2)

(e)

f(2 -

(c)

f(-1)

(f)

f(f(1)

(Hint:

of

ti

X353

.316

''This is the value


at

f(1).)

A1-1
-

If

5.

2x

is a function defined by g(x) =

find, if possible,

4-77
(a)

g(0)

(d)

g(2)

(b)

g(1)

(e)

g( -3)

(f)

g(15).

*(c)

Which o \the following mappings represent functions?

6.

(a)

(d)

(c)

7.

Given the functions f : x----x and 'g


.

number, are

x7-....

..,

and

the same function?

x
x

8.

and

the same function?

(it.)f : x I- 2x

10:

and ,g : x--0.

What number or numbers have the image

( 6)

h': x --- ,1)711.;.

x---1- x2

--:-11.

x - 2
Why or why not?

If

a : x--. Ix - 41

(e)

0 : x----[x]

Which of the following statements are always true for any function

-asgumingthat xl

and

x2

atb...tin the domain of

(a)

If

xi = x2, 'then .f(xl) = f(x2).

(b)

If

xi / k2, then' f(xl) / 1(x2).

(d)

If

f(xl) = f(x2), .then

xl = x2.

(d)

If

f(x1),/ f(x2),

xl / x2.

then

x ,is

10 under the follOwing mappings?

(d)
N.,

(b)

is a real

Given the functions f : x--.


-,. x + 2
real, are

9.

Why or-why not?

If

1'6 2

fi

f,

Insert page 355


11.

A1-1

If, 1(x) = Ix',

which of the following statements are true for all real

numbers

t?

and

is -an odd function.

1(x2) =

f(x)2

f(x - t) < f(x) - f(t)

f(x + t) <f(x) + f(t)


12.
.

Which of the followkfirfUnctions are even, which are odd,


a4d which are
neither even nor'odd?
(a)

(e)

(f)

(g)

t : x

(h)

x---4-x3 + 4
.

'.(b)

(c)

2x2 +
2

x--..-x

- 4x + 4

x-x3

- 2x

(d)

x=---..-

'

2x,'+ 1

...

13.

Which of the fo_lowing graphs could repesen't functions?


`-..)
Y
:

(a) ;

I-Y

(c)

-0
r

(b)
6

3363

(f)

14.

Suppose that

x--e-f(;)

is the function whose graph is shown.

Sketch the graphs of

15.

(a)

(b)

(c)

(d)

A function

is defined by
for

x /

for

x =

f(x) : {1234
0

Identify this function and sketch its graph.


16.

ir

Sketch the graph of each function, specifying iil''.domain and ranget.


(a),

f : x

(b)

(c)

f: x--.- il - xi

(i

(d)'

C : x--1- 1 -- lx1

(j)

f :_sx--..x[x]

(k)

(/)

Fc2
ixi

(g)

(h)

sgn, x

St]

f: x ---o....Ca
x-,,,

(e)
(f)

.f.

2 xo- ixi - x

.f

x :--r,... -1x1 + ix -11

(Hint: Consider separately

x < 0, 0 < x < 1,, and


x > 1.)
I

356

3 t'S

x--4-1x2 - 2x:- 31
4240,

the 'three :,possibilities:

: xo- 11 - x
x21

Al-I
)

Sketch the graphs of the functions in Exercises 17 to 19.


-which are periodic, indicate their periods.

For thOse functions

Indicate those functions which are

even or. odd.


17.

2x2

EA

f-:

2x2

42]

18.

x -

[c -

2x2 - 2[X]2

(a). f

ax

4f

a > 0

Ca)2
[2x]

Dx].

x(1/5' + 1) -

.1 + sgn x
2

[x]

This function is also called the Heaviside

unit function and is designated


(b)

(c)

(d)

f4

(F)

20.

.(g)

If

x--44.11(x)

x
:

H(x,- 2)

(x - 2)2

4i(X)

x --a. H(x) + H(x - 2) t H(x - 4)

H(x2

and

2.)

( sgn x)(x - 1)2 t Egn(x

.f

+ H(x - 2)

(f) f : X
Ah

are periodic functions of periods

(m, n integers), show that alto and

21.

m and

n.

m and .n,

respectively

f g are also periodic.

. ,examples to showethat the peTni of _f + g

than both of

Ir

2)]x2

Give

oan either be greater or less

Repeat the same for the product

f g.

(a), Can a funCtion be both even and odd?


(b)

What can you say about the evenness or oddness of the product of:
(1),

an,eve!T'function by an even.fUnction?

(2). an 'even function by an odd function?


(3)

(c)

an odd function by an odd functio

Show that every function whose domain


contains

ontains

-x

whenever it

can be expressed as the sum of ah evep function plus

an odd function./

357

365

22.

Find functions

f(x)

satisfying
2,(x)

Suggestion:
23.

f(_x) = 1

(called a functional equation).

Use 21(c).

Prove that no periodicunction other than a constant can be a rational


function.

functions.)

(Note:

A rational function is the ratio of two polynomial

el

358:

a t3

4:2
A1-2.

Composite Functions

Given two functions

and

with domai s whose intersection is non-

empty, we can construct new functions by usir

any of the elementary rational,

operations--altion, subtraction, multiplication, division--on'the given


l'unctiona.

Thus, the sum of

and

f+g

is defined to be the function

x--;---.-f:),17.((.)x)

which has for dotain those elements contained in the intersection.of the
domains of

f and g.
Similarly there are definitions for the difference,
product, and quotient of two functions; there is, in fact, a whole algebra

of functions, just 8.6 there is the familiar algebra of real numbers.

In this algebra of functions there is one operation that has no counterpart in the algebra of numbers:

the operation of composition.

This operation

isbest explained by'examples.


Let
g

Old

2x + 1 4"

--4-x

2.

We observe that
g(l) =, 3. and

f(3)

g(2) = 5

f(5) = 25,

and, in general, the value of


f(g(x))

and

at

g(x)

is

= f(2x + 1) = (2x + 1)2.

We have constructed it new function which maps

(2x + 1).' this function}, defined by the mapping

by fg,

is called a composite of

and

g.

represent the value of the function

fg, by

Either symbol means the'value of

at

onto the square of'


x---...f(g(x))

and denoted

Hereafter we shall usually


fg(x)

rather than f(g(4.

g(x).*

The symbol fg, denoting the composite of the functions t and g


must not be confused with the product of the funCtions. In this text we
distinguish the latter by use of t
of for.multiplication; i.e., f .g.

3 6 7:

A1-2
An immediate question arises as to the or in which two functions are
composed:

is the composition of functips a commutative operation; i.e., in

general, are
that

gf(x)

fg(2) = f(5)

fg(x)

arid

equal?

In the example above we have seen

and we calculate.

= 25,

gf(2):

fg(2)

gf(2) = g(4) = 9

Thil counterexample is sufficient to prove that in general


The operation of composition applied to two functions
produces two different composite functions
order in which they

fg

and

gf,

generally

depending upon the

are composed.

A word of caution must b /injected at this point.


defined only if

gf(x) # fg(x).

and

is in the domain of

and

The number

fg(x)

is

.is in the domain of

4(x)

f.

For example, if
f(x)

and g(x)'43x - 9,

= -Li'

then

fg(x) = f(3x -'9) =

and the domain of

fg

is the set of,real numbers

nonnegative; hence the domain is the set of all

for which

3x - 9

is

x > 3.

For the other composition of the same functions

and

g,

we have

gf(x) = g(-1/) =3IX'- 9


,which is defined for all nonnegative real numbers

(x.

We define composition of functions formally.

The composite

fg

of two functions
fg

The domain of
of

fg

for which

and

fg( x) = f (g(

is the set of all elements


is in the domain of

g(x)

is the function

g.

x
f.

in the domain'

The operation

of forming a composite of two functions is called composition.

The definition may be extended to the composition of three or more


functions.

Thus, if

h are functions, one composite is

and

f, fig,

fgh
6

In order to evaluate
h(x),

fgh(x) =

fgh(x),

and finally the value of

we first find
f

at

(g(h(x)))

.h(x),

then the value of

gh(x).

36o

368

(l

at

te
,-

-r:

% Al-r2

.
0

-,

Exercises, A1-2

a.

Given that

x---. -x

and

': x---x

le0 4F(2) + g(2)

.
-,

'

(e)

+ 1. for all real

find

x,

"

f(x) .4.4.rgSx')., :

l*

(b)

g(2)

f(2)

:(f) ff(x)

ro

g(x).

.
.,

2.

3.

fg(2).

(g)

(d)

gf(2)

(h)... gf(x)

If

f(x) = 3x + 2

(a)

fg(x).

(b)

gf(x).

and

fg(x) (and

..

find

g(x). = 5,

f(x) = 2x +1 and

If
(a)

g(x) = x2, find'

gf(x).

'

(b) A For what values/


4.

-fex)

(c)

if any, arf :fg(x)

x,

For each pair of functions

g,

and

and

gf(x)

equal?

find, the composite functions

fg,

and gf and specify the domain (and range, if possible) of each.


f

( a)

g: x

x
1

(b)

(c)

x
x

2x - 6

'

: x-x2

x-

- 4

'

ir

(a)

x\--;.. x2

(e)

f :

x2

(f)

f x 5(2- 1 ,,.g

g : x

x A - x

g, :

16C

5.

Given that

f(x) = x2-+ 3

and

g(x).= 1/)77.7,

solve the equation

'fg(x) = grflx).

6.'' Solve problem 5 taking


7.

Describe functions

g(x)
and

such that gf will equal


'

(a)

3(x + 2) - 4.
-%

(b)
(c)

(2X

5)?

'

(a)
(e)

- 41,

(42;

2x - 5

am=
J

361.

.a6 9

<o

,0
.

For each Air or :fUnctions

' and, gf

.f

g ;find the composite functions.fg

and

Also,
r.
sketch the,graph of each, 'and give the period (fundamental) of those

and specify the domain (and range; if pPssible) of each.

which are peribdic.


f

(a)

x---4sgn(x - 2)

4,

x----s-IXI

(b)
-9.

x---4=2 sgn(x - 2)-- 1

What can yoou say about the evenness or oddness ofthe composite.of
an, even function of an even function?

(a)

an even function of an odd function?

At(b)

10.

(c)

an odd function of an odd function?

(d)

an odd function of an even function?,

If the function

is periodic, what can you say about the periodic

character of the composite functions

11.' T? the functions


fg

and,

and

gf

gf

assuming these exist


Illustrate by examples.

are each periodic, then the composite func-

(a,se?Imed to exist) are also periodic.

of*either one be less than that of both


12.

and

is an arbitrary functiOn (ilot periodic)?

and, g

tions

fg

A sequence. at), al, a2,

an+1

an,

.,

and

Can the period'

g?

"is defined by the equation

f(an), n = (i) 1) 2, 3,

wher

is a given function and

and

is a given number.

a0

If

ao = 0

.then

al

f(a0)

11

a2 = f(al) = ff(a0) =

a3 = f(a2) = Mai) =,fff(a0) = )2 +

12717

Show that for any


(a)

an < 2

(b)

N.

> 2

.2

n-1 '

n > O.
4'

I
*.f

th

A1-2
13.

e0

If c+1 = gan); n = 0, 1, 2,
of

and n,

g
f

(b)

(c)

(d)

(e)

g,

for tne folloNing functi'Ons

find

,n

a's a function.

a + bx.

(a)

xm.

irxT:

z
-

x)-1.

ti

se.

..1.4

363

all
V

A1-3
A1.3.

Inverse Functions
Recall the vertical line test for the graph of a function (Section.A1 -1):

if every line which is parallel to the y -axis intersects p graph in at most

one point, then the graph is that of a function.

Thus in FigUrp A1-3a, (i)

and (ii) illustrate graphs of functions, (iIi) is the graph of'a relation that
is not a function.'
1

(i)

(ii)

Figure A1-3a
This figure also illustrates an important distinction between two classes of
functions:

for graph (i

there is at least one line parallel to the x-axis

which intersects the graph in more than one point; this is not the case for
graph (ii).

The latter is typical of a class of functions called one-to-one

'

each element in the domain is mapped into one and only one image

functions:

iii the range, and each element in tl* range corresponds to one and only one

preimage in the domain.

In other words, a funCtionof this kind establishes

a'one-to-one correspondence between the domain and the range of the function:
,

A function

is one-to-one if w enever

f(xl) = f(x2):, the

xi = x2.

Note the distinction between the definition of function and this defirli-

The former states that any function

tion.
xl.,,='x

2'

.function

then f(xly= f(x2),


is such that

has the property that 4

whereas the latter states that a one=to-one

f(xl) = f(x2)

if and only if

xi. =-x2. .

The class of one-to-one functions is important because fo; each ember


,

of ihis
class we can specify a function that, in a, loose way of speaking,
,
.

-----.

undoes #10-7a6V7:5-r7.70 Win 4unctI-64.--#04, fo4example, if


,

II.-

04 ri,

g; i
.

is the

0.1

3-1

4,

,
,

,,

364

3 7 2

7,--

......k

104

C],

....

.:..,

.
;,

A1-3
function which maps each real number onto its double, then there is a function
g,. called 'the inverse' of

number onto its half:: f

If a. function
g

x,

which reverses this mapping and takes each real

f,

2x;

f(x)

y
2,
1

is one-to-one, then the function

whose domain is the range of

the inverse of

is called

f,

f.

The functions

and

represent the same association but considered

from opposite directions; the domain of


of

is the domain of

inverse is

It

f; Furthermore,

and the range

is itself one -to -one and its

f.

Is instructive to lOok at the composites of two funbtions

inverse to one another.


into

is the range of f

If

maps

in other words, if,,y

x;

''gf(x)= g(y) = x,

f(x),

into

then

then

y,

maps

and

back

x = E(y). Fgrice,

for all

for all

y in the range of

in the domain of

f,

and

0 fJ
fay) = f(x) = Y.,

Observe that thp restriction of the domain of


of

1\

to coincide with the range

part of the definition of ,the inverse.

Example Al -3a.

Consider the one-to-one function

what is its inverse?

Here

must add

3,"

In orderto-reverse this' procedure, we

and then divide by .2.

.is tile function

<,

This'suggests that the inverse of

in thd domain of

1(x)
g' is the inverse of

f.

into

Furthermore, in the Opposite direction,

f.

in the domain of

as expected.

f(x)

(2x - 3) + 3

= g(2X - 3)

'

-I--

g `maps
.4

By substitution,

fg(x) .'f(2472) -

to prove this fact, we nattst show that

satisfies the definition of inverse; i.ee, show that

for all x

f.: x--0. 2x - 3;

is described by the instruction, "Take a. number,

double it, and then subtract

for all

f%

g%

Hence,

is t4 inverse.of the functi n

'

365

3 73

Insert page 366,

Al -3

The graph bf the inver e

graph of

f.

If

maps

of a function

into

is easily found from the

then.g maps

b,

(a,b)

is on the graph of

Figure4A1-3b shops three poidts

is on the graph of

h of a function

(4,,40) 01, on the gr

f,

It

a.

,e,

if and only if

f,

g..

into
A

follows that the point

(b,a)

(1,-3), (2,1),

and

and their corresponding points, obtained

by intercha 1140/P
of coordinates, on the graph of

g.

y =1g(x)

= x
4

-Ye

(1,2)

= f(x).

/,

(4,2)

-4

-3

-2

(2,1)

-1

// 0

-1

-2\

o.

(1,-3)

-3

Figure A1-3b
FroM this figure we see that the points
with res ect to the line

y = x;

(a,b) k and

two poin s is perpendicular to," and bisected by, the line


-(b,a)

the reflection of

(a,b)

(b)a)

are symmetric,

that is, the line segment determined by these

inthaline 1y

y = x. We call

..= x.
,

mt5,

Example Al-;
g :

2
x ---)..x ,.- 2.

.Consider the functions

The function

cannot be the inverse of


ing the domhin of

xi> 0

x--... 17-77, .x > -2 and

f- is one -to -one;

as it stands.

is not and, hence,

This can be restri ted by restrict-

to, x > 0; i.e., the id;rerselof

(Figure A1-3c).

inverse to oge another:

51

is

The composite functions verify that, f

: x

and

.X2
g' a'e

- 2,

fg

fg(x) = Ax2 - 2) + 2 =

x > 0;

gf

gf(x) = (477)2

x >c2.

- 2.= x,

T./

, Y

/
x > -2

x2 - 2

x > 0

-1

/ -1

//
O

` Figure Al -3c

The relationship between the coordinateS of a point ' (a,b)


coordinates of its reflection

(1?,a)

in the line

= x

and the

suggests a formal

method tor obtaining an equation' of the inverse of a given function assume

that the in

exists.

Example A1-3c

'Consider2the function
4

f ; x

y = 3x + 5

for all real

x.

Yf 'we ,d4tercigginge

and
.

-,

(3.)

in the equation

y = 3x 4. 5,

, r
.

we obtair7

1'

(2)

x = 3y + 5.

- For every 'pair of numbers

f"

(a,b)

is,in the solution set of (2).

in the solution set o


Hence, (2). is an equat

(l), a pair

(b,a)

n defining -implicitly

441

the inverse of the given ,function


we 'solve (2) for
1 e\

in terms of x

In order to obtain. the explicit .form,

and obtain

.1

367

0'17 r
t)
s)

A1-3

The inverse of f

is, therefore,

g :

gi(a) = a

You shopld verify the fact that


and that

fg(b) = b.

Example Al -3d,

for any

for all. real- x.

3 5

for any

(range of

in the domain

in the domain of

f,

f).

If the given equation defines a quadratic function, the


In the first place, the

problem of finding an inverse is more coipplicated.


.

given function must be restricted to a domain which gives a one-to-one


function; in the second place, the technical details of interchanging the
variables

and

are more

in the given equation and then ,solving for


1

involved.

Consider the function


^

2
f

+ 2x + 3

for example, we..restrict

function fl
fl

is

(r

to the domain

and opening upward.

(-1,2)

whose graph is a parabola with vertex at

(x

then we have 'a

x > -1),

gl.

and this will be the domain of

We proceed to.find a formula defining


2

The range of

gl.

which is one-to-one and hence has an inverse


y =4i'1(x) 2: 2),

If,

We are given

+ 2x + 3,

y = x

nd We,int rchange the variables to.obtain'


61

+ 2y + 3.

-= :pc = y

We now solve-foi y

in the quadratic equation

+ 2y + (3 - x) = 02

obtaining
or

y"'=. LI + )x -52

y = -1.-IT72 .'

;e4
0

ation

Which of these formulas defines the

gl?

Since

,-

here represents

any element in tile range of the inverse func'tiox, and'sinde the rage must be

the same set of Lumbers

as

t he domain of J fl,

we see that .x > ,1,

ig

required. 'Hem

4111

ggf

a.

/4

pEN=1=1.
.

A1-3

y = -1 + 167:7
defines the inverse function

whosedomain

is

(x :x > 2).

(Note, again,'that this is the range of


,

fi;)

It is helpful to sketch the graphs of the two inverse functions in Order


to see more clearly the relationships between their domains and,ranges.
Figure A1-3d.)

more clearly
as many diffe

In fact, if you graph the original function

you may_see4

ow its doma.n may be restricted in infinitely many ways to give *


nt one-to-one functions,, each of which has a unique inverse

function.

f,

(See'

;(

r.

r.

369

377

A1-3

ercises
/

1.

What is the reflectiOn'of

ria e

line

y = f(x) = 3x

Write An equation defining the inverse of

in the line

y :-..

x?

f.
",/'

et

2.

Which,points are their own reflections in the ling y = x?

What is the

grapheof all such paints?


.3.

Find the slope of the segment from

(a)

(a,b)': to

the segment is perpendicular to the line

Prove that th

(b)

What is the reflection of

(a)

(b). y = 0?

'(e)

to

y = x.
(b,a). is bisected by the

in thl line

(1,1)

x = 0?

(a)

y
x = -3?

--

= -x?

(c)
5.

(a,b)

and prove that

y = X.'

line
4.

segment from

(b,a),

Describe any function or functions you can think of which are their own

inverses.

6. }in equation or an expression (phrase) 'is Said to be symmetric in


y

and

if the equations or the expressions remain unaltered by interchanging

x and y;

e.g.,

x -'xy + y.
about the

+ y ,= 0,

x.3

+ y3.= 3xy,

lx - yl = lx + yl,

It follows thaegraphs of symmetric equations are symmetric


y = x line.

Geometrically; we can consider the line y = x,

behaving as a. mirror, i.e., for any portion of the graph there must also

be a portion whichisithe mirror image.


The equation 'x
is

+ y

bviously symmetric with r spect

to the line

y = x.

What (4h r

axes of'Symmetry Zmirror type


does it have?

r.

/
7.

The expression
a'+ b-+

- b

obviously symmetric in
and
Hint:

+
a

re.'+

and

b+ la - to! :ITC
Show that it is also symmetric in

c:
67104 six cases (i) a

c,

p.1)

a < c <

b, < a

<t,,C,

ato.
'

370

I'

87
4

A1-3
Fnd,the inverse of eadh ,function.

9.

(a)

(b)

(c)

3x + 6
- 5
2

- 3

Which of the follOwing functions have inverses?

Describe each, inverse by

means of a graph or equation'and give its domain and range.

-x2

,(a) "f :
CP"

(b)

(c)

10./As we have seen,

( d)

(e)

4: x---.-xixl

(f)

[x]

sgn x

for all rea4 t does not have an inverse..

Do the following:
(a)

Sketch graphs of,


x < 0,

(b)

I',

: x ---x

for

x > 0 sand ,_f

and determine the inverses of

f 1,

--0-x2

: x

for

'ane ft.
2

What relationship exists among the domains of

and V?

fl,

f,

(f1

is called the restriction of

to the domain

(x

x > 0)

';

'

and 7,f

is similarly the restriction of

to the domain

(x': x < 0).)


11.

Sketch a graph of

(a)

f :

and show that

does not have

an inverse._

Divide the domain of

(b)

f
(c)

into two parts such that the restriction of

to eitAr part has an inverse.

Write an equation defining each'inverse of part (b) aria sketch the


graphs.
1

12.

DO Problem 11 for

13:

Giv9 that
fg(x)

: x--.- x

f(x) = 3x =72- and

ge(x).

- 4x.

g(x) = -2x + k,, find

For this value of

are

and

such that

inverse to one

another? 'Give reasons for your answers.'


14.

Show that
for

S.

x _> 1

f :

x2 - 4x + 5 for

x---...

yl in the domain of

of

x > 2

and gliix.. 2.+ 1/7:-.77717

are inv'erse to one another by showing that

f .

'ig,

and that gf(x) = x

for all

If f(x) =12x-3 + 1 7

/..

for all

in the domain

'
.

15.

fg(y) = y

find at least two functions

f0x) = gf(x)

e'

371

su h that

'

8--

A1-4.

Monotone Functions

increasing, oft&functIons

If we examine the behavior, for x


7olt

as

1G and

g :'x --b. sin x,

we note that the values of -f

g, are sometimes increasing and kme-

increases, while the values of

Geometrically this means that the graph of

times decreasing.

increase

is con-

tinually qsing as we survey it from left to right (the direction of increasing


x), ,whereas the graph of

like a wave, is now rising, now falling.

g,

The

graph _of a function may also contain horizontal portions (parallel to the

x-axis), where the values of the function remain constant on an interval.


.

function such as

illustrates 'this, and also points up the fact

that the graph of such a function need not be continuous.

Example A1-4a.

The function
-x

defined by
2
,

0 < x < 1;
1 < x < 2;

h(x)

7-

x3

x,

0.

has the graph shown in Fgui'e


Ad

It is easy to see that the function decreases as


interyal

(1,2]

increases except on the

on which it remains constant.

Taken as a class, the increasing, decreasing, strictly. increasing, and


strictly decreasing functions

called monotone (compare with- monotonous).;

becaesethe changes in the values of the functions as

increases are

always in one direction.


irection.

Let

be a function defined on an ;interval

yl ..f(x1),

y2

f(x2)

'pair,of numbers

(2)

21

x2

in 'I,

with

and let

If, for each

I.

y_satfsfy the, ineclualitY

then

is a strictly decreasing function;

(3)

Y1 < Y2,

then f

14, an 'increasing function;,

{4)

y1

then

is a degreasing function. -

Y2'

the

< x.
2'

is a strictly increasing function;

.2?

Y1 > Y

in

then

.< y

xl, 2c2

and

corresponding values of
y

for

Briefly, this definition states that a function which preserves order


.

relations is increasing; a function which reverses order relations is


decreasing. -Note.particularly that a Strictly increasing functi on is a special,
'case of an increasing function; siMilarly"a strictly decreasing function is-:a
speciril case of a decreasing function.

A funCtionwhich is either increasing or deCreasing is called

monotone. Wunction w

is tither strictly increasing or


r

strictlydecreasing is e lled-strictly 'monotone.

Of

For examp

.k

the function

f EXample Al-lfa'iss monotone over its

entire domain and ,strictly monotoiieori.:the.nosed.'interyal 02X: < 1

well as on the interval

as

x >2.

The graph'of a strictly monotone functiOns

the function

must be one-to-one, hence must have an inverse;/4


.

t:

s. 1

i ,1;.Y

c.

'

In some texts the term "nondecreasIng" ,.s useo instead 46, "increasing";
"nonincreasing" is used instead df "decreasing4,.., ,ii Volume 1 'of this book we
usually drop the phrase "str,intly" from these defiriitions Y ,u4ing (1) or (3)
'las the definition of "increab.ing."
0, 1'.
...s..
,

'

'

03
4.

/.

N;

J
.

'

Al -1+

If,a function is ,strictly monotone, then it has an inverse,

THEOREM

which is strictly monotone in'the same sense.

We treat tne case for

Proof.

strictly increasing; the proof for

If

strictly decreasing is entirely similar.


f(xl) < f(x2); -'that is;

1 ,

Hence,

f(xl) # f(x2).

x2,

take

xi <'x2,

is one-to-olie,-and

f-

then

has an inverse
f(x)---44x

defined for all 4alues

domain of g and
less than

yl < y2,

then

Y1 = f(x1),

and

are in the

and

y2 = f(x2)

must be

g(y2).

g(y1) = xl < x2

Therefore,

(Why?)

x2.

f.

is strictly increasing, for if

Finally,

in the range of

f(x)

Example Alltb. 'The,function


,-.

...-

/76

f : x .,--..xn,

n a natural number, is strictly moftotOne (increasing) for all real


Hence,'

f. has the inverse function

x > O.

,*

1K >0.

x,

g :

For an arbitrary element

.which is alsq an increasing function.

domaIh

of

g;', we denote

by

g(y)

r147;:

thus (1) may be rewritten\

jg

(2)

y > 0.'

CoMpailng (1) and (2), we see that


of the equation

ri,g

n2`'

we call

xn = Yr;

in the

is the unique. positive solution- x,


'

the n-th root of

for all real

'y

If the natural number

is odd, then tlae function. f

strongly monotone for all real

x,. as is its inverse 'function.

that every.real number_ has a unique n-th root or

odd.

This means

For` example,,

'

is:

for n odd and

real,' 9/7171
n

If

is even,

incrtaginifor.all real

dothaim x >0 and

f2

x'7.

x >0.

If

is ddcreasing for all real, x < 0,


f

is the, restriction of f

is the rest iction of


I

these kunCtions has an ,inverse, n

ily .

f "to x < 0,

and

to the

theh eaCh,of

A1-4

and

n,for

n even and all real y > 0.

For n *even,. the positive n-th.root


of
a nonnegative real, number is sometimes called its principal
n -1h root.. The,
symbol 9
always meacs,the principal n-th root.

rr.

Exercises A1-4
A

1.

o'

Prove that
' -(Hint:

2
1

Let

> x

x ---a. x2

for ,x > 0

xi > x2 > 0;

then

- x2 >.0.
2 ;

From this shoW that

2
.)

2.

is a strictly increasing function-.

..

Which of the following functiOns are decreasing?


_

decreasing?

strictly in

real numbers unless oth


.f1

(b)

x----a-c,

easing?

7ise restricted.

& constant

(h)

f8

(i)

f9

(J)

fit)

+ IX'-

--0,1x1
3

(d)

(e)

f 5 :'

strictly

In each case the domain is the set of

f2 :

:(c)

0
increasing'?

(k) .fis

X-- IX

-441

gl

(2)
(r)

f6

(g)

f
7

x < 0

x > 0

(4c g2

f'3f4(

X.'-'6'f4f3(X)

3;

For each function in Probla

2 which is not monotone, divide, its

into parts such thatIthe re triction of

a monotone or strictly.mono onefunctiom4


4

totany.of the,he parts,gives

A1-4
C

4.

We are given that the function.


is increasing,
#,

is strictly increasing,
g,

.g4

14Aecreasing,
is strictly detrpasing,

in a common domain., What is the monotone, character, if any, of the


.7 0

followingfunctions:,
(a)

+ f

f
1

(b)

f2 + gi.
(k)

(c) ,g1 + g,.


(d),,

g2 4 fl,.,

(e)

fl

(1.)

f2

6(2)* glf#21..

f2

(m).. glg2'

-Y g^c

(g)

ga.

g2

-(I;)

g2

fl

g2fi.

(P)

flg2'

,-

or# if

(0)

.
.
"

7
at

4.

...

it has coordinates

(r., e + g)

also.

(rye)

= 0'

(Lr,8)

(r,e + n)

ft

Figure Al -5b

In a cartesian
pair of coerdinates

coordinate system every ppint in the plane.has a unique


(x,y),

is not tae; a given point

In a polar coordinate system, by contrast, this


representation.
in the.plane does not have a unique

In both coordiP' in Figure A1-5a)


in,polar coordinates (see point
specifies a unique point in
nate`systeTs, however, a given pair of coordinate's
the,plane.

A relation between

and

y may be represented by a graph in a .

r and e may be replesented


cartesian coordinate plane. A relation in
and only
coordinate system; a point lies on the'graph if
,by a graph ip a polar
satisfies the given relation:
if it.has at least one coordinate pair which

by equations. ifffiolar
We discuss the graphs of a few functions defined

coOrdinates.
'43

.370 8 (i

A1-5'

..'.

.....

47,;, 'if. A is one angle,

IThere are infinitely many such angles for each point


Thus; a point may be
are the others.
(n = 1, 2', 3, ...)

then 8,.2ng

,4entlfied by infinitely many pairsof polar, coordinates.


(4,21),

(FigureA1-5a), point P with polar coordinates,


(4 , ZE),

(4, .- 27-1),

and,, in

gener4

3 3
The pole (origin) is a special.qase:
pair (0,A),
e

(4

, 3 +,

2ng)

For example
also has coordinates

for any integer

n.

to it we absign.as polar coordinates any


".

any real number.

it is customaty to
When we assign polar coordinates to locate a,point,
is located
(-1-,.8)
r > 0, the point
ff.119w 'r alp tobe negative.
with respect to the origin (Figure AL -5b);
symmetrically'to the, point (r,A)
it has coordinates

,-

(r', e + g)

also.

,e)

8 = 0'

(r,A + 1)'
Figure Al -5b

in the plane. has a unique


In a cartesian coordinate system every ppint
by contrast, this
pair of. coardinates (x,y). In a polar coordinate system,
have a uniqpe representation.
is not tPUe; a given point in the.Plane does not
P' in Figure A1-5a), In both coordiin,polar coordinates (see point
specifies a unique point in
natessysteTs, however, a given pair of coordinate's

the,plane.

,AI

A relation between

and

y may be represented by a graph in a

r and 8 may be reprresented


cartesian coordinate plane. A relation in
only
coordinate system; a point lies on the'graph if and
iby a graph ip a polar
satisfies the given rethion:
if Lt.has at least one coordinate pair which
by equations, id1Polar
Wd discuss the graphs of a few functions defined

coOrdinates.

'

A1:5

The graph of the eqpation


,

r = c

.
contairis:all point

(cie), e

anY'Nlleal,

hding 4t; center at the pole.

number; it is a circle of radius

The equation

r = -c .describes the same

1.

circle.

The points for which

e = c
.

lie on the line'passing through the pole which formi an angle of

c
radians
with the polar''axie;-each point of the line has coordinates (r,c) for some
.
.
real r. For r positive, the points form
.
the ray in direction e, for r
negative,' the ray has, direction 8 + It.
The0line has infinitely many equations e . c + nrc,. n an integer.
.

The circular-functions of` e


polps'coordinates be

are 'especially conveniently represented in

use the entire graph is traced out,in one period.

We

shall illustrate a

ifocedure for 'sketching a' graph of such a function.-psing

polar.coordinate graph
I

r.: Note that the function specifies the graph; a


.,.

function

however, cannot be recovered from its graph in polar coordinates:,

Example Al -5a,. Sketch a...graph of the function def4ined-by-

r =4 cos e.
.

Since

r .is a function of

A- the 'correspond.ing

at e=

valUe '0

valospf
rt
-f

to '1

Hence, in this.interval,
t:,

Since

=2.

increases from -0

and the curVefor

the point
-

to

(4 cos(e

<4 <

5-

and,then decreases.to
1) ,e) 'is the same as

is the entire graph.


4

To 's16
-

values of

1' the graith-ofsthe function, we calculate


:0 ( -

e 'and calculate

:r.- We,ktiow that the cosine increases from,the


at e = 0 and then decreases to 0 at e

cos(e +.x) = -cos e,

(-4 cos 0,,e),

we consider values of

e,

for a few convenient

x
T,E, etc.),

locate the corresponding Points on polar


coordinatepaper, and sketch the graph (Figure Al -5c); it appears.to be a
circle and we shall presently verify that it is.

4i

%.

379

3 87

..

A1-5 ,

(
Figure Al -5c

in the plane has both rectangular end polar coordi-

Since each point- P


nates (Figure Al -5d), for

r > 0,

we

have froth the trigonometric functions


of angles

(1)

x = r cos e, y

r sin e.

We leave it to'you to verify that


equations (1) hold for

r-< 0.

Thus

the rectangulaand polar coordinates


of-eack-point in the plane are related
by (1).

(2)

It folloWs that .
x2

2
= r .

Figure A1-5d
O

380

,388

. A1-5
<

Now we re-examine the function defiried by

and ptov6that its graph is It circle.

r = 4 cop e

(Example A1-4a)

We shall do so by transforming the

given equation into an equation involving rectangular coardinatqs


Nbil the given equation

(3)

r = 4 cos a

and

y,

has the same graph as the equation

r = 4r cos e,
r2

for if

r 1 0,

we may divide both members of. the equation by

given equation

= 0

to obtain the

corresponds to the fact that the pole is on both

graphs.

This may notle immed' tely obvious since only 'Certain pairs of.
..
coordinates representing the' le will satisfy the equation r'= 4 cos e.
e

Forexample, both

(0,0)

of.these pairs satisfies

and (02) represent the pole; yet only the latter


0

r = 4 cos e.

We use (1) and (2) to obtain from (3) that


2

x. + y

or

(x - 2)2 + y2 = 4,
911

an equivalent equation in rectangular coordinates,,

equation of the circle with center at


r

(2,0

We recognize this as an

and radius

2,

verifying tne

graph in Figure A1-5c.

Example Al -5b. -.Find an equation in polar coordinates of the curve whose.


I
2
2 2
y2)2
(x2
y2).
.
equation in cartesian coordinate's is' kx
(x2 + y )
=
.

Applying Equations (1) and (2),weJnave


t

r4 ...a2r2(cos.20

2 2

= a r

sin2e)

cos 2e

This is equivalent to'


r

Since

'r2 ,= a

= 0

(th4cole)

and r

cos 20' is satisfied by

(9,t),

cos 29.'

a set of'polar coordinates

for the pole, we see that r2 =\0 contributes no points not in the
2
2
ra cos 2e. Hence', the latter lean equation in polar.form
whie

381

389
S

of
is the

A1-5

of the given one,

caly the

The-graph of this equation is

lemniscate of Bernoulli and is displayed,ln Figure Al-5e,

Figure, Al-5e

We now develop an equation which, for suitable choice Of a parameter',


For this

will represent either a parabola, an ellipse, or a hyperbola.

purpose -we need the definition of these curves

(conic sections)' in terms of

Every conieCtion (other than the

focus., directrix, and eccentricity.

,circle) maybe defined to be thetet (locus) of all points

such thiit the

and a fixed point

ratio ofthe distance between

(the focus) to the

;'distance between
et:ant

and a fixed line

(the directrix) is EipcNitive con-

e;'called.the eccentricity of the conic section. _If

section is a parabola, if
a hyperbola

the conic

it is an'ellipse,.and if 'e > 1

0 < e < 1

i,

is

*
I

In order to dgrive an evation in


polar coordinates of aconic section,
it is convenient to place the focus
.

/ F at the pole (origin)


direetrio .12

and the

perpendicular to the

extension of the polar axis at distance

p > 0 ?from t}\e pole, as s own

in Figure A1-5f.

(Other orients

ons

= o
,

1.0,COS*e

-are possible; see.EXereises Al=5;


%

Nos: 8-104

isaby

Of the conic section.

FigUre Al-5f.

.
This Curve is defined as the set (locus) ofpoints''P. such that the
product of the"distapces of P from two fixed points is the square. of half
the: distance %etWeerrthe two f1Xed points.
f

382

A1-5

1.

I..
Tft lei

/.

(r,9)

FP '= r' and

be a'n'y pair of'bolar coordinates of',P

ifOr which- r > 0;

DP

+ r cos 6 (Figure A1-5f). The definition of the


FP
conic sections requires that, Bis.
e or
- e.
SolV,Igg for r
p + r cos 8

1,11A1

we 'obtain
ep,

(4)

r,

e, .

e cos 0

wilizh we take to be the standard form of the polar equation of conic sections..
.

' having focus and directrix oriented as in Figure Al -7f.

if

e < 1

(ellipse or parabola), then

r > 0;

if

From Equation-(4),

e > 1, (hgerboli), 'r

may b'e neggtive and these values give us the branch pf the hyperbole lying to
the left.of the directriSc.

,,
.

\
Describe and sketch the graph otthe equation

Example Al_....
-5!'.

16

114

5.- 3 cos e'''

may put this equation in the standard form


16
r

r - 1 cos e
5
.

16.

1 -

,r)s 4

from which

and pi=

e =

Since

e < 1,

the graph is an ellipse with

'3

focus

F1' at' the pdlesand Major axis on the polar axle:. By giiiing

values
2,n).

0 ,-and

g, we find the ends of thj major axis to be

(8,0)

.9- the

and

Thus the length Of the major axis is

10,
the center of the ellipse
is the point (3,0).; and the other focus is, the "point F (6 0).- Since
'
16 ?
p =
(the distance betyeen. a focus and corresponding directrix of the
3

ellipse); the equation of the directrix

pole is

r cos e = -

16
3

the direqriX 12

21

corresponding to the focus at the-

s(sge Exercises A1-5, No. ea.), and the equation of

corresponding to

F2(6,0

is

r cos 9 =

3
.

Uhen

e =

then

r =

,16

16

, and we have the point

(latus rectum)

through F1.

y,

(-5-, .51

at one end of the focal chord

The other.andwnt-has polar coordinates

16 .311

(75 2 )f

these points help us to skdtch the ellipse,as shown

-4

'3&3

419

3p,

A1-5
--

Exercises A1-5
l'. ,Find all .polar coordinates of each of the following. points:

i).

(a)

(6

(b)

(-6 ,,'-i).

(c) (62 - ?).

(a)(-6, - 'i).
%

2.

A)

Find, rectangular 'coordinates of the joints iri Exercise i.

Find polar coordinates of each of the following points given in rectangu.


lar coordinates:
ea(

'

(4, -4).

(a)
'

(e)

(.-3

(f)

(-3 , 4).

, 0).

(b)

(:-2-2'a

(c)

(-2

(d)

,.

i).,

-2iD.

*(g)

(9, -10) 7

(h)

-1,

1)

; -,/).'

(,/'

,,

4.

Given the cartesian coordinates


polar coordinates

(x,y)

of ca point, formulate unique

(r,e). for Q < e < v. '(Hit:


t:

use arccos 1

r.

,,'

5.

Ane

Determine the polar: coordinates bf the three vertices of an equilateral

triangle if a side of the triangle has length

L,

the centroid ot the

triangle coincides with tlie pole, and one angular coordinate of a vertex
is
-

radians.

Find equations inpolar coordinates of th,foidowing curves:


(a) .x = c,

a constant.

(b)

'a constant.

y = c,

%,

(c)

ax +by = c.

(c1)

x2 +.(y1 k)2

(f)

7.

xx

-W.
10

= k2,

..

aa

= 'a2,

'Fineequations in rectangular coordinates of the following curves:


(a)

r . a.

(b)

r, sin:e

(c)

2a,sin B.

(d) r
(e)

=-5.
1

1,- cos e

r i.2 tan B.

385

393_
I-

Al-5'
8.

Derive an equation in polar.coordinates for conic sections with a focus


at the. pole and directrix perpendicular tothe polar axis and
.

units

to the right-etthe pole.

9.

Repeat

Number(8 if the

ireetrix.is parallel to the polar axis and

units above the focus at the pole.

10.

is

Repeat Number 8 if the directrix

parallel to the polar axis and

units below thg focus at the pole.


Discuss and sketch each of the following curves.in polar coordinates.

11.

Example Ai.--5c and Nos.

8', 9,

10.)

8
r'

1 7 cos

.*
5

12

36
-

sin

1.4.

16

(d) "r

5+3
sin

(e)

"-

- 3 cos

pin 0-

= 1 - r

1.2., Certain types c7f symmetry of curves in-polar coordinate's are readily
detected.

For example, a curve is symmetric about the pol4 if the equa-

tit?n is unchanged when


1

is replaced by

-r.

What kind of symmetry

occurs if an equation is,unchanged when


(a)

is replaced by

-0?

(b)

is replaced by

7(-

(c)

and

(d)

1.s replaced by

13.

0?

CI

are replaced by

-r

and

'N.

respectively?

n + 0?

Without actually sketching the graphs, describe the symmetries of the


graphs of the following equations:
r2 = 4 sin 20.

. (a)

(
k

-e,

(b)-,r(1 - c6s19).= 10.


(c) ,r = cos

201

Iair

38 9

Al-5

Sketch the follol,iing curves in polar coordinates:


(a)

(b)

r= a.(1,- co's 0).

= ae.

(d)

r = a sin20 cos 8.

(e)

re = a.

-,

(c) r 'a sin 20.


,

15. In each of the .following, find'all


pairs/of equations.

points of intersectiQh of the given

(Recall that the polar repiesentation of a point is

x.

'snot 41.11..te.)

(a),` r= 2

2 sin 0,

'r = -2 sin420,
(c)

A'

r = 4(1 4., cos'

r = 2

2 cos.

;
r = 2 cos 0
0,

- cos

e)

'p

L
38

395.

Is.

A2-1
4

Appendix 2
POLYNONIAIIS

A2-1.

'

Signiance of Polynomi,pls
The importenceof,polynomials in applicatiots to engineering and the

'natural sciences, as well as in the body of mathematics itself, is not an


accident.

The utility of polynomial6 is based largely on mathematicq proper-.

ties that; for all practical purposes, PermAt the replacement or much more com:plicated functions by polynomial functions in a host of situatfolls.

We shall

enumerate some of these properties:.


(a)

Polynomial functions are among tEe'simplest fudttions to manipulate


formally.

The sum, product, and compositetof polynomial functions,

the determination of slope End area, and the location of zeros and
!k*

maxima andtnima are all'within the reach of elementary methods.

0)1

Polynomial.funotions are among the simplest functions to evaluate.


It is quite easy to find:the value of ,f(x),
x -*a

+ alx + a2x2 .+

given

+ anxn
k

with a specific 'set of coefficients

number for

x.

a0, al, ...an and a specific

Nothingmore than multiplication and addition is

involved, and the-computation'can be shortened by using the method:


of synthet,ic.substitution..

The foregoing two properties or polynothial functionsare those that mat e

them valuab

'as replacements for mare.complicated-functions.

(c) requently an experimental scientist makes a series of measurements,'


'plots them as points, and tiren tries to find a reasonably simple

continuous curve that will pass through these points.

The graph, of

a polynomial function can,always be used for this purpose, and because it has no sharp changes of direction, and only a:limited number
of upA and downs, it is in many.ways the best curve for'the. purpose.
'4er

Thus, for the purpose of fitting a.continuous,graph to a finite number of


points, we wou\III, prefer,to#Nork with polynomials and we need not, look beyond

the polynomials, as we shall prove.


follows:

We can state the prOhlem formally as

389
o.

396
4if

4--

.A2-1

Given

and corresponding values

x1, x2, ..., xn

distinct numbers

that a function is supposed to assume, it is /5ossible,to,find

y2,*..., yn

y1,
-a polynomial,: function of degree at most

i = 01." n.

'.points

n'

whose graph contains the

Y4Ltshave already done this for

you

n = 2:

or(constant function whose graph contained two given points

founda

(x3:,y1), (2,y2), x2./ x1.

the result is'

is also-differe'ntfrom. yr,

y2

If

4t,

it is a constant function.

y2 = yr,

a linear function; if

One way of doing this is, to assume a.polynomia1,9f,the stated form,


,4
2
n-1
,
+
f(x) = ao + alx + a2x +
.

,r

f
.

and write the

.,

equations

yi, i = 1, 2, ..., n.

f(xi)

1
.

'-o,$This giVes.

n lineae equations in the

unknowns

...

0'

l'

and

a'

4-1,

in these circumstances such a system will always have a solution.


%-

,,,

Example A211a.

ft

Suppose that we want the graph of a function to pass

through the points

(-2,2),'(143), (2,-1),

and OM. We know that therys

a polynomial graph of degree no greafer tlian, 3

poi&s.

Assume, therefore,

,-

aix + a2x 2 +-a3x3.

f(x) = a0
Then, if the graph of

which gAs through these

is to go hrough the given points, we must have


and

f(-2) = 2, f(1) = 3, f(2) =,-1,

f(4) = 1;

that it,

ao - tai +.4a2 - 8a3 = 2,


a0 +
a

a
,

Solving these, we find, a0

0
0

al +

a2 +

+ 2a 14 "4a
16a

4: 4a
1

a3 =

+'8a3 = -1,
2
2

+ 64a 0 to

20

3)

31

37

a2

and

, al

11

a3

Hence
4.

f(X)

= 44160

62X

37X2..4

The labor of solving systems of linear equations such as thede can be


rather discouraging, especially if there are many equations.

For this reason,

various methods have been worked,out for organizing and reducing


vplyed.

labor in-

One of the most important of these methods, called the Lagrange Intl"

pOlation Formula, is based On the fglloWi4 simple line of reasoning.


f

390
6 Z)

We can

A2-1
easily write down a formilia for a polynomial of degree

at *n - 1

xis.

-of the,

that is zero

ti

ti

Figure A2-la

A set of values to be taken on by a polynomial function.


instance, that we have four points

(xl,y,1); (x2,y2), (x3,i3)

as in Figure A2-16.

The polynomial

(1)

g (x) = C,(x -,2)(x - x )(x - x

has zerosat.i x
81

(x ),= y 1
1

2'

3'

and

By proper choice of

x4.

Let us do so!

Take

C1

such that

Suppose, for
and

(x4,y4)

we can take

Ci,

40.

= g (x ) = C
1

X2)(X1

x4)',

x3)(xl
4

that is, take

(2)

Cl - (x1

If we substitute

C1

yi
x2)(xi - x3)(xl - x4)

from (2)into (1), we get


(x - x2)(x.- x3). (x - xl;)

, 81(x)

-(37;

= y

(x1 - x,c )(x,

- x3),(x1 - x4)

It
y,

If

at

/ 0

Equation (3) defines a polytomial of degree

that has the value

XLI'

and is zero at

x2, x3,

and

x4.

Similarly, onp finds that

.(x

(4)

f35(x) =

x4)

x)(x

(X2 - xl)(x2

x3)(x2 - x1.4)

'

4
3

391

3 9
e

A2-1
N

( 5)

g3(X) = Y3

(x -

- x2) (x - x4)

- x1)(x3 - x2)(k3 - x4)

(x

'

,.1-

and
*

(x - xl)(x - x2)(x - x3)


. (6)

g4(x)

(xi:- x15(x4 - x2)(x4 = x3)

Y4

are alsg'polynomitlp, ,each having ,the property tlot it is zero at three of the

four given values of

x, 'and-is the appropriate

at the fourth

This

x.

F.

is shown in the talplp below.

The Lagrange Interpolation Formult /Illustrated.

Values of

dingy

Corr
Value of

g (x)

Value of

x4,-.-

Y-4.

Y4

'

0 .

P2(x)

`N.

`Value of

g3(x)

Value of

g4(x)

3
-10

' 0

y
1

Y4

'

If we form the sum

.1;

=, g1(x)+ g2(x) + g3(x) + g4(x),

g(x)

'(T)
.'

L.

..,

then it is clear from the table that

,1

g(xi) = ;Y1 + 0 + 0 +0 =,y1,


g(x2) = 0

+ Y2 + + p

.-

T2'

g(X3) = 0 + 0 + y3 + 0 = y

g(x4) = 0 + 0 + 0+ y4 = y4.
.,

41r.

--.

From .Egillatlons (3), (4), (5), and (6) it is also dealt that
in
.*'

whose degree is At most

polynomial of degree < 3,

3.

_
is a polynomial

Hence Equation (7) tells us how to find a

whose graph contain s the given points.

Example A2 -lb.

tains the points

Find a polynomial of degree at most

(-1,2), (0,0).1 (2,-1),

and

whose graph con-

(4,2).

We find that
,-

g (x)' =
-

1.

(x

4)

2(-1 - 0)(-1 - 2)(-1 - 4)

-g(x) = 0,

J
392

399

2x(x - 2)(x.- 4)
-1,

)
A2-1

1)(1)2(x - 4)

(x

g3(x)

(x t 1)(x)(x - 2)

g4(x)

(()14c

))((lct.

and
i

2
...-gfx),,,,, - 15 x(x - 2)(x - 4) + 1-(x + 1)(x)(x - 4) + 2,(x + 1)(x)(x - 2)
,
12
20
.

- ix).

-1".(x2

The right hand sides of Equations (3), (4),.(5), and (6Y have the
.

4 ;

following structure:
4

Ni(x)

gi(x) = yi

'

i = 1, 2, 3, 4:

,....s,

-Tile numerator of the fraction, is the_product_of.41.but .one, of the factors


"(x 1 xi), (x - .x2), (x :: g3);

and the missing factor is

(x - x ).

and

(x - x4),

The denominator is the value of the_

numerator at

x4t--...x

41)/.. N (x.).

This same structure.would still hold if we had.more (or fewer) points given.
(d)

Instead of a finite set of pdints to whixh a Simple continuous func-'


tion is to be fitted, a mathematicitn4s_sometimes 2gonfronted with a
continudus but very 'complicated function that he. wOUld-14e-tA,,,..-

approximate by a simpler function.

Fortunately,. there is an extremely

powerful theoresof higher, mathematics that enlarges the breadth of


aploliCation oflgiynomials to this situation.

In a_sense this theorem

16;.i*

permits the "fitting" of a polynomial graph'tb any continuous graph.


In'other words, any continuous function whatever can be aeproxithated

by a polynomial function over a'finite interval of its domain, with


preassigned accuracy.
is continuous over

Mote specifically, if the functiori, x -4f(x)

a < x < b,

exists,a polynomial function


If(x)

- g(x)I < c

and
g

for all

such that
x

in

Akt.

393-

is any positive number, there

400.

a < x < b.

A2 -1

-% ,;
t\C4

../
//

vs

.
Figure A2 -lc

A strip between

f(x) - c

and

f(r) + c.

This is known as the Weierstrass Approximation Theorem.


pretation of the theorem

ir indicated in Figure ,A2-1c

continuous curve,

may 'have

The febmetric inter-

..

and minima between x= a

and

But Suppose that we introduce a

Xhe graph of

is a

jarp corners or even infl.nitely ,many maxima

x = b.

atci.p,

No polynomial graph behaves like that


centered on the graph'

f,- extending

between the.graphs of the functions

st

x f(x) -.,c
and
X

where

-4.f.

(X) +

is any preassigned positive number, however small.

Then the theorem

guarantees that there is Al polynomial function


g : x -;,g(x),

whose graph on, a < x < b *lies entirely inside "this strip.
cise meaning of the statement:

o.

This is the pre-

"Any continuous function whatever can be

approximated by a polynomial function over a finite interval of its domain,


with pfeassigned accuracy."

Exercises A2-,1
1.

Carry out the computations in 'Example A'2 -la, above.

Simplifyl the expression for

3.

in Example A2-1b,.. above.,_

Find a polynlixiiial function of degree less than or equal to

graph contains

4.

g(x)

points

and

whole

(4,2), (0,-1),,(2,3).

Find a polynomial function whose graph contains the points

(2,91-,---(3;307

'( 1,6).

O7,

394

4'0
TO

(0,1), (1,0),

, A2-2
vt;

A2-2.

Number of Zeros
One could get, the impression that every polynomial function of degree

n > 0 'has exactly

zeros.

This is 'not quite right4, what we say in Sedtion

n zeros.

1-9 is that every such-function has at most

Let us exhibit a poly-

.nomial function for which the numberof zeros is less than the degree:
V
quadratic function

f
has only one zero, namely
factors

x - 3,

x2

3.

But since the quadratic has two identical

-, 6x + 9 = (x - 3)2.

14e say that the zero-

has multiplicity two.

Wedefine the multiplicity of a zero

expogntof the highest power,of

f(4) = (x - r)
where. ,q(x)

is a zero of

is a polynomial, and if
f

The

of multiplicity

of a polynomial

that diVides

f(x).

to be the

k > 0,

q(x),

noes not divide

x - r

That is, if

q(x),

then

k.

The proof of the general theorem abput the number of zeros of a'poly:
nomial function depends on !She fact that every such function has at least one
zero.

.7.

This fact, often referred tq as Gauss's Theorem, is stated as follows:

pie

The Fundamental Theorem of Algebra.

Every polynomial function of

degree greater than zero has at least one zero, real or complex.

This is the simplest form of the Fundamental Theordm of Algebra.

(As a

matter of fact, the theorem isicorrect even if some or all of the coefficients
of the polynomial are complex numbers.)

The first known proof of the theorem was published by the great.German
mathematician Carl Friedrich Gauss (1777 = 1855) in 1799.
has written an inteieSting'accoUni of Gauss.

(Eric Temple Bell

See World of Mathematics, Simon

and Schuster, 1956, Volume 1, paget7295-339, or E: T. Bell, Men of Mathematics,


Simon and_Schuster, 1937, pages 218 -269.)

The proof was cdntained in'Gauspis


A

doctoral dissertation, published when he was

A translation of his second

22.

proof (1816) is in A Source Book in Mathematics, by David Eugene Smith, McGraw-

Rill Book Co., 1929,pages 292-310.

Gauss gave a total of four different ,.,

proofs of the theorem, the last in 1850. 'None of the proofs is sufficiently
9

395

14Q2

12-2.,

elementary to be given here.

There is a. proof in Birkhoff and MacLane, A Sur-

Lei of Modern Algebra, Macmillan, 19531 pages 107-109.

A pro3f is also given

in L.. E. Dickson? New First Course in the theory of Equations, John Wiley and
Sons, 1939..

We now state and prove the general theorem.

The General Form of the Fundamental Theorem of Algebra.


a polynomial function of degree
and at most

Then

n > O.

Let

f be

has at least one

complex zeros, and the sum of the multiplicities of

the zeros is exactly

n.

From Section 1-9 we know that

has at least one zero, 'say

(recall the Factor Theore) there is a polynomial

rl.

Then

of degreern - 1

q(x)

such that
k:I^e

f(x) = (x - r )q(x).

(1)

If

.o.fs q

1,

is

is of degree-zero and we have finished.

n - 1

and'is.positive...Then

could'happen that

r2 = r1)

If

n > i,

the degree

has at least one zero, r2

(it

and

q(x) = (x - r2)s(x),

(2),

where

s *is of degree

n - 2.

Combining (1) and (2) gives

1,

(3)
If

f(x) = (x
n = 2,

then

ri)(X - r2)s(x).

in Equation (3) is of degree zerostand we have finished.

Otherwise, the proCess may be continued until we arrive at the final stage,
(10

f(x) = (x - ri)(x

where the degree of

r2)

n - n = O.

is

(x - rn)z(x),

Hence,

z(x)

is a constant.

Compari-

son of the expanded form of Equation (4) with the equivalent form
f(x) = anxn + a

*shows that
(5) '

z(x) = a

/ O.

n-1

xn-1 +

+..a

0,
ti

Hence,

f(x) =

(x - t )(x -'r2)
.n

(x - r ).

396

x + a

A2-2,

Now, if we substitute any complex; number


r

in.place of

2"

different from 'r-

r2,

In.Equation (5), we get


te

f(r);:= a n(r.r r 1)(r - '2)

(r - r )..

Since every factor is different fr4m!zero, the product cannot be zero.


no number except
;

r2,

r
1,

r
'

is a zero of

f,

and

Hence,

hasat most

zeros.

Since it is possible.that soine of the -r is


zeros of

f may be less,than

may be eqUal, the numbei. of

n.. ,But Equatioli 5) shows that

n factors of the form. x - r.

has exactly

and therefore the sum of.the multiplicities

'of the zeros must be

`n.
Example A2-2.a..
f

x -4x5 + x

5x,

has zeros of multipli.city greater than one.

+ 8x - 4

Find the zeros and indicate the


-

multiplicity of each.
Since the coefficient of the term of" highest degree is

any rational zeros of

we know that

1,

f 'must be integers thdt are factors of

4.

'

Section 1-7.)

(Refer to:.

Uiing synthetic substitution and the'polknomial of reduced,

degree obtained each time a zero is found, we discover that


multiplicity three and

-2

is a zero of multiplicity two.

is a zero of

Note that the sum

of the multiplicities is fide, which is also the degree of the given.polynOml:al.

It may be helpful to show a practical way for putting down the synthetic
substitutions by which we obtained the zeros and their multiplicities.

This

-is.done'in Table A2-2.

Table A2.-2

Finding the Zeros of

5x

.4- x

-1

-4

-3

-4

-3

-4

-a

-4

ta,

4._

x --4x

- x

+ 8x - 4

fs a zero

1
1....1...4

0.

-4

-0

'-4

3.-

-4

of

three.

40'4

of

multiplicity

397

in

-it',

A2-2
The entries

,3, -4, 4

1,

1,*4, 4

sixth row, and

in the third row,

degreA four, three, and two, respectively.


2

x --4x

2
x

+ 4x + 4

+ 4x + 4

has

1, 3, 0, -4

in the

.
In the last-row
are coefficients of polynomialsof

The quadratic function

as a zero of multiplicity two since'

-2

(x + 2)2.

Thus,'the zeros, of

(of multiplicity three) and

are .1

(of

-2

multiplicity two).
two):
."

viP

The graph of

is. shown in Figdre A2-2a in 'order to give you some idea

of its shape in the neighborhood of the zeros. -2

and

(points

and

ti

f (x)

arc

Ffgure
Graph_Of

f :,x :4Z

8
4'

398
94 4:

L.

sr

t..)

A2=2

\'
.

>,
'

.-

, The Fundamental TheorelA of Algebra implies that the.range of-any-_-,poncon-

stant polynoMial function includes zero when its domaih is the set of ail

cm-

.;

plex numbers.

The range does not always include zero "wh' en the domain-is the
,

set of real numbers.: For example, if

f :.x -*.y

x2 + 1,

R,
p
-AT"

then the range of

is the set.
(y

When the domain of


f

is > 0,

y > 1).

is the' set of complex numbers, and the degree of

then its range is also the set of all complex numbers.

suppose that
plex number.

f' is a polynomial of degree

n > 0 'and

a + ib.

Foe,

is any com-

Then the equation


f(x)

a + ib

is equivalent to
(6)

a - ib = 0.

f(x)

'

This is a polynomial equation of degree


of Algebra, Equation (6) has a solution.
complex number

that is mapped by

hence, by the Fundamental Theorem

n;

That is, th4re exists at least one

f into

a + ib:

'

f(x) = a + ib.

Moreover, there may be aa-many as


map into

a + ib,

will be exactly

n_

different' numbers in the domain that,

and the sum of the multiplicities of the solutions of (6)

n:
.1

The Fundamental Theorem does not tell Us how to find even one of the zeros
of

f.

Yt just guarantees that they exist.

The general problem of finding a

cdmplex zero o on arbitrary Polynomial is ,quite difficult.

'In the '1930's the

Bell Telephone Laboratories built a machine, the Isogiaph, for solving such
,

problems when the desee


g
is

1Q kr less.

See The Isograph -- A Md6'anica1

Root-Finder, by R. L. Dietzold, Bell Labs ReCord, 16, December, 1937, page 130.

Nowadays, electronic computers are used to do this*job, and many others.


Numerous applications of computers in science and industry are discussed in a
series of articles in the book The Computing Laboratory in the University,
University of Wisconsin Press, Madison, Wisconsin, 1957, edited by Preston C.
Hammer.
'

The followyg quotation is taken from a book called Mathematics


and CoMputers,ty GeorgekRiptibitz and Jules A. Larrivee, McGraw-Hill Book
4.01

Co., Net; York, 1957, page 37:

399

406

A2-2.

"There is an interesting use for the roots of the 'characteristic equation' of a vibrating system in the dynamics of electromagnetic and mechanical
systems Where many.of the properties of amplifiers, filters, servos, airfo21s,
and other devices must be determined.

If any one of the complex roots of

-characteristic equation for a system has 0 positive real part,the system


will be'unstable:

amplifiers will howl, servos will oscillate.uncontrolla-

, bly, and bridges will collapse under the stresses exerted by the winds.
..

The

prediction of such behavior isof great importance to designers of the amplifies that boost your voice-are ktdcrOsies the country over telephone lines,
and the servos that point guns at an attacking,plane."

Exercises A2-2

Assume that the equations given below are the'characteristic eqUations of

1.

some mechanical or electrical system.

According to the quotation' front

,-"tstabitz and Larrivee, are the systems stable or Astable?


3

- x

(a)

(b)

x3.- 3x2 + 4x - 2,= 0,

(c)

x3 + 3x2 + 4x + 2 = 0

(d)

x311 x2 - 2 = 0,

(e)

x3 + 6x2 +.13x + 10 = O.

"S.

+ 2 = Q,,

The following equatiips have multiple roots.

12.

Find them and, in eachcase,'

snow trat the s.m oP the malti.%licities of the roots equals...the degree of

the

olynomial.

(01 x3 - 3x - 2 = 0,
(b)

x3 - 3X + 2 = 0,
13

x4

5x3

9x2

7x + 2 = O.

Find the roots and their multiplicities of each of the following equa-

3.

tions.
.

Compare the solution%ets of the two equations.

(a)

x5 + 4x4 + x3

(b)

X5

X4

5X

10x2 - 4x + 8 =

+ 8x -4 =0

0'

4 0,7

'

A2-2
4.

A number system is said to be algebraically closed if, and only if, every
polynomial equatipn of degree > 0,
68E: solution in that system.

with coefficients in that system,

Which of the following number systems are,

and which are no:t, algebraically closed?


(a)

The integers:

Give reasons for your answers.

-1, 0, 1, 2; 3,

(b). ThetrationaI numbers.

Thfreal numbers:

5,

(d)

The pure imaginary numbers b

(e)

The complex numbers.

os,

You may have heard that it was necessary for matheMatic4ans to invent
.and,other.complex numbers in order'.to solve some ,quadratic equatiohc.

Do you4suppose that they needed to inve.nt,something that' might be called

"super-complex" numbers t6 expressXach things ap


pp?

Gives reasons for youransyers.

,,10.

--14

71.1'

tw

401

i0g
11

Y-1, 6, and so ''

A2-4,

A2-3.

Complex Zeros

We know that a iladraticequation


A

ax + bx + c = 0,

(1).

has roots given by the quadratic formula

4.

x-

'(2)

The coefficients

a, b,

and

-)

t 42 -

C
.

2a

if O.) are here assumed to be real numbers.

The quantity under the radical in (2) 16 called the discriminant-.

The roots are

determines the nature of the roots of (1).

imaginary if

- 4ac j 0,

- 4at < 0.

What are the roots

EXagple

a.

bg = 4ac > 0,

real and unequarif


real and equal if

Its sign

cff

2'

= 0'?

'

The roots are4

-1 - i 4

-1 + i 4
2

'

We notice that these roots are complex conjugates; that 1,e, they have the
.1,

form
A
11

1.r+ ty

u - iv,

and

and v=

where .0

and

are real., In this example,

,/-

,,

2 '

Let

Is it just a coincidence that these roots are complex con:lugates?


us look at (2), and suppose thatfa, b,, and
discriminant is negative, say

-d

2
.

are real nuottb76s and that the

Then the roots of

ax

+ bx + c = 0

are

, d,
b
b
.d.
conjugates. Thus, if a,
- ik). Thes'e are
+ i(2a) and 2a
2a
2a
2a
t.
c are real and if the roots of (1) are imaginary, then these rbotsand
b,
.

are Complex conjugates.,,,TE/r3is true of polynomials of any degree, as we

shall now prove.

(In the following theorem, the letters 'a

and

represent
.

the real and imaginary parts of a complex root of an equation of any degree,

and do ngt refer to the coeffIcientSin a Oadraticekpression.)

*4.5106

:,.._

4
I

1402

409

...1,r

A2-3'
Complex-conjugates Theorem.

coefficients, and if
imaginary part

If. f(x)

a + tb

b /-0,

is a polynomial with real

is a complex root 'of

then

f(x) = 0,

is also a root.

- ib

with

(Another way of saying this is that if' f(ac4-4) = 0, ,with


a

and

real and

b # 0,

then

f(p

ib) = 0.)

We give two proofs of this result.

First PrQof.

The key to this proof is the use pfd the quadratic polynomial
tt- ,
that is the product of x + it)
and x
(a - ib). We show ,that it
.

dies f(x).

We can then conclude that

f(a - it) = 0

and we have com-

pft-ted the proof.


Thd..3.- let

(3)

fx

p(x)

1(4

(a + ib)][x- (a - ib)]
a)

- it2lf.(x

+ ib)

a)

;
K

.N

Note that' p(X)

with real lioeffients

ic; a quadratic pAyhomial

Now

a polynomial ic divided by a quadratic, a remaindet of degree les, thab


obtained,
q(x)

Henc4, if

and a remainder

'k

where

f(x)

h, k,

is divided by

f(x) = p(x)

This is an identity in
p(a + 1.0 = 0.

is

of dowiree

(but no greater),

.L,

and all the coeffic Ats

(4)

we get a polynomial quotient

p( ),

ossi

r(o* xli= hx +

whet

are real.

Thus,

+ k.

q(x)

f(a + ib) = 0, and from Equatpn

By hypothesisa

Therefors, if we substitute. a + ib

for

in Equation

(4), we get
0 = 0 + ha + ihb + k.

Since real and imaginary parts most both be


4-

0,

we have

ha + k

(5)

and

hb = O.

(6)

-Since

b / 0

tion (5) gives

0
* , 4 0 -P`'

(by,);lypothesis), Equation (6) requires fiat


k = 0.

h = 0.

Then Equa-

Therefore, the remainde7441ix + k in Equation (4) is

zero, and

4.10,

fr.

A2-3

Since

p(a - i

f(x) 4 p(x)

q(x).

by Equation (3), it follows from Equation (7)/that

= 0

172.

f(a - ib) = 0.

Let

Second Proof.

f(x) = a xn + an-lx

(8)

n-1

and suppose that


Equation (8),

f(a + ib) = 0.

we can expand

+.a lx + a

When we substitute

(a + ib)

(a + ib)3,

0'

a '+:tb

for

in

pnd sb on, by the Bino-

mial Theorem.

We can prove the complex-conjugatps theorem, however, without

actually carrying out all of these` expansions, if we observe how the terms
behave.

Consider the first few poWers of


(a + ib)

a +

= a + ib,
.,:
)

(a +'ib)

- a

+ 2aib +

(a2

iO3

(a

a3

.2.
b )

i2b2

'

+ i(2ab),

3a2ib 4:3a(i2b2),.4. i3b3

= (a3 - 3ab2) + i(3ab - b3).

Now observe where


b

b 'occurs in the above expanded forms.

In the real parts,

either does not occur at all, or it)occurs only to even powers.

imaginary part40- b

that all evri powers of 'i

In the

This follows prom the fact

always occurs to odd powers.

are real and all odd powers are imaginary.

If we

change the_sign of

b,

w:e therefore leave the real part unchanged and chadge

the signyof the imaginary part.

f(a- ib) = u - iv.

Thus, if

f(a + ib) = u +- iv,

then

tut by hypothesis,

'f(a + ib) = 0;
so that

u + iv = 0,

and therefore

u = V = 0..

f(a - ib)

Hence

0. r.

.404
o

A2-3
What is the degree of a polynomial function

,ExamRle A2-4a.

of mini-

and

2 + i, 1,

mum 'degree if,

3 - 2i 'are zeros of

f?

6
I

If it is not required ,that the coefficients of the polynomial be real,

then we may take


'f(x) =,(x - (2 -4 i)1[
= x

+ (-6

In this case, the degree of


can have

zeros, so

coefficients of

l][x - (3

i)x2 7 (13 - 2i)x 4- (-8 + i) .

is

3.

NO polynomial function of lower degree

is the answer.

However, if it is required that the

be real, thert the answer to tile question is

f(x)

For

5.

41

then the conjugates of ,2 + i

and

3 - 2i..

polynomial function of degree less tan 5


2 + i, 2

(9)?,

must also be zeros of


can have the

f.

No

zeros

1, 3 - 2i, 3 + 2i.
11

But

(10). [x

(2 + ink

4 a polynomial of degree

i)][x - l][x -.(3 - 20][X,

(2

5,

(3

2i)]

with real ,coefficients, that does have the

numbers listed in (9) as its zeros.,

Exercises A2-3
1.

Multiply the ,fa tors in (10) above to show that the expression does have
real coefficients.

What is the coefficient of


%-

What is the constant term?

in your answer?

Compare these with the sum and the product

of the zeros listed in 4(9).


2.

Write a*Iklynomial'function of mininillin degree that has


(a)

if imaginary coefficients are allowed,

(b)

if the coefficients must be:real.

2 + 3i

as a zero,

A2",1

3.

Find all roots of the following equations:


r,

(al

x3 - 1 = 0

(b)

x3 + 1 = 0

(c)

x3

(d)

(e)

x4

(f)

(g)

x6

x2 + 2,C= 8

+ 5x
-

2x3 + 10x2.- 18x + 9 = 0

+ 2x
-

+ 4 = 0

+ 3x

2x5-+ 3x4

+ 4X
-

+ 3x

+ 2x t 1 '=-0

(-

4x3 + 3x2 - 2x + 1 = 0

What is the degree of the polynomial equatoion of minimum degree with

4.

real coefficients having

2 + i, -2 + i, 2 - 1, 3 + i, -3 + i

as roots?
I

arl- b1,

Consider-the set of numbers of the form

5.

a - big

rational\ Then

where

is called the conjugate surd ,of

and

a + big.

are

Prove

the follo4dg theorem on conjugate surds:


If

a + b1 is a root of

f(x) = 0,

that if, u = v1 = 0,

and

then

apd

Otherwise, we 'could solve for v = numbers.

is a polynomial with rational coefficients, and if

f(x)

a - big

is also a root.

are rational, then


,

(Note

u = v = 0.

the quotient of two rational

But we know that V is irrational.)

6. .Find a polynomial with rational coefficients and minimum.degree'hpving

3 + 2V as a zero.

7. /State and prove a theorem similar to that in Exercise 5 abOve for numbers

of the form

a + b 1. Is there a comparable theorem abOut roots Of the

'ford a + 1)4 ?" Give reasons for your answers.


Write a polynomial:function of minimum degree that has

8.

-1

and

3 -

as zeros, if
, ...

irrational oerficients are allowed;

(b)

thecoefficie is must be rational.


0

4
9.

,t

)
('a)

Find,a,polynomial of

inimum degree with rational coefficicents having

If as a zero.
10.

What is the degree of a polynomial, of minimum degree with (a) real, and
(b) rational coeffici ents having

'

as a zero?

(1)

i + ig

(2)

1 + i1 as a zero?

(3)

if +

as a zero?

406.

413
.