CLP-1 Differential Calculus Problem Book2

CLP-1 D IFFERENTIAL C ALCULUS
E XERCISES
Elyse Y EAGER
Joel F ELDMAN Andrew R ECHNITZER
T HIS DOCUMENT WAS TYPESET ON T UESDAY 23 RD A UGUST, 2022.

§§ Legal stuff
• Copyright © 2016–21 Joel Feldman, Andrew Rechnitzer and Elyse Yeager.
• This work is licensed under the Creative Commons Attribution-NonCommercial-
ShareAlike 4.0 International License. You can view a copy of the license at
https://creativecommons.org/licenses/by-nc-sa/4.0/.
• Links to the source files can be found at the text webpage
2
H OW TO USE THIS BOOK
Ĳ Introduction
First of all, welcome to Calculus!
This book is written as a companion to the CLP-1 Differential Calculus textbook.
§§ How to Work Questions

This book is organized into four sections: Questions, Hints, Answers, and Solutions. As
you are working problems, resist the temptation to prematurely peek at the back! It’s
important to allow yourself to struggle for a time with the material. Even professional
mathematicians don’t always know right away how to solve a problem. The art is in
gathering your thoughts and figuring out a strategy to use what you know to find out
what you don’t.
If you find yourself at a real impasse, go ahead and look for a hint in the Hints section.
Think about it for a while, and don’t be afraid to read back in the notes to look for a key
idea that will help you proceed. If you still can’t solve the problem, well, we included the
Solutions section for a reason! As you’re reading the solutions, try hard to understand
why we took the steps we did, instead of memorizing step-by-step how to solve that one
particular problem.
If you struggled with a question quite a lot, it’s probably a good idea to return to it in a few
days. That might have been enough time for you to internalize the necessary ideas, and
you might find it easily conquerable. Pat yourself on the back–sometimes math makes you
feel good! If you’re still having troubles, read over the solution again, with an emphasis
on understanding why each step makes sense.
One of the reasons so many students are required to study calculus is the hope that it will
improve their problem-solving skills. In this class, you will learn lots of concepts, and
be asked to apply them in a variety of situations. Often, this will involve answering one
really big problem by breaking it up into manageable chunks, solving those chunks, then
i
putting the pieces back together. When you see a particularly long question, remain calm
and look for a way to break it into pieces you can handle.
§§ Working with Friends

Study buddies are fantastic! If you don’t already have friends in your class, you can ask
your neighbours in lecture to form a group. Often, a question that you might bang your
head against for an hour can be easily cleared up by a friend who sees what you’ve missed.
Regular study times make sure you don’t procrastinate too much, and friends help you
maintain a positive attitude when you might otherwise succumb to frustration. Struggle
in mathematics is desirable, but suffering is not.
When working in a group, make sure you try out problems on your own before coming
together to discuss with others. Learning is a process, and getting answers to questions
that you haven’t considered on your own can rob you of the practice you need to master
skills and concepts, and the tenacity you need to develop to become a competent problem-
solver.
§§ Types of Questions
Q[1]: Questions outlined in blue make up the representative question set. This set of
questions is intended to cover the most essential ideas in each section. These questions
are usually highly typical of what you’d see on an exam, although some of them are
atypical but carry an important moral. If you find yourself unconfident with the idea
behind one of these, it’s probably a good idea to practice similar questions.
This representative question set is our suggestion for a minimal selection of questions to
work on. You are highly encouraged to work on more.
Q[2](˚): In addition to original problems, this book contains problems pulled from quizzes
and exams given at UBC for Math 100 and 180 (first-semester calculus) and Math 120
(honours first-semester calculus). These problems are marked with a star. The authors
would like to acknowledge the contributions of the many people who collaborated to
produce these exams over the years.
Instructions and other comments that are attached to more than one question are written in this font. The
questions are organized into Stage 1, Stage 2, and Stage 3.
§§ Stage 1
The first category is meant to test and improve your understanding of basic underlying
concepts. These often do not involve much calculation. They range in difficulty from
very basic reviews of definitions to questions that require you to be thoughtful about the
concepts covered in the section.
ii
§§ Stage 2
Questions in this category are for practicing skills. It’s not enough to understand the philo-
sophical grounding of an idea: you have to be able to apply it in appropriate situations.
This takes practice!
§§ Stage 3
The last questions in each section go a little farther than Stage 2. Often they will combine
more than one idea, incorporate review material, or ask you to apply your understanding
of a concept to a new situation.
In exams, as in life, you will encounter questions of varying difficulty. A good skill to
practice is recognizing the level of difficulty a problem poses. Exams will have some easy
questions, some standard questions, and some harder questions.
iii
iv
C ONTENTS
How to use this book i
I The questions 1
1 Limits 3
1.1 Drawing Tangents and a First Limit . . . . . . . . . . . . . . . . . . . . . . . 3
1.2 Another Limit and Computing Velocity . . . . . . . . . . . . . . . . . . . . . 4
1.3 The Limit of a Function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
1.4 Calculating Limits with Limit Laws . . . . . . . . . . . . . . . . . . . . . . . 7
1.5 Limits at Infinity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
1.6 Continuity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
2 Derivatives 19
2.1 Revisiting tangent lines . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
2.2 Definition of the derivative . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
2.3 Interpretations of the derivative . . . . . . . . . . . . . . . . . . . . . . . . . . 26
2.4 Arithmetic of derivatives - a differentiation toolbox . . . . . . . . . . . . . . 27
2.5 Proofs of the arithmetic of derivatives . . . . . . . . . . . . . . . . . . . . . . 29
2.6 Using the arithmetic of derivatives - examples . . . . . . . . . . . . . . . . . 29
2.7 Derivatives of Exponential Functions . . . . . . . . . . . . . . . . . . . . . . . 31
2.8 Derivatives of trigonometric functions . . . . . . . . . . . . . . . . . . . . . . 33
2.9 One more tool - the chain rule . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
2.10 The natural logarithm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
2.11 Implicit Differentiation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
2.12 Inverse Trigonometric Functions . . . . . . . . . . . . . . . . . . . . . . . . . 43
2.13 The Mean Value Theorem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
2.14 Higher Order Derivatives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
3 Applications of derivatives 53
3.1 Velocity and acceleration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
v
CONTENTS CONTENTS
3.2 Related Rates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55

3.3 Exponential Growth and Decay . . . . . . . . . . . . . . . . . . . . . . . . . . 63
3.3.1 Carbon Dating . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
3.3.2 Newton’s Law of Cooling . . . . . . . . . . . . . . . . . . . . . . . . . 64
3.3.3 Population Growth . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66
3.3.4 Further problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
3.4 Taylor polynomials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
3.4.1 Zeroeth Approximation . . . . . . . . . . . . . . . . . . . . . . . . . . 68
3.4.2 First Approximation . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
3.4.3 Second Approximation . . . . . . . . . . . . . . . . . . . . . . . . . . 71
3.4.4 Still Better Approximations . . . . . . . . . . . . . . . . . . . . . . . . 73
3.4.5 Some Examples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74
3.4.6 Estimating Changes and ∆x, ∆y notation . . . . . . . . . . . . . . . . 75
3.4.7 Further Examples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76
3.4.8 The error in Taylor polynomial approximation . . . . . . . . . . . . . 78
3.4.9 Further problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80
3.5 Optimization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82
3.5.1 Local and global maxima and minima . . . . . . . . . . . . . . . . . . 82
3.5.2 Finding global maxima and minima . . . . . . . . . . . . . . . . . . . 84
3.5.3 Max/min examples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
3.6 Sketching Graphs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88
3.6.1 Domain, intercepts and asymptotes . . . . . . . . . . . . . . . . . . . 88
3.6.2 First derivative - increasing or decreasing . . . . . . . . . . . . . . . . 90
3.6.3 Second derivative - concavity . . . . . . . . . . . . . . . . . . . . . . . 92
3.6.4 Symmetries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
3.6.5 A checklist for sketching . . . . . . . . . . . . . . . . . . . . . . . . . . 96
3.6.6 Sketching examples . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96
3.7 L’Hôpital’s Rule and indeterminate forms . . . . . . . . . . . . . . . . . . . . 100
4 Towards Integral Calculus 103

4.1 Introduction to antiderivatives . . . . . . . . . . . . . . . . . . . . . . . . . . 103
II Hints to problems 107

1.3 The Limit of a Function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109
1.5 Limits at Infinity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112
1.6 Continuity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113
2.1 Revisiting tangent lines . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114
2.3 Interpretations of the derivative . . . . . . . . . . . . . . . . . . . . . . . . . 116
2.7 Derivatives of Exponential Functions . . . . . . . . . . . . . . . . . . . . . . 117
vi
CONTENTS CONTENTS
2.8 Derivatives of trigonometric functions . . . . . . . . . . . . . . . . . . . . . 118

2.9 One more tool - the chain rule . . . . . . . . . . . . . . . . . . . . . . . . . . 119
2.10 The natural logarithm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 120
2.11 Implicit Differentiation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122
2.12 Inverse Trigonometric Functions . . . . . . . . . . . . . . . . . . . . . . . . 123
2.13 The Mean Value Theorem . . . . . . . . . . . . . . . . . . . . . . . . . . . . 124
2.14 Higher Order Derivatives . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125
3.1 Velocity and acceleration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126
3.2 Related Rates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127
3.3.1 Carbon Dating . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 129
3.3.2 Newton’s Law of Cooling . . . . . . . . . . . . . . . . . . . . . . . . . . . . 130
3.3.3 Population Growth . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 130
3.3.4 Further problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131
3.4.1 Zeroeth Approximation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131
3.4.2 First Approximation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 132
3.4.3 Second Approximation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 132
3.4.4 Still Better Approximations . . . . . . . . . . . . . . . . . . . . . . . . . . . 133
3.4.5 Some Examples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133
3.4.6 Estimating Changes and ∆x, ∆y notation . . . . . . . . . . . . . . . . . . . 134
3.4.7 Further Examples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 134
3.4.8 The error in Taylor polynomial approximation . . . . . . . . . . . . . . . . 135
3.5.1 Local and global maxima and minima . . . . . . . . . . . . . . . . . . . . . 137
3.5.2 Finding global maxima and minima . . . . . . . . . . . . . . . . . . . . . . 137
3.5.3 Max/min examples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 138
3.6.1 Domain, intercepts and asymptotes . . . . . . . . . . . . . . . . . . . . . . 139
3.6.2 First derivative - increasing or decreasing . . . . . . . . . . . . . . . . . . . 139
3.6.3 Second derivative - concavity . . . . . . . . . . . . . . . . . . . . . . . . . . 140
3.6.4 Symmetries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 140
3.6.5 A checklist for sketching . . . . . . . . . . . . . . . . . . . . . . . . . . . . 141
3.6.6 Sketching examples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 141
3.7 L’Hôpital’s Rule and indeterminate forms . . . . . . . . . . . . . . . . . . . 142
III Answers to problems 145

1.6 Continuity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 153
vii
CONTENTS CONTENTS

3.2 Related Rates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 181
3.3.1 Carbon Dating . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 182
3.4.5 Some Examples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 187
3.5.3 Max/min examples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 196
3.6.4 Symmetries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 200
IV Solutions to problems 215

1.6 Continuity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 249
viii
CONTENTS CONTENTS

3.2 Related Rates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 368
3.3.1 Carbon Dating . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 390
3.4.5 Some Examples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 423
3.5.3 Max/min examples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 468
3.6.4 Symmetries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 486
ix
CONTENTS CONTENTS
x
Part I
T HE QUESTIONS
1
Chapter 1
L IMITS
1.1Ĳ Drawing Tangents and a First Limit

Exercises
Jump to HINTS, ANSWERS, SOLUTIONS or TABLE OF CONTENTS.
§§ Stage 1
Q[1]: On the graph below, draw:

(a) The tangent line to y = f ( x ) at P,
(b) the tangent line to y = f ( x ) at Q, and
(c) the secant line to y = f ( x ) through P and Q.
y
y = f (x)
Q
P
x
Q[2]: Suppose a curve y = f ( x ) has tangent line y = 2x + 3 at the point x = 2.

(a) True or False: f (2) = 7
(b) True or False: f (3) = 9
Q[3]: Let L be the tangent line to a curve y = f ( x ) at some point P. How many times will
L intersect the curve y = f ( x )?
3
L IMITS 1.2 A NOTHER L IMIT AND C OMPUTING V ELOCITY
1.2Ĳ Another Limit and Computing Velocity

Exercises
§§ Stage 1
Q[1]: As they are used in this section, what is the difference between speed and velocity?
Q[2]: Speed can never be negative; can it be zero?
Q[3]: Suppose you wake up in the morning in your room, then you walk two kilometres
to school, walk another two kilometres to lunch, walk four kilometres to a coffee shop to
study, then return to your room until the next morning. In the 24 hours from morning to
morning, what was your average velocity? (In CLP-1, we are considering functions of
one variable. So, at this stage, think of our whole world as being contained in the x-axis.)
Q[4]: Suppose you drop an object, and it falls for a few seconds. Which is larger: its speed
at the one second mark, or its average speed from the zero second mark to the one second
mark?
Q[5]: The position of an object at time t is given by s(t). Then its average velocity over the
s(b) ´ s( a)
time interval t = a to t = b is given by . Explain why this fraction also gives
bá
the slope of the secant line of the curve y = s(t) from the point t = a to the point t = b.
Q[6]: Below is the graph of the position of an object at time t. For what periods of time is
the object’s velocity positive?
y
y = s(t)
t
1 2 3 4 5 6 7
§§ Stage 2
Q[7]: Suppose the position of a body at time t (measured in seconds) is given by
s(t) = 3t2 + 5.
4
L IMITS 1.3 T HE L IMIT OF A F UNCTION
(a) What is the average velocity of the object from 3 seconds to 5 seconds?
(b) What is the velocity of the object at time t = 1?
Q[8]: Suppose
? the position of a body at time t (measured in seconds) is given by
s(t) = t.
(a) What is the average velocity of the object from t = 1 second to t = 9 seconds?
(b) What is the velocity of the object at time t = 1?
(c) What is the velocity of the object at time t = 9?
1.3Ĳ The Limit of a Function

Exercises
§§ Stage 1
Q[1]: Given the function shown below, evaluate the following:
(a) lim f ( x )
xÑ´2
(b) lim f ( x )
xÑ0
(c) lim f ( x )
xÑ2
y = f (x)
2
x
´2 2
´2
Q[2]: Given the function shown below, evaluate lim f ( x ).

xÑ0
5
L IMITS 1.3 T HE L IMIT OF A F UNCTION
y y = f (x)
´1
Q[3]: Given the function shown below, evaluate:

(a) lim f ( x )
xÑ´1´
(b) lim f ( x )
xÑ´1+
(c) lim f ( x )
xÑ´1
(d) lim f ( x )
xÑ´2+
(e) lim f ( x )
xÑ2´
y
y = f (x)
2
x
´2 ´1 1 2
´2
Q[4]: Draw a curve y = f ( x ) with lim f ( x ) = f (3) = 10.

xÑ3
Q[5]: Draw a curve y = f ( x ) with lim f ( x ) = 10 and f (3) = 0.

xÑ3
Q[6]: Suppose lim f ( x ) = 10. True or false: f (3) = 10.

xÑ3
Q[7]: Suppose f (3) = 10. True or false: lim f ( x ) = 10.

xÑ3
6
L IMITS 1.4 C ALCULATING L IMITS WITH L IMIT L AWS
Q[8]: Suppose f ( x ) is a function defined on all real numbers, and lim f ( x ) = 16. What
xÑ´2
is lim f ( x )?
xÑ´2´
Q[9]: Suppose f ( x ) is a function defined on all real numbers, and lim f ( x ) = 16. What
xÑ´2´
is lim f ( x )?
xÑ´2
§§ Stage 2
In Questions 10 through 17, evaluate the given limits. If you aren’t sure where to begin,
it’s nice to start by drawing the function.
Q[10]: lim sin t

tÑ0
Q[11]: lim log x

xÑ0+
Q[12]: lim y2
yÑ3
1
Q[13]: lim
xÑ0´ x
1
Q[14]: lim
xÑ0 x
1
Q[15]: lim
xÑ0 x2
1
Q[16]: lim
xÑ3 10
"
sin x x ď 2.9
Q[17]: lim f ( x ), where f ( x ) = .
xÑ3 x2 x ą 2.9
1.4Ĳ Calculating Limits with Limit Laws

Exercises
7
§§ Stage 1
Q[1]: Suppose lim f ( x ) = 0 and lim g( x ) = 0. Which of the following limits can you
xÑa xÑa
compute, given this information?
f (x)
(a) lim
xÑa 2
2
(b) lim
xÑa f ( x )
f (x)
(c) lim
xÑa g ( x )
(d) lim f ( x ) g( x )
xÑa
Q[2]: Give two functions f ( x ) and g( x ) that satisfy lim f ( x ) = lim g( x ) = 0 and
xÑ3 xÑ3
f (x)
lim = 10.
xÑ3 g ( x )
Q[3]: Give two functions f ( x ) and g( x ) that satisfy lim f ( x ) = lim g( x ) = 0 and
xÑ3 xÑ3
f (x)
lim = 0.
xÑ3 g ( x )
f (x)
Q[4]: Give two functions f ( x ) and g( x ) that satisfy lim f ( x ) = lim g( x ) = 0 and lim =
xÑ3 xÑ3 xÑ3 g( x )
8.
f (x)
Q[5]: Suppose lim f ( x ) = lim g( x ) = 0. What are the possible values of lim ?
xÑa xÑa xÑa g( x )
§§ Stage 2
For Questions 6 through 41, evaluate the given limits.
2(t ´ 10)2
Q[6]: lim
tÑ10 t
(y + 1)(y + 2)(y + 3)
Q[7]: lim
yÑ0 cos y
4
4x ´ 2
Q[8]: lim
xÑ3 x+2

1´t
Q[9](˚): lim
tÑ´3 cos(t)
8
(2 + h )2 ´ 4
Q[10](˚): lim
hÑ0 2h

t´5
Q[11](˚): lim
tÑ´2 t+4
a
Q[12](˚): lim 5x3 + 4
xÑ1

t´2
Q[13](˚): lim
tÑ´1 t+3
log(1 + x ) ´ x
Q[14](˚): lim
xÑ1 x2

x´2
Q[15](˚): lim
xÑ2 x2 ´ 4
x2 ´ 4x
Q[16](˚): lim
xÑ4 x 2 ´ 16
x2 + x ´ 6
Q[17](˚): lim
xÑ2 x´2
x2 ´ 9
Q[18](˚): lim
xÑ´3 x + 3
1
Q[19]: lim t4 ´ 3t3 + t
tÑ2 2
?
x2 + 8 ´ 3
Q[20](˚): lim .
xÑ´1 x+1
? ?
x + 7 ´ 11 ´ x
Q[21](˚): lim .
xÑ2 2x ´ 4
? ?
x+2´ 4´x
Q[22](˚): lim
xÑ1 x´1
? ?
x´2´ 4´x
Q[23](˚): lim .
xÑ3 x´3
3t ´ 3
Q[24](˚): lim ? .
tÑ1 2 ´ 5 ´ t
9

2 3
Q[25]: lim ´x cos
xÑ0 x

1 1
x4 sin x + 5x2 cos x +2
Q[26]: lim
xÑ0 ( x ´ 2)2

2 1
Q[27](˚): lim x sin
xÑ0 x
2w2 ´ 50
Q[28]: lim
wÑ5 ( w ´ 5)( w ´ 1)
r
Q[29]: lim
rÑ´5 r 2 + 10r + 25
d
x3 + x2 + x + 1
Q[30]: lim
xÑ´1 3x + 3
x2 + 2x + 1
Q[31]: lim
xÑ0 3x 5 ´ 5x3
t2 x2 + 2tx + 1
Q[32]: lim , where x is a positive constant
tÑ7 t2 ´ 14t + 49
Q[33]: lim x5 ´ 32x + 15, where x is a constant

dÑ0
" 2 #
x2 ´ 3x + 2
Q[34]: lim ( x ´ 1)2 sin + 15
xÑ1 x2 ´ 2x + 1
Q[35](˚): Evaluate
lim x1/101 sin x´100

xÑ0
or explain why this limit does not exist.
x2 ´ 4
Q[36](˚): lim
xÑ2 x 2 ´ 2x
( x ´ 5)2
Q[37]: lim
xÑ5 x + 5
1 1
3t2
+ t2 ´1
Q[38]: Evaluate lim .
tÑ 12 2t ´ 1

|x|
Q[39]: Evaluate lim 3 + .
xÑ0 x
10
|3d + 12|
Q[40]: Evaluate lim
dÑ´4 d + 4
5x ´ 9
xÑ0 |x| + 2
x f (x) + 3
Q[42]: Suppose lim f ( x ) = ´1. Evaluate lim .
xÑ´1 xÑ´1 2 f ( x ) + 1
x2 + ax + 3
Q[43](˚): Find the value of the constant a for which lim exists.
xÑ´2 x2 + x ´ 2
Q[44]: Suppose f ( x ) = 2x and g( x ) = 1x . Evaluate the following limits.

(a) lim f ( x )
xÑ0
(b) lim g( x )
xÑ0
(c) lim f ( x ) g( x )
xÑ0
f (x)
(d) lim
xÑ0 g ( x )
(e) lim [ f ( x ) + g( x )]
xÑ2
f (x) + 1
(f) lim
xÑ0 g ( x + 1)
§§ Stage 3
1
Q[45]: The curve y = f ( x ) is shown in the graph below. Sketch the graph of y = .
f (x)
y
y = f (x)
1
x
1
Q[46]: The graphs of functions f ( x ) and g( x ) are shown in the graphs below. Use these
f (x)
to sketch the graph of .
g( x )
11
y y
y = f (x)
1 1 y = g( x )
x x
1 1
Q[47]: Suppose the position of a white ball, at time t, is given by s(t), and the position of a
red ball is given by 2s(t). Using the definition from Section 1.2 of the velocity of a particle,
and the limit laws from this section, answer the following question: if the white ball has
velocity 5 at time t = 1, what is the velocity of the red ball?
Q[48]: Let f ( x ) = 1x and g( x ) = ´1

x .
(a) Evaluate lim f ( x ) and lim g( x ).
xÑ0 xÑ0
(b) Evaluate lim [ f ( x ) + g( x )]
xÑ0
(c) Is it always true that lim [ f ( x ) + g( x )] = lim f ( x ) + lim g( x )?
xÑa xÑa xÑa
Q[49]: Suppose $ 2
& x +3 , x ą 0
f (x) = 0 , x=0
% 2
x ´3 , x ă 0
(a) Evaluate lim f ( x ).
xÑ0´
(b) Evaluate lim f ( x ).

xÑ0+
(c) Evaluate lim f ( x ).

xÑ0
Q[50]: Suppose $
’ x2 + 8x + 16
’
& 2 , x ą ´4
f (x) = x + 30x ´ 4
’
’
% 3
x + 8x2 + 16x , x ď ´4
(a) Evaluate lim f ( x ).
xÑ´4´
(b) Evaluate lim f ( x ).

xÑ´4+
(c) Evaluate lim f ( x ).

xÑ´4
12
L IMITS 1.5 L IMITS AT I NFINITY
1.5Ĳ Limits at Infinity

Exercises
§§ Stage 1
Q[1]: Give a polynomial f ( x ) with the property that both lim f ( x ) and lim f ( x ) are
xÑ8 xÑ´8
(finite) real numbers.
Q[2]: Give a polynomial f ( x ) that satisfies lim f ( x ) ‰ lim f ( x ).

xÑ8 xÑ´8
§§ Stage 2
Q[3]: Evaluate lim 2´x
xÑ8
Q[4]: Evaluate lim 2x

xÑ8
Q[5]: Evaluate lim 2x

xÑ´8
Q[6]: Evaluate lim cos x

xÑ´8
Q[7]: Evaluate lim x ´ 3x5 + 100x2 .

xÑ8
?
3x8 + 7x4 + 10
xÑ8 x4 ´ 2x2 + 1
h? ? i
Q[9](˚): lim x2 + 5x ´ x2 ´x
xÑ8
3x
Q[10](˚): Evaluate lim ? .
xÑ´8 4x2 + x ´ 2x
1 ´ x ´ x2
Q[11](˚): Evaluate lim .
xÑ´8 2x2 ´ 7
?
Q[12](˚): Evaluate lim x2 + x ´ x
xÑ8
13
L IMITS 1.5 L IMITS AT I NFINITY
5x2 ´ 3x + 1
xÑ+8 3x2 + x + 7
?
4x+2
xÑ+8 3 x + 4
4x3 + x
xÑ+8 7x3 + x 2 ´ 2
? ?
x + x ´ 4 x4 + 5
3 2
Q[16]: Evaluate lim
xÑ´8 x+1
5x2 + 10
xÑ+8 3x 3 + 2x2 + x
x+1
Q[18]: Evaluate lim ? .
xÑ´8 x2
x+1
Q[19]: Evaluate lim ?
xÑ8 x2

π |x| 1
Q[20](˚): Find the limit lim sin + .
xÑ´8 2 x x
3x + 5
Q[21](˚): Evaluate lim ? .
xÑ´8 x2 + 5 ´ x
5x + 7
Q[22](˚): Evaluate lim ?
xÑ´8 4x2 + 15 ´ x
3x7 + x5 ´ 15
xÑ´8 4x2 + 32x
a
Q[24](˚): Evaluate lim n2 + 5n ´ n .
nÑ8
a2 ´ 1a
aÑ0+ a´1
2x + 8
xÑ3 1 + x21´9
x´3
§§ Stage 3
Q[27]: Give a rational function f ( x ) with the properties that lim f ( x ) ‰ lim f ( x ), and
xÑ8 xÑ´8
both limits are (finite) real numbers.
14
L IMITS 1.6 C ONTINUITY
Q[28]: Suppose the concentration of a substance in your body t hours after injection is
given by some formula c(t), and lim c(t) ‰ 0. What kind of substance might have been
tÑ8
injected?
1.6Ĳ Continuity
Exercises
§§ Stage 1
Q[1]: Give an example of a function (you can write a formula, or sketch a graph) that has
infinitely many infinite discontinuities.
Q[2]: When I was born, I was less than one meter tall. Now, I am more than one meter
tall. What is the conclusion of the Intermediate Value Theorem about my height?
Q[3]: Give an example (by sketch or formula) of a function f ( x ), defined on the interval
[0, 2], with f (0) = 0, f (2) = 2, and f ( x ) never equal to 1. Why does this not contradict
the Intermediate Value Theorem?
Q[4]: Is the following a valid statement?

Suppose f is a continuous function over [10, 20], f (10) = 13, and
f (20) = ´13. Then f has a zero between x = 10 and x = 20.

Suppose f is a continuous function over [10, 20], f (10) = 13, and
f (20) = ´13. Then f (15) = 0.

Suppose f is a function over [10, 20], f (10) = 13, and f (20) = ´13, and f
takes on every value between ´13 and 13. Then f is continuous.
Q[7]: Suppose f (t) is continuous at t = 5. True or false: t = 5 is in the domain of f (t).

Q[8]: Suppose lim f (t) = 17, and suppose f (t) is continuous at t = 5. True or false:
tÑ5
f (5) = 17.
Q[9]: Suppose lim f (t) = 17. True or false: f (5) = 17.
tÑ5
x f (x)
Q[10]: Suppose f ( x ) and g( x ) are continuous at x = 0, and let h( x ) = . What is
g2 ( x ) + 1
lim h( x )?
xÑ0+
15
§§ Stage 2
Q[11]: Find a constant k so that the function

#
x sin 1x when x ‰ 0
a( x ) =
k when x = 0
is continuous at x = 0.
Q[12]: Use the Intermediate Value Theorem to show that the function f ( x ) = x3 + x2 +
x + 1 takes on the value 12345 at least once in its domain.
1
Q[13](˚): Describe all points for which the function is continuous: f ( x ) = 2 .
x ´1
1
Q[14](˚): Describe all points for which this function is continuous: f ( x ) = ? .
x2 ´1
1
Q[15](˚): Describe all points for which this function is continuous: a .
1 + cos( x )
1
Q[16](˚): Describe all points for which this function is continuous: f ( x ) = .
sin x
Q[17](˚): Find all values of c such that the following function is continuous at x = c:
"
8 ´ cx if x ď c
f (x) =
x2 if x ą c
Use the definition of continuity to justify your answer.
Q[18](˚): Find all values of c such that the following function is continuous everywhere:
#
x2 + c x ě 0
f (x) =
cos cx x ă 0
Q[19](˚): Find all values of c such that the following function is continuous:
#
x2 ´ 4 if x ă c
f (x) =
3x if x ě c .
Q[20](˚): Find all values of c such that the following function is continuous:
"
6 ´ cx if x ď 2c
f (x) =
x2 if x ą 2c
16
§§ Stage 3
Q[21]: Show that there exists at least one real number x satisfying sin x = x ´ 1
Q[22](˚): Show that there exists at least one real number c such that 3c = c2 .
Q[23](˚): Show that there exists at least one real number c such that 2 tan(c) = c + 1.
a
Q[24](˚): Show that there exists at least one real number c such that cos(πc) = sin(2πc) +
1
2.
1 3
Q[25](˚): Show that there exists at least one real number c such that 2
= c+ .
(cos πc) 2
Q[26]: Use the intermediate value theorem to find an interval of length one containing a
root of f ( x ) = x7 ´ 15x6 + 9x2 ´ 18x + 15.
?
Q[27]: Use the intermediate value theorem to give a decimal approximation of 3 7 that is
correct to at least two decimal places. You may use a calculator, but only to add, subtract,
multiply, and divide.
Q[28]: Suppose f ( x ) and g( x ) are functions that are continuous over the interval [ a, b],
with f ( a) ď g( a) and g(b) ď f (b). Show that there exists some c P [ a, b] with f (c) = g(c).
17
18
Chapter 2
D ERIVATIVES
2.1Ĳ Revisiting tangent lines
Exercises
§§ Stage 1
Q[1]: Shown below is the graph y = f ( x ). If we choose a point Q on the graph to the left
of the y-axis, is the slope of the secant line through P and Q positive or negative? If we
choose a point Q on the graph to the right of the y-axis, is the slope of the secant line
through P and Q positive or negative?
y = f (x)
19
D ERIVATIVES 2.1 R EVISITING TANGENT LINES
Q[2]: Shown below is the graph y = f ( x ).

(a) If we want the slope of the secant line through P and Q to increase, should we slide Q
closer to P, or further away?
(b) Which is larger, the slope of the tangent line at P, or the slope of the secant line
through P and Q?
y
P
Q[3]: Group the functions below into collections whose secant lines from x = ´2 to x = 2
all have the same slopes.
y y y
x x x
´2 2 ´2 2 ´2 2
(a) (b) (c)
y y y
x x x
´2 2 ´2 2 ´2 2
(d) (e) (f)
§§ Stage 2
Q[4]: Give your best approximation of the slope of the tangent line to the graph below at
the point x = 5.
20
D ERIVATIVES 2.1 R EVISITING TANGENT LINES
x
1 5
Q[5]: On the graph below, sketch the tangent line to y = f ( x ) at P. Then, find two points
Q and R on the graph so that the secant line through Q and R has the same slope as the
tangent line at P.
y
y = f (x)
21
D ERIVATIVES 2.2 D EFINITION OF THE DERIVATIVE
Q[6]: Mark the points where the curve shown below has a tangent line with slope 0.
y
y = f (x)
(Later on, we’ll learn how these points tell us a lot about the shape of a graph.)
2.2Ĳ Definition of the derivative

Exercises
§§ Stage 1
df
Q[1]: The function f ( x ) is shown. Select all options below that describe its derivative, :
dx
(a) constant (b) increasing (c) decreasing
(d) always positive (e) always negative
y
y = f (x)
df
dx
22
x
y = f (x)
df
dx
y
y = f (x)
Q[4](˚): State, in terms of a limit, what it means for f ( x ) = x3 to be differentiable at x = 0.
23
Q[5]: For which values of x does f 1 ( x ) not exist?

y
x
1
Q[6]: Suppose f ( x ) is a function defined at x = a with
f ( a + h) ´ f ( a) f ( a + h) ´ f ( a)
lim = lim = 1.
hÑ0´ h hÑ0 + h
True or false: f 1 ( a) = 1.
Q[7]: Suppose f ( x ) is a function defined at x = a with
lim f 1 ( x ) = lim f 1 ( x ) = 1.
xÑa´ xÑa+
True or false: f 1 ( a) = 1.
Q[8]: Suppose s(t) is a function, with t measured in seconds, and s measured in metres.
What are the units of s1 (t)?
§§ Stage 2
Q[9]: Use the definition of the derivative to find the equation of the tangent line to the
curve y( x ) = x3 + 5 at the point (1, 6).
Q[10]: Use the definition of the derivative to find the derivative of f ( x ) = 1x .
Q[11](˚): Let f ( x ) = x|x|. Using the definition of the derivative, show that f ( x ) is differ-
entiable at x = 0.
Q[12](˚): Use the definition of the derivative to compute the derivative of the function
2
f ( x ) = x+ 1.
24
Q[13](˚): Use the definition of the derivative to compute the derivative of the function
f ( x ) = x21+3 .
Q[14]: Use the definition of the derivative to find the slope of the tangent line to the curve
f ( x ) = x log10 (2x + 10) at the point x = 0.
1
Q[15](˚): Compute the derivative of f ( x ) = x2
directly from the definition.
Q[16](˚): Find the values of the constants a and b for which

" 2
x xď2
f (x) =
ax + b x ą 2
is differentiable everywhere.
Remark: In the text, you have already learned the derivatives of x2 and ax + b. In this
question, you are only asked to find the values of a and b—not to justify how you got
them—so you don’t have to use the definition of the derivative. However, on an exam,
you might be asked to justify your answer, in which case you would show how to
differentiate the two branches of f ( x ) using the definition of a derivative.
?
Q[17](˚): Use the definition of the derivative to compute f 1 ( x ) if f ( x ) = 1 + x. Where
does f 1 ( x ) exist?
§§ Stage 3
Q[18]: Use the definition of the derivative to find the velocity of an object whose position
is given by the function s(t) = t4 ´ t2 .
Q[19](˚): Determine whether the derivative of following function exists at x = 0.

#
x cos x if x ě 0
f (x) = ? 2 4
x +x if x ă 0
You must justify your answer using the definition of a derivative.
Q[20](˚): Determine whether the derivative of the following function exists at x = 0

#
x cos x if x ď 0
f (x) = ?
1 + x ´ 1 if x ą 0

# 3
x ´ 7x2 if x ď 0
f (x) =
x3 cos 1x if x ą 0
25
D ERIVATIVES 2.3 I NTERPRETATIONS OF THE DERIVATIVE

$
&4x2 ´ 8x + 4 if x ď 1
f (x) = 1
%( x ´ 1)2 sin if x ą 1
x´1
Q[23]: Sketch a function f ( x ) with f 1 (0) = ´1 that takes the following values:
x ´1 ´ 12 ´ 14 ´ 81 0 1
8
1
4
1
2 1
f(x) ´1 ´ 12 ´ 14 ´ 81 0 1
8
1
4
1
2 1
Remark: you can’t always guess the behaviour of a function from its points, even if the
points seem to be making a clear pattern.
Q[24]: Let p( x ) = f ( x ) + g( x ), for some functions f and g whose derivatives exist. Use
limit laws and the definition of a derivative to show that p1 ( x ) = f 1 ( x ) + g1 ( x ).
Remark: this is called the sum rule, and we’ll learn more about it in Lemma 2.4.1.
Q[25]: Let f ( x ) = 2x, g( x ) = x, and p( x ) = f ( x ) ¨ g( x ).
(a) Find f 1 ( x ) and g1 ( x ).
(b) Find p1 ( x ).
(c) Is p1 ( x ) = f 1 ( x ) ¨ g1 ( x )?
In Theorem 2.4.3, you’ll learn a rule for calculating the derivative of a product of two
functions.
Q[26](˚): There are two distinct straight lines that pass through the point (1, ´3) and are
tangent to the curve y = x2 . Find equations for these two lines.
Remark: the point (1, ´3) does not lie on the curve y = x2 .
Q[27](˚): For which values of a is the function

"
0 xď0
f (x) = a 1
x sin x x ą 0
differentiable at 0?
2.3Ĳ Interpretations of the derivative

Exercises
26
D ERIVATIVES 2.4 A RITHMETIC OF DERIVATIVES - A DIFFERENTIATION TOOLBOX
§§ Stage 2
Q[1]: Suppose h(t) gives the height at time t of the water at a dam, where the units of t
are hours and the units of h are meters.
(a) What is the physical interpretation of the slope of the secant line through the points
(0, h(0)) and (24, h(24))?
(b) What is the physical interpretation of the slope of the tangent line to the curve
y = h(t) at the point (0, h(0))?
Q[2]: Suppose p(t) is a function that gives the profit generated by selling t widgets. What
is the practical interpretation of p1 (t)?
Q[3]: T (d) gives the temperature of water at a particular location d metres below the
surface. What is the physical interpretation of T 1 (d)? Would you expect the magnitude of
T 1 (d) to be larger when d is near 0, or when d is very large?
Q[4]: C (w) gives the calories in w grams of a particular dish. What does C1 (w) describe?
Q[5]: The velocity of a moving object at time t is given by v(t). What is v1 (t)?
Q[6]: The function T ( j) gives the temperature in degrees Celsius of a cup of water after j
joules of heat have been added. What is T 1 ( j)?
Q[7]: A population of bacteria, left for a fixed amount of time at temperature T, grows to
P( T ) individuals. Interpret P1 ( T ).
§§ Stage 3
Q[8]: You hammer a small nail into a wooden wagon wheel. R(t) gives the number of
rotations the nail has undergone t seconds after the wagon started to roll. Give an equation
for how quickly the nail is rotating, measured in degrees per second.
Q[9]: A population of bacteria, left for a fixed amount of time at temperature T, grows to
P( T ) individuals. There is one ideal temperature where the bacteria population grows
largest, and the closer the sample is to that temperature, the larger the population is
(unless the temperature is so extreme that it causes all the bacteria to die by freezing or
boiling). How will P1 ( T ) tell you whether you are colder or hotter than the ideal
temperature?
2.4Ĳ Arithmetic of derivatives - a differentiation toolbox

Exercises
27
D ERIVATIVES 2.4 A RITHMETIC OF DERIVATIVES - A DIFFERENTIATION TOOLBOX
§§ Stage 1
d
Q[1]: True or false: t f ( x ) + g( x )u = f 1 ( x ) + g1 ( x ) when f and g are differentiable
dx
functions.
d
Q[2]: True or false: t f ( x ) g( x )u = f 1 ( x ) g1 ( x ) when f and g are differentiable functions.
dx
" *
d f (x) f 1 (x) f ( x ) g1 ( x )
Q[3]: True or false: = ´ when f and g are differentiable
dx g( x ) g( x ) g2 ( x )
functions.
Q[4]: Let f be a differentiable function. Use at least three different rules to differentiate
g( x ) = 3 f ( x ), and verify that they all give the same answer.
§§ Stage 2
Q[5]: Differentiate f ( x ) = 3x2 + 4x1/2 for x ą 0.
?
Q[6]: Use the product rule to differentiate f ( x ) = (2x + 5)(8 x ´ 9x ).
1
Q[7](˚): Find the equation of the tangent line to the graph of y = x3 at x = .
2
Q[8](˚): A particle moves along the x–axis so that its position at time t is given by
x = t3 ´ 4t2 + 1 .
(a) At t = 2, what is the particle’s speed?
(b) At t = 2, in what direction is the particle moving?
(c) At t = 2, is the particle’s speed increasing or decreasing?
2x ´ 1
Q[9](˚): Calculate and simplify the derivative of
2x + 1
2
3x + 1
Q[10]: What is the slope of the graph y = when x = 1?
3x ´ 2
1
Q[11]: Find the equation of the tangent line to the curve f ( x ) = ? at the point
x+1
1, 12 .
28
D ERIVATIVES 2.5 P ROOFS OF THE ARITHMETIC OF DERIVATIVES
§§ Stage 3
Q[12]: A town is founded in the year 2000. After t years, it has had b(t) births and d(t)
deaths. Nobody enters or leaves the town except by birth or death (whoa). Give an
expression for the rate the population of the town is growing.
Q[13](˚): Find all points on the curve y = 3x2 where the tangent line passes through (2, 9).
a ? !
100180 + y ´ 100180
Q[14](˚): Evaluate lim by interpreting the limit as a
yÑ0 y
derivative.
Q[15]: A rectangle is growing. At time t = 0, it is a square with side length 1 metre. Its
width increases at a constant rate of 2 metres per second, and its length increases at a
constant rate of 5 metres per second. How fast is its area increasing at time t ą 0?
Q[16]: Let f ( x ) = x2 g( x ) for some differentiable function g( x ). What is f 1 (0)?
g( x )
Q[17]: Verify that differentiating f ( x ) = using the quotient rule gives the same
h( x )
g( x ) k( x )
answer as differentiating f ( x ) = ¨ using the product rule and the quotient rule.
k( x ) h( x )
2.5Ĳ Proofs of the arithmetic of derivatives
No exercises for Section 2.5

Jump to TABLE OF CONTENTS.
2.6Ĳ Using the arithmetic of derivatives - examples
Exercises
29
D ERIVATIVES 2.6 U SING THE ARITHMETIC OF DERIVATIVES - EXAMPLES
§§ Stage 1
Q[1]: Spot and correct the error(s) in the following calculation.
2x
f (x) =
x+1
2( x + 1) + 2x
f 1 (x) =
( x + 1)2
2( x + 1)
=
( x + 1)2
2
=
x+1
d x
Q[2]: True or false: t2 u = x2x´1 .
dx
§§ Stage 2
Q[3]: Differentiate f ( x ) = 23 x6 + 5x4 + 12x2 + 9 and factor the result.
Q[4]: Differentiate s(t) = 3t4 + 5t3 ´ 1t .

Q[5]: Differentiate x (y) = 2y + 1y ¨ y3 .
?
x+1
Q[6]: Differentiate T ( x ) = .
x2 +3

7x + 2
Q[7](˚): Compute the derivative of .
x2 + 3
Q[8]: What is f 1 (0), when f ( x ) = (3x3 + 4x2 + x + 1)(2x + 5)?
3x3 + 1
Q[9]: Differentiate f ( x ) = .
x2 + 5x
3x2 + 5

Q[10](˚): Compute the derivative of
2´x
2 ´ x2

3x2 + 5
3
2x + 1
x+2
30
D ERIVATIVES 2.7 D ERIVATIVES OF E XPONENTIAL F UNCTIONS
?
x
Q[13](˚): For what values of x does the derivative of exist? Explain your answer.
1 ´ x2
? ?
Q[14]: Differentiate f ( x ) = (3 5 x + 15 3 x + 8) 3x2 + 8x ´ 5 .

? ?
( x2 + 5x + 1)( x + 3 x )
x
1
Q[16]: Find all x-values where f ( x ) = 1 5
has a horizontal tangent line.
5x + x4 ´ 53 x3
§§ Stage 3
Q[17](˚): Find an equation of a line that is tangent to both of the curves y = x2 and
y = x2 ´ 2x + 2 (at different points).
Q[18](˚): Find all lines that are tangent to both of the curves y = x2 and y = ´x2 + 2x ´ 5.
Illustrate your answer with a sketch.
x2015 ´ 22015

xÑ2 x´2
2.7Ĳ Derivatives of Exponential Functions
Exercises
31
D ERIVATIVES 2.7 D ERIVATIVES OF E XPONENTIAL F UNCTIONS
§§ Stage 1
Q[1]: Match the curves in the graph to the following functions:

x
1
( a) y = (b) y = 1x (c) y = 2x (d) y = 2´x (e) y = 3x
2
A y B C
Q[2]: The graph below shows an exponential function f ( x ) = a x and its derivative f 1 ( x ).
Choose all the options that describe the constant a.
( a) a ă 0 (b) a ą 0 (c) a ă 1 (d) a ą 1 (e) a ă e (f) a ą e
y y = f (x)
y = f 1 (x)
d x
Q[3]: True or false: te u = xe x´1
dx
Q[4]: A population of bacteria is described by P(t) = 100e0.2t , for 0 ď t ď 10. Over this
time period, is the population increasing or decreasing?
32
D ERIVATIVES 2.8 D ERIVATIVES OF TRIGONOMETRIC FUNCTIONS
We will learn more about the uses of exponential functions to describe real-world phe-
nomena in Section 3.3.
§§ Stage 2
ex
Q[5]: Find the derivative of f ( x ) = .
2x
Q[6]: Differentiate f ( x ) = e2x .
Q[7]: Differentiate f ( x ) = e a+ x , where a is a constant.
Q[8]: For which values of x is the function f ( x ) = xe x increasing?
Q[9]: Differentiate e´x .
Q[10]: Differentiate f ( x ) = (e x + 1)(e x ´ 1).
Q[11]: A particle’s position is given by
s ( t ) = t2 e t .
When is the particle moving in the negative direction?
§§ Stage 3
Q[12]: Let g( x ) = f ( x )e x , for a differentiable function f ( x ). Give a simplified formula for

g1 ( x ).
Functions of the form g( x ) are relatively common. If you remember this formula, you
can save yourself some time when you need to differentiate them. We will explore this
more in Question 19, Section 2.14.
Q[13]: Which of the following functions describe a straight line?
( a) y = e3 log x + 1 (b) 2y + 5 = e3+log x (c) y = e2x + 4 (d) y = elog x 3e + log 2
Q[14](˚): Find constants a, b so that the following function is differentiable:

" 2
ax + b x ď 1
f (x) =
ex xą1
2.8Ĳ Derivatives of trigonometric functions

Exercises
33
§§ Stage 1
Q[1]: Graph sine and cosine on the same axes, from x = ´2π to x = 2π. Mark the points
where sin x has a horizontal tangent. What do these points correspond to, on the graph
of cosine?
Q[2]: Graph sine and cosine on the same axes, from x = ´2π to x = 2π. Mark the points
where sin x has a tangent line of maximum (positive) slope. What do these points
correspond to, on the graph of cosine?
§§ Stage 2
Q[3]: Differentiate f ( x ) = sin x + cos x + tan x.
Q[4]: For which values of x does the function f ( x ) = sin x + cos x have a horizontal
tangent?
Q[5]: Differentiate f ( x ) = sin2 x + cos2 x.
Q[6]: Differentiate f ( x ) = 2 sin x cos x.

Q[7]: Differentiate f ( x ) = e x cot x.
2 sin x + 3 tan x
Q[8]: Differentiate f ( x ) =
cos x + tan x
5 sec x + 1
ex
Q[10]: Differentiate f ( x ) = (e x + cot x )(5x6 ´ csc x ).

Q[11]: Differentiate f (θ ) = sin π2 ´ θ .
Q[12]: Differentiate f ( x ) = sin(´x ) + cos(´x ).
cos θ + sin θ
Q[13]: Differentiate s(θ ) = .
cos θ ´ sin θ
Q[14](˚): Find the values of the constants a and b for which
"
cos( x ) x ď 0
f (x) =
ax + b x ą 0
is differentiable everywhere.
34
π
Q[15](˚): Find the equation of the line tangent to the graph of y = cos( x ) + 2x at x = .
2
§§ Stage 3

cos( x ) ´ cos(2015)
xÑ2015 x ´ 2015

cos( x ) ´ 1/2
xÑπ/3 x ´ π/3

sin( x )
xÑπ x´π
Q[19]: Show how you can use the quotient rule to find the derivative of tangent, if you
already know the derivatives of sine and cosine.
Q[20](˚): The derivative of the function
"
ax + b for x ă 0
f (x) = 6 cos x
2+sin x +cos x for x ě 0
exists for all x. Determine the values of the constants a and b.
Q[21](˚): For which values of x does the derivative of f ( x ) = tan x exist?
10 sin( x )
Q[22](˚): For what values of x does the derivative of 2 exist? Explain your
x +x´6
answer.
x2 + 6x + 5
Q[23](˚): For what values of x does the derivative of exist? Explain your
sin( x )
answer.
π
Q[24](˚): Find the equation of the line tangent to the graph of y = tan( x ) at x = .
4
Q[25](˚): Find the equation of the line tangent to the graph of y = sin( x ) + cos( x ) + e x at
x = 0.
Q[26]: For which values of x does the function f ( x ) = e x sin x have a horizontal tangent
line?
Q[27]: Let
sin x
"
x , x‰0
f (x) =
1 , x=0
Find f 1 (0), or show that it does not exist.
§§ Application of Understanding
Q[28](˚): Differentiate the function
h( x ) = sin(|x|)
35
D ERIVATIVES 2.9 O NE MORE TOOL - THE CHAIN RULE
and give the domain where the derivative exists.
Q[29](˚): For the function #

0 xď0
f (x) = ?( x )
sin
x
xą0
which of the following statements is correct?
i. f is undefined at x = 0.
ii. f is neither continuous nor differentiable at x = 0.
iii. f is continuous but not differentiable at x = 0.
iv. f is differentiable but not continuous at x = 0.
v. f is both continuous and differentiable at x = 0.
99
sin x27 + 2x5 e x
xÑ0 sin5 x
2.9Ĳ One more tool - the chain rule

Exercises
§§ Stage 1
Q[1]: Suppose the amount of kelp in a harbour depends on the number of urchins.
Urchins eat kelp: when there are more urchins, there is less kelp, and when there are
fewer urchins, there is more kelp. Suppose further that the number of urchins in the
harbour depends on the number of otters, who find urchins extremely tasty: the more
otters there are, the fewer urchins there are.
Let O, U, and K be the populations of otters, urchins, and kelp, respectively.
dK
(a) Is dU positive or negative?
dU
(b) Is dO positive or negative?
dK
(c) Is dO positive or negative?
Remark: An urchin barren is an area where unchecked sea urchin grazing has decimated
the kelp population, which in turn causes the other species that shelter in the kelp forests
to leave. Introducing otters to urchin barrens is one intervention to increase biodiversity.
A short video with a more complex view of otters and urchins in Canadian waters is
available on YouTube: https://youtu.be/ASJ82wyHisE
Q[2]: Suppose A, B, C, D, and E are functions describing an interrelated system, with

the following signs: dA dB dC dD dA
dB ą 0, dC ą 0, dD ă 0, and dE ą 0. Is dE positive or negative?
36
§§ Stage 2
Q[3]: Evaluate the derivative of f ( x ) = cos(5x + 3).
5
Q[4]: Evaluate the derivative of f ( x ) = x2 + 2 .
17
Q[5]: Evaluate the derivative of T (k) = 4k4 + 2k2 + 1 .
d
x2 + 1
Q[6]: Evaluate the derivative of f ( x ) = .
x2 ´ 1
2
Q[7]: Evaluate the derivative of f ( x ) = ecos( x ) .

Q[8](˚): Evaluate f 1 (2) if f ( x ) = g x/h( x ) , h(2) = 2, h1 (2) = 3, g1 (1) = 4.
Q[9](˚): Find the derivative of e x cos( x) .

2
Q[10](˚): Evaluate f 1 ( x ) if f ( x ) = e x +cos x .
c
1 x´1
Q[11](˚): Evaluate f ( x ) if f ( x ) = .
x+2

1 a
f (x) = + x2 ´ 1
x2
and give the domain where the derivative exists.
sin 5x
Q[13](˚): Evaluate the derivative of f ( x ) =
1 + x2
Q[14]: Evaluate the derivative of f ( x ) = sec(e2x+7 ).
π
Q[15]: Find the tangent line to the curve y = tan2 x + 1 cos2 x at the point x = .

4
3 ´7t2 +8t
Q[16]: The position of a particle at time t is given by s(t) = et . For which values of
t is the velocity of the particle zero?
2
Q[17]: What is the slope of the tangent line to the curve y = tan e x at the point x = 1?
Q[18](˚): Differentiate y = e4x tan x. You do not need to simplify your answer.
37
x3
Q[19](˚): Evaluate the derivative of the following function at x = 1: f ( x ) = .
1 + e3x
2 (x)
Q[20](˚): Differentiate esin .
Q[21](˚): Compute the derivative of y = sin e5x

2
Q[22](˚): Find the derivative of ecos( x ) .
?
Q[23](˚): Compute the derivative of y = cos x2 +

x2 + 1
Q[24](˚): Evaluate the derivative.
y = (1 + x2 ) cos2 x
Q[25](˚): Evaluate the derivative.

e3x
y=
1 + x2
Q[26](˚): Find g1 (2) if g( x ) = x3 h( x2 ), where h(4) = 2 and h1 (4) = ´2.
2 ´1) /2
Q[27](˚): At what points ( x, y) does the curve y = xe´( x have a horizontal tangent?
Q[28]: A particle starts moving at time t = 1, and its position thereafter is given by

1
s(t) = sin .
t
When is the particle moving in the negative direction?
ex
Q[29]: Compute the derivative of f ( x ) = .
cos3 (5x ´ 7)
d ! 2x )
Q[30](˚): Evaluate xe cos 4x .
dx
§§ Stage 3
Q[31]: A particle moves along the Cartesian plane from time t = ´π/2 to time t = π/2.
The x-coordinate of the particle at time t is given by x = cos t, and the y-coordinate is
given by y = sin t, so the particle traces a curve in the plane. When does the tangent line
to that curve have slope ´1?
2
Q[32](˚): Show that, for all x ą 0, e x+ x ą 1 + x.
Q[33]: We know that sin(2x ) = 2 sin x cos x. What other trig identity can you derive from
this, using differentiation?
38
D ERIVATIVES 2.10 T HE NATURAL LOGARITHM
d
2
3 ecsc x
Q[34]: Evaluate the derivative of f ( x ) = ? . You do not have to simplify
x3 ´ 9 tan x
your answer.
Q[35]: Suppose a particle is moving in the Cartesian plane over time. For any real
number t ě 0, the coordinate of the particle at time t is given by (sin t, cos2 t).
(a) Sketch a graph of the curve traced by the particle in the plane by plotting points, and
describe how the particle moves along it over time.
10π
(b) What is the slope of the curve traced by the particle at time t = ?
3
2.10Ĳ The natural logarithm

Exercises
Jump to HINTS, ANSWERS, SOLUTIONS or TABLE OF CONTENTS. Reminder: in these notes,
we use log x to mean loge x, which is also commonly written elsewhere as ln x.
§§ Stage 1
Q[1]: The volume in decibels (dB) of a sound is given by the formula:

P
V ( P) = 10 log10
S
where P is the intensity of the sound and S is the intensity of a standard baseline sound.
(That is: S is some constant.)
How much noise will ten speakers make, if each speaker produces 3dB of noise? What
about one hundred speakers?
Q[2]: An investment of $1000 with an interest rate of 5% per year grows to
A(t) = 1000et/20
dollars after t years. When will the investment double?
Q[3]: Which of the following expressions, if any, is equivalent to log cos2 x ?

(a) 2 log(cos x ) (b) 2 log | cos x| (c) log2 (cos x ) (d) log(cos x2 ))
§§ Stage 2
Q[4]: Differentiate f ( x ) = log(10x ).
39
D ERIVATIVES 2.10 T HE NATURAL LOGARITHM
Q[5]: Differentiate f ( x ) = log( x2 ).
Q[6]: Differentiate f ( x ) = log( x2 + x ).
Q[7]: Differentiate f ( x ) = log10 x.
log x
Q[8](˚): Find the derivative of y = .
x3
d
Q[9]: Evaluate log(sec θ ).
dθ
Q[10]: Differentiate the function f ( x ) = ecos(log x) .

Q[11](˚): Evaluate the derivative. You do not need to simplify your answer.
a
y = log( x2 + x4 + 1)
a
Q[12](˚): Differentiate ´ log(cos x ).
?
Q[13](˚): Calculate and simplify the derivative of log x + x2 + 4 .
2 ?
Q[14](˚): Evaluate the derivative of g( x ) = log(e x + 1 + x4 ).
2x ´ 1
Q[15](˚): Evaluate the derivative of the following function at x = 1: g( x ) = log .
2x + 1
d 
( x2 + 5)3
Q[16]: Evaluate the derivative of the function f ( x ) = log  .
x4 + 10

Q[17]: Evaluate f 1 (2) if f ( x ) = log g xh( x ) , h(2) = 2, h1 (2) = 3, g(4) = 3, g1 (4) = 5.
g( x ) = π x + xπ .
Q[19]: Differentiate f ( x ) = x x .
Q[20](˚): Find f 1 ( x ) if f ( x ) = x x + log10 x.

c
4 4 2
4 ( x + 12)( x ´ x + 2)
x3
40
D ERIVATIVES 2.11 I MPLICIT D IFFERENTIATION
Q[22]: Differentiate f ( x ) = ( x + 1)( x2 + 1)2 ( x3 + 1)3 ( x4 + 1)4 ( x5 + 1)5 .
5x2 + 10x + 15

1
3x4 + 4x3 + 5 10( x + 1)
Q[24](˚): Let f ( x ) = (cos x )sin x , with domain 0 ă x ă π

2. Find f 1 ( x ).
Q[25](˚): Find the derivative of (tan( x )) x , when x is in the interval (0, π/2).
2 +1)
Q[26](˚): Find f 1 ( x ) if f ( x ) = ( x2 + 1)( x
Q[27](˚): Differentiate f ( x ) = ( x2 + 1)sin( x) .
3 (x)
Q[28](˚): Let f ( x ) = xcos , with domain (0, 8). Find f 1 ( x ).
2 ´3
Q[29](˚): Differentiate f ( x ) = (3 + sin( x )) x .
§§ Stage 3
d ! )
Q[30]: Let f ( x ) and g( x ) be differentiable functions, with f ( x ) ą 0. Evaluate [ f ( x )] g(x) .
dx
Q[31]: Let f ( x ) be a function whose range includes only positive numbers. Show that the
curves y = f ( x ) and y = log( f ( x )) have horizontal tangent lines at the same values of x.
2.11Ĳ Implicit Differentiation

Exercises
§§ Stage 1
Q[1]: If we implicitly differentiate x2 + y2 = 1, we get the equation 2x + 2yy1 = 0. In the
step where we differentiate y2 to obtain 2yy1 , which rule(s) below are we using?
(a) power rule (b) chain rule (c) quotient rule

(d) derivatives of exponential functions
41
D ERIVATIVES 2.11 I MPLICIT D IFFERENTIATION
dy
Q[2]: Using the picture below, estimate at the three points where the curve crosses the
dx
y-axis.
y
x
´4 ´3 ´2 ´1 1 2 3 4
´1
´2
´3
´4
Remark: for this curve, one value of x may correspond to multiple values of y. So, we
cannot express this curve as y = f ( x ) for any function x. This is one typical situation
where we might use implicit differentiation.
Q[3]: Consider the unit circle, formed by all points ( x, y) that satisfy x2 + y2 = 1.
y
(a) Is there a function f ( x ) so that y = f ( x ) completely describes the unit circle? That is,
so that the points ( x, y) that make the equation y = f ( x ) true are exactly the same
points that make the equation x2 + y2 = 1 true?
(b) Is there a function f 1 ( x ) so that y = f 1 ( x ) completely describes the slope of the unit
circle? That is, so that for every point ( x, y) on the unit circle, the slope of the tangent
line to the circle at that point is given by f 1 ( x )?
42
D ERIVATIVES 2.12 I NVERSE T RIGONOMETRIC F UNCTIONS
dy
(c) Use implicit differentiation to find an expression for . Simplify until the
dx
expression is a function in terms of x only (not y), or explain why this is impossible.
§§ Stage 2
dy
Q[4](˚): Find if xy + e x + ey = 1.
dx
dy
Q[5](˚): If ey = xy2 + x, compute .
dx
Q[6](˚): If x2 tan(πy/4) + 2x log(y) = 16, then find y1 at the points where y = 1.
dy
Q[7](˚): If x3 + y4 = cos( x2 + y) compute dx .
Q[8](˚): If x2 ey + 4x cos(y) = 5, then find y1 at the points where y = 0.

dy
Q[9](˚): If x2 + y2 = sin( x + y) compute dx .
Q[10](˚): If x2 cos(y) + 2xey = 8, then find y1 at the points where y = 0.
Q[11]: At what points on the ellipse x2 + 3y2 = 1 is the tangent line parallel to the line
y = x?
?
Q[12](˚): For the curve defined by the equation xy = x2 y ´ 2, find the slope of the
tangent line at the point (1, 4).
dy
Q[13](˚): If x2 y2 + x sin(y) = 4, find .
dx
§§ Stage 3
Q[14](˚): If x2 + (y + 1)ey = 5, then find y1 at the points where y = 0.
Q[15]: For what values of x do the circle x2 + y2 = 1 and the ellipse x2 + 3y2 = 1 have
parallel tangent lines?
2.12Ĳ Inverse Trigonometric Functions

Exercises
43
§§ Stage 1
Q[1]: Give the domains of each of the following functions.
(a) f ( x ) = arcsin(cos x ) (b) g( x ) = arccsc(cos x ) (c) h( x ) = sin(arccos x )
Q[2]: A particle starts moving at time t = 10, and it bobs up and down, so that its height at
time t ě 10 is given by cos t. True or false: the particle has height 1 at time t = arccos(1).
Q[3]: The curve y = f ( x ) is shown below, for some function f . Restrict f to the largest
possible interval containing 0 over which it is one–to–one, and sketch the curve
y = f ´1 ( x ).
y
Q[4]: Let a be some constant. Where does the curve y = ax + cos x have a horizontal
tangent line?
Q[5]: Define a function f ( x ) = arcsin x + arccsc x. What is the domain of f ( x )? Where is
f ( x ) differentiable?
§§ Stage 2
x
Q[6]: Differentiate f ( x ) = arcsin . What is the domain of f ( x )?
3
arccos t
Q[7]: Differentiate f (t) = . What is the domain of f (t)?
t2 ´ 1
Q[8]: Differentiate f ( x ) = arcsec(´x2 ´ 2). What is the domain of f ( x )?

1 x
Q[9]: Differentiate f ( x ) = arctan , where a is a nonzero constant.
a a
What is the domain of f ( x )?
?
Q[10]: Differentiate f ( x ) = x arcsin x + 1 ´ x2 . What is the domain of f ( x )?
44
Q[11]: For which values of x is the tangent line to y = arctan( x2 ) horizontal?
d
Q[12]: Evaluate tarcsin x + arccos xu.
dx
1

Q[13](˚): Find the derivative of y = arcsin x .
1

Q[14](˚): Find the derivative of y = arctan x .
Q[15](˚): Calculate and simplify the derivative of (1 + x2 ) arctan x.
d
Q[16]: Show that tsin (arctan( x ))u = ( x2 + 1)´3/2 .
dx
d ´1
Q[17]: Show that tcot (arcsin( x ))u = ? .
dx x 1 ´ x2
2
Q[18](˚): Determine all points on the curve y = arcsin x where the tangent line is parallel
to the line y = 2x + 9.
Q[19]: For which values of x does the function f ( x ) = arctan(csc x ) have a horizontal
tangent line?
§§ Stage 3
Q[20](˚): Let f ( x ) = x + cos x, and let g(y) = f ´1 (y) be the inverse function. Determine
g1 ( y ).
1
Q[21](˚): f ( x ) = 2x ´ sin( x ) is one–to–one. Find f ´1 (π ´ 1).
1
Q[22](˚): f ( x ) = e x + x is one–to–one. Find f ´1 (e + 1).
Q[23]: Differentiate f ( x ) = [sin x + 2]arcsec x . What is the domain of this function?
1
Q[24]: Suppose you can’t remember whether the derivative of arcsine is ? or
1 ´ x2
1
? . Describe how the domain of arcsine suggests that one of these is wrong.
x2 ´ 1
π
Q[25]: Evaluate lim ( x ´ 1)´1 arctan x ´ .
xÑ1 4
5x ´ 9
Q[26]: Suppose f (2x + 1) = . Evaluate f ´1 (7).
3x + 7
45
D ERIVATIVES 2.13 T HE M EAN VALUE T HEOREM
2x + 3
Q[27]: Suppose f ´1 (4x ´ 1) = . Evaluate f (0).
x+1
Q[28]: Suppose a curve is defined implicitly by
arcsin( x + 2y) = x2 + y2
Solve for y1 in terms of x and y.
2.13Ĳ The Mean Value Theorem

Exercises
§§ Stage 1
Q[1]: Suppose a particular caribou has a top speed of 70 kph, and in one year it migrates
5000 km. What do you know about the amount of time the caribou spent travelling
during its migration?
Q[2]: Suppose a migrating sandhill crane flew 240 kilometres in one day. What does the
mean value theorem tell you about its speed during that day?
Q[3]: Below is the graph of y = f ( x ), where x is continuous on [ a, b] and differentiable on

( a, b). Mark on the graph the approximate location of a value c guaranteed by the mean
value theorem.
y
x
a b
Q[4]: Give a function f ( x ) with the properties that:

• f ( x ) is differentiable on the open interval 0 ă x ă 10
46
D ERIVATIVES 2.13 T HE M EAN VALUE T HEOREM
• f (0) = 0, f (10) = 10
but for all c P (0, 10), f 1 (c) = 0.
Q[5]: For each of the parts below, sketch a function f ( x ) (different in each part) that is
continuous and differentiable over all real numbers, with f (1) = f (2) = 0, and with the
listed property, or explain why no such function exists.
(a) f 1 (c) = 0 for no point c P (1, 2)
(b) f 1 (c) = 0 for exactly one point c P (1, 2)
(c) f 1 (c) = 0 for exactly five points c P (1, 2)
(d) f 1 (c) = 0 for infinitely many points c P (1, 2)
a
Q[6]: Suppose you want to show that a point exists where the function f ( x ) = |x| has a
1
tangent line with slope 13 . Find the mistake(s) in the following work, and give a correct
proof.
The function f ( x ) is continuous and differentiable over all real numbers, so

the mean value theorem applies. f (´4) = 2 and f (9) = 3, so by the mean
3´2 1
value theorem, there exists some c P (´4, 9) such that f 1 ( x ) = = .
9 ´ (´4) 13
§§ Stage 2
Q[7](˚): Let f ( x ) = x2 ´ 2πx + cos( x ) ´ 1. Show that there exists a real number c such
that f 1 (c) = 0.
Q[8](˚): Let f ( x ) = e x + (1 ´ e) x2 ´ 1. Show that there exists a real number c such that
f 1 (c) = 0.
a
Q[9](˚): Let f ( x ) = 3 + sin( x ) + ( x ´ π )2 . Show that there exists a real number c such
that f 1 (c) = 0.
Q[10](˚): Let f ( x ) = x cos( x ) ´ x sin( x ). Show that there exists a real number c such that
f 1 (c) = 0.
Q[11]: How many roots does the function f ( x ) = 3x ´ sin x have?
(4x + 1)4
Q[12]: How many roots does the function f ( x ) = + x have?
16
Q[13]: How many roots does the function f ( x ) = x3 + sin x5 have?

Q[14]: How many positive-valued solutions does the equation e x = 4 cos(2x ) have?
47
D ERIVATIVES 2.14 H IGHER O RDER D ERIVATIVES
Q[15](˚): Let f ( x ) = 3x5 ´ 10x3 + 15x + a, where a is some constant.

(a) Prove that, regardless of the value a, f 1 ( x ) ą 0 for all x in (´1, 1).
(b) Prove that, regardless of the value a, f ( x ) = 3x5 ´ 10x3 + 15x + a has at most one
root in [´1, 1].
Q[16](˚): Find the point promised by the Mean Value Theorem for the function e x on the
interval [0, T ].
Q[17]: Use Corollary 2.13.11 and Theorem 2.12.7 in the CLP-1 text to show that arcsec x =
C ´ arccsc x for some constant C. Find C.
§§ Stage 3
1
Q[18](˚): Suppose f (0) = 0 and f 1 ( x ) = . Prove that f (100) ă 100.
1 + e´ f ( x )
Remark: an equation relating a function to its own derivative is called a differential equa-
tion. We’ll see some very basic differential equations in Section 3.3.
Q[19]: Let f ( x ) = 2x + sin x. What is the largest interval containing x = 0 over which
f ( x ) is one–to–one? What are the domain and range of f ´1 ( x )?
x
Q[20]: Let f ( x ) =
+ sin x. What is the largest interval containing x = 0 over which f ( x )
2
is one–to–one? What are the domain and range of f ´1 ( x ), if we restrict f to this interval?
Q[21]: Suppose f ( x ) and g( x ) are functions that are continuous over the interval [ a, b] and
differentiable over the interval ( a, b). Suppose further that f ( a) ă g( a) and g(b) ă f (b).
Show that there exists some c P [ a, b] with f 1 (c) ą g1 (c).
Q[22]: Suppose f ( x ) is a function that is differentiable over all real numbers, and f 1 ( x )
has precisely two roots. What is the maximum number of distinct roots that f ( x ) may
have?
Q[23]: How many roots does f ( x ) = sin x + x2 + 5x + 1 have?
2.14Ĳ Higher Order Derivatives

Exercises
§§ Stage 1
Q[1]: What is the 180th derivative of the function f ( x ) = e x ?
48
Q[2]: Suppose f ( x ) is a differentiable function, with f 1 ( x ) ą 0 and f 2 ( x ) ą 0 for every

x P ( a, b). Which of the following must be true?
(i) f ( x ) is positive over ( a, b)
(ii) f ( x ) is increasing over ( a, b)
(iii) f ( x ) is increasing at a constant rate over ( a, b)
(iv) f ( x ) is increasing faster and faster over ( a, b)
(v) f 3 ( x ) ą 0 for some x P ( a, b)
Q[3]: Let f ( x ) = ax15 for some constant a. Which value of a results in f (15) ( x ) = 3?
Q[4]: Find the mistake(s) in the following work, and provide a corrected answer.
d2 y
Suppose ´14x2 + 2xy + y2 = 1. We find at the point (1, 3). Differentiating
dx2
implicitly:
´28x + 2y + 2xy1 + 2yy1 = 0
Plugging in x = 1, y = 3:
´28 + 6 + 2y1 + 6y1 = 0

11
y1 =
4
Differentiating:
y2 = 0
§§ Stage 2
Q[5]: Let f ( x ) = (log x ´ 1) x. Evaluate f 2 ( x ).
d2
Q[6]: Evaluate tarctan xu.
dx2
d2 y
Q[7]: The unit circle consists of all point x2 + y2 = 1. Give an expression for in terms
dx2
of y.
et
Q[8]: Suppose the position of a particle at time t is given by s(t) = . Find the
t2 + 1
acceleration of the particle at time t = 1.
d3
Q[9]: Evaluate tlog(5x2 ´ 12)u.
dx3
Q[10]: The height of a particle at time t seconds is given by h(t) = ´ cos t. Is the particle
49
speeding up or slowing down at t = 1?
Q[11]: The height of a particle at time t seconds is given by h(t) = t3 ´ t2 ´ 5t + 10. Is the
particle’s motion getting faster or slower at t = 1?
Q[12]: Suppose a curve is defined implicitly by
x2 + x + y = sin( xy)
d2 y
What is at the point (0, 0)?
dx2
Q[13]: Which statements below are true, and which false?

d4
(a) sin x = sin x
dx4
d4
(b) cos x = cos x
dx4
d4
(c) tan x = tan x
dx4
§§ Stage 3
Q[14]: A function f ( x ) satisfies f 1 ( x ) ă 0 and f 2 ( x ) ą 0 over ( a, b). Which of the

following curves below might represent y = f ( x )?
y y y
x x x
a b a b a b
(i) (ii) (iii)

y y
x x
a b a b
(iv) (v)
Q[15]: Let f ( x ) = 2x . What is f (n) ( x ), if n is a whole number?
50
Q[16]: Let f ( x ) = ax3 + bx2 + cx + d, where a, b, c, and d are nonzero constants. What is
dn f
the smallest integer n so that = 0 for all x?
dx n
Q[17](˚):
2 3
f ( x ) = e x+x h( x ) = 1 + x + x2
2
(a) Find the first and second derivatives of both functions
(b) Evaluate both functions and their first and second derivatives at 0.
(c) Show that for all x ą 0, f ( x ) ą h( x ).
Remark: for some applications, we only need to know that a function is “big enough.”
Since f ( x ) is a difficult function to evaluate, it may be useful to know that it is bigger than
h( x ) when x is positive.
Q[18](˚): The equation x3 y + y3 = 10x defines y implicitly as a function of x near the
point (1, 2).
(a) Compute y1 at this point.
(b) It can be shown that y2 is negative when x = 1. Use this fact and your answer to (a)
to make a sketch showing the relationship of the curve to its tangent line at (1, 2).
Q[19]: Let g( x ) = f ( x )e x . In Question 12, Section 2.7, we learned that
g1 ( x ) = [ f ( x ) + f 1 ( x )]e x .
(a) What is g2 ( x )?
(b) What is g3 ( x )?
(c) Based on your answers above, guess a formula for g(4) ( x ). Check it by differentiating.
Q[20]: Suppose f ( x ) is a function whose first n derivatives exist over all real numbers,
and f (n) ( x ) has precisely m roots. What is the maximum number of roots that f ( x ) may
have?
Q[21]: How many roots does the function f ( x ) = ( x + 1) log( x + 1) + sin x ´ x2 have?
Q[22](˚): Let f ( x ) = x|x|.

(a) Show that f ( x ) is differentiable at x = 0, and find f 1 (0).
(b) Find the second derivative of f ( x ). Explicitly state, with justification, the point(s) at
which f 2 ( x ) does not exist, if any.
51
52
Chapter 3
A PPLICATIONS OF DERIVATIVES
3.1Ĳ Velocity and acceleration

Exercises
§§ Stage 1
Q[1]: Suppose you throw a ball straight up in the air, and its height from t = 0 to t = 4 is
given by h(t) = ´4.9t2 + 19.6t. True or false: at time t = 2, the acceleration of the ball is 0.
Q[2]: Suppose an object is moving with a constant acceleration. It takes ten seconds to
accelerate from 1 ms to 2 ms . How long does it take to accelerate from 2 ms to 3 ms ? How
long does it take to accelerate from 3 ms to 13 ms ?
Q[3]: Let s(t) be the position of a particle at time t. True or false: if s2 ( a) ą 0 for some a,
then the particle’s speed is increasing when t = a.
Q[4]: Let s(t) be the position of a particle at time t. True or false: if s1 ( a) ą 0 and s2 ( a) ą 0
for some a, then the particle’s speed is increasing when t = a.
§§ Stage 2
For this section, we will ask you a number of questions that have to do with objects falling on Earth. Unless
otherwise stated, you should assume that an object falling through the air has an acceleration due to gravity
of 9.8 meters per second per second.
53
A PPLICATIONS OF DERIVATIVES 3.1 V ELOCITY AND ACCELERATION
Q[5]: A flower pot rolls out of a window 10m above the ground. How fast is it falling just
as it smacks into the ground?
Q[6]: You want to know how deep a well is, so you drop a stone down and count the
seconds until you hear it hit bottom.
(a) If the stone took x seconds to hit bottom, how deep is the well?
(b) Suppose you think you dropped the stone down the well, but actually you tossed it
down, so instead of an initial velocity of 0 metres per second, you accidentally
imparted an initial speed of 1 metres per second. What is the actual depth of the well,
if the stone fell for x seconds?
Q[7]: You toss a key to your friend, standing two metres away. The keys initially move
towards your friend at 2 metres per second, but slow at a rate of 0.25 metres per second
per second. How much time does your friend have to react to catch the keys? That is–how
long are the keys flying before they reach your friend?
km
Q[8]: A car is driving at 100 kph, and it brakes with a deceleration of 50 000 . How long
hr2
does the car take to come to a complete stop?
Q[9]: You are driving at 120 kph, and need to stop in 100 metres. How much deceleration
do your brakes need to provide? You may assume the brakes cause a constant
deceleration.
Q[10]: You are driving at 100 kph, and apply the brakes steadily, so that your car deceler-
ates at a constant rate and comes to a stop in exactly 7 seconds. What was your speed one
second before you stopped?
Q[11]: About 8.5 minutes after liftoff, the US space shuttle has reached orbital velocity, 17
500 miles per hour. Assuming its acceleration was constant, how far did it travel in those
8.5 minutes?
(Source: https://www.nasa.gov/mission_pages/shuttle/shuttlemissions/sts121/launch/qa-leinbach.html)
Q[12]: A pitching machine has a dial to adjust the speed of the pitch. You rotate it so that
it pitches the ball straight up in the air. How fast should the ball exit the machine, in
order to stay in the air exactly 10 seconds?
You may assume that the ball exits from ground level, and is acted on only by gravity,
which causes a constant deceleration of 9.8 metres per second.
Q[13]: A peregrine falcon can dive at a speed of 325 kph. If you were to drop a stone, how
high up would you have to be so that the stone reached the same speed in its fall?
Q[14]: You shoot a cannon ball into the air with initial velocity v0 , and then gravity
brings it back down (neglecting all other forces). If the cannon ball made it to a height of
100m, what was v0 ?
Q[15]: Suppose you are driving at 120 kph, and you start to brake at a deceleration of 50000
54
A PPLICATIONS OF DERIVATIVES 3.2 R ELATED R ATES
kph per hour. For three seconds you steadily increase your deceleration to 60000 kph per
hour. (That is, for three seconds, the rate of change of your deceleration is constant.) How
fast are you driving at the end of those three seconds?
§§ Stage 3
Q[16]: You jump up from the side of a trampoline with an initial upward velocity of 1
metre per second. While you are in the air, your deceleration is a constant 9.8 metres per
second per second due to gravity. Once you hit the trampoline, as you fall your speed
decreases by 4.9 metres per second per second. How many seconds pass between the
peak of your jump and the lowest part of your fall on the trampoline?
Q[17]: Suppose an object is moving so that its velocity doubles every second. Give an
expression for the acceleration of the object.
3.2Ĳ Related Rates
Exercises
55
§§ Stage 1
Q[1]: Suppose the quantities P and Q are related by the formula P = Q3 . P and Q are
changing with respect to time, t. Given this information, which of the following are
problems you could solve?
dP dQ dP
i. Given (0), find (0). (Remember: the notation (0) means the derivative of P
dt dt dt
with respect to t at the time t = 0.)
dP dQ
ii. Given (0) and the value of Q when t = 0, find (0).
dt dt
dQ dP
iii. Given (0), find (0).
dt dt
dQ dP
iv. Given (0) and the value of P when t = 0, find (0).
dt dt
§§ Stage 2
For problems 2 through 4, the relationship between several variables is explicitly given. Use this information
to relate their rates of change.
(
2 + y2 = 1 in the xy–plane. At
Q[2](˚):
? A point
? is moving on the unit circle ( x, y ) : x
(2/ 5, 1/ 5), its y–coordinate is increasing at rate 3. What is the rate of change of its
x–coordinate?
Q[3](˚): The quantities P, Q and R are functions of time and are related by the equation
R = PQ. Assume that P is increasing instantaneously at the rate of 8% per year and
P1
that Q is decreasing instantaneously at the rate of 2% per year. That is, = 0.08 and
P
Q 1
= ´0.02. Determine the percentage rate of change for R.
Q
Q[4](˚): Three quantities, F, P and Q all depend upon time t and are related by the
equation
P
F=
Q
(a) Assume that at a particular moment in time P = 25 and P is increasing at the
instantaneous rate of 5 units/min. At the same moment, Q = 5 and Q is increasing at
the instantaneous rate of 1 unit/min. What is the instantaneous rate of change in F at
this moment?
(b) Assume that at another moment in time P is increasing at the instantaneous rate of
10% and Q is decreasing at the instantaneous rate 5%. What can you conclude about
the rate of change of F at this moment?
For Questions 5 through 9, look for a way to use the Pythagorean Theorem.
Q[5](˚): Two particles move in the Cartesian plane. Particle A travels on the x-axis starting
at (10, 0) and moving towards the origin with a speed of 2 units per second. Particle B
56
travels on the y-axis starting at (0, 12) and moving towards the origin with a speed of 3
units per second. What is the rate of change of the distance between the two particles
when particle A reaches the point (4, 0)?
Q[6](˚): Two particles A and B are placed on the Cartesian plane at (0, 0) and (3, 0) re-
spectively. At time 0, both start to move in the +y direction. Particle A moves at 3 units
per second, while B moves at 2 units per second. How fast is the distance between the
particles changing when particle A is at a distance of 5 units from B.
Q[7](˚): Ship A is 400 miles directly south of Hawaii and is sailing south at 20
miles/hour. Ship B is 300 miles directly east of Hawaii and is sailing west at 15
miles/hour. At what rate is the distance between the ships changing?
Q[8](˚): Two tall sticks are vertically planted into the ground, separated by a distance of
30 cm. We simultaneously put two snails at the base of each stick. The two snails then
begin to climb their respective sticks. The first snail is moving with a speed of 25 cm per
minute, while the second snail is moving with a speed of 15 cm per minute. What is the
rate of change of the distance between the two snails when the first snail reaches 100 cm
above the ground?
Q[9](˚): A 20m long extension ladder leaning against a wall starts collapsing in on itself
at a rate of 2m/s, while the foot of the ladder remains a constant 5m from the wall. How
fast is the ladder moving down the wall after 3.5 seconds?
For Questions 10 through 14, look for tricks from trigonometry.
Q[10]: A watering trough has a cross section shaped like an isosceles trapezoid. The
trough is 2 metres long, 50 cm high, 1 metre wide at the top, and 60 cm wide at the
bottom.
2m
1m
50 cm
60 cm
A pig is drinking water from the trough at a rate of 3 litres per minute. When the height
of the water is 25 cm, how fast is the height decreasing?
Q[11]: A tank is 5 metres long, and has a trapezoidal cross section with the dimensions
shown below.
57
1m
1.25 m
5m
3m 3m
A hose is filling the tank up at a rate of one litre per second. How fast is the height of the
water increasing when the water is 10 centimetres deep?
Q[12]: A rocket is blasting off, 2 kilometres away from you. You and the rocket start at
the same height. The height of the rocket in kilometres, t hours after liftoff, is given by
h(t) = 61750t2
How fast (in radians per second) is your line of sight rotating to keep looking at the rocket,
one minute after liftoff?
Q[13](˚): A high speed train is traveling at 2 km/min along a straight track. The train is
moving away from a movie camera which is located 0.5 km from the track.
(a) How fast is the distance between the train and the camera increasing when they are
1.3 km apart?
(b) Assuming that the camera is always pointed at the train, how fast (in radians per
min) is the camera rotating when the train and the camera are 1.3 km apart?
Q[14]: A clock has a minute hand that is 10 cm long, and an hour hand that is 5 cm long.
Let D be the distance between the tips of the two hands. How fast is D decreasing at 4:00?
For Questions 15 through 20, you’ll need to know formulas for volume or area.
58
Q[15](˚): Find the rate of change of the area of the annulus t( x, y) : r2 ď x2 + y2 ď R2 u.

(i.e. the points inside the circle of radius R but outside the circle of radius r) if
dR cm dr cm
R = 3 cm, r = 1 cm, =2 , and =7 .
dt s dt s
R
r
Q[16]: Two spheres are centred at the same point. The radius R of the bigger sphere at
time t is given by R(t) = 10 + 2t, while the radius r of the smaller sphere is given by
r (t) = 6t, t ě 0. How fast is the volume between the spheres (inside the big sphere and
outside the small sphere) changing when the bigger sphere has a radius twice as large as
the smaller?
Q[17]: You attach two sticks together at their ends, and stick the other ends in the mud.
One stick is 150 cm long, and the other is 200 cm.
cm 200
0 cm
1.4 m 15
The structure starts out being 1.4 metres high at its peak, but the sticks slide, and the
height decreases at a constant rate of three centimetres per minute. How quickly is the
area of the triangle (formed by the two sticks and the level ground) changing when the
height of the structure is 120 cm?
Q[18]: The circular lid of a salt shaker has radius 8. There is a cut-out to allow the salt to
pour out of the lid, and a door that rotates around to cover the cut-out. The door is a
quarter-circle of radius 7 cm. The cut-out has the shape of a quarter-annulus with outer
radius 6 cm and inner radius 1 cm. If the uncovered area of the cut-out is A cm2 , then the
salt flows out at 15 A cm3 per second.
59
salt shaker lid salt shaker lid

cut-out uncovered cut-out partially covered
Recall: an annulus is the set of points inside one circle and outside another, like a flat
doughnut (see Question 15).
annulus quarter annulus
While pouring out salt, you spin the door around the lid at a constant rate of π6 radians
per second, covering more and more of the cut-out. When exactly half of the cut-out is
covered, how fast is the flow of salt changing?
Q[19]: A cylindrical sewer pipe with radius 1 metre has a vertical rectangular door that
slides in front of it to block the flow of water, as shown below. If the uncovered area of
the pipe is A m2 , then the flow of water through the pipe is 51 A cubic metres per second.
The door slides over the pipe, moving vertically at a rate of 1 centimetre per second.
How fast is the flow of water changing when the door covers the top 25 centimetres of
the pipe?
60
Q[20]: A martini glass is shaped like a cone, with top diameter 10 cm and side length 10
cm.
10
10 10
When the liquid in the glass is 7 cm high, it is evaporating at a rate of 5 mL per minute.
How fast is the height of the liquid decreasing?
§§ Stage 3
Q[21]: A floating buoy is anchored to the bottom of a river. As the river flows, the buoy is
pulled in the direction of flow until its 2-metre rope is taut. A sensor at the anchor reads
the angle θ between the rope and the riverbed, as shown in the diagram below. This data
is used to measure the depth D of water in the river, which depends on time.
D
θ
π dθ rad
(a) If θ = and = 0.25 , how fast is the depth D of the water changing?
4 dt hr
dθ dD
(b) A measurement shows = 0, but ‰ 0. Under what circumstances does this
dt dt
occur?
dθ dD
(c) A measurement shows ą 0, but ă 0. Under what circumstances does this
dt dt
occur?
Q[22]: A point is moving in the xy-plane along the quadrilateral shown below.
61
x
1
(a) When the point is at (0, ´2), it is moving to the right. An observer stationed at the
origin must turn at a rate of one radian per second to keep looking directly at the
point. How fast is the point moving?
(b) When the point is at (0, 2), its x-coordinate is increasing at a rate of one unit per
second. How fast it its y-coordinate changing? How fast is the point moving?
Q[23]: You have a cylindrical water bottle 20 cm high, filled with water. Its cross section
is a circle of radius 5. You slowly smoosh the sides, so the cross section becomes an
ellipse with major axis (widest part) 2a and minor axis (skinniest part) 2b.
5
b
5 a
20 20
After t seconds of smooshing the bottle, a = 5 + t cm. The perimeter of the cross section
is unchanged as the bottle deforms. The perimeter of an ellipse is actually quite difficult
to calculate, but we will use an approximation derived by Ramanujan and assume that
the perimeter p of our ellipse is
b
p « π 3( a + b ) ´ ( a + 3b)(3a + b) .
The area of an ellipse is πab.

(a) Give an equation that relates a and b (and no other variables).
62
A PPLICATIONS OF DERIVATIVES 3.3 E XPONENTIAL G ROWTH AND D ECAY
(b) Give an expression for the volume of the bottle as it is being smooshed, in terms of a
and b (and no other variables).
(c) Suppose the bottle was full when its cross section was a circle. How fast is the water
spilling out when a is twice as big as b?
Q[24]: The quantities A, B, C, and D all depend on time, and are related by the formula

AB = log C2 + D2 + 1 .
At time t = 10, the following values are known:

• A=0
dA
• = 2 units per second
dt
What is B when t = 10?
3.3Ĳ Exponential Growth and Decay

3.3.1 §§ Carbon Dating
Exercises
§§ Stage 1
Q[1]: Which of the following is a differential equation for an unknown function y of x?

dy dy dx
(a) y = (b) = 3 [ y ´ 5] (c) y = 3 y ´ (d) e x = ey + 1 (e) y = 10e x
dx dx dx
Q[2]: Which of the following functions Q(t) satisfy the differential equation
dQ
Q(t) = 5 ?
dt
(a) Q(t) = 0 (b) Q(t) = 5et (c) Q(t) = e5t (d) Q(t) = et/5 (e) Q(t) = et/5 + 1
Q[3]: Suppose a sample starts out with C grams of a radioactive isotope, and the amount
of the radioactive isotope left in the sample at time t is given by
Q(t) = Ce´kt
for some positive constant k. When will Q(t) = 0?
63
§§ Stage 2
Q[4](˚): Consider a function of the form f ( x ) = Aekx where A and k are constants. If
f (0) = 5 and f (7) = π, find the constants A and k.
dy
Q[5](˚): Find the function y(t) if + 3y = 0, y(1) = 2.
dt
Q[6]: A sample of bone belongs to an animal that died 10,000 years ago. If the bone
contained 5 µg of Carbon-14 when the animal died, how much Carbon-14 do you expect
it to have now?
Q[7]: A sample containing one gram of Radium-226 was stored in a lab 100 years ago;
now the sample only contains 0.9576 grams of Radium-226. What is the half-life of
Radium-226?
Q[8](˚): The mass of a sample of Polonium–210, initially 6 grams, decreases at a rate pro-
portional to the mass. After one year, 1 gram remains. What is the half–life (the time it
takes for the sample to decay to half its original mass)?
Q[9]: Radium-221 has a half-life of 30 seconds. How long does it take for only 0.01% of an
original sample to be left?
§§ Stage 3
Q[10]: Polonium-210 has a half life of 138 days. What percentage of a sample of
Polonium-210 decays in a day?
Q[11]: A sample of ore is found to contain 7.2 ˘ 0.3 µg of Uranium-232, the half-life of
which is between 68.8 and 70 years. How much Uranium-232 will remain undecayed in
the sample in 10 years?
3.3.2 §§ Newton’s Law of Cooling

Exercises
§§ Stage 1
Q[1]: Which of the following functions T (t) satisfy the differential equation
dT
= 5 [ T ´ 20]?
dt
(a) T (t) = 20 (b) T (t) = 20e5t ´ 20 (c) T (t) = e5t + 20 (d) T (t) = 20e5t + 20
64
Q[2]: At time t = 0, an object is placed in a room, of temperature A. After t seconds,

Newton’s Law of Cooling gives the temperature of the object is as
T (t) = 35eKt ´ 10
What is the temperature of the room? Is the room warmer or colder than the object?
Q[3]: A warm object is placed in a cold room. The temperature of the object, over time,
approaches the temperature of the room it is in. The temperature of the object at time t is
given by
T (t) = [ T (0) ´ A]eKt + A.
Can K be a positive number? Can K be a negative number? Can K be zero?
Q[4]: Suppose an object obeys Newton’s Law of Cooling, and its temperature is given by
T (t) = [ T (0) ´ A]ekt + A

for some constant k. At what time is T (t) = A?
§§ Stage 2
Q[5]: A piece of copper at room temperature (25˝ ) is placed in a boiling pot of water. After
10 seconds, it has heated to 90˝ . When will it be 99.9˝ ?
Q[6]: Today is a chilly day. We heated up a stone to 500˝ C in a bonfire, then took it out
and left it outside, where the temperature is 0˝ C. After 10 minutes outside of the bonfire,
the stone had cooled to a still-untouchable 100˝ C. Now the stone is at a cozy 50˝ C. How
long ago was the stone taken out of the fire?
§§ Stage 3
Q[7](˚): Isaac Newton drinks his coffee with cream. To be exact, 9 parts coffee to 1 part
cream. His landlady pours him a cup of coffee at 95˝ C into which Newton stirs cream
taken from the icebox at 5˝ C. When he drinks the mixture ten minutes later, he notes that
it has cooled to 54˝ C. Newton wonders if his coffee would be hotter (and by how much)
if he waited until just before drinking it to add the cream. Analyze this question,
assuming that:
(i) The temperature of the dining room is constant at 22˝ C.
(ii) When a volume V1 of liquid at temperature T1 is mixed with a volume V2 at
V T + V2 T2
temperature T2 , the temperature of the mixture is 1 1 .
V1 + V2
(iii) Newton’s Law of Cooling: The temperature of an object cools at a rate proportional
to the difference in temperature between the object and its surroundings.
(iv) The constant of proportionality is the same for the cup of coffee with cream as for
the cup of pure coffee.
65
Q[8](˚): The temperature of a glass of iced tea is initially 5˝ . After 5 minutes, the tea has
heated to 10˝ in a room where the air temperature is 30˝ .
(a) Use Newton’s law of cooling to obtain a differential equation for the temperature
T (t) at time t.
(b) Determine when the tea will reach a temperature of 20˝ .
Q[9]: Suppose an object is changing temperature according to Newton’s Law of Cooling,

and its temperature at time t is given by
T (t) = 0.8kt + 15
Is k positive or negative?
3.3.3 §§ Population Growth

Exercises
§§ Stage 1
Q[1]: Let a population at time t be given by the Malthusian model,
P(t) = P(0)ebt for some positive constant b.
Evaluate lim P(t). Does this model make sense for large values of t?
tÑ8
§§ Stage 2
Q[2]: In the 1950s, pure-bred wood bison were thought to be extinct. However, a small
population was found in Canada. For decades, a captive breeding program has been
working to increase their numbers, and from time to time wood bison are released to the
wild. Suppose in 2015, a released herd numbered 121 animals, and a year later, there
were 136* . If the wood bison adhere to the Malthusian model (a big assumption!), and if
there are no more releases of captive animals, how many animals will the herd have in
2020?
* These numbers are loosely based on animals actually released near Shageluk, Alaska in 2015. Watch
the first batch being released here.
Q[3]: A founding colony of 1,000 bacteria is placed in a petri dish of yummy bacteria food.
After an hour, the population has doubled. Assuming the Malthusian model, how long
will it take for the colony to triple its original population?
66
Q[4]: A single pair of rats comes to an island after a shipwreck. They multiply according
to the Malthusian model. In 1928, there were 1,000 rats on the island, and the next year
there were 1500. When was the shipwreck?
Q[5]: A farmer wants to farm cochineals, which are insects used to make red dye. The
farmer raises a small number of cochineals as a test. In three months, a test population of
cochineals will increase from 200 individuals to 1000, given ample space and food.
The farmer’s plan is to start with an initial population of P(0) cochineals, and after a year
have 1 000 000 + P(0) cochineals, so that one million can be harvested, and P(0) saved to
start breeding again. What initial population P(0) does the Malthusian model suggest?
§§ Stage 3
Q[6]: Let f (t) = 100ekt , for some constant k.

(a) If f (t) is the amount of a decaying radioactive isotope in a sample at time t, what is
the amount of the isotope in the sample when t = 0? What is the sign of k?
(b) If f (t) is the number of individuals in a population that is growing according to the
Malthusian model, how many individuals are there when t = 0? What is the sign of
k?
(c) If f (t) is the temperature of an object at time t, given by Newton’s Law of Cooling,
what is the ambient temperature surrounding the object? What is the sign of k?
3.3.4 §§ Further problems

Exercises
§§ More Section 3.3 Problems

Q[1](˚): Find f (2) if f 1 ( x ) = π f ( x ) for all x, and f (0) = 2.
dT
Q[2]: Which functions T (t) satisfy the differential equation = 7T + 9?
dt
Q[3](˚): It takes 8 days for 20% of a particular radioactive material to decay. How long
does it take for 100 grams of the material to decay to 40 grams?
Q[4]: A glass of boiling water is left in a room. After 15 minutes, it has cooled to 85˝ C,
and after 30 minutes it is 73˝ C. What temperature is the room?
67
A PPLICATIONS OF DERIVATIVES 3.4 TAYLOR POLYNOMIALS
Q[5](˚): A 25-year-old graduate of UBC is given $50,000 which is invested at 5% per year
compounded continuously. The graduate also intends to deposit money continuously at
the rate of $2000 per year. Assuming that the interest rate remains 5%, the amount A(t)
of money at time t satisfies the equation
dA
= 0.05A + 2000
dt
(a) Solve this equation and determine the amount of money in the account when the
graduate is 65.
(b) At age 65, the graduate will withdraw money continuously at the rate of W dollars
per year. If the money must last until the person is 85, what is the largest possible
value of W?
Q[6](˚): An investor puts $120,000 which into a bank account which pays 6% annual
interest, compounded continuously. She plans to withdraw money continuously from
the account at the rate of $9000 per year. If A(t) is the amount of money at time t, then
dA
= 0.06A ´ 9000
dt
(a) Solve this equation for A(t).
(b) When will the money run out?
Q[7](˚): A particular bacterial culture grows at a rate proportional to the number of bac-
teria present. If the size of the culture triples every nine hours, how long does it take the
culture to double?
Q[8](˚): An object falls under gravity near the surface of the earth and its motion is
impeded by air resistance proportional to its speed. Its velocity v satisfies the differential
equation
dv
= ´g ´ kv
dt
where g and k are positive constants.
(a) Find the velocity of the object as a function of time t, given that it was v0 at t = 0.
(b) Find lim v(t).

tÑ8
3.4Ĳ Taylor polynomials

3.4.1 §§ Zeroeth Approximation
Exercises
68
§§ Stage 1
Q[1]: The graph below shows three curves. The black curve is y = f ( x ), the red curve is
y = g( x ) = 1 + 2 sin(1 + x ), and the blue curve is y = h( x ) = 0.7. If you want to estimate
f (0), what might cause you to use g(0)? What might cause you to use h(0)?
y
y = g( x )
y = f (x)
y = h( x )
x
§§ Stage 2
In this and following sections, we will ask you to approximate the value of several constants, such as
log(0.93). A valid question to consider is why we would ask for approximations of these constants that take
lots of time, and are less accurate than what you get from a calculator.
One answer to this question is historical: people were approximating logarithms before they had calculators,
and these are some of the ways they did that. Pretend you’re on a desert island without any of your usual
devices and that you want to make a number of quick and dirty approximate evaluations.
Another reason to make these approximations is technical: how does the calculator get such a good approx-
imation of log(0.93)? The techniques you will learn later on in this chapter give very accurate formulas for
approximating functions like log x and sin x, which are sometimes used in calculators.
A third reason to make simple approximations of expressions that a calculator could evaluate is to provide a
reality check. If you have a ballpark guess for your answer, and your calculator gives you something wildly
different, you know to double-check that you typed everything in correctly.
For now, questions like Question 2 through Question 4 are simply for you to practice the fundamental ideas
we’re learning.
Q[2]: Use a constant approximation to estimate the value of log( x ) when x = 0.93.
Sketch the curve y = f ( x ) and your constant approximation.
(Remember that in CLP-1 we use log x to mean the natural logarithm of x, loge x.)
69
Q[3]: Use a constant approximation to estimate arcsin(0.1).
?
Q[4]: Use a constant approximation to estimate 3 tan(1).
§§ Stage 3
Q[5]: Use a constant approximation to estimate the value of 10.13 . Your estimation
should be something you can calculate in your head.
3.4.2 §§ First Approximation

Exercises
§§ Stage 1
Q[1]: Suppose f ( x ) is a function, and we calculated its linear approximation near x = 5

to be f ( x ) « 3x ´ 9.
(a) What is f (5)?
(b) What is f 1 (5)?
(c) What is f (0)?
Q[2]: The curve y = f ( x ) is shown below. Sketch the linear approximation of f ( x ) about
x = 2.
y
y = f (x)
x
2
Q[3]: What is the linear approximation of the function f ( x ) = 2x + 5 about x = a?
70
§§ Stage 2
Q[4]: Use a linear approximation to estimate log( x ) when x = 0.93. Sketch the curve
y = f ( x ) and your linear approximation.
?
Q[5]: Use a linear approximation to estimate 5.
?
5
Q[6]: Use a linear approximation to estimate 30
§§ Stage 3
Q[7]: Use a linear approximation to estimate 10.13 , then compare your estimation with
the actual value.
Q[8]: Imagine f ( x ) is some function, and you want to estimate f (b). To do this, you
choose a value a and take an approximation (linear or constant) of f ( x ) about a. Give an
example of a function f ( x ), and values a and b, where the constant approximation gives a
more accurate estimation of f (b) than the linear approximation.
Q[9]: The function

?
1 4π ´ 27
L( x ) = x +
4 12
is the linear approximation of f ( x ) = arctan x about what point x = a?
3.4.3 §§ Second Approximation

Exercises
§§ Stage 1
Q[1]: The quadratic approximation of a function f ( x ) about x = 3 is
f ( x ) « ´x2 + 6x
What are the values of f (3), f 1 (3), f 2 (3), and f 3 (3)?
Q[2]: Give a quadratic approximation of f ( x ) = 2x + 5 about x = a.
71
§§ Stage 2
Q[3]: Use a quadratic approximation to estimate log(0.93).


1
Q[4]: Use a quadratic approximation to estimate cos .
15
Q[5]: Calculate the quadratic approximation of f ( x ) = e2x about x = 0.

4
Q[6]: Use a quadratic approximation to estimate 5 3 .
Q[7]: Evaluate the expressions below.
30
ÿ
(a) 1
n =5
3 h
ÿ i
(b) 2( n + 3) ´ n2
n =1
10
ÿ 1 1
(c) ´
n n+1
n =1
4
ÿ 5 ¨ 2n
(d)
n =1
4n +1
Q[8]: Write the following in sigma notation:

(a) 1 + 2 + 3 + 4 + 5
(b) 2 + 4 + 6 + 8
(c) 3 + 5 + 7 + 9 + 11
(d) 9 + 16 + 25 + 36 + 49
(e) 9 + 4 + 16 + 5 + 25 + 6 + 36 + 7 + 49 + 8
(f) 8 + 15 + 24 + 35 + 48
(g) 3 ´ 6 + 9 ´ 12 + 15 ´ 18
§§ Stage 3
Q[9]: Use a quadratic approximation of f ( x ) = 2 arcsin x about x = 0 to approximate

f (1). What number are you approximating?
Q[10]: Use a quadratic approximation of e x to estimate e as a decimal.
72
Q[11]: Group the expressions below into collections of equivalent expressions.

10
ÿ
(a) 2n
n =1
ÿ10
(b) 2n
n =1
ÿ10
(c) n2
n =1
10
ÿ
(d) 2 n
n =1
ÿ11
(e) 2 ( n ´ 1)
n =2
14
ÿ
(f) ( n ´ 4)2
n =5
10
4n +1

1 ÿ
(g)
4 2n
n =1
3.4.4 §§ Still Better Approximations

Exercises
§§ Stage 1
Q[1]: The 3rd degree Taylor polynomial for a function f ( x ) about x = 1 is
T3 ( x ) = x3 ´ 5x2 + 9x
What is f 2 (1)?
Q[2]: The nth degree Taylor polynomial for f ( x ) about x = 5 is

n
ÿ 2k + 1
Tn ( x ) = ( x ´ 5) k
3k ´ 9
k =0
What is f (10) (5)?
§§ Stage 3
Q[3]: The 4th-degree Maclaurin polynomial for f ( x ) is
T4 ( x ) = x4 ´ x3 + x2 ´ x + 1
73
What is the third-degree Maclaurin polynomial for f ( x )?
Q[4]: The 4th degree Taylor polynomial for f ( x ) about x = 1 is
T4 ( x ) = x4 + x3 ´ 9
What is the third degree Taylor polynomial for f ( x ) about x = 1?
Q[5]: For any even number n, suppose the nth degree Taylor polynomial for f ( x ) about
x = 5 is
n/2
ÿ 2k + 1
( x ´ 5)2k
3k ´ 9
k =0
What is f (10) (5)?

11
Q[6]: The third-degree Taylor polynomial for f ( x ) = x3 2 log x ´ about x = a is
3
2? 3 ?
T3 ( x ) = ´ e + 3ex ´ 6 ex2 + x3
3
What is a?
3.4.5 §§ Some Examples

Exercises
§§ Stage 2
Q[1]: Give the 16th degree Maclaurin polynomial for f ( x ) = sin x + cos x.
Q[2]: Give the 100th degree Taylor polynomial for s(t) = 4.9t2 ´ t + 10 about t = 5.
Q[3]: Write the nth-degree Taylor polynomial for f ( x ) = 2x about x = 1 in sigma notation.
Q[4]: Find the 6th degree Taylor polynomial of f ( x ) = x2 log x + 2x2 + 5 about x = 1,
remembering that log x is the natural logarithm of x, loge x.
1
Q[5]: Give the nth degree Maclaurin polynomial for in sigma notation.
1´x
74
§§ Stage 3
Q[6]: Calculate the 3rd-degree Taylor Polynomial for f ( x ) = x x about x = 1.
Q[7]: Use a 5th-degree Maclaurin polynomial for 6 arctan x to approximate π.
Q[8]: Write the 100th-degree Taylor polynomial for f ( x ) = x (log x ´ 1) about x = 1 in

sigma notation.
π
Q[9]: Write the (2n)th-degree Taylor polynomial for f ( x ) = sin x about x = in sigma
4
notation.
Q[10]: Estimate the sum below

1 1 1 1
1+ + + +¨¨¨+
2 3! 4! 157!
by interpreting it as a Maclaurin polynomial.
Q[11]: Estimate the sum below

100 2k
(´1)k

ÿ 5π
2k! 4
k =0
by interpreting it as a Maclaurin polynomial.
3.4.6 §§ Estimating Changes and ∆x, ∆y notation
Exercises
75
§§ Stage 1
Q[1]: In the picture below, label the following:
f (x) f ( x + ∆x ) ∆x ∆y
y = f (x)
x
x x + ∆x
Q[2]: At this point in the book, every homework problem takes you about 5 minutes. Use
the terms you learned in this section to answer the question: if you spend 15 minutes
more, how many more homework problems will you finish?
§§ Stage 2
Q[3]: Let f ( x ) = arctan x.

(a) Use a linear approximation to estimate f (5.1) ´ f (5).
(b) Use a quadratic approximation to estimate f (5.1) ´ f (5).
Q[4]: When diving off a cliff from x metres above the water, your speed as you hit the
water is given by
? m
s( x ) = 19.6x
sec
Your last dive was from a height of 4 metres.
(a) Use a linear approximation of ∆y to estimate how much faster you will be falling
when you hit the water if you jump from a height of 5 metres.
(b) A diver makes three jumps: the first is from x metres, the second from x + ∆x metres,
and the third from x + 2∆x metres, for some fixed positive values of x and ∆x. Which
is bigger: the increase in terminal speed from the first to the second jump, or the
increase in terminal speed from the second to the third jump?
3.4.7 §§ Further Examples

Exercises
76
§§ Stage 1
Q[1]: Let f ( x ) = 7x2 ´ 3x + 4. Suppose we measure x to be x0 = 2 but that the real value
of x is x0 + ∆x. Suppose further that the error in our measurement is ∆x = 1. Let ∆y be
the change in f ( x ) corresponding to a change of ∆x in x0 . That is,
∆y = f ( x0 + ∆x ) ´ f ( x0 ).
True or false: ∆y = f 1 (2)(1) = 25
Q[2]: Suppose the exact amount you are supposed to tip is $5.83, but you approximate
and tip $6. What is the absolute error in your tip? What is the percent error in your tip?
Q[3]: Suppose f ( x ) = 3x2 ´ 5. If you measure x to be 10, but its actual value is 11,
estimate the resulting error in f ( x ) using the linear approximation, and then the quadratic
approximation.
§§ Stage 2
Q[4]: A circular pen is being built on a farm. The pen must contain A0 square metres,
with an error of no more than 2%. Estimate the largest percentage error allowable on the
radius.
Q[5]: A circle with radius 3 has a sector cut out of it. It’s a smallish sector, no more than a
quarter of the circle. You want to find out the area of the sector.
θ d
(a) Suppose the angle of the sector is θ. What is the area of the sector?
(b) Unfortunately, you don’t have a protractor, only a ruler. So, you measure the chord
made by the sector (marked d in the diagram above). What is θ in terms of d?
(c) Suppose you measured d = 0.7, but actually d = 0.68. Estimate the absolute error in
your calculation of the area removed.
Q[6]: A conical tank, standing on its pointy end, has height 2 metres and radius 0.5
metres. Estimate change in volume of the water in the tank associated to a change in the
77
height of the water from 50 cm to 45 cm.
0.5
§§ Stage 3
Q[7]: A sample begins with precisely 1 µg of a radioactive isotope, and after 3 years is
measured to have 0.9 µg remaining. If this measurement is correct to within 0.05 µg,
estimate the corresponding accuracy of the half-life calculated using it.
3.4.8 §§ The error in Taylor polynomial approximation

Exercises
§§ Stage 1
Q[1]: Suppose f ( x ) is a function that we approximated by F ( x ). Further, suppose

f (10) = ´3, while our approximation was F (10) = 5. Let R( x ) = f ( x ) ´ F ( x ).
(a) True or false: |R(10)| ď 7
(b) True or false: |R(10)| ď 8
(c) True or false: |R(10)| ď 9
(d) True or false: |R(10)| ď 100
Q[2]: Let f ( x ) = e x , and let T3 ( x ) be the third-degree Maclaurin polynomial for f ( x ),
1 1
T3 ( x ) = 1 + x + x2 + x3
2 3!
Use Equation 3.4.33 to give a reasonable bound on the error | f (2) ´ T3 (2)|. Then, find the
error | f (2) ´ T3 (2)| using a calculator.
Q[3]: Let f ( x ) = 5x3 ´ 24x2 + ex ´ π 4 , and let T5 ( x ) be the fifth-degree Taylor polynomial
for f ( x ) about x = 1. Give the best bound you can on the error | f (37) ´ T (37)|.
Q[4]: You and your friend both want to approximate sin(33). Your friend uses the first-
degree Maclaurin polynomial for f ( x ) = sin x, while you use the zeroth-degree (constant)
78
Maclaurin polynomial for f ( x ) = sin x. Who has a better approximation, you or your
friend?
§§ Stage 2
Q[5]: Suppose a function f ( x ) has sixth derivative
6!(2x ´ 5)
f (6) ( x ) = .
x+3
Let T5 ( x ) be the 5th-degree Taylor polynomial for f ( x ) about x = 11.
Give a bound for the error | f (11.5) ´ T5 (11.5)|.
Q[6]: Let f ( x ) = tan x, and let T2 ( x ) be the second-degree Taylor polynomial for f ( x )
about x = 0. Give a reasonable bound on the error | f (0.1) ´ T (0.1)| using Equation 3.4.33.
Q[7]: Let f ( x ) = log(1 ´ x ), and let T5 ( x ) be the fifth-degree

Maclaurin
polynomial
for
f ( x ). Use Equation 3.4.33 to give a bound on the error | f ´ 41 ´ T5 ´ 14 |.
(Remember log x = loge x, the natural logarithm of x.)
?
Q[8]: Let f ( x ) = 5 x, and let T3 ( x ) be the third-degree Taylor polynomial for f ( x ) about
x = 32. Give a bound on the error | f (30) ´ T3 (30)|.
Q[9]: Let

1
f ( x ) = sin ,
x
1
and let T1 ( x ) be the first-degree Taylor polynomial for f ( x ) about x = . Give a bound
π
on the error | f (0.01) ´ T1 (0.01)|, using Equation 3.4.33. You may leave your answer in
terms of π.
Then, give a reasonable bound on the error | f (0.01) ´ T1 (0.01)|.
Q[10]: Let f ( x ) = arcsin x, and let T2 ( x ) be the

ˇ second-degree
ˇ Maclaurin polynomial for
f ( x ). Give a reasonable bound on the error ˇ f 2 ´ T2 12 ˇ using Equation 3.4.33. What
ˇ 1 ˇ
ˇ ˇ
is the exact value of the error ˇ f 12 ´ T2 21 ˇ?
ˇ ˇ
§§ Stage 3
Q[11]: Let f ( x ) = log( x ), and let Tn ( x ) be the nth-degree Taylor polynomial for f ( x )
about x = 1. You use Tn (1.1) to estimate log(1.1). If your estimation needs to have an
error of no more than 10´4 , what is an acceptable value of n to use?
79
?
Q[12]: Give an estimation of 7 2200 using a Taylor polynomial. Your estimation should
have an error of less than 0.001.
Q[13]: Use Equation 3.4.33 to show that

4241 4243
ď sin(1) ď
5040 5040
Q[14]: In this question, we use the remainder of a Maclaurin polynomial to approximate

e.
(a) Write out the 4th degree Maclaurin polynomial T4 ( x ) of the function e x .
(b) Compute T4 (1).
326 325
(c) Use your answer from b to conclude ăeă .
120 119
3.4.9 §§ Further problems

Exercises
§§ More Problems for Section 3.4

§§ Stage 1
Q[1](˚): Consider a function f ( x ) whose third-degree Maclaurin polynomial is 4 + 3x2 +
1 3 1 2
2 x . What is f (0)? What is f (0)?
Q[2](˚): Consider a function h( x ) whose third-degree Maclaurin polynomial is 1 + 4x ´
1 2 2 3
x + x . What is h(3) (0)?
3 3
Q[3](˚): The third-degree Taylor polynomial of h( x ) about x = 2 is
1
3 + ( x ´ 2) + 2( x ´ 2)3 .
2
What is h1 (2)? What is h2 (2)?
§§ Stage 2
Q[4](˚): The function f ( x ) has the property that f (3) = 2, f 1 (3) = 4 and f 2 (3) = ´10.
(a) Use the linear approximation to f ( x ) centred at x = 3 to approximate f (2.98).
(b) Use the quadratic approximation to f ( x ) centred at x = 3 to approximate f (2.98).
Q[5](˚): Use the tangent line to the graph of y = x1/3 at x = 8 to find an approximate
value for 101/3 . Is the approximation too large or too small?
80
?
Q[6](˚): Estimate 2 using a linear approximation.
?
Q[7](˚): Estimate 3 26 using a linear approximation.
Q[8](˚): Estimate (10.1)5 using a linear approximation.

101π
Q[9](˚): Estimate sin using a linear approximation. (Leave your answer in terms
100
of π.)
π
Q[10](˚): Use a linear approximation to estimate arctan(1.1), using arctan 1 = .
4
Q[11](˚): Use a linear approximation to estimate (2.001)3 . Write your answer in the form
n/1000 where n is an integer.
Q[12](˚): Using a suitable linear approximation, estimate (8.06)2/3 . Give your answer as
a fraction in which both the numerator and denominator are integers.
Q[13](˚): Find the third–order Taylor polynomial for f ( x ) = (1 ´ 3x )´1/3 around x = 0.
x
Q[14](˚): Consider a function f ( x ) which has f (3) ( x ) = . Show that when we
22 ´ x2
approximate f (2) using its second degree Taylor polynomial at a = 1, the absolute value
1
of the error is less than 50 = 0.02.
cos( x2 )
Q[15](˚): Consider a function f ( x ) which has f (4) ( x ) = . Show that when we
3´x
approximate f (0.5) using its third-degree Maclaurin polynomial, the absolute value of
1
the error is less than 500 = 0.002.
e´x
Q[16](˚): Consider a function f ( x ) which has f (3) ( x ) =. Show that when we
8 + x2
approximate f (1) using its second degree Maclaurin polynomial, the absolute value of
the error is less than 1/40.
Q[17](˚):
(a) By using an appropriate linear approximation for f ( x ) = x1/3 , estimate 52/3 .
(b) Improve your answer in (a) by making a quadratic approximation.
(c) Obtain an error estimate for your answer in (a) (not just by comparing with your
calculator’s answer for 52/3 ).
§§ Stage 3
Q[18]: The 4th degree Maclaurin polynomial for f ( x ) is
T4 ( x ) = 5x2 ´ 9
81
A PPLICATIONS OF DERIVATIVES 3.5 O PTIMIZATION
What is the third degree Maclaurin polynomial for f ( x )?
Q[19](˚): The equation y4 + xy = x2 ´ 1 defines y implicitly as a function of x near the

point x = 2, y = 1.
(a) Use the tangent line approximation at the given point to estimate the value of y when
x = 2.1.
(b) Use the quadratic approximation at the given point to estimate the value of y when
x = 2.1.
(c) Make a sketch showing how the curve relates to the tangent line at the given point.
Q[20](˚): The equation x4 + y + xy4 = 1 defines y implicitly as a function of x near the

point x = ´1, y = 1.
(a) Use the tangent line approximation at the given point to estimate the value of y when
x = ´0.9.
(b) Use the quadratic approximation at the given point to get another estimate of y when
x = ´0.9.
(c) Make a sketch showing how the curve relates to the tangent line at the given point.
Q[21](˚): Given that log 10 « 2.30259, estimate log 10.3 using a suitable tangent line ap-
proximation. Give an upper and lower bound for the error in your approximation by
using a suitable error estimate.
x
Q[22](˚): Consider f ( x ) = ee .
(a) Give the linear approximation for f near x = 0 (call this L( x )).
(b) Give the quadratic approximation for f near x = 0 (call this Q( x )).
(c) Prove that L( x ) ă Q( x ) ă f ( x ) for all x ą 0.
(d) Find an interval of length at most 0.01 that is guaranteed to contain the number e0.1 .
3.5Ĳ Optimization
3.5.1 §§ Local and global maxima and minima
Exercises
82
§§ Stage 1
Q[1]: Identify every critical point and every singular point of f ( x ) shown on the graph
below.
Which correspond to local extrema?
y
y = f (x)
Q[2]: Identify every critical point and every singular point of f ( x ) on the graph below.
Which correspond to local extrema? Which correspond to global extrema over the
interval shown?
y
y = f (x)
Q[3]: Draw a graph y = f ( x ) where f (2) is a local maximum, but it is not a global maxi-
mum.
83
§§ Stage 2
x´1
Q[4]: Suppose f ( x ) = .
x2 + 3
(a) Find all critical points.
(b) Find all singular points.
(c) What are the possible points where local extrema of f ( x ) may exist?
§§ Stage 3
Q[5]: Below are a number of curves, all of which have a singular point at x = 2. For each,
label whether x = 2 is a local maximum, a local minimum, or neither.
y y y y
x x x x
2 2 2 2
Q[6]: Draw a graph y = f ( x ) where f (2) is a local maximum, but x = 2 is not a critical
point and is not an endpoint.
Q[7]:
b
f (x) = |( x ´ 5)( x + 7)|
Find all critical points and all singular points of f ( x ). You do not have to specify whether
a point is critical or singular.
Q[8]: Suppose f ( x ) is the constant function f ( x ) = 4. What are the critical points and
singular points of f ( x )? What are its local and global maxima and minima?
3.5.2 §§ Finding global maxima and minima

Exercises
§§ Stage 1
Q[1]: Sketch a function f ( x ) such that:
• f ( x ) is defined over all real numbers
• f ( x ) has a global max but no global min.
84

• f ( x ) is always positive
• f ( x ) has no global max and no global min.

• f ( x ) has a global minimum at x = 5
• f ( x ) has a global minimum at x = ´5, too.
§§ Stage 2
Q[4]: f ( x ) = x2 + 6x ´ 10. Find all global extrema on the interval [´5, 5]
2 3
Q[5]: f ( x ) = x ´ 2x2 ´ 30x + 7. Find all global extrema on the interval [´4, 0].
3
3.5.3 §§ Max/min examples

Exercises
§§ Stage 2
For Questions 1 through 3, the quantity to optimize is already given to you as a function of a single variable.
Q[1](˚): Find the global maximum and the global minimum for f ( x ) = x5 ´ 5x + 2 on the
interval [´2, 0].
Q[2](˚): Find the global maximum and the global minimum for f ( x ) = x5 ´ 5x ´ 10 on
the interval [0, 2].
Q[3](˚): Find the global maximum and the global minimum for f ( x ) = 2x3 ´ 6x2 ´ 2 on
the interval [1, 4].
For Questions 4 and 5, you can decide whether a critical point is a local extremum by considering the
derivative of the function.
Q[4](˚): Consider the function h( x ) = x3 ´ 12x + 4. What are the coordinates of the local
maximum of h( x )? What are the coordinates of the local minimum of h( x )?
Q[5](˚): Consider the function h( x ) = 2x3 ´ 24x + 1. What are the coordinates of the local
maximum of h( x )? What are the coordinates of the local minimum of h( x )?
85
For Questions 6 through 13, you will have to find an expression for the quantity you want to optimize as a
function of a single variable.
Q[6](˚): You are in a dune buggy at a point P in the desert, 12 km due south of the nearest
point A on a straight east-west road. You want to get to a town B on the road 18 km east
of A. If your dune buggy can travel at an average speed of 15 km/hr through the desert
and 30 km/hr along the road, towards what point Q on the road should you head to
minimize your travel time from P to B?
A Q B
12 km
Q[7](˚): A closed three dimensional box is to be constructed in such a way that its volume
is 4500 cm3 . It is also specified that the length of the base is 3 times the width of the
base. Find the dimensions of the box that satisfies these conditions and has the minimum
possible surface area. Justify your answer.
Q[8](˚): A closed rectangular container with a square base is to be made from two
different materials. The material for the base costs $5 per square metre, while the
material for the other five sides costs $1 per square metre. Find the dimensions of the
container which has the largest possible volume if the total cost of materials is $72.
Q[9](˚): Find a point X on the positive x–axis and a point Y on the positive y–axis such
that (taking O = (0, 0))
(i) The triangle XOY contains the first quadrant portion of the unit circle x2 + y2 = 1
and
(ii) the area of the triangle XOY is as small as possible.
A complete and careful mathematical justification of property (i) is required.
86
Q[10](˚): A rectangle is inscribed in a semicircle of radius R so that one side of the

rectangle lies along a diameter of the semicircle. Find the largest possible perimeter of
such a rectangle, if it exists, or explain why it does not. Do the same for the smallest
possible perimeter.
Q[11](˚): Find the maximal possible volume of a cylinder with surface area A.*
* Food is often packaged in cylinders, and companies wouldn’t want to waste the metal they are made
out of. So, you might expect the dimensions you find in this problem to describe a tin of, say, cat food.
Read here about why this isn’t the case.
Q[12](˚): What is the largest possible area of a window, with perimeter P, in the shape of
a rectangle with a semicircle on top (so the diameter of the semicircle equals the width of
the rectangle)?
Q[13](˚): Consider an open-top rectangular baking pan with base dimensions x

centimetres by y centimetres and height z centimetres that is made from A square
centimetres of tin plate. Suppose y = px for some fixed constant p.
(a) Find the dimensions of the baking pan with the maximum capacity (i.e., maximum
volume). Prove that your answer yields the baking pan with maximum capacity.
Your answer will depend on the value of p.
(b) Find the value of the constant p that yields the baking pan with maximum capacity
and give the dimensions of the resulting baking pan. Prove that your answer yields
the baking pan with maximum capacity.
§§ Stage 3
Q[14](˚): Let f ( x ) = x x for x ą 0.

(a) Find f 1 ( x ).
(b) At what value of x does the curve y = f ( x ) have a horizontal tangent line?
(c) Does the function f have a local maximum, a local minimum, or neither of these at
the point x found in part (b)?
87
A PPLICATIONS OF DERIVATIVES 3.6 S KETCHING G RAPHS
Q[15](˚): A length of wire is cut into two pieces, one of which is bent to form a circle, the
other to form a square. How should the wire be cut if the area enclosed by the two curves
is maximized? How should the wire be cut if the area enclosed by the two curves is
minimized? Justify your answers.
3.6Ĳ Sketching Graphs
3.6.1 §§ Domain, intercepts and asymptotes
Exercises
§§ Stage 1
Q[1]: Suppose f ( x ) is a function given by
g( x )
f (x) =
x2 ´ 9
where g( x ) is also a function. True or false: f ( x ) has a vertical asymptote at x = ´3.
88
§§ Stage 2
Q[2]: Match the functions f ( x ), g( x ), h( x ), and k ( x ) to the curves y = A( x ) through

y = D ( x?). ? ? ?
f ( x ) = x2 + 1 g( x ) = x2 ´ 1 h( x ) = x2 + 4 k ( x ) = x2 ´ 4
y y
y = B( x )
y = A( x )
2 2
1 1
x x
´2 ´1 1 2 ´2 ´1 1 2
y y
y = C(x)
y = D(x)
2 2
1 1
x x
´2 ´1 1 2 ´2 ´1 1 2
89
Q[3]: Below is the graph of b

y = f (x) = log2 ( x + p)
(a) What is p?
(b) What is b (marked on the graph)?
(c) What is the x-intercept of f ( x )?
Remember log( x + p) is the natural logarithm of x + p, loge ( x + p).
y
x
´8 b ´6 ´5 ´4 ´3 ´2 ´1 1 2
x (2x + 1)( x ´ 7)
Q[4]: Find all asymptotes of f ( x ) = .
3x3 ´ 81
Q[5]: Find all asymptotes of f ( x ) = 103x´7 .
3.6.2 §§ First derivative - increasing or decreasing
Exercises
90
§§ Stage 1
Q[1]: Match each function graphed below to its derivative from the list. (For example,
which function on the list corresponds to A1 ( x )?)
The y-axes have been scaled to make the curve’s behaviour clear, so the vertical scales
differ from graph to graph.
l ( x ) = ( x ´ 2)4 m ( x ) = ( x ´ 2)4 ( x + 2) n ( x ) = ( x ´ 2)2 ( x + 2)2
o ( x ) = ( x ´ 2)( x + 2)3 p ( x ) = ( x + 2)4
y y = B( x )
y
y = A( x )
x
x
´2 2
´2 2
y = C(x)
x
´2 2
y y
y = D(x) y = E( x )
x x
´2 2 ´2 2
91
§§ Stage 2
ex
Q[2](˚): Find the largest open interval on which f ( x ) = is increasing.
x+3
?
x´1
Q[3](˚): Find the largest open interval on which f ( x ) = is increasing.
2x + 4
Q[4](˚): Find the largest open interval on which f ( x ) = 2 arctan( x ) ´ log(1 + x2 ) is

increasing.
3.6.3 §§ Second derivative - concavity

Exercises
§§ Stage 1
Q[1]: On the graph below, mark the intervals where f 2 ( x ) ą 0 (i.e. f ( x ) is concave up)
and where f 2 ( x ) ă 0 (i.e. f ( x ) is concave down).
y
Q[2]: Sketch a curve that is:

• concave up when |x| ą 5,
• concave down when |x| ă 5,
• increasing when x ă 0, and
• decreasing when x ą 0.
92
Q[3]: Suppose f ( x ) is a function whose second derivative exists and is continuous for all
real numbers.
True or false: if f 2 (3) = 0, then x = 3 is an inflection point of f ( x ).
Remark: compare to Question 7
§§ Stage 2
Q[4](˚): Find all inflection points for the graph of f ( x ) = 3x5 ´ 5x4 + 13x.
§§ Stage 3
Questions 5 through 7 ask you to show that certain things are true. Give a clear explanation using concepts
and theorems from the CLP-1 text.
Q[5](˚): Let
x5 5x3
f (x) = + ´ 10x2 + 500x + 1000
20 6
Show that f ( x ) has exactly one inflection point.
Q[6](˚): Let f ( x ) be a function whose first two derivatives exist everywhere, and
f 2 ( x ) ą 0 for all x.
(a) Show that f ( x ) has at most one critical point and that any critical point is an absolute
minimum for f ( x ).
(b) Show that the maximum value of f ( x ) on any finite interval occurs at one of the
endpoints of the interval.
Q[7]: Suppose f ( x ) is a function whose second derivative exists and is continuous for all
real numbers, and x = 3 is an inflection point of f ( x ). Use the Intermediate Value
Theorem to show that f 2 (3) = 0.
Remark: compare to Question 3.
3.6.4 §§ Symmetries
Exercises
93
§§ Stage 1
Q[1]: What symmetries (even, odd, periodic) does the function graphed below have?
y
y = f (x)
Q[2]: What symmetries (even, odd, periodic) does the function graphed below have?
y = f (x)
Q[3]: Suppose f ( x ) is an even function defined for all real numbers. Below is the curve
y = f ( x ) when x ą 0. Complete the sketch of the curve.
94
Q[4]: Suppose f ( x ) is an odd function defined for all real numbers. Below is the curve
y = f ( x ) when x ą 0. Complete the sketch of the curve.
y
§§ Stage 2
Q[5]:
x4 ´ x6
f (x) = 2
ex
Show that f ( x ) is even.
95
Q[6]:
x
f ( x ) = sin( x ) + cos
2
Show that f ( x ) is periodic.
In Questions 7 through 10, find the symmetries of a function from its equation.
Q[7]:
f ( x ) = x4 + 5x2 + cos x3
What symmetries (even, odd, periodic) does f ( x ) have?
Q[8]:
f ( x ) = x5 + 5x4
What symmetries (even, odd, periodic) does f ( x ) have?
Q[9]:
f ( x ) = tan (πx )
What is the period of f ( x )?
§§ Stage 3
Q[10]:
f ( x ) = tan (3x ) + sin (4x )
What is the period of f ( x )?
3.6.5 §§ A checklist for sketching

No exercises for Section 3.6.5
Jump to TABLE OF CONTENTS.
3.6.6 §§ Sketching examples

Exercises
96
§§ Stage 2
?
Q[1](˚): Let f ( x ) = x 3 ´ x.
(a) Find the domain of f ( x ).
(b) Determine the x-coordinates of the local maxima and minima (if any) and intervals
where f ( x ) is increasing or decreasing.
(c) Determine intervals where f ( x ) is concave upwards or downwards, and the x
coordinates of inflection points (if any). You may use, without verifying it, the
formula f 2 ( x ) = (3x ´ 12)(3 ´ x )´3/2 /4.
(d) There is a point at which the tangent line to the curve y = f ( x ) is vertical. Find this
point.
(e) Sketch the graph y = f ( x ), showing the features given in items (a) to (d) above and
giving the ( x, y) coordinates for all points occurring above.
In Questions 2 through 4, you will sketch the graphs of rational functions.
Q[2](˚): Sketch the graph of
x3 ´ 2
f (x) = .
x4
Indicate the critical points, local and absolute maxima and minima, vertical and horizontal
asymptotes, inflection points and regions where the curve is concave upward or down-
ward.
x4
Q[3](˚): The first and second derivatives of the function f ( x ) = are:
1 + x3
4x3 + x6 12x2 ´ 6x5

f 1 (x) = and f 2 (x) =
(1 + x 3 )2 (1 + x 3 )3
Graph f ( x ). Include local and absolute maxima and minima, regions where f ( x ) is in-
creasing or decreasing, regions where the curve is concave upward or downward, and
any asymptotes.
x3
Q[4](˚): The first and second derivatives of the function f ( x ) = are:
1 ´ x2
1 3x2 ´ x4 2 6x + 2x3
f (x) = and f (x) =
(1 ´ x 2 )2 (1 ´ x 2 )3
Graph f ( x ). Include local and absolute maxima and minima, regions where the curve is
concave upward or downward, and any asymptotes.
97
Q[5](˚): The function f ( x ) is defined by

#
ex xă0
f (x) = x 2 +3
3( x +1)
xě0
(a) Explain why f ( x ) is continuous everywhere.

(b) Determine all of the following if they are present:
i. x–coordinates of local maxima and minima, intervals where f ( x ) is increasing or
decreasing;
ii. intervals where f ( x ) is concave upwards or downwards;
iii. equations of any horizontal or vertical asymptotes.
(c) Sketch the graph of y = f ( x ), giving the ( x, y) coordinates for all points of interest
above.
In Questions 6 and 7, you will sketch the graphs of functions with an exponential component. In the next
section, you will learn how to find their horizontal asymptotes, but for now these are given to you.
Q[6](˚): The function f ( x ) and its derivative are given below:
2 2
f ( x ) = (1 + 2x )e´x and f 1 ( x ) = 2(1 ´ x ´ 2x2 )e´x
Sketch the graph of f ( x ). Indicate the critical points, local and/or absolute maxima and
minima, and asymptotes. Without actually calculating the inflection points, indicate on
the graph their approximate location.
Note: lim f ( x ) = 0.
xÑ˘8
2 /2
Q[7](˚): Consider the function f ( x ) = xe´x .
Note: lim f ( x ) = 0.
xÑ˘8
(a) Find all inflection points and intervals of increase, decrease, convexity up, and
convexity down.
2
You may use without proof the formula f 2 ( x ) = ( x3 ´ 3x )e´x /2 .
(b) Find local and global minima and maxima.
(c) Use all the above to draw a graph for f . Indicate all special points on the graph.
In Questions 8 and 9, you will sketch the graphs of functions that have a trigonometric component.
Q[8]: Use the techniques from this section to sketch the graph of f ( x ) = x + 2 sin x.
Q[9](˚):
f ( x ) = 4 sin x ´ 2 cos 2x
Graph the equation y = f ( x ), including all important features. (In particular, find all
local maxima and minima and all inflection points.) Additionally, find the maximum and
minimum values of f ( x ) on the interval [0, π ].
98
c
x+1
3
Q[10]: Sketch the curve y = .
x2
´( x + 2) 2
You may use the facts y1 ( x ) = 5/3 and y 2 ( x ) = 4x + 16x + 10 .
3x ( x + 1)2/3 9x8/3 ( x + 1)5/3
§§ Stage 3
Q[11](˚): A function f ( x ) defined on the whole real number line satisfies the following
conditions
f (0) = 0 f (2) = 2 lim f ( x ) = 0 f 1 ( x ) = K (2x ´ x2 )e´x
xÑ+8
for some positive constant K. (Read carefully: you are given the derivative of f ( x ), not
f ( x ) itself.)
(a) Determine the intervals on which f is increasing and decreasing and the location of
any local maximum and minimum values of f .
(b) Determine the intervals on which f is concave up or down and the x–coordinates of
any inflection points of f .
(c) Determine lim f ( x ).
xÑ´8
(d) Sketch the graph of y = f ( x ), showing any asymptotes and the information
determined in parts (a) and (b).
Q[12](˚): Let f ( x ) = e´x , x ě 0.
(a) Sketch the graph of the equation y = f ( x ). Indicate any local extrema and inflection
points.
(b) Sketch the graph of the inverse function y = g( x ) = f ´1 ( x ).
(c) Find the domain and range of the inverse function g( x ) = f ´1 ( x ).
(d) Evaluate g1 ( 21 ).
Q[13](˚):
(a) Sketch the graph of y = f ( x ) = x5 ´ x, indicating asymptotes, local maxima and
minima, inflection points, and where the graph is concave up/concave down.
(b) Consider the function f ( x ) = x5 ´ x + k, where k is a constant, ´8 ă k ă 8. How
many roots does the function have? (Your answer might depend on the value of k.)
Q[14](˚): The hyperbolic trigonometric functions sinh( x ) and cosh( x ) are defined by
e x ´ e´x e x + e´x
sinh( x ) = cosh( x ) =
2 2
They have many properties that are similar to corresponding properties of sin( x ) and
cos( x ). In particular, it is easy to see that
d d
sinh( x ) = cosh( x ) cosh( x ) = sinh( x ) cosh2 ( x ) ´ sinh2 ( x ) = 1
dx dx
99
A PPLICATIONS OF DERIVATIVES 3.7 L’H ÔPITAL’ S R ULE AND INDETERMINATE FORMS
You may use these properties in your solution to this question.
(a) Sketch the graphs of sinh( x ) and cosh( x ).
(b) Define inverse hyperbolic trigonometric functions sinh´1 ( x ) and cosh´1 ( x ), carefully
specifing their domains of definition. Sketch the graphs of sinh´1 ( x ) and cosh´1 ( x ).
d ! ´1
)
(c) Find cosh ( x ) .
dx
3.7Ĳ L’Hôpital’s Rule and indeterminate forms

Exercises
§§ Stage 1
In Questions 1 to 3, you are asked to give pairs of functions that combine to make indeterminate forms.
Remember that an indeterminate form is indeterminate precisely because its limit can take on a number of
values.
Q[1]: Give two functions f ( x ) and g( x ) with the following properties:

(i) lim f ( x ) = 8
xÑ8
(ii) lim g( x ) = 8
xÑ8
f (x)
(iii) lim = 2.5
xÑ8 g( x )

(i) lim f ( x ) = 8
xÑ8
(ii) lim g( x ) = 8
xÑ8
f (x)
(iii) lim =0
xÑ8 g( x )

(i) lim f ( x ) = 1
xÑ8
(ii) lim g( x ) = 8
xÑ8
(iii) lim [ f ( x )] g( x) = 5
xÑ8
100
§§ Stage 2
x3 ´ e x´1
xÑ1 sin( πx )
log x
Q[5](˚): Evaluate lim . (Remember: in these notes, log means logarithm base e.)
xÑ0+ x
Q[6](˚): Evaluate lim (log x )2 e´x .

xÑ8
Q[7](˚): Evaluate lim x2 e´x .

xÑ8
x ´ x cos x
xÑ0 x ´ sin x
?
x6 + 4x4
Q[9]: Evaluate lim 2 .
xÑ0 x cos x
(log x )2
xÑ8 x
1 ´ cos x
xÑ0 sin2 x
x
xÑ0 sec x
csc x ¨ tan x ¨ ( x2 + 5)
xÑ0 ex
a a
Q[14]: Evaluate lim 2x2 + 1 ´ x2 + x.
xÑ8
sin( x3 + 3x2 )
xÑ0 sin2 x
log( x3 )
xÑ1 x 2 ´ 1
2
e´1/x
xÑ0 x4
xe x
xÑ0 tan(3x )
x2
a
Q[19]: Evaluate lim sin2 x.
xÑ0
101
?
x2
Q[20]: Evaluate lim cos x.
xÑ0
Q[21]: Evaluate lim e x log x .

xÑ0+
h ix
2
Q[22]: Evaluate lim ´ log( x ) .
xÑ0
1 + cx ´ cos x
Q[23](˚): Find c so that lim 2 exists.
xÑ0 ex ´ 1
2
ek sin( x ) ´ (1 + 2x2 )
Q[24](˚): Evaluate lim , where k is a constant.
xÑ0 x4
§§ Stage 3
Q[25]: Suppose an algorithm, given an input with with n variables, will terminate in at
most S(n) = 5n4 ´ 13n3 ´ 4n + log(n) steps. A researcher writes that the algorithm will
terminate in roughly at most A(n) = 5n4 steps. Show that the percentage error involved
in using A(n) instead of S(n) tends to zero as n gets very large. What happens to the
absolute error?
Remark: this is a very common kind of approximation. When people deal with functions
that give very large numbers, often they don’t care about the exact large number–they only
want a ballpark. So, a complicated function might be replaced by an easier function that
doesn’t give a large relative error. If you would like to know more about the ways people
describe functions that give very large numbers, you can read about “Big O Notation” in
Section 3.6.3 of the CLP101 notes.
102
Chapter 4
T OWARDS I NTEGRAL C ALCULUS
4.1Ĳ Introduction to antiderivatives
Exercises
§§ Stage 1
Q[1]: Let f ( x ) be a function with derivative f 1 ( x ). What is the most general antiderivative
of f 1 ( x )?
103
T OWARDS I NTEGRAL C ALCULUS 4.1 I NTRODUCTION TO ANTIDERIVATIVES
Q[2]: On the graph below, the black curve is y = f ( x ). Which of the coloured curves is an
antiderivative of f ( x )?
y
x
f (x)
B( x )
C(x)
A( x )
§§ Stage 2
In Questions 3 through 12, you are asked to find the antiderivative of a function. Phrased like this, we mean
the most general antiderivative. These will all include some added constant. The table after Example 4.1.3
might be of help.
Q[3]: Find the antiderivative of f ( x ) = 3x2 + 5x4 + 10x ´ 9.
3 7
Q[4]: Find the antiderivative of f ( x ) = x ´ 18x4 + x.
5
? 9
Q[5]: Find the antiderivative of f ( x ) = 4 3 x ´ .
2x2.7
1
Q[6]: Find the antiderivative of f ( x ) = ? .
7 x
Q[7]: Find the antiderivative of f ( x ) = e5x+11 .
104
Q[8]: Find the antiderivative of f ( x ) = 3 sin(5x ) + 7 cos(13x ).
Q[9]: Find the antiderivative of f ( x ) = sec2 ( x + 1).
1
Q[10]: Find the antiderivative of f ( x ) = .
x+2
7
3 ´ 3x2
1
Q[12]: Find the antiderivative of f ( x ) =
1 + 25x2
In Questions 13 through 16, you are asked to find a specific antiderivative of a function. In this case, you
should be able to solve for the entire function–no unknown constants floating about.
Q[13]: Find the function f ( x ) with f 1 ( x ) = 3x2 ´ 9x + 4 and f (1) = 10.
Q[14]: Find the function f ( x ) with f 1 ( x ) = cos(2x ) and f (π ) = π.
1
Q[15]: Find the function f ( x ) with f 1 ( x ) = and f (´1) = 0.
x
1 π
Q[16]: Find the function f ( x ) with f 1 ( x ) = ? + 1 and f (1) = ´ .
1´x 2 2
In Questions 17 through 19, we will explore some simple situations where antiderivatives might arise.
Q[17]: Suppose a population of bacteria at time t (measured in hours) is growing at a rate

of 100e2t individuals per hour. Starting at time t = 0, how long will it take the initial
colony to increase by 300 individuals?
Q[18]: Your bank account at time t (measured in years) is growing at a rate of
t
1500e 50
dollars per year. How much money is in your account at time t?

Q[19]: At time t during a particular day, 0 ď t ď 24, your house consumes energy at a
rate of π
0.5 sin t + 0.25
24
kW. (Your consumption was smallest in the middle of the night, and peaked at noon.)
How much energy did your house consume in that day?
105
§§ Stage 3
?
Q[20](˚): Let f ( x ) = 2 sin´1 x and g( x ) = sin´1 (2x ´ 1). Find the derivative of f ( x ) ´
g( x ) and simplify your answer. What does the answer imply about the relation between
f ( x ) and g( x )?
For Questions 21 through 26, you are again asked to find the antiderivatives of certain functions. In general,
finding antiderivatives can be extremely difficult–indeed, it will form the main topic of next semester’s
calculus course. However, you can work out the antiderivatives of the functions below using what you’ve
learned so far about derivatives.
Q[21]: Find the antiderivative of f ( x ) = 2 cos(2x ) cos(3x ) ´ 3 sin(2x ) sin(3x ).
( x2 + 1)e x ´ e x (2x )
Q[22]: Find the antiderivative of f ( x ) = .
( x 2 + 1)2
3
Q[23]: Find the antiderivative of f ( x ) = 3x2 e x .
Q[24]: Find the antiderivative of f ( x ) = 5x sin( x2 ).

Q[25]: Find the antiderivative of f ( x ) = elog x .
7
3 ´ x2
Q[27]: Imagine forming a solid by revolving the parabola y = x2 + 1 around the x-axis,
as in the picture below.
y
y = x2 + 1
Use the method of Example 4.1.7 to find the volume of such an object if the segment of the
parabola that we rotate runs from x = ´H to x = H.
106
Part II
H INTS TO PROBLEMS
107
Hints for Exercises 1.1. — Jump to TABLE OF CONTENTS.
H-2: The tangent line to a curve at point P passes through P.
H-3: Try drawing tangent lines to the following curves, at the given points P:
y
y y = f (x) y P
y = f (x)
x
P
x x
P y = f (x)

H-3: Where did you start, and where did you end?
H-4: Is the object falling faster and faster, slower and slower, or at a constant rate?
H-5: Slope is change in vertical component over change in horizontal component.
H-6: Sign of velocity gives direction of motion: the velocity is positive at time t if s(t) is
increasing at time t.
H-7: Velocity is distance over time.
? ? ?
a´b2
H-8: Use that c = c ¨ ?aa+
a´b a´b b
+b
= ?
c( a+b)
.

H-2: Consider the difference between a limit and a one-sided limit.
H-3: Pay careful attention to which limits are one-sided and which are not.
H-5: The function doesn’t have to be continuous.
H-6: See Question 5
H-7: See Question 5
H-8: What is the relationship between the limit and the two one-sided limits?
H-9: What is the relationship between the limit and the two one-sided limits?
H-14: What are the one-sided limits?
1
H-16: Think about what it means that x does not appear in the function f ( x ) = .
10
109
H-17: We only care about what happens really, really close to x = 3.

H-2: Try to make two functions with factors that will cancel.
H-3: Try to make g( x ) cancel out.
H-5: See Questions 2, 3, and 4.
H-6: Find the limit of the numerator and denominator separately.
H-7: Break it up into smaller pieces, evaluate the limits of the pieces.
4x ´ 2
H-8: First find the limit of the “inside” function, .
x+2
H-9: Is cos(´3) zero?
H-10: Expand, then simplify.
H-14: Try the simplest method first.
H-15: Factor the denominator.
H-16: Factor the numerator and the denominator.
H-17: Factor the numerator.
H-18: Simplify first by factoring the numerator.
H-19: The function is a polynomial.
H-20: Multiply
? both the numerator and the denominator by the conjugate of the
numerator, x2 + 8 + 3.
H-21: Multiply
? both the
? numerator and the denominator by the conjugate of the
numerator, x + 7 + 11 ´ x.
H-22: Multiply
? both the
numerator, x + 2 + 4 ´ x.
H-23: Multiply
? both the
numerator, x ´ 2 + 4 ´ x.
H-24: Multiply both
? the numerator and the denominator by the conjugate of the
denominator, 2 + 5 ´ t.
H-25: Consider the factors x2 and cos 3x separately. Review the squeeze theorem.

H-26: Look for a reason to ignore the trig. Review the squeeze theorem.
H-27: As in the previous questions, we want to use the Squeeze Theorem. If x ă 0, then
´x is positive, so x ă ´x. Use this fact when you bound your expressions.
H-28: Factor the numerator.
H-29: Factor the denominator; pay attention to signs.
110
H-30: First find the limit of the “inside” function.
H-31: Factor; pay attention to signs.
H-32: Look for perfect squares
H-33: Think about what effect changing d has on the function x5 ´ 32x + 15.
H-34: There’s an easy way.
H-35: What can you do to safely ignore the sine function?
H-36: Factor
H-37: If you’re looking at the hints for this one, it’s probably easier than you think.
H-38: You’ll want to simplify this, since t = 12 is not in the domain of the function. One
way to start your simplification is to add the fractions in the numerator by finding a
common denominator.
|x|
H-39: If you’re not sure how behaves, try plugging in a few values of x, like x = ˘1
x
and x = ˘2.
|X|
H-40: Look to Question 39 to see how a function of the form behaves.
X
H-41: Is anything weird happening to this function at x = 0?
H-42: Use the limit laws.
H-43: The denominator goes to zero; what must the numerator go to?
H-45: Try plotting points. If you can’t divide by f ( x ), take a limit.
H-46: There is a close relationship between f and g. Fill in the following table:
f (x)
x f (x) g( x )
g( x )
´3
´2
´1
´0
1
2
3
s (1 + h ) ´ s (1)
H-47: Velocity of white ball when t = 1 is lim .
hÑ0 h
H-49: When you’re evaluating lim f ( x ), you’re only considering values of x that are less
xÑ0´
than 0.
H-50: When you’re considering lim f ( x ), you’re only considering values of x that are
xÑ´4´
less than ´4.
111
When you’re considering lim f ( x ), think about the domain of the rational function in
xÑ´4+
the top line.

H-1: It might not look like a traditional polynomial.
H-2: The degree of the polynomial matters.
H-3: What does a negative exponent do?
H-4: You can think about the behaviour of this function by remembering how you first
learned to describe exponentiation.
H-5: The exponent will be a negative number.
H-6: What single number is the function approaching?
H-7: The highest-order term dominates when x is large.
H-8: Factor the highest power of x out of both the numerator and the denominator. You
can factor through square roots (carefully).
? ?
H-9: Multiply and divide by the conjugate, x2 + 5x + x2 ´ x.
H-10: Divide both the numerator and the denominator by the highest power of x that is
in the denominator.
?
Remember that is defined to be the positive square root. Consequently, if x ă 0, then
?
2
x , which is positive, is not the same as x, which is negative.
H-11: Factor out the highest power of the denominator.
? ?
H-12: The conjugate of ( x 2 + x ´ x ) is ( x2 + x + x ).
?
x2 + x + x
Multiply by 1 = ? to coax your function into a fraction.
x2 + x + x
in the denominator.
in the denominator.
in the denominator.
H-16: Divide both the numerator and the denominator by x (which is the largest power
of x in the denominator). In the numerator, move the resulting factor of 1/x inside the
two roots. Be careful about the signs when you do so. Even and odd roots behave
differently– see Question 10.
in the denominator.
112
H-18: Divide both the numerator and the denominator
? by the highest power of x that is
2
in the denominator. It is not always true that x = x.
H-19: Simplify.
H-20: What is a simpler version of |x| when you know x ă 0?
H-22: Divide both the numerator ? by the highest power of x that is
? and the denominator
in the denominator. When is x = x, and when is x = ´x?
in the denominator. Pay careful attention to signs.
?
H-24: Multiply and divide the expression by its conjugate, n2 + 5n + n .
H-25: Consider what happens to the function as a becomes very, very small. You
shouldn’t need to do much calculation.
H-26: Since x = 3 is not in the domain of the function, we need to be a little creative. Try
simplifying the function.
H-27: This is a bit of a trick question. Consider what happens to a rational function as
x Ñ ˘8 in each of these three cases:
• the degree of the numerator is smaller than the degree of the denominator,
• the degree of the numerator is the same as the degree of the denominator, and
• the degree of the numerator is larger than the degree of the denominator.
H-28: We tend to conflate “infinity” with “some really large number.”

H-1: Try a repeating pattern.
H-2: f is my height.
H-3: The intermediate value theorem only works for a certain kind of function.
H-7: Compare what is given to you to the definition of continuity.
H-8: Compare what is given to you to the definition of continuity.
H-9: What if the function is discontinuous?
H-10: What is h(0)?
H-11: Use the definition of continuity.
H-12: If this is your password, you might want to change it.
H-13: Find the domain: when is the denominator zero?
H-14: When is the denominator zero? When is the argument of the square root negative?
H-15: When is the denominator zero? When is the argument of the square root negative?
113
H-16: There are infinitely many points where it is not continuous.
H-17: x = c is the important point.
H-18: The important place is x = 0.
H-19: The important point is x = c.
H-20: The important point is x = 2c.
H-21: Consider the function f ( x ) = sin x ´ x + 1.
H-22: Consider the function f ( x ) = 3x ´ x2 , and how it relates to the problem and the
IVT.
H-23: Consider the function 2 tan x ´ x ´ 1 and its roots.
a
H-24: Consider the function f ( x ) = cos(πx ) ´ sin(2πx ) ´ 1/2, and be careful about
where it is continuous.
H-25: Consider the function f ( x ) = 1/ cos2 (πx ) ´ x ´ 32 , paying attention to where it is
continuous.
H-26: We want f ( x ) to be 0; 0 is between a positive number and a negative number. Try
evaluating f ( x ) for some integer values of x.
?
H-27: 3 7 is the value where x3 = 7.
H-28: You need to consider separately the cases where f ( a) ă g( a) and f ( a) = g( a). Let
h( x ) = f ( x ) ´ g( x ). What is h(c)?

H-2: You can use (a) to explain (b).
H-3: Your calculations for slope of the secant lines will all have the same denominators;
to save yourself some time, you can focus on the numerators.
H-4: You can do this by calculating several secant lines. You can also do this by getting
out a ruler and trying to draw the tangent line very carefully.
H-5: There are many possible values for Q and R.
H-6: A line with slope 0 is horizontal.

H-1: What are the properties of f 1 when f is a line?
H-2: Be very careful not to confuse f and f 1 .
H-3: Be very careful not to confuse f and f 1 .
H-5: The slope has to look “the same” from the left and the right.
114
H-6: Use the definition of the derivative, and what you know about limits.
H-7: Consider continuity.
H-8: Look at the definition of the derivative. Your answer will be a fraction.
H-9: You need a point (given), and a slope (derivative).
H-10: You’ll need to add some fractions.
H-11: You don’t have to take the limit from the left and right separately–things will
cancel nicely.
H-12: You might have to add fractions
H-14: Your limit should be easy.
H-15: add fractions
H-16: For f to be differentiable at x = 2, two things must be true: it must be continuous
at x = 2, and the derivative from the right must equal the derivative from the left.
f ( x ) to the definition
H-17: After you plug in ? ? of a derivative, you’ll want to multiply and
divide by the conjugate 1 + x + h + 1 + x.
H-18: From Section 1.2, compare the definition of velocity to the definition of a
derivative. When you’re finding the derivative, you’ll need to cancel a lot on the
numerator, which you can do by expanding the polynomials.
and right. The fact that f (0) = 0 is useful
H-19: You’ll need to look at limits from the left ?
for your computation. Recall that if x ă 0 then x2 = |x| = ´x.
H-20: You’ll need to look at limits from the left and right. The fact that f (0) = 0 is useful
for your computation.
H-23: There’s lots of room between 0 and 18 ; see what you can do with it.
H-24: Set up your usual limit, then split it into two pieces
H-25: You don’t need the definition of the derivative for a line.
H-26: A generic point on the curve has coordinates (α, α2 ). In terms of α, what is the
equation of the tangent line to the curve at the point (α, α2 )? What does it mean for
(1, ´3) to be on that line?
H-27: Remember for a constant n,
$
& 0 ną0
n
lim h = 1 n=0
hÑ0
DNE n ă 0
%
115
H-1: Think about units.
H-8: There are 360 degrees in one rotation.
H-9: P1 (t) was discussed in Question 7.

H-1: Look at the Sum rule
H-2: Try an example, like f ( x ) = g( x ) = x.
H-3: Simplify
H-4: g( x ) = f ( x ) + f ( x ) + f ( x )
H-5: Use linearity and the known derivatives of x2 and x1/2 .
d ?
H-6: You have already seen dx t xu.
H-7: The equation of a line can be determined using a point, and the slope. The
derivative of x3 can be found by writing x3 = ( x )( x2 ).
H-8: Be careful to distinguish between speed and velocity.
H-10: How do you take care of that power?
H-11: You know how to take the derivative of a reciprocal; this might be faster than using
the quotient rule.
H-12: Population growth is rate of change of population.
H-14: Interpret it as a derivative that you know how to compute.
H-15: The answer is not 10 square metres per second.
H-16: You don’t need to know g(0) or g1 (0).
No exercises for Section 2.5. — Jump to TABLE OF CONTENTS

H-1: Check signs
H-2: Read Lemma 2.6.9 carefully.
H-3: First, factor an x out of the derivative. What’s left over looks like a quadratic
equation, if you take x2 to be your variable, instead of x.
1
H-4: t = t´1
116
3 be confused by the role reversal of x and y: x is just the name of
H-5: First simplify. Don’t
1
the function 2y + y ¨ y , which is a function of the variable y. You are to differentiate
with respect to y.
?
H-6: x = x1/2
H-8: You don’t need to multiply through.
H-9: You can use the quotient rule.
H-13: There are two pieces of the given function that could cause problems.
?
H-14: 3 x = x1/3
H-15: Simplify first
H-17: Let m be the slope of such a tangent line, and let P1 and P2 be the points where the
tangent line is tangent to the two curves, respectively. There are three equations m fulfils:
it has the same slope as the curves at the given points, and it is the slope of the line
passing through the points P1 and P2 .
H-18: A line has equation y = mx + b, for some constants m and b. What has to be true
for y = mb + x to be tangent to the first curve at the point x = α, and to the second at the
point x = β?
H-19: Compare this to one of the forms given in the text for the definition of the
derivative.

H-1: Two of the functions are the same.
H-2:
H-3: When can you use the power rule?
H-4: What is the shape of the curve e ax , when a is a positive consant?
H-5: Quotient rule
H-6: e2x = (e x )2
H-7: e a+ x = e a e x
H-8: Figure out where the derivative is positive.
1
H-9: e´x = ex
H-10: Product rule will work nicely here. Alternately, review the result of Question 6.
H-11: To find the sign of a product, compare the signs of each factor. The function et is
always positive.
H-12: After you differentiate, factor out e x .
H-13: Simplify
117
H-14: In order to be differentiable, a function should be continuous. To determine the
differentiability of the function at x = 1, use the definition of the derivative.

H-1: A horizontal tangent line is where the graph appears to “level off.”
H-2: You are going to mark there points on the sine graph where the graph is the
steepest, going up.
H-3: You need to memorize the derivatives of sine, cosine, and tangent.
H-4: There are infinitely many values. You need to describe them all.
H-5: Simplify first.
H-6: The identity won’t help you.
H-8: Quotient rule
H-11: Use an identity.
H-12: How can you move the negative signs to a location that you can more easily deal
with?
H-13: Apply the quotient rule.
H-14: The only spot to worry about is when x = 0. For f ( x ) to be differentiable, it must
be continuous, so first find the value of b that makes f continuous at x = 0. Then, find the
value of a that makes the derivatives from the left and right of x = 0 equal to each other.
derivative.
derivative.
derivative.
sin θ
H-19: tan θ =
cos θ
H-20: In order for a derivative to exist, the function must be continuous, and the
derivative from the left must equal the derivative from the right.
H-21: There are infinitely many places where it does not exist.
f ( h ) ´ f (0)
H-27: You can set up the derivative using the limit definition: f 1 (0) = lim . If
hÑ0 h
the limit exists, it gives you f 1 (0); if the limit does not exist, you conclude f 1 (0) does not
exist.
To evaluate the limit, recall that when we differentiated sine, we learned that for h near 0,
sin h
cos h ď ď1
h
118
"
x xě0
H-28: Recall |x| = . To determine whether h( x ) is differentiable at x = 0,
´x x ă 0
use the definition of the derivative.
H-29: To decide whether the function is differentiable, use the definition of the derivative.
sin x
H-30: In this chapter, we learned lim = 1. If you divide the numerator and
xÑ0 x
denominator by x5 , you can make use of this knowledge.

H-1: For parts (a) and (b), remember the definition of a derivative:
dK K (U + h ) ´ K (U )
= lim .
dU hÑ0 h
When h is positive, U + h is an increased urchin population; what is the sign of
K (U + h ) ´ K (U ) ?
For part (c), use the chain rule!
H-2: Remember that Leibniz notation suggests fractional cancellation.
H-3: If g( x ) = cos x and h( x ) = 5x + 3, then f ( x ) = g(h( x )). So we apply the chain rule,
with “outside” function cos x and “inside” function 5x + 3.
H-4: You can expand this into a polynomial, but it’s easier to use the chain rule. If
g( x ) = x5 , and h( x ) = x2 + 2, then f ( x ) = g(h( x )).
H-5: You can expand this into a polynomial, but it’s easier to use the chain rule. If
g(k) = k17 , and h(k) = 4k4 + 2k2 + 1, then T (k) = g(h(k )).
? x2 + 1
H-6: If we define g( x ) = x and h( x ) = , then f ( x ) = g(h( x )).
x2 ´ 1
d ? d ! 1/2 ) 1 ´1/2 1
To differentiate the square root function: t xu = x = x = ? .
dx dx 2 2 x
H-7: You’ll need to use the chain rule twice.
H-8: Use the chain rule.
1 ?
H-12: Recall 2 = x´2 and x2 ´ 1 = ( x2 ´ 1)1/2 .
x
H-14: If we let g( x ) = sec x and h( x ) = e2x+7 , then f ( x ) = g(h( x )), so by the chain rule,
f 1 ( x ) = g1 (h( x )) ¨ h1 ( x ). However, in order to evaluate h1 ( x ), we’ll need to use the chain
rule again.
H-15: What trig identity can you use to simplify the first factor in the equation?
119
H-16: Velocity is the derivative of position with respect to time. In this case, the velocity
of the particle is given by s1 (t).
H-17: The slope of the tangent line is the derivative.
You’ll need to use the chain rule twice.
H-18: Start with the product rule, then use the chain rule to differentiate e4x .
H-19: Start with the quotient rule; you’ll need the chain rule only to differentiate e3x .
H-20: More than one chain rule needed here.
H-21: More than one chain rule application is needed here.
H-24: What rule do you need, besides chain? Also, remember that cos2 x = [cos x ]2 .
H-27: The product of two functions is zero exactly when at least one of the functions is
zero.
1
H-28: If t ě 1, then 0 ă t ď 1.
H-29: The notation cos3 (5x ´ 7) means [cos(5x ´ 7)]3 . So, if g( x ) = x3 and
h( x ) = cos(5x ´ 7), then g(h( x )) = [cos(5x + 7)]3 = cos3 (5x + 7).
H-30: In Example 2.6.6, we generalized the product rule to three factors:
d
t f ( x ) g ( x ) h ( x )u = f 1 ( x ) g ( x ) h ( x ) + f ( x ) g1 ( x ) h ( x ) + f ( x ) g ( x ) h1 ( x )
dx
This isn’t strictly necessary, but it will simplify your computations.

H-31: At time t, the particle is at the point x (t), y(t) , with x (t) = cos t and y(t) = sin t.
Over time, the particle traces out a curve; let’s call that curve y = f ( x ). Then

y(t) = f x (t) , so the slope of the curve at the point x (t), y(t) is f 1 x (t) . You are to

determine the values of t for which f 1 x (t) = ´1.
2
H-32: Set f ( x ) = e x+ x and g( x ) = 1 + x. Compare f (0) and g(0), and compare f 1 ( x ) and
g1 ( x ).
H-33: If sin 2x and 2 sin x cos x are the same, then they also have the same derivatives.
H-34: This is a long, nasty problem, but it doesn’t use anything you haven’t seen before.
Be methodical, and break the question into as many parts as you have to. At the end, be
proud of yourself for your problem-solving abilities and tenaciousness!
H-35: To sketch the curve, you can start by plotting points. Alternately, consider x2 + y.

H-1: Each speaker produces 3dB of noise, so if P is the power of one speaker,
P

3 = V ( P) = 10 log10 S . Use this to find V (10P) and V (100P).
120
H-2: The question asks you when A(t) = 2000. So, solve 2000 = 1000et/20 for t.
H-3: What happens when cos x is a negative number?
H-4: There are two easy ways: use the chain rule, or simplify first.
H-5: There are two easy ways: use the chain rule, or simplify first.
H-6: Don’t be fooled by a common mistake: log( x2 + x ) is not the same as log( x2 ) + log x.
H-7: Use the base-change formula to convert this to natural logarithm (base e).
H-10: Use the chain rule twice.
H-11: You’ll need to use the chain rule twice.
H-13: Use the chain rule to differentiate.
H-14: You can differentiate this by using the chain rule several times.
H-15: Using logarithm rules before you differentiate will make this easier.
H-16: Using logarithm rules before you differentiate will make this easier.
H-17: First, differentiate using the chain rule and any other necessary rules. Then, plug
in x = 2.
d x
H-18: In the text, you are given the derivative a , where a is a constant.
dx
H-19: You’ll need to use logarithmic differentiation. Set g( x ) = log( f ( x )), and find g1 ( x ).
d x
Then, use that to find f 1 ( x ). This is the method used in the text to find a .
dx
log a
H-20: Use Question 19 and the base-change formula, logb ( a) = .
log b
H-21: To make this easier, use logarithmic differentiation. Set g( x ) = log( f ( x )), and find
d x
g1 ( x ). Then, use that to find f 1 ( x ). This is the method used in the text to find a , and
dx
again in Question 19.
H-22: To make this easier, use logarithmic differentiation. Set g( x ) = log( f ( x )), and find
d x
g1 ( x ). Then, use that to find f 1 ( x ). This is the method used in the text to find a , and
dx
again in Question 19.
H-23: It’s not going to come out nicely, but there’s a better way than blindly applying
quotient and product rules, or expanding giant polynomials.
d x
Then, use that to find f 1 ( x ). This is the method used in the text to find a , and again in
dx
Question (19).
121
d x
dx
Question (19).
d x
dx
Question (19).
H-27: You’ll need to use logarithmic differentiation. Differentiate log( f ( x )), then solve
d x
for f 1 ( x ). This is the method used in the text to find a .
dx
d x
dx
d x
dx
d ! )
H-30: Evaluate log [ f ( x )] g( x) .
dx
H-31: Differentiate y = log( f ( x )). When is the derivative equal to zero?

H-1: Where did the y1 come from?
H-2: The three points to look at are (0, ´4), (0, 0), and (0, 4). What does the slope of the
tangent line look like there?
H-3: A function must pass the vertical line test: one input cannot result in two different
outputs.
H-4: Remember that y is a function of x. Use implicit differentiation, then collect all the
dy dy
terms containing on one side of the equation to solve for .
dx dx
H-5: Differentiate implicitly, then solve for y1 .
H-6: Remember that y is a function of x. You can determine explicitly the values of x for
which y( x ) = 1.
H-8: Plug in y = 0 at a strategic point in your work to simplify your computation.
H-10: Plug in y = 0 at a strategic point in your work to simplify your computation.
H-11: If the tangent line has slope y1 , and it is parallel to y = x, then y1 = 1.
H-12: You don’t need to solve for y1 in general: only at a single point.
H-13: After you differentiate implicitly, get all the terms containing y1 onto one side so
you can solve for y1 .
122
dy
H-14: You don’t need to solve for dx for all values of x–only when y = 0.

H-1: Remember that only certain numbers can come out of sine and cosine, but any
numbers can go in.
H-2: What is the range of the arccosine function?
H-3: A one–to–one function passes the horizontal line test. To graph the inverse of a
function, reflect it across the line y = x.
H-4: Your answer will depend on a. The arcsine function alone won’t give you every
value.
H-5: In order for x to be in the domain of f , you must be able to plug x into both arcsine
and arccosecant.
H-6: For the domain of f , remember the domain of arcsine is [´1, 1].
H-7: The domain of arccos(t) is [´1, 1], but you also have to make sure you aren’t
dividing by zero.
d 1
H-8: tarcsec xu = ? , and the domain of arcsec x is |x| ě 1.
dx |x| x2 ´ 1
H-9: The domain of arctan( x ) is all real numbers.
?
H-10: The domain of arcsin x is [´1, 1], and the domain of x is x ě 0.
H-11: This occurs only once.
H-12: The answer is a very simple expression.
H-13: chain rule
H-16: You can simplify the expression before you differentiate to remove the
trigonometric functions. If arctan x = θ, then fill in the sides of the triangle below using
the definition of arctangent and the Pythagorean theorem:
With the sides labeled, you can figure out sin (arctan x ) = sin (θ ).
H-17: You can simplify the expression before you differentiate to remove the
trigonometric functions. If arcsin x = θ, then fill in the sides of the triangle below using
the definition of arctangent and the Pythagorean theorem:
123
θ
With the sides labeled, you can figure out cot (arcsin x ) = cot (θ ).
H-18: What is the slope of the line y = 2x + 9?
H-19: Differentiate using the chain rule.

H-20: If g(y) = f ´1 (y), then f ( g(y)) = f f ´1 (y) = y. Differentiate this last equality
using the chain rule.
H-21: To simplify notation, let g(y) = f ´1 (y). Simplify and differentiate g( f ( x )).
H-22: To simplify notation, let g(y) = f ´1 (y). Simplify and differentiate g( f ( x )).
H-23: Use logarithmic differentiation.
H-24: Where are those functions defined?
derivative.
H-26: f ´1 (7) is the number y that satisfies f (y) = 7.
H-27: If f ´1 (y) = 0, that means f (0) = y. So, we’re looking for the number that we plug
into f ´1 to get 0.
H-28: As usual, after you differentiate implicitly, get all the terms containing y1 onto one
side of the equation, so you can factor out y1 .

H-1: How long would it take the caribou to travel 5000 km, travelling at its top speed?
H-2: Let f ( x ) be the position of the crane, where x is the hour of the day.
H-3: For an example, look at Figure 2.13.2.
H-4: How does this question differ from the statement of the mean value theorem?
H-5:
H-6: Where is f ( x ) differentiable?
H-7: To use Rolle’s Theorem, you will want two values where the function is zero. If
you’re stuck finding one of them, think about when x2 ´ 2πx is equal to zero.
H-11: To show that there are exactly n roots, you need to not only show that n exist, but
also that there are not more than n.
124
H-12: To show that there are exactly n roots, you need to not only show that n exist, but
also that there are not more than n. If you can’t explicitly find the root(s), you can use the
intermediate value theorem to show they exist.
ˇ
H-13: If f ( x ) = 0, then |x3 | = ˇsin x5 ˇ ď 1. When |x| ă 1, is cos( x5 ) positive or negative?
ˇ
H-14: Let f ( x ) = e x ´ 4 cos(2x ), and use Rolle’s Theorem. What is the interval where
f ( x ) can have a positive root?
H-15: For (b), what does Rolle’s Theorem tell you has to happen in order for f ( x ) to have
more than one root in [´1, 1]?
H-16: Since f ( x ) = e x is a continuous and differentiable function, the MVT promises that
there exists some number c such that
f ( T ) ´ f (0)
f 1 (c) = .
T
Find that c, in terms of T.
H-17: Let f ( x ) = arcsec x + arccsc c ´ C. What is f 1 ( x )?
H-18: Show that f is differentiable by showing that f 1 ( x ) exists for every x. Then, the
Mean Value Theorem applies. What is the largest f 1 ( x ) can be, for any x? If f (100) ă 100,
what does the MVT tell you must be true of f 1 (c) for some c?
H-19: In order for f ´1 ( x ) to be defined over an interval, f ( x ) must be one–to–one over
that interval.
H-20: In order for f ´1 ( x ) to be defined over an interval, f ( x ) must be one–to–one over
that interval.
H-21: Let h( x ) = f ( x ) ´ g( x ). What does the Mean Value Theorem tell you about the
derivative of h?
H-22: Rolle’s Theorem relates the roots of a function to the roots of its derivative.
H-23: To show that there are exactly n distinct roots, you need to not only show that n
exist, but also that there are not more than n.

H-1: If you know the first derivative, this should be easy.
H-2: Exactly two of the statements must be true.
H-3: Use factorials, as in Example 2.14.2.
H-4: The problem isn’t with any of the algebra.
d 1
H-5: Recall log x = .
dx x
d 1
H-6: Recall tarctan xu = = (1 + x2 )´1 .
dx 1 + x2
125
H-7: Use implicit differentiation.
H-8: The acceleration is given by s2 (t).
H-9: Remember to use the chain rule.
H-10: h1 (t) gives the velocity of the particle, and h2 (t) gives its acceleration–the rate the
velocity is changing.
H-11: h1 (t) gives the velocity of the particle, and h2 (t) gives its acceleration–the rate the
velocity is changing. Be wary of signs–as in legends, they may be misleading.
H-12: You don’t need to solve for y2 in general–only when x = y = 0. To do this, you also
need to find y1 at the point (0, 0).
H-13: To show that two functions are unequal, you can show that one input results in
different outputs.
H-14: Only one of the curves could possibly represent y = f ( x ).
d x
H-15: Remember t2 u = 2x log 2.
dx
H-16: Differentiate a few times until you get zero, remembering that a, b, c, and d are all
constants.
H-17: Use a similar method to Question 32, Section 2.9.
H-18: For (b), you know a point where the curve and tangent line intersect, and you
know what the tangent line looks like. What do the derivatives tell you about the shape
of the curve?
H-19: Review Pascal’s Triangle.
H-20: Rolle’s Theorem relates the roots of a function to the roots of its derivative. So, the
fifth derivative tells us something about the fourth, the fourth derivative tells us
something about the third, and so on.
H-21: You’ll want to use Rolle’s Theorem, but the first derivative won’t be very
tractable–use the idea behind Question 20.
H-22: You can re-write this function as a piecewise function, with branches x ě 0 and
x ă 0. To figure out the derivatives at x = 0, use the definition of a derivative.

H-1: Is the velocity changing at t = 2?
H-2: The acceleration (rate of change of velocity) is constant.
H-3: Remember the difference between speed and velocity.
H-4: How is this different from the wording of Question 3?
126
H-5: The equation of an object falling from rest on the earth is derived in Example 3.1.2.
It would be difficult to use exactly the version given for s(t), but using the same logic,
you can find an equation for the height of the flower pot at time t.
m
H-6: Remember that a falling object has an acceleration of 9.8 s2
.
H-7: Acceleration is constant, so finding a formula for the distance your keys have
travelled is a similar problem to finding a formula for something falling.
H-8: See Example 3.1.3.
H-9: Be careful to match up the units.
H-10: Think about what it means for the car to decelerate at a constant rate. You might
also review Question 2.
H-11: Let a be the acceleration of the shuttle. Start by finding a, then find the position
function of the shuttle.
H-12: Review Example 3.1.2, but account for the fact that your initial velocity is not zero.
H-13: Be very careful with units. The acceleration of gravity you’re used to is 9.8 metres
per second squared, so you might want to convert 325 kpm to metres per second.
m
H-14: Since gravity alone brings it down, its acceleration is a constant ´9.8 s2
.
H-15: First, find an equation for a(t), the acceleration of the car, noting that a1 (t) is
constant. Then, use this to find an equation for the velocity of the car. Be careful about
seconds versus hours.
H-16: We recommend using two different functions to describe your height:
h1 (t) while you are in the air, not yet touching the trampoline, and
h2 (t) while you are in the trampoline, going down.
Both h1 (t) and h2 (t) are quadratic equations, since your acceleration is constant over
both intervals, but be very careful about signs.
H-17: First, find an expression for the speed of the object. You can let v0 be its velocity at
time t = 0.

H-1: If you know P, you can figure out Q.
H-2: Since the point moves along the unit circle, we know that x2 + y2 = 1, where x and
y are functions of time.
H-3: You’ll need some implicit differentiation: what should your variable be?
Example 3.2.3 shows how to work with percentage rate of change.
H-4: For (b), refer to Example 3.2.3 for percentage rate of change.
H-5: Pay attention to direction, and what it means for the sign (plus/minus) of the
velocities of the particles.
127
H-6: You’ll want to think about the difference in the y-coordinates of the two particles.
H-7: Draw a picture, and be careful about signs.
H-8: You’ll want to think about the difference in height of the two snails.
H-9: The length of the ladder is changing.

H-10: If a trapezoid has height h and (parallel) bases b1 and b2 , then its area is h b1 +2 b2 .
To figure out how wide the top of the water is when the water is at height h, you can cut
the trapezoid up into a rectangle and two triangles, and make use of similar triangles.
H-11: Be careful with units. One litre is 1000 cm3 , which is not the same as 10 m3 .
H-12: You, the rocket, and the rocket’s original position form a right triangle.
H-13: Your picture should be a triangle.
H-14: Let θ be the angle between the two hands. Using the Law of Cosines, you can get
dθ
an expression for D in terms of θ. To find , use what you know about how fast clock
dt
hands move.
H-15: The area in the annulus is the area of the outer circle minus the area of the inner
circle.
4
H-16: The volume of a sphere with radius R is πr3 .
3
H-17: The area of a triangle is half its base times its height. To find the base, split the
triangle into two right triangles.
H-18: The easiest way to figure out the area of the sector of an annulus (or a circle) is to
figure out the area of the entire annulus, then multiply by what proportion of the entire
1
annulus the sector is. For example, if your sector is 10 of the entire annulus, then its area
1
is 10 of the area of the entire annulus. (See Section A.4 to see how this works out for
circles.)
H-19: Think about the ways in which this problem is similar to and different from
Example 3.2.6 and Question 18.
H-20: The volume of a cone with height h and radius r is 31 πr2 h. Also, one millilitre is the
same as one cubic centimetre.
H-21: If you were to install the buoy, how would you choose the length of rope? For
which values of θ do sin θ and cos θ have different signs? How would those values of θ
look on the diagram?
H-22: At both points of interest, the point is moving along a straight line. From the
diagram, you can figure out the equation of that line.
For the question “How fast is the point moving?” in part (b), remember that the velocity
of an object can be found by differentiating (with respect to time) the equation that gives
the position of the object. The complicating factors in this case are that (1) the position of
128
our object is not given as a function of time, and (2) the position of our object is given in
two dimensions, not one.
H-23: (a) Since the perimeter of the cross section of the bottle does not change, p (the
perimeter of the ellipse) is the same as the perimeter of the circle of radius 5.
(b) The volume of the bottle will be the area of its cross section times its height. This is
always the case when you have some two-dimensional shape, and turn it into a
three-dimensional object by “pulling” the shape straight up. (For example, you can think
of a cylinder as a circle that has been “pulled” straight up. To understand why this
formula works, think about what is means to measure the area of a shape in square
centimetres, and the volume of an object in cubic centimetres.)
db
(c) You can use what you know about a and the formula from (a) to find b and . Then
dt
use the formula from (b).
H-24: If A = 0, you can figure out C and D from the relationship given.
Hints for Exercises 3.3.1. — Jump to TABLE OF CONTENTS.

H-1: Review the definition of a differential equation at the beginning of this section.
H-2: You can test whether a given function solves a differential equation by substituting
the function into the equation.
H-3: Solve 0 = Ce´kt for t.
H-4: No calculus here–just a review of the algebra of exponentials.
H-5: Use Theorem 3.3.2.
H-6: From the text, we see the half-life of Carbon-14 is 5730 years. A microgram (µg) is
one-millionth of a gram, but you don’t need to know that to solve this problem.
H-7: The quantity of Radium-226 in the sample at time t will be Q(t) = Ce´kt for some
positive constants C and k. You can use the given information to find C and e´k .
In the following work, remember we use log to mean natural logarithm, loge .
H-8: The fact that the mass of the sample decreases at a rate proportional to its mass tells
us that, if Q(t) is the mass of Polonium-201, the following differential equation holds:
dQ
= ´kQ(t)
dt
where k is some positive constant. Compare this to Theorem 3.3.2.
H-9: The amount of Radium-221 in a sample at time t will be Q(t) = Ce´kt for some
positive constants C and k. You can leave C as a variable–it’s the original amount in the
sample, which isn’t specified. What you want to find is the value of t such that
Q(t) = 0.0001Q(0) = 0.0001C.
H-10: You don’t need to know the original amount of Polonium-210 in order to answer
this question: you can leave it as some constant C, or you can call it 100%.
129
H-11: Try to find the most possible and least possible remaining Uranium-232, given the
bounds in the problem.

H-1: You can refer to Corollary 3.3.8, but you can also just differentiate the various
dT
proposed functions and see whether, in fact, is the same as 5[ T ´ 20].
dt
H-2: From Newton’s Law of Cooling and Corollary 3.3.8, the temperature of the object
will be
T (t) = [ T (0) ´ A]eKt + A
where A is the ambient temperature, T (0) is the initial temperature of the copper, and K
is some constant.
H-3: What is lim eKt when K is positive, negative, or zero?
tÑ8
H-4: Solve A = [ T (0) ´ A]ekt for t.

H-5: From Newton’s Law of Cooling and Corollary 3.3.8, we know the temperature of
the copper will be
T (t) = [ T (0) ´ A]eKt + A
where A is the ambient temperature, T (0) is the initial temperature of the copper, and K
is some constant. Use the given information to find an expression for T (t) not involving
any unknown constants.
H-6: From Newton’s Law of Cooling and Corollary 3.3.8, we know the temperature of
the stone t minutes after it leaves the fire is
T (t) = [ T (0) ´ A]eKt + A
where A is the ambient temperature, T (0) is the temperature of the stone the instant it
left the fire, and K is some constant.
H-8: Newton’s Law of Cooling models the temperature of the tea after t minutes as
T (t) = [ T (0) ´ A]eKt + A
where A is the ambient temperature, T (0) is the initial temperature of the tea, and K is
some constant.
H-9: What is lim T (t)?
tÑ8

H-1: P(0) is also (probably) a positive constant.
H-2: The assumption that the animals grow according to the Malthusian model tells us
that their population t years after 2015 is given by P(t) = 121ebt for some constant b.
130
H-3: The Malthusian model says that the population of bacteria t hours after being
placed in the dish will be P(t) = 1000ebt for some constant b.
H-4: If 1928 is a years after the shipwreck, you might want to make use of the fact that
eb(a+1) = eba eb .
H-5: If the population has a net birthrate per individual per unit time of b, then the
Malthusian model predicts that the number of individuals at time t will be
P(t) = P(0)ebt . You can use the test population to find eb .
H-6: One way to investigate the sign of k is to think about f 1 (t): is it positive or negative?

H-1: Use Theorem 3.3.2 to figure out what f ( x ) looks like.
H-2: To use Corollary 3.3.8, you need to re-write the differential equation as

dT 9
=7 T´ ´ .
dt 7
H-3: The amount of the material at time t will be Q(t) = Ce´kt

for some constants C and k.
2
H-4: In your calculations, it might come in handy that e30K = e15K .
H-5: The differential equation in the problem has the same form as the differential
equation from Newton’s Law of Cooling.
H-6: We know the form of the solution A(t) from Corollary 3.3.8.
H-7: If a function’s rate of change is proportional to the function itself, what does the
function looks like?
H-8: The equation from Newton’s Law of Cooling, in Corollary 3.3.8, has a similar form
to the differential equation in this question.

H-1: An approximation should be something you can actually figure out–otherwise it’s
no use.
H-2: You’ll need some constant a to approximation log(0.93) « log( a). This a should
have two properties: it should be close to 0.93, and you should be able to easily evaluate
log( a).
H-3: You’ll need some constant a to approximate arcsin(0.1) « arcsin( a). This a should
have two properties: it should be close to 0.1, and you should be able to easily evaluate
arcsin( a).
131
? ?
H-4: You’ll need some constant a to approximate 3 tan(1) « 3 tan( a). This a should
have
? two properties: it should be close to 1, and you should be able to easily evaluate
3 tan( a).
H-5: We could figure out 10.13 exactly, if we wanted, with pen and paper. Since we’re
asking for an approximation, we aren’t after perfect accuracy. Rather, we’re after ease of
calculation.

H-1: The linear approximation L( x ) is chosen so that f (5) = L(5) and f 1 (5) = L1 (5).
H-2: The graph of the linear approximation is a line, passing through (2, f (2)), with
slope f 1 (2).
H-3: It’s an extremely accurate approximation.
H-4: You’ll need to centre your approximation about some x = a, which should have two
properties: you can easily compute log( a), and a is close to 0.93.
?
H-5: Approximate the function f ( x ) = x.
?
H-6: Approximate the function f ( x ) = 5 x.
H-7: Approximate the function f ( x ) = x3 .
H-8: One possible choice of f ( x ) is f ( x ) = sin x.
H-9: Compare the derivatives.

H-1: If Q( x ) is the quadratic approximation of f about 3, then Q(3) = f (3),
Q1 (3) = f 1 (3), and Q2 (3) = f 2 (3).
H-2: It is a very good approximation.
H-3: Approximate f ( x ) = log x.
H-4: You’ll probably want to centre your approximation about x = 0.
H-5: The quadratic approximation of a function f ( x ) about x = a is
1 2
f ( x ) « f ( a) + f 1 ( a)( x ´ a) + f ( a)( x ´ a)2
2
H-6: One way to go about this is to approximate the function f ( x ) = 5 ¨ x1/3 , because
then 54/3 = 5 ¨ 51/3 = f (5).
H-7: For (c), look for cancellations.
H-8: Compare (c) to (b).
Compare (e) and (f) to (d).
To get an alternating sign, consider powers of (´1).
132
H-9: You can evaluate f (1) exactly.
d 1
Recall arcsin x = ? .
dx 1 ´ x2
H-10: Let f ( x ) = e x , and use the quadratic approximation of f ( x ) about x = 0 (given in
your text, or you can reproduce it) to approximate f (1).
3
ÿ 7
ÿ
H-11: Be wary of indices: for example n= ( n ´ 4).
n =1 n =5

H-1: T32 ( x ) and f 2 ( x ) agree when x = 1.
H-2: The nth degree Taylor polynomial for f ( x ) about x = 5 is
n
ÿ f ( k ) (5)
Tn ( x ) = ( x ´ 5) k
k!
k =0
Match up the terms.

H-3: The fourth-degree Maclaurin polynomial for f ( x ) is
1 2 1 1
T4 ( x ) = f (0) + f 1 (0) x + f ( 0 ) x 2 + f 3 ( 0 ) x 3 + f (4) ( 0 ) x 4
2 3! 4!
while the third-degree Maclaurin polynomial for f ( x ) is
1 2 1
T3 ( x ) = f (0) + f 1 (0) x + f (0) x 2 + f 3 (0) x 3
2 3!
H-4: The third-degree Taylor polynomial for f ( x ) about x = 1 is
1 2 1
T3 ( x ) = f (1) + f 1 (1)( x ´ 1) + f (1)( x ´ 1)2 + f 3 (1)( x ´ 1)3
2 3!
How can you recover f (1), f 1 (1), f 2 (1), and f 3 (1) from T4 ( x )?
H-5: Compare the given polynomial to the more standard form of the nth degree Taylor
polynomial,
n
ÿ 1 (k)
f (5)( x ´ 5)k
k!
k =0
and notice that the term you want (containing f (10) (5)) corresponds to k = 10 in the
standard form, but is not the term corresponding to k = 10 in the polynomial given in the
question.
H-6: T33 ( a) = f 3 ( a)
133
H-1: The derivatives of f ( x ) repeat themselves.
H-2: You are approximating a polynomial with a polynomial.
d
H-3: Recall t2x u = 2x log 2, where log 2 is the constant loge 2.
dx
H-4: Just keep differentiating–it gets easier!
H-5: Start by differentiating, and finding the pattern for f (k) (0). Remember the chain
rule!
H-6: You’ll need to differentiate x x . This is accomplished using logarithmic
differentiation, covered in Section 2.10.

1
H-7: What is 6 arctan ? ?
3
H-8: After a few derivatives, this will be very similar to Example 3.4.13.
H-9: Treat the even and odd powers separately.
H-10: Compare this to the Maclaurin polynomial for e x .
H-11: Compare this to the Maclaurin polynomial for cosine.

H-1: ∆x and ∆y represent changes in x and y, respectively, while f ( x ) and f ( x + ∆x ) are
the y-values the function takes.
H-2: Let f ( x ) be the number of problems finished after x minutes of work.
H-3: ∆y = f (5.1) ´ f (5)
H-4: Use the approximation ∆y « s1 (4)∆x when x is near 4.

H-1: Is the linear approximation exact, or approximate?
H-2: When an exact value Q0 is measured as Q0 + ∆Q, Definition 3.4.25 gives us the
|∆Q|
absolute error as |∆Q|, and the percentage error as 100 .
Q0
H-3: Let ∆y is the change in f ( x ) associated to a change in x from a to a + ∆x. The linear
approximation tells us
∆y « f 1 ( a)∆x
while the quadratic approximation tells us
1 2
∆y « f 1 ( a)∆x + f ( a) (∆x )2
2
134
H-4: The exact area desired is A0 . Let the corresponding exact radius desired be r0 . The
linear approximation tells us ∆A « A1 (r0 )∆r. Use this relationship, and what you know
about the error allowable in A, to find the error allowable in r.
H-5: For part (b), cut the triangle (with angle θ and side d) into two right triangles.
H-6: The volume of a cone of height h and radius r is 31 πr2 h.
H-7: Remember that the amount of the isotope present at time t is Q(t) = Q(0)e´kt for
some constant k. The measured quantity after 3 years will allow you to replace k in the
equation, then solving Q(t) = 12 Q(0) for t will give you the half-life of the isotope.

H-1: R(10) = f (10) ´ F (10) = ´3 ´ 5
H-2: Equation 3.4.33 tells us
ˇ ˇ
ˇ f (4) ( c ) ˇ
4ˇ
| f (2) ´ T3 (2)| = ˇ (2 ´ 0) ˇ
ˇ
ˇ 4! ˇ
for some c strictly between 0 and 2.

H-3: You are approximating a third-degree polynomial with a fifth-degree Taylor
polynomial. You should be able to tell how good your approximation will be without a
long calculation.
H-4: Draw a picture–it should be clear how the two approximations behave.
H-5: In this case, Equation 3.4.33 tells us that
ˇ ˇ
ˇ f (6) ( c ) ˇ
6ˇ
| f (11.5) ´ T5 (11.5)| = ˇ (11.5 ´ 11) ˇ
ˇ
ˇ 6! ˇ
for some c strictly between 11 and 11.5.

ˇ ˇ
ˇ f (3) ( c ) ˇ
H-6: In this case, Equation 3.4.33 tells us that | f (0.1) ´ T2 (0.1)| = ˇ
ˇ 3
(0.1 ´ 0) ˇ for
ˇ
ˇ 3! ˇ
some c strictly between 0 and 0.1.
H-7: In our case, Equation 3.4.33 tells us
ˇ ˇ ˇˇ (6) 6 ˇˇ
ˇ 1 1 ˇ ˇ f (c) 1
ˇf ´ ´ T5 ´ ˇˇ = ˇ ´ ´0 ˇ
ˇ
ˇ 4 4 ˇ 6! 4 ˇ
1
for some c between ´ and 0.
4
ˇ ˇ
ˇ f (4) ( c ) ˇ
H-8: In this case, Equation 3.4.33 tells us that | f (30) ´ T3 (30)| = ˇ
ˇ 4
(30 ´ 32) ˇ for
ˇ
ˇ 4! ˇ
some c strictly between 30 and 32.
135
H-9: In our case, Equationˇ 3.4.33 tells us ˇ
ˇ f (2) ( c ) 2 ˇ 1
1
| f (0.01) ´ T1 (0.01)| = ˇ 0.01 ´ π ˇ for some c between 0.01 and .
ˇ ˇ
ˇ 2! ˇ π
ˇ 1 1 ˇ ˇ f (c) 1
H-10: Using Equation 3.4.33, ˇˇ f ´ T2 ˇ=ˇ ´ 0 ˇ for some c in
ˇ
2 2 ˇ ˇ 3! 2 ˇ

1
0, .
2
H-11: It helps to have a formula for f (n) ( x ). You can figure it out by taking several
derivatives and noticing the pattern, but also this has been given previously in the text.
1
H-12: You can approximate the function f ( x ) = x 7 .
It’s a good bit of trivia to know 37 = 2187.
A low-degree Taylor approximation will give you a good enough estimation. If you
guess a degree, and take that Taylor polynomial, the error will probably be less than 0.001
(but you still need to check).
H-13: Use the 6th-degree Maclaurin approximation for f ( x ) = sin x.
H-14: For part (c), after you plug in the appropriate values to Equation 3.4.33, simplify
the upper and lower bounds for e separately. In particular, for the upper bound, you’ll
have to solve for e.

H-1: Compare the given polynomial to the definition of a Maclaurin polynomial.
H-3: Compare the given polynomial to the definition of a Taylor polynomial.
H-5: You can use the error formula to determine whether the approximation is too large
or too small.
?
H-6: Use the function f ( x ) = x.
H-7: Use the function f ( x ) = x1/3 . What is a good choice of centre?
H-8: Try using the function f ( x ) = x5 .
H-9: If you use the function f ( x ) = sin( x ), what is a good centre?
d 1
H-10: Recall tarctan xu = .
dx 1 + x2
H-11: Try using the function f ( x ) = (2 + x )3 .
H-12: You can try using f ( x ) = (8 + x )1/3 . What is a suitable centre for your
approximation?
H-13: This is the same as the Maclaurin polynomial.
H-14: This is a straightforward application of Equation 3.4.33.
136
H-17: 52/3 = f (52 ) = f (25)
H-18: The fourth-degree Maclaurin polynomial for f ( x ) is
1 2 1 1
T4 ( x ) = f (0) + f 1 (0) x + f ( 0 ) x 2 + f 3 ( 0 ) x 3 + f (4) ( 0 ) x 4
2 3! 4!
1 2 1
T3 ( x ) = f (0) + f 1 (0) x + f (0) x 2 + f 3 (0) x 3
2 3!
H-19: For part (c), think about what the quadratic approximation looks like–is it pointing
up or down?
H-22: For (c), you can write f ( x ) as the sum of Q( x ) and its error term.
For (d), you can use the linear approximation of e x centred at 0, with its error term when
x = 0.1.

H-1: Estimate f 1 (0).
H-2: If the graph is discontinuous at a point, it is not differentiable at that point.
H-3: Try making a little bump at x = 2, the letting the function get quite large somewhere
else.
H-4: Critical points are those values of x for which f 1 ( x ) = 0.
Singular points are those values of x for which f ( x ) is not differentiable.
H-5: We’re only after local extrema, not global. Let f ( x ) be our function. If there is some
interval around x = 2 where nothing is bigger than f (2), then f (2) is a local maximum,
whether or not it is a maximum overall.
H-6: By Theorem 3.5.4, if x = 2 not a critical point, then it must be a singular point.
" global minima of f ( x ) in your head.

H-7: You should be able to figure out the
X Xě0
Remember with absolute values, |X| = .
´X X ă 0
H-8: Review the definitions of critical points and extrema: Definition 3.5.6 and
Definition 3.5.3.

H-1: One way to avoid a global minimum is to have lim f ( x ) = ´8. Since f ( x ) keeps
xÑ8
getting lower and lower, there is no one value that is the lowest.
H-2: Try allowing the function to approach the x-axis without ever touching it.
H-3: Since the global minimum value occurs at x = 5 and x = ´5, it must be true that
f (5) = f (´5).
137
H-4: Global extrema will either occur at critical points in the interval (´5, 5) or at the
endpoints x = 5, x = ´5.
H-5: You only need to consider critical points that are in the interval (´4, 0).

H-1: Factor the derivative.
H-2: Remember to test endpoints.
H-4: One way to decide whether a critical point x = c is a local extremum is to consider
the first derivative. For example: if f 1 ( x ) is negative for all x just to the left of c, and
positive for all x just to the right of c, then f ( x ) decreases up till c, then increases after c,
so f ( x ) has a local minimum at c.
H-5: One way to decide whether a critical point x = c is a local extremum is to consider
the first derivative. For example: if f 1 ( x ) is negative for all x just to the left of c, and
positive for all x just to the right of c, then f ( x ) decreases up till c, then increases after c,
so f ( x ) has a local minimum at c.
H-6: Start with a formula for travel time from P to B. You might want to assign a variable
to the distance from A where your buggy first reaches the road.
H-7: A box has three dimensions; make variables for them, and write the relations given
in the problem in terms of these variables.
H-8: Find a formula for the cost of the base, and another formula for the cost of the other
sides. The total cost is the sum of these two formulas.
H-9: The setup is this:
y
x
O X
H-10: Put the whole system on xy-axes, so that you can easily describe the pieces using
( x, y)-coordinates.
H-11: The surface area consists of two discs and a strip. Find the areas of these pieces.
The volume of a cylinder with radius r and height h is πr2 h.
138
r
h h
H-12: If the circle has radius r, and the entire window has perimeter P, what is the height
of the rectangle?
H-14: Use logarithmic differentiation to find f 1 ( x ).
H-15: When you are finding the global extrema of a function, remember to check
endpoints as well as critical points.

H-1: What happens if g( x ) = x + 3?
H-2: Use domains and intercepts to distinguish between the functions.
H-3: To find p, the equation f (0) = 2 gives you two possible values of p. Consider the
domain of f ( x ) to decide between them.
H-4: Check for horizontal asymptotes by evaluating lim f ( x ), and check for vertical
xÑ˘8
asymptotes by finding any value of x near which f ( x ) blows up.
H-5: Check for horizontal asymptotes by evaluating lim f ( x ), and check for vertical
xÑ˘8
asymptotes by finding any value of x near which f ( x ) blows up.

H-1: For each of the graphs, consider where the derivative is positive, negative, and zero.
H-2: Where is f 1 ( x ) ą 0?
H-3: Consider the signs of the numerator and the denominator of f 1 ( x ).
139
d 1
H-4: Remember tarctan xu = .
dx 1 + x2

H-1: There are two intervals where the function is concave up, and two where it is
concave down.
H-2: Try allowing your graph to have horizontal asymptotes. For example, let the
function get closer and closer to the x-axis (or another horizontal line) without touching
it.
H-3: Consider f ( x ) = ( x ´ 3)4 .
H-5: You must show it has at least one inflection point (try the Intermediate Value
Theorem), and at most one inflection point (consider whether the second derivative is
increasing or decreasing).
H-6: Use (a) in proving (b).
H-7: Since x = 3 is an inflection point, we know the concavity of f ( x ) changes at x = 3.
That is, there is some interval around 3, with endpoints a and b, such that
• f 2 ( a) ă 0 and f 2 ( x ) ă 0 for every x between a and 3, and
• f 2 (b) ą 0 and f 2 ( x ) ą 0 for every x between b and 3.
Use the IVT to show that f 2 ( x ) = 3 for some x between a and b; then show that this value
of x can’t be anything except x = 3.

H-1: This function is symmetric across the y-axis.
H-2: There are two.
H-3: Since the function is even, you only have to reflect the portion shown across the
y-axis to complete the sketch.
H-4: Since the function is odd, to complete the sketch, reflect the portion shown across
the y-axis, then the x-axis.
H-5: A function is even if f (´x ) = f ( x ).
H-6: Its period is not 2π.
H-7: Simplify f (´x ) to see whether it is the same as f ( x ), ´ f ( x ), or neither.
H-8: Simplify f (´x ) to see whether it is the same as f ( x ), ´ f ( x ), or neither.
H-9: Find the smallest value k such that f ( x + k ) = f ( x ) for any x in the domain of f .
You may use the fact that the period of g( X ) = tan X is π.
140
H-10: It is true that f ( x ) = f ( x + 2π ) for every x in the domain of f ( x ), but the period is
not 2π.
No exercises for Section 3.6.5. — Jump to TABLE OF CONTENTS

H-1: You’ll find the intervals of increase and decrease. These will give you a basic outline
of the behaviour of the function. Use concavity to refine your picture.
H-2: The local maximum is also a global maximum.
H-3: The sign of the first derivative is determined entirely by the numerator, but the sign
of the second derivative depends on both the numerator and the denominator.
H-4: The function is odd.
H-5: The function is continuous at x = 0, but its derivative is not.
H-6: Since you aren’t asked to find the intervals of concavity exactly, sketch the intervals
of increase and decrease, and turn them into a smooth curve. You might not get exactly
the intervals of concavity that are given in the solution, but there should be the same
number of intervals as the solution, and they should have the same positions relative to
the local extrema.
H-7: Use intervals of increase and decrease, concavity, and asymptotes to sketch the
curve.
H-8: Although the function exhibits a certain kind of repeating behaviour, it is not
periodic.
H-9: The period of this function is 2π. So, it’s enough to graph the curve y = f ( x ) over
the interval [´π, π ], because that figure will simply repeat.
Use trigonometric identities to write f 2 ( x ) = ´4(4 sin2 x + sin x ´ 2). Then you can find
where f 2 ( x ) = 0 by setting y = sin x and solving 0 = 4y2 + y ´ 2.
H-10: There is one point where the curve is continuous but has a vertical tangent line.
H-11: Use lim f 1 ( x ) to determine lim f ( x ).
xÑ´8 xÑ´8
H-12: Once you have the graph of a function, reflect it over the line y = x to graph its
inverse. Be careful of the fact that f ( x ) is only defined in this problem for x ě 0.
H-14: For (a), don’t be intimidated by the new names: we can graph these functions
using the methods learned in this section.
For (b), remember that to define an inverse of a function, we need to restrict the domain
of that function to an interval where it is one-to-one. Then to graph the inverse, we can
simply reflect the original function over the line y = x.
141
For (c), set y( x ) = cosh´1 ( x ), so cosh(y( x )) = x. The differentiate using the chain rule.
To get your final answer in terms of x (instead of y), use the identity
cosh2 (y) ´ sinh2 (y) = 1.

H-1: Try making one function a multiple of the other.
H-2: Try making one function a multiple of the other, but not a constant multiple.
H-3: Try modifying the function from Example 3.7.20.
H-4: Plugging in x = 1 to the numerator and denominator makes both zero. This is
exactly one of the indeterminate forms where l’Hôpital’s rule can be directly applied.
H-5: Is this an indeterminate form?
H-6: First, rearrange the expression to a more natural form (without a negative
exponent).
H-7: If at first you don’t succeed, try, try again.
H-8: Keep at it!
H-9: Rather than use l’Hôpital, try factoring out x2 from the numerator and denominator.
H-10: Keep going!
H-12: Try plugging in x = 0. Is this an indeterminate form?
H-13: Simplify the trigonometric part first.
H-14: Rationalize, then remember your training.
H-15: If it is too difficult to take a derivative for l’Hôpital’s Rule, try splitting up the
function into smaller chunks and evaluating their limits independently.
H-17: Try manipulating the function to get it into a nicer form
x2
a 1
H-19: lim sin2 x = (sin2 x ) x2 ; what form is this?
xÑ0
?
x2
1
H-20: lim cos x = lim (cos x ) x2
xÑ0 xÑ0
H-21: logarithms
H-22: Introduce yet another logarithm.
H-23: If the denominator tends to zero, and the limit exists, what must be the limit of the
numerator?
H-24: Start with one application of l’Hôpital’s Rule. After that, you need to consider
three distinct cases: k ą 2, k ă 2, and k = 2.
142
ˇ exactápprox ˇ
ˇ ˇ
H-25: Percentage error: 100 ˇ exact ˇ. Absolute error: |exact ´ approx|. (We’ll see
these definitions again in 3.4.25.)

H-1: The function f ( x ) is an antiderivative of f 1 ( x ), but it is not the most general one.
H-2: When f ( x ) is positive, its antiderivative F ( x ) is increasing. When f ( x ) is negative,
its antiderivative F ( x ) is decreasing. When f ( x ) = 0, F ( x ) has a horizontal tangent line.
1 n +1 .
H-3: For any constant n ‰ ´1, an antiderivative of x n is n +1 x
1 n +1 .
H-4: For any constant n ‰ ´1, an antiderivative of x n is n +1 x
1 n +1 .
H-5: For any constant n ‰ ´1, an antiderivative of x n is n +1 x The constant n does not
have to be an integer.
?
H-6: What is the derivative of x?
H-7: The derivative of e5x+11 is close to, but not exactly the same as, f ( x ). Don’t be afraid
to just make a guess. But be sure to check by differentiating your guess. If the derivative
isn’t what you want, you will often still learn enough to be able to then guess the correct
antiderivative.
H-8: From the table in the text, an antiderivative of sin x is ´ cos x, and an antiderivative
of cos x is sin x.
H-9: What is the derivative of tangent?
1
H-10: What is an antiderivative of ?
x

7 7 1
H-11: ? =? ?
3 ´ 3x2 3 1 ´ x2
H-12: How is this similar to the derivative of the arctangent function?
H-13: First, find the antiderivative of f 1 ( x ). Your answer will have an unknown +C in it.
Figure out which value of C gives f (1) = 10.
H-14: Remember that one antiderivative of cos x is sin x (not ´ sin x).
1
H-15: An antiderivative of is log( x ) + C, but this is not the most general antiderivative.
x
H-16: What is the derivative of the arcsine function?
H-17: If P(t) is the population at time t, then the information given in the problem is
P1 (t) = 100e2t .
H-18: You can’t know exactly–there will be a constant involved.
H-19: If P(t) is the amount of power in kW-hours the house has used since time t = 0,
where t is measured in hours, then the information given is that
143
π
P1 ( t ) = 0.5 sin t + 0.25 kW.
24
H-20: The derivatives should match. Remember sin´1 is another way of writing arcsine.
H-21: Think about the product rule.
H-22: Think about the quotient rule for derivatives.
H-23: Notice that the derivative of x3 is 3x2 .
H-24: Think about the chain rule for derivatives. You might need to multiply your first
guess by a constant.
H-25: Simplify.
H-26: This problem is similar to Question 11, but you’ll ! have to do some harder
1
factoring. Try getting f ( x ) into the form a a for some constants a and b.
1 ´ (bx )2
H-27: Following Example 4.1.7, let V ( H ) be the volume of the solid formed by rotating
the segment of the parabola from x = ´H to x = H. The plan is to evaluate
V ( H ) ´ V (h)
V 1 ( H ) = lim and then antidifferentiate V 1 ( H ) to find V ( H ). Since you
hÑH H´h
don’t know V ( H ) ´ V (h) (yet), first find upper and lower bounds on it when h ă H.
These bounds can be the volumes of two cylinders, one with radius H (and what height?)
and the other with radius h.
144
Part III
A NSWERS TO PROBLEMS
145
Answers to Exercises 1.1 — Jump to TABLE OF CONTENTS
A-1:
b
y
y = f (x)
c
Q
P
x
A-2:
(a) True
(b) In general, this is false. For most functions f ( x ) this will be false, but there are some
functions for which it is true.
A-3: At least once.

A-1: Speed is nonnegative; velocity has a sign (positive or negative) that indicates
direction.
A-2: Yes–an object that is not moving has speed 0.
A-3: 0 kph
A-4: The speed at the one second mark is larger than the average speed.
change in vertical component
A-5: The slope of a curve is given by . The change in the
change in horizontal component
vertical component is exactly s(b) ´ s( a), and the change in the horizontal component is
exactly b ´ a.
A-6: (0, 2) Y (6, 7)
A-7:
(a) 24 units per second.
(b) 6 units per second
A-8:
1
(a) 4 units per second
147
1
(b) 2 units per second
1
(c) 6 units per second
Remark: the average velocity is not the average of the two instantaneous velocities.

A-1:
(a) lim f ( x ) = 1
xÑ´2
(b) lim f ( x ) = 0
xÑ0
(c) lim f ( x ) = 2
xÑ2
A-2: DNE
A-3:
(a) lim f ( x ) = 2
xÑ´1´
(b) lim f ( x ) = ´2
xÑ´1+
(c) lim f ( x ) = DNE

xÑ´1
(d) lim f ( x ) = 0
xÑ´2+
(e) lim f ( x ) = 0
xÑ2´
A-4: Many answers are possible; here is one.
y
y = f (x)
10
x
3
148
y
y = f (x)
10
x
3
A-6: In general, this is false.

A-7: False
A-8: lim f ( x ) = 16
xÑ´2´
A-9: Not enough information to say.

A-10: lim sin t = 0
tÑ0
A-11: lim log x = ´8

xÑ0+
A-12: lim y2 = 9
yÑ3
1
A-13: lim = ´8
xÑ0´ x
1
A-14: lim = DNE
xÑ0 x
1
A-15: lim =8
xÑ0 x2
1 1
A-16: lim =
xÑ3 10 10
A-17: 9

A-1: (a) and (d)
A-2: There are many possible answers; one is f ( x ) = 10( x ´ 3), g( x ) = x ´ 3.
A-3: There are many possible answers; one is f ( x ) = ( x ´ 3)2 and g( x ) = x ´ 3. Another
is f ( x ) = 0 and g( x ) = x ´ 3.
A-4: There are many possible answers; one is f ( x ) = x ´ 3, g( x ) = ( x ´ 3)3 .
149
A-5: Any real number; positive infinity; negative infinity; does not exist.
A-6: 0
A-7: 6
A-8: 16
A-9: 4/ cos(3)
A-10: 2
A-11: ´7/2
A-12: 3
A-13: ´ 23
A-14: log(2) ´ 1
1
A-15: 4
1
A-16:
2
A-17: 5
A-18: ´6
A-19: ´14
A-20: ´ 31
1
A-21: 6
A-22: ?1
3
A-23: 1
A-24: 12
A-25: 0
1
A-26: 2
A-27: 0
A-28: 5
A-29: ´8
c
2
A-30:
3
A-31: DNE
A-32: 8
A-33: x5 ´ 32x + 15
A-34: 0
150
A-35: 0
A-36: 2
A-37: 0
32
A-38: ´
9
A-39: DNE
A-40: DNE
9
A-41: ´
2
A-42: ´4
7
A-43: a =
2
A-44:
(a) lim f ( x ) = 0
xÑ0
(b) lim g( x ) = DNE

xÑ0
(c) lim f ( x ) g( x ) = 2
xÑ0
f (x)
(d) lim =0
xÑ0 g( x )
9
(e) lim f ( x ) + g( x ) =
xÑ2 2
f (x) + 1
(f) lim =1
xÑ0 g ( x + 1)
A-45:
y
x
1
Pictures may vary somewhat; the important points are the values of the function at
integer values of x, and the vertical asymptotes.
151
A-46:
y
f (x)
y=
g( x )
x
1
A-47: 10
A-48: (a) DNE , DNE (b) 0 (c) No: it is only true when both lim f ( x ) and lim g( x )
xÑa xÑa
exist.
A-49: (a) lim f ( x ) = ´3 (b) lim f ( x ) = 3 (c) lim f ( x ) = DNE
xÑ0´ xÑ0+ xÑ0
A-50: (a) lim f ( x ) = 0 (b) lim f ( x ) = 0 (c) lim f ( x ) = 0

xÑ´4´ xÑ´4+ xÑ´4

A-1: There are many answers: any constant polynomial has this property. One answer is
f ( x ) = 1.
A-2: There are many answers: any odd-degree polynomial has this property. One answer
is f ( x ) = x.
A-3: 0
A-4: 8
A-5: 0
A-6: DNE
A-7: ´8
?
A-8: 3
A-9: 3
A-10: ´ 34
1
A-11: ´
2
1
A-12: 2
152
5
A-13: 3
A-14: 0
4
A-15: 7
A-16: 1
A-17: 0
A-18: ´1
A-19: 1
A-20: ´1
A-21: ´ 23
A-22: ´ 53
A-23: ´8
5
A-24:
2
a2 ´ 1a
A-25: lim =8
aÑ0+ a´1
2x + 8
A-26: lim 1 1
=0
xÑ3
x´3 + 2
x ´9
A-27: No such rational function exists.
A-28: This is the amount of the substance that will linger long-term. Since it’s nonzero,
the substance would be something that would stay in your body. Something like “tattoo
ink” is a reasonable answer, while “penicillin” is not.

A-1: Many answers are possible; the tangent function behaves like this.
A-2: At some time between my birth and now, I was exactly one meter tall.
"
0 when 0 ď x ď 1
A-3: One example is f ( x ) = . The IVT only guarantees f (c) = 1
2 when 1 ă x ď 2
for some c in [0, 2] when f is continuous over [0, 2]. If f is not continuous, the IVT says
nothing.
153
y
A-4: Yes
A-5: No
A-6: No
A-7: True.
A-8: True.
A-9: In general, false.
A-10: lim h( x ) = 0
xÑ0+
A-11: k = 0
A-12: Since f is a polynomial, it is continuous over all real numbers. f (0) = 1 ă 12345
and f (12345) = 123453 + 123452 + 12345 + 1 ą 12345 (since all terms are positive). So by
the IVT, f (c) = 12345 for some c between 0 and 12345.
A-13: (´8, ´1) Y (´1, 1) Y (1, +8)
A-14: (´8, ´1) Y (1, +8)
A-15: The function is continuous except at x = ˘π, ˘3π, ˘5π, . . . .
A-16: x ‰ nπ, where n is any integer
A-17: ˘2
A-18: c = 1
A-19: ´1, 4
A-20: c = 1, c = ´1
A-21: This isn’t the kind of equality that we can just solve; we’ll need a trick, and that
trick is the IVT. The general idea is to show that sin x is somewhere bigger, and
somewhere smaller, than x ´ 1. However, since the IVT can only show us that a function
is equal to a constant, we need to slightly adjust our language. Showing sin x = x ´ 1 is
equivalent to showing sin x ´ x + 1 = 0, so let f ( x ) = sin x ´ x + 1, and let’s show that it
has a real root.
First, we need to note that f ( x ) is continuous (otherwise we can’t use the IVT). Now, we
need to find a value of x for which it is positive, and for which it’s negative. By checking
a few values, we find f (0) is positive, and f (100) is negative. So, by the IVT, there exists
a value of x (between 0 and 100) for which f ( x ) = 0. Therefore, there exists a value of x
for which sin x = x ´ 1.
154
A-22: We let f ( x ) = 3x ´ x2 . Then f ( x ) is a continuous function, since both 3x and x2 are
continuous for all real numbers.
We want a value a such that f ( a) ą 0. We see that a = 0 works since
f (0) = 30 ´ 0 = 1 ą 0.
We want a value b such that f (b) ă 0. We see that b = ´1 works since
1
f (´1) = ´ 1 ă 0.
3
So, because f ( x ) is continuous on (´8, 8) and f (0) ą 0 while f (´1) ă 0, then the
Intermediate Value Theorem guarantees the existence of a real number c P (´1, 0) such
that f (c) = 0.
A-23: We let f ( x ) = 2 tan( x ) ´ x ´ 1. Then f ( x ) is a continuous function on the interval
(´π/2, π/2) since tan( x ) = sin( x )/ cos( x ) is continuous on this interval, while x + 1 is a
polynomial and therefore continuous for all real numbers.
We find a value a P (´π/2, π/2) such that f ( a) ă 0. We observe immediately that a = 0
works since
f (0) = 2 tan(0) ´ 0 ´ 1 = 0 ´ 1 = ´1 ă 0.
We find a value b P (´π/2, π/2) such that f (b) ą 0. We see that b = π/4 works since
f (π/4) = 2 tan(π/4) ´ π/4 ´ 1 = 2 ´ π/4 ´ 1 = 1 ´ π/4 = (4 ´ π )/4 ą 0,
because 3 ă π ă 4.
So, because f ( x ) is continuous on [0, π/4] and f (0) ă 0 while f (π/4) ą 0, then the
Intermediate Value Theorem guarantees the existence of a real number c P (0, π/4) such
that f (c) = 0.
a
A-24: Let f ( x ) = cos(πx ) ´ sin(2πx ) ´ 1/2. This function is continuous provided
cos(πx ) ě 0. This is true for 0 ď x ď 21 .
Now f takes positive values on [0, 1/2]:
b ?
f (0) = cos(0) ´ sin(0) ´ 1/2 = 1 ´ 1/2 = 1/2.
And f takes negative values on [0, 1/2]:

b
f (1/2) = cos(π/2) ´ sin(π ) ´ 1/2 = 0 ´ 0 ´ 1/2 = ´1/2
? ?
(Notice that f (1/3) = ( 2 ´ 3)/2 ´ 1/2 also works)
So, because f ( x ) is continuous on [0, 1/2) and f (0) ą 0 while f (1/2) ă 0, then the
Intermediate Value Theorem guarantees the existence of a real number c P (0, 1/2) such
that f (c) = 0.
155
1 3
A-25: We let f ( x ) = ´ x ´ . Then f ( x ) is a continuous function on the interval
cos2 (πx ) 2
(´1/2, 1/2) since cos x is continuous everywhere and non-zero on that interval.
The function f takes negative values. For example, when x = 0:
1 3 3 1
f (0) = ´ 0 ´ = 1 ´ = ´ ă 0.
cos2 (0) 2 2 2
It also takes positive values, for instance when x = 1/4:
1 1 3
f (1/4) = 2
´ ´
(cos π/4) 4 2
1 1+6
= ´
1/2 4
= 2 ´ 7/4 = 1/4 ą 0.
By the IVT there is c, 0 ă c ă 1/4 such that f (c) = 0, in which case
1 3
= c + .
(cos πc)2 2
A-26: [0, 1] is the easiest answer to find. Also acceptable are [´2, ´1] and [14, 15].
A-27: 1.91
A-28:
• If f ( a) = g( a), or f (b) = g(b), then we simply take c = a or c = b.
• Suppose f ( a) ‰ g( a) and f (b) ‰ g(b). Then f ( a) ă g( a) and g(b) ă f (b), so if we
define h( x ) = f ( x ) ´ g( x ), then h( a) ă 0 and h(b) ą 0. Since h is the difference of
two functions that are continuous over [ a, b], also h is continuous over [ a, b]. So, by
the Intermediate Value Theorem, there exists some c P ( a, b) with h(c) = 0; that is,
f ( c ) = g ( c ).

A-1: If Q is to the left of the y axis, the secant line has positive slope; if Q is to the right of
the y axis, the secant line has negative slope.
A-2: (a) closer (b) the tangent line has the larger slope
A-3: {(a), (c), (e)}, {(b),(f)}, {(d)}
A-4: Something like 1.5. A reasonable answer would be between 1 and 2.
A-5: There is only one tangent line to f ( x ) at P (shown in blue), but there are infinitely
many choices of Q and R (one possibility shown in red).
156
y
y = f (x)
x
R
Q
P
A-6:
y
y = f (x)

A-1: (a), (d)
A-2: (e)
A-3: (b)
A-4: By definition, f ( x ) = x3 is differentiable at x = 0 if the limit
f ( h ) ´ f (0) h3 ´ 0
lim = lim
hÑ0 h hÑ0 h
exists.
A-5: x = ´1 and x = 3
A-6: True. (Contrast to Question 7.)
A-7: In general, false. (Contrast to Question 6.)
157
A-8: metres per second
A-9: y ´ 6 = 3( x ´ 1), or y = 3x + 3
´1
A-10:
x2
A-11: By definition
f ( h ) ´ f (0) h|h|
f 1 (0) = lim = lim = lim |h| = 0
hÑ0 h hÑ0 h hÑ0
In particular, the limit exists, so the derivative exists (and is equal to zero).
´2
A-12:
( x + 1)2
´2x
A-13:
[ x2+ 3]2
A-14: 1
2
A-15: f 1 ( x ) = ´
x3
A-16: a = 4, b = ´4
1
A-17: f 1 ( x ) = ? when x ą ´1; f 1 ( x ) does not exist when x ď ´1.
2 1+x
A-18: v(t) = 4t3 ´ 2t
A-19: No, it does not.
A-20: No, it does not.
A-21: Yes, it is.
A-22: Yes, it is.
158
y
1
x
1
A-24:
p( x + h) ´ p( x )
p1 ( x ) = lim
hÑ0 h
f ( x + h) + g( x + h) ´ f ( x ) ´ g( x )
= lim
hÑ0 h
f ( x + h) ´ f ( x ) + g( x + h) ´ g( x )
= lim
hÑ0 h

f ( x + h) ´ f ( x ) g( x + h) ´ g( x )
= lim +
hÑ0 h h

f ( x + h) ´ f ( x ) g( x + h) ´ g( x )
(˚) = lim + lim
hÑ0 h hÑ0 h
1 1
= f (x) + g (x)
At step (˚), we use the limit law that lim [ F ( x ) + G ( x )] = lim F ( x ) + lim G ( x ), as long as
xÑa xÑa xÑa
lim F ( x ) and lim G ( x ) exist. Because the problem states that f 1 ( x ) and g1 ( x ) exist, we
xÑa xÑa
f ( x + h) ´ f ( x ) g( x + h) ´ g( x )
know that lim and lim exist, so our work is valid.
hÑ0 h hÑ0 h
A-25: (a) f 1 (x) = 2 and g1 ( x ) = 1 (b) p1 ( x ) = 4x (c) no
A-26: y = 6x ´ 9 and y = ´2x ´ 1
A-27: a ą 1

A-1: (a) The average rate of change of the height of the water over the single day starting
m
at t = 0, measured in hr .
159
(b) The instantaneous rate of change of the height of the water at the time t = 0.
A-2: Profit per additional widget sold, when t widgets are being sold. This is called the
marginal profit per widget, when t widgets are being sold.
A-3: T 1 (d) measures how quickly the temperature is changing per unit change of depth,
measured in degrees per metre. |T 1 (d)| will probably be largest when d is near zero,
unless there are hot springs or other underwater heat sources.
A-4: Calories per additional gram, when there are w grams.
A-5: The acceleration of the object.
A-6: Degrees Celsius temperature change per joule of heat added. (This is closely related
to heat capacity and to specific heat — there’s a nice explanation of this on Wikipedia.)
A-7: Number of bacteria added per degree. That is: the number of extra bacteria
(possibly negative) that will exist in the population by raising the temperature by one
degree.
A-8: 360R1 (t)
A-9: If P1 (t) is positive, your sample is below the ideal temperature, and if P1 (t) is
negative, your sample is above the ideal temperature. If P1 (t) = 0, you don’t know
whether the sample is exactly at the ideal temperature, or way above or below it with no
living bacteria.

A-1: True
A-2: False, in general
A-3: True
A-4: If you’re creative, you can find lots of ways to differentiate!
Constant multiple: g1 ( x ) = 3 f 1 ( x ).
d
Product rule: g1 ( x ) = dx t3u f ( x ) + 3 f 1 ( x ) = 0 f ( x ) + 3 f 1 ( x ) = 3 f 1 ( x ).
d
Sum rule: g1 ( x ) = dx t f ( x )"+ f (* x ) + f ( x )u = f 1 ( x ) + f 1 ( x ) + f 1 ( x ) = 3 f 1 ( x ).
1 1 1 1
d f (x) 3 f ( x )´ f ( x )(0) 3 f (x) 1
Quotient rule: g1 ( x ) = dx 1 = 1 = 1 = 9 1 1
3 f ( x ) = 3 f ( x ).
3 9 9
All rules give g1 ( x ) = 3 f 1 ( x ).
A-5: f 1 ( x ) = 6x + ?2x
?
A-6: ´36x + 24 x + ?20x ´ 45

A-7: y ´ 8 = 4 ¨ x ´ 2 , or y = 34 x ´ 41
1 3 1
A-8: (a) 4 (b) left (c) decreasing

1 4
A-9: 2
, or
( x + 1/2) (2x + 1)2
160
A-10: ´72
A-11: y ´ 12 = ´ 18 ( x ´ 1), or y = ´ 18 x + 5
8
A-12: b1 (t) ´ d1 (t)
A-13: (1, 3), (3, 27)
1
A-14: ?
2 100180
A-15: 20t + 7 square metres per second.
A-16: 0
A-17:
g( x )
First expression, f ( x ) = :
h( x )
h ( x ) g1 ( x ) ´ g ( x ) h1 ( x )
f 1 (x) =
h2 ( x )
g( x ) k( x )
Second expresson, f ( x ) = ¨ :
k( x ) h( x )
k ( x ) g1 ( x ) ´ g ( x ) k 1 ( x ) h ( x ) k 1 ( x ) ´ k ( x ) h1 ( x )

k( x) g( x )
f 1 (x) = +
k2 ( x ) h( x ) k( x) h2 ( x )
k ( x ) g1 ( x ) ´ g ( x ) k 1 ( x ) g ( x ) h ( x ) k 1 ( x ) ´ g ( x ) k ( x ) h1 ( x )
= +
k( x )h( x ) k ( x ) h2 ( x )
h ( x ) k ( x ) g1 ( x ) ´ h ( x ) g ( x ) k 1 ( x ) g ( x ) h ( x ) k 1 ( x ) ´ g ( x ) k ( x ) h1 ( x )
= +
k ( x ) h2 ( x ) k ( x ) h2 ( x )
h ( x ) k ( x ) g1 ( x ) ´ h ( x ) g ( x ) k 1 ( x ) + g ( x ) h ( x ) k 1 ( x ) ´ g ( x ) k ( x ) h1 ( x )
=
k ( x ) h2 ( x )
h ( x ) k ( x ) g1 ( x ) ´ g ( x ) k ( x ) h1 ( x )
=
k ( x ) h2 ( x )
h ( x ) g1 ( x ) ´ g ( x ) h1 ( x )
=
h2 ( x )
and this is exactly what we got from differentiating the first expression.
161
A-1: In the quotient rule, there is a minus, not a plus. Also, 2( x + 1) + 2x is not the same
as 2( x + 1).
The correct version is:
2x
f (x) =
x+1
2( x + 1)´2x
f 1 (x) =
( x + 1)2
2
=
( x + 1)2
A-2: False
A-3: 4x ( x2 + 2)( x2 + 3)
1
A-4: 12t3 + 15t2 + t2
A-5: x1 (y) = 8y3 + 2y

1
?
( x2 + 3) ?
2 x
´ ( x + 1)(2x )
A-6: T 1 ( x ) =
( x2 + 3)2
21 ´ 4x ´ 7x2
A-7:
( x 2 + 3)2
A-8: 7
3x4 + 30x3 ´ 2x ´ 5
A-9:
( x2 + 5x )2
´3x2 + 12x + 5
A-10:
(2 ´ x )2
´22x
A-11:
(3x2 + 5)2
4x3 + 12x2 ´ 1
A-12:
( x + 2)2
A-13: The derivative of the function is
?
(1 ´ x2 ) ¨ 2?1 x ´ x ¨ (´2x ) (1 ´ x2 ) ´ 2x ¨ (´2x )
= ?
(1 ´ x 2 )2 2 x (1 ´ x 2 )2
The derivative is undefined if either x ă 0 or x = 0, ˘1 (since the square-root is
undefined for x ă 0 and the denominator is zero when x = 0, 1, ´1. Putting this together
— the derivative exists for x ą 0, x ‰ 1.
´4 ´2
? ?
A-14: 53 x 5 + 5x 3 3x2 + 8x ´ 5 + (3 5 x + 15 3 x + 8) (6x + 8)

1 ´1/2 ´2/3 2 ´1 ´3/2 2 ´5/3
A-15: f ( x ) = (2x + 5)( x +x ) + ( x + 5x + 1) 2 x ´ 3x
A-16: x = ´5 and x = 1
162
1
A-17: y = x ´
4
A-18:
y
x
−1 2
y = 4x ´ 4 and y = ´2x ´ 1
A-19: 2015 ¨ 22014

A-1: A-( a) and (d), B-(e), C-(c), D-(b)
A-2: (b), (d), (e)
A-3: False
A-4: increasing
( x ´ 1) e x
A-5:
2x2
A-6: 2e2x
A-7: e a+ x
A-8: x ą ´1
A-9: é´x
A-10: 2e2x
A-11: When t is in the interval (´2, 0).
A-12: g1 ( x ) = [ f ( x ) + f 1 ( x )]e x
A-13: (b) and (d)
e
A-14: a = b =
2

A-1:
163
y y = cos x
y = sin x
x
´π π
The graph f ( x ) = sin x has horizontal tangent lines precisely at those points where
cos x = 0.
A-2:
y y = cos x
y = sin x
x
´π π
The graph f ( x ) = sin x has maximum slope at those points where cos x has a maximum.
That is, where cos x = 1.
A-3: f 1 ( x ) = cos x ´ sin x + sec2 x
π
A-4: x = 4 + πn, for any integer n.
A-5: 0
A-6: f 1 ( x ) = 2(cos2 x ´ sin2 x )
A-7: f 1 ( x ) = e x (cot x ´ csc2 x )
2 + 3 sec x + 2 sin x ´ 2 tan x sec x + 3 sin x tan x
A-8: f 1 ( x ) =
(cos x + tan x )2
5 sec x tan x ´ 5 sec x ´ 1
A-9: f 1 ( x ) =
ex
A-10: f 1 ( x ) = (e x + cot x )(30x5 + csc x cot x ) + (e x ´ csc2 x )(5x6 ´ csc x )
A-11: ´ sin(θ )
A-12: f 1 ( x ) = ´ cos x ´ sin x
cos θ + sin θ 2

A-13: +1
cos θ ´ sin θ
A-14: a = 0, b = 1.
A-15: y ´ π = 1 ¨ ( x ´ π/2)
A-16: ´ sin(2015)
?
A-17: ´ 3/2
A-18: ´1
164
A-19:
sin θ
tan θ =
cos θ
So, using the quotient rule,
d cos θ cos θ ´ sin θ (´ sin θ ) cos2 θ + sin2 θ

ttan θu = =
dθ cos2 θ cos2 θ
2
1
= = sec2 θ
cos θ
A-20: a = ´ 32 , b = 2
π
A-21: All values of x except x = 2 + nπ, for any integer n.
A-22: The function is differentiable whenever x2 + x ´ 6 ‰ 0 since the derivative equals
10 cos( x ) ¨ ( x2 + x ´ 6) ´ 10 sin( x ) ¨ (2x + 1)

,
( x 2 + x ´ 6)2
which is well-defined unless x2 + x ´ 6 = 0. We solve x2 + x ´ 6 = ( x ´ 2)( x + 3) = 0,

and get x = 2 and x = ´3. So, the function is differentiable for all real values x except for
x = 2 and for x = ´3.
A-23: The function is differentiable whenever sin( x ) ‰ 0 since the derivative equals
sin( x ) ¨ (2x + 6) ´ cos( x ) ¨ ( x2 + 6x + 5)

,
(sin x )2
which is well-defined unless sin x = 0. This happens when x is an integer multiple of π.

So, the function is differentiable for all real values x except x = nπ,, where n is any
integer.
A-24: y ´ 1 = 2 ¨ ( x ´ π/4)
A-25: y = 2x + 2
3π
A-26: x = 4 + nπ for any integer n.
A-27: f 1 (0) = 0
"
cos x x ą 0
A-28: h1 ( x ) = It exists for all x ‰ 0.
´ cos x x ă 0
A-29: iii
A-30: 2

dK dU dK
A-1: (a) dU is negative (b) dO is negative (c) dO is positive
165
A-2: negative
A-3: ´5 sin(5x + 3)
A-4: 10x ( x2 + 2)4
A-5: 17(4k4 + 2k2 + 1)16 ¨ (16k3 + 4k )
´2x
A-6: ?
( x2 ´1) x4 ´1
2
A-7: écos( x ) ¨ sin( x2 ) ¨ 2x
A-8: ´4
A-9: [cos x ´ x sin x ]e x cos( x)
2 +cos( x )
A-10: [2x ´ sin x ]e x
3
A-11: ? ? 3
2 x´1 x +2
2 x
A-12: f 1 ( x ) = ´ + ? is defined for x in (´8, 1) Y (1, 8).
x3 x2 ´ 1
(1 + x2 )(5 cos 5x ) ´ (sin 5x )(2x )
A-13: f 1 ( x ) =
(1 + x 2 )2
A-14: 2e2x+7 sec(e2x+7 ) tan(e2x+7 )
A-15: y = 1
2
A-16: t = 3 and t = 4
A-17: 2e sec2 (e)
A-18: y1 = 4e4x tan x + e4x sec2 x
3
A-19:
(1 + e3 )2
2 (x)
A-20: 2 sin( x ) ¨ cos( x ) ¨ esin
A-21: cos e5x ¨ e5x ¨ 5

2
A-22: écos( x ) ¨ sin( x2 ) ¨ 2x

? x
A-23: y1 = ´ sin x2 +

x2 + 1 2x + ?
2
x +1
A-24: y1 = 2x cos2 x ´ 2(1 + x2 ) sin x cos x
e3x (3x2 ´ 2x + 3)
A-25: y1 =
(1 + x 2 )2
A-26: ´40
A-27: (1, 1) and (´1, ´1).
166
A-28: Always
A-29: e x sec3 (5x ´ 7)(1 + 15 tan(5x ´ 7))
A-30: e2x cos 4x + 2xe2x cos 4x ´ 4xe2x sin 4x
π
A-31: t =
4
2
A-32: Let f ( x ) = e x+ x and g( x ) = 1 + x. Then f (0) = g(0) = 1.
2
f 1 ( x ) = (1 + 2x )e x+ x and g1 ( x ) = 1. When x ą 0,
2 2 2 2
f 1 ( x ) = (1 + 2x )e x+ x ą 1 ¨ e x+ x = e x+ x ą e0+0 = 1 = g1 ( x ).
Since f (0) = g(0), and f 1 ( x ) ą g1 ( x ) for all x ą 0, that means f and g start at the same
place, but f always grows faster. Therefore, f ( x ) ą g( x ) for all x ą 0.
A-33: cos(2x ) = cos2 x ´ sin2 x
A-34:
? !2
1 x 3 ´ 9 tan x 3
f 1 (x) = 2 ¨
3 ecsc x
? 
2 2 3x 2 tan x ?
3 csc x 2 2 csc x 3 2
 x ´ 9 tan x (´2x )e csc( x ) cot( x ) ´ e + x ´ 9 sec x 
?
2 x3 ´9
 
 (tan2 x )( x3 ´ 9) 
A-35: (a)
y
(0,1)
t=0,π,2π

´ ?1 , 12 ?1 , 1
2 2 2
t=5π/4,7π/4 t=π/4,3π/4
x
(´1,0) (1,0)
t=3π/2 t=π/2
The particle traces the curve y = 1 ´ x2 restricted to domain [´1, 1]. At t = 0, the particle
is at the top of the curve, (1, 0). Then it moves to the right, and goes back and forth along
the ?curve, repeating its path every 2π units of time.
(b) 3
167
A-1: Ten speakers: 13 dB. One hundred speakers: 23 dB.
A-2: 20 log 2 « 14 years
A-3: (b)
1
A-4: f 1 ( x ) =
x
2
A-5: f 1 ( x ) =
x
2x + 1
A-6: f 1 ( x ) =
x2 + x
1
A-7: f 1 ( x ) =
x log 10
1 ´ 3 log x
A-8: y1 =
x4
d
A-9: log(sec θ ) = tan θ
dθ
écos(log x) sin(log x )
A-10: f 1 (x) =
x
3
2x + ?4x4
2 x +1
A-11: y1 = ?
x2 + x4 + 1
tan x
A-12: a
2 ´ log(cos x )
?
x2 + 4 + x 1
A-13: ? = ?
x x2 + 4 + x2 + 4 x2 + 4
?
2xe x2 1 + x 4 + 2x3
A-14: g1 ( x ) = 2 ?
e x 1 + x4 + 1 + x4
4
A-15:
3
3x 2x3
A-16: f 1 ( x ) = ´
x2 + 5 x4 + 10
40
A-17:
3
A-18: g1 ( x ) = π x log π + πx π´1
A-19: f 1 ( x ) = x x (log x + 1)
1
A-20: x x (log x + 1) +
x log 10
168
c !
( x4 + 12)( x4 ´ x2 4x3 4x3 ´ 2x

1 4 + 2) 3
A-21: f 1 (x) = + 4 ´
4 x3 4 2
x + 12 x ´ x + 2 x
A-22: h i
1 4x 2 16x3 25x4
f 1 ( x ) = ( x + 1)( x2 + 1)2 ( x3 + 1)3 ( x4 + 1)4 ( x5 + 1)5 x+ 1 + x 2 +1
+ x9x
3 +1 + x 4 +1
+ x 5 +1
x2 + 2x + 3 6x2

1 1
A-23: ´ ´
3x4 + 4x3 + 5 x + 2x + 3 3x + 4x + 5 2( x + 1)2
2 4 3
A-24: f 1 ( x ) = (cos x )sin x [(cos x ) log(cos x ) ´ sin x tan x ]

d x x
x
A-25: t(tan x ) u = (tan x ) log(tan x ) +
dx sin x cos x
2 +1
A-26: 2x ( x2 + 1) x (1 + log( x2 + 1))

x
A-27: f 1 ( x ) = ( x2 + 1)sin( x) ¨ cos x ¨ log( x2 + 1) + 2xx2sin +1
cos3 ( x )

3
A-28: xcos ( x) ¨ ´3 cos2 ( x ) sin( x ) log( x ) +
x
( x2 ´ 3) cos( x )

x 2 ´3
A-29: (3 + sin( x )) ¨ 2x log(3 + sin( x )) +
3 + sin( x )
g( x ) f 1 ( x )

d ! g( x )
)
g( x ) 1
A-30: [ f ( x )] = [ f ( x )] g ( x ) log( f ( x )) +
dx f (x)
f 1 (x)
A-31: Let g( x ) := log( f ( x )). Notice g1 ( x ) = f ( x) .
In order to show that the two curves have horizontal tangent lines at the same values of
x, we will show two things: first, that if f ( x ) has a horizontal tangent line at some value
of x, then also g( x ) has a horizontal tangent line at that value of x. Second, we will show
that if g( x ) has a horizontal tangent line at some value of x, then also f ( x ) has a
horizontal tangent line at that value of x.
Suppose f ( x ) has a horizontal tangent line where x = x0 for some point x0 . This means
f 1 (x ) f 1 (x )
f 1 ( x0 ) = 0. Then g1 ( x0 ) = f ( x 0) . Since f ( x0 ) ‰ 0, f ( x 0) = f (0x ) = 0, so g( x ) also has a
0 0 0
horizontal tangent line when x = x0 . This shows that whenever f has a horizontal
tangent line, g has one too.
Now suppose g( x ) has a horizontal tangent line where x = x0 for some point x0 . This
f 1 (x )
means g1 ( x0 ) = 0. Then g1 ( x0 ) = f ( x 0) = 0, so f 1 ( x0 ) exists and is equal to zero.
0
Therefore, f ( x ) also has a horizontal tangent line when x = x0 . This shows that
whenever g has a horizontal tangent line, f has one too.

A-1: (a) and (b)
dy dy
A-2: At (0, 4) and (0, ´4), is 0; at (0, 0), does not exist.
dx dx
169
A-3: (a) no (b) no
dy x dy
= ´ . It is not possible to write as a function of x, because (as stated in (b)) one
dx y dx
dy
value of x may give two values of . For instance, when x = π/4, at the point
dx
π 1 dy π ´1
,? the circle has slope = ´1, while at the point ,? the circle has slope
4 2 dx 4 2
dy
= 1.
dx
dy ex + y
A-4: =´ y
dx e +x
dy y2 + 1
A-5: = y
dx e ´ 2xy
1 1
A-6: At ( x, y) = (4, 1), y1 = ´ . At ( x, y) = (´4, 1), y1 = .
π+1 π´1
2x sin( x2 + y) + 3x2
A-7: ´
4y3 + sin( x2 + y)
6
A-8: At ( x, y) = (1, 0), y1 = ´6, and at ( x, y) = (´5, 0), y1 = 25 .
dy cos( x + y) ´ 2x
A-9: dx =
2y ´ cos( x + y)
A-10: At ( x, y) = (2, 0) we have y1 = ´ 32 , and at ( x, y) = (´4, 0) we have y1 = ´ 34 .

? ?
3 ´1 ´ 3 1
A-11: , ? , , ?
2 2 3 2 2 3
28
A-12: ´
3
dy 2xy2 + sin y
A-13: =´ 2
dx 2x y + x cos y
A-14: At ( x, y) = (2, 0), y1 = ´2. At ( x, y) = (´2, 0), y1 = 2.
A-15: x = 0, x = 1, x = ´1

A-1: (a) (´8, 8) (b) all integer multiples of π (c) [´1, 1]
A-2: False
A-3:
170
y
y = f ´1 ( x )
x
1
A-4:
• If |a| ą 1, there is no point where the curve has horizontal tangent line.
aπ
• If |a| = 1, the curve has a horizontal tangent line where x = 2πn +
2
for any integer n.
• If |a| ă 1, the curve has a horizontal tangent line where x = 2πn + arcsin( a) or
x = (2n + 1)π ´ arcsin( a) for any integer n.
A-5: Domain: x = ˘1. Not differentiable anywhere.
1
A-6: f 1 ( x ) = ? ; domain of f is [´3, 3].
9 ´ x2
2
´ ?t ´1 2 ´ 2t arccos t
1´t
A-7: f 1 (t) = , and the domain of f (t) is (´1, 1).
( t2 ´ 1)2
´2x
A-8: The domain of f ( x ) is all real numbers, and f 1 ( x ) = ? .
( x2 + 2) x4 + 4x2 + 3
1
A-9: f 1 ( x ) = and the domain of f ( x ) is all real numbers.
a2 + x 2
A-10: f 1 ( x ) = arcsin x, and the domain of f ( x ) is [´1, 1].
A-11: x = 0
d
A-12: tarcsin x + arccos xu = 0
dx
´1
A-13: y1 = b
1
x2 1 ´ x2
´1
A-14: y1 =
1 + x2
A-15: 2x arctan x + 1
171
A-16: Let θ = arctan x. Then θ is the angle of a right triangle that gives tan θ = x. In
particular, the ratio of the opposite side to the adjacent side is x. So, we have a triangle
that looks like this:
?
x2 + 1
x
θ
1
where the length of the hypotenuse came from the Pythagorean Theorem. Now,
opp x
sin (arctan x ) = sin θ = =? 2
hyp x +1
From here, we differentiate using the quotient rule:

?
" * x2 + 1 ´ x ?2x
d x 2 x 2 +1
? =
dx x2 + 1 x2 + 1
? 2
 ?
x2 + 1 ´ ? x2 x2 + 1
x +1 
= ¨?
x2 + 1 x2 + 1
( x 2 + 1) ´ x 2
=
( x2 + 1)3/2
1
= 2 3/2
= ( x2 + 1)´3/2
( x + 1)
A-17: Let θ = arcsin x. Then θ is the angle of a right triangle that gives sin θ = x. In
particular, the ratio of the opposite side to the hypotenuse is x. So, we have a triangle
1
x
θ
?
1 ´ x2
where the length of the adjacent side came from the Pythagorean Theorem. Now,
?
adj 1 ´ x2
cot (arcsin x ) = cot θ = =
opp x
172
#? ?
2
+ x ?´2x 2 ´ 1 ´ x2
d 1´x 2 1´x
=
dx x x2
´x2 ´ (1 ´ x2 )
= ?
x2 1 ´ x2
´1
= ?
x2 1 ´ x2
?
3 π

A-18: ( x, y) = ˘ 2 , 3
(2n + 1)π
A-19: x = for any integer n
2
1
A-20: g1 (y) =
1 ´ sin g(y)
1
A-21:
2
1
A-22:
e+1

log[sin x + 2] arcsec x ¨ cos x
A-23: f 1 (x) = [sin x + 2]arcsec x ? + . The domain of f ( x ) is
|x| x2 ´ 1 sin x + 2
|x| ě 1.
1
A-24: The function ? exists only for those values of x with x2 ´ 1 ą 0: that is, the
x2 ´1
1
domain of ? is |x| ą 1. However, the domain of arcsine is |x| ď 1. So, there is not
x2 ´ 1
1
one single value of x where arcsin x and ? are both defined.
x2 ´ 1
1
If the derivative of arcsin( x ) were given by ? , then the derivative of arcsin( x )
´1 x2
would not exist anywhere, so we would probably just write “derivative does not exist,”
instead of making up a function with a mismatched domain. Also, the function
f ( x ) = arcsin( x ) is a smooth curve–its derivative exists at every point strictly inside its
domain. (Remember not all curves are like this: for instance, g( x ) = |x| does not have a
derivative at x = 0, but x = 0 is strictly inside its domain.) So, it’s a pretty good bet that
1
the derivative of arcsine is not ? .
x2 ´ 1
1
A-25:
2
25
A-26: f ´1 (7) = ´
4
A-27: f (0) = ´7
173
a
2x 1 ´ ( x + 2y)2 ´ 1 2x cos( x2 + y2 ) ´ 1
A-28: y1 = a , or equivalently, y1 =
2 ´ 2y 1 ´ ( x + 2y)2 2 ´ 2y cos( x2 + y2 )

A-1: The caribou spent at least about 71 and a half hours travelling during its migration
(probably much more) in one year.
A-2: At some point during the day, the crane was travelling at exactly 10 kph.
A-3: One possible answer:
y
x
a c b
Another possible answer:

y
x
a c b
"
0 x ‰ 10
A-4: One possible answer: f ( x ) =
10 x = 10
174
"
10 x ‰ 0
Another answer: f ( x ) =
0 x=0
$
& 5 x ‰ 0, 10
Yet another answer: f ( x ) = 10 x = 10
0 x=0
%
A-5: (a) No such function is possible: Rolle’s Theorem guarantees f 1 (c) = 0 for at least
one point c P (1, 2).
For the other functions, examples are below, but many answers are possible.
(b) (c) (d)

y y y
x x x
1 2 1 2 1 2
A-6: The function f ( x ) is continuous over all real numbers, but it is only differentiable
when x ‰ 0. So, if we want to apply the MVT, our interval must consist of only positive
numbers or only negative numbers: the interval (´4, 13) is not valid.
It is possible to use the mean value theorem to prove what we want: if a = 1 and b = 144,
then f ( x ) is differentiable over the interval (1, 144) (since 0 is not contained in that
interval), and f ( x ) is continuous
a everywhere,
a so by the mean value theorem there exists
|144| ´ |1| 11 1
some point c where f 1 ( x ) = = = .
144 ´ 1 143 13
That being said, an easier way to prove that a point exists is to simply find it–without
? 1 169 1
using the MVT. When x ą 0, f ( x ) = x, so f 1 ( x ) = ? . Then f 1 = .
2 x 4 13
A-7: We note that f (0) = f (2π ) = 0. Then using the Mean Value Theorem (note that the
function is differentiable for all real numbers), we conclude that there exists c in (0, 2π )
such that
f (2π ) ´ f (0)
f 1 (c) = = 0.
2π ´ 0
A-8: We note that f (0) = f (1) = 0. Then using the Mean Value Theorem (note that the
function is differentiable for all real numbers), we get that there exists c P (0, 1) such that
f (1) ´ f (0)
f 1 (c) = = 0.
1´0
?
A-9: We note that f (0) = f (2π ) = 3 + π 2 . Then using the Mean Value Theorem (note
that the function is differentiable for all real numbers since 3 + sin x ą 0), we get that
there exists c P (0, 2π ) such that
f (2π ) ´ f (0)
f 1 (c) = = 0.
2π ´ 0
175
A-10: We note that f (0) = 0 and f (π/4) = 0. Then using the Mean Value Theorem (note
that the function is differentiable for all real numbers), we get that there exists
c P (0, π/4) such that
f (π/4) ´ f (0)
f 1 (c) = = 0.
π/4 ´ 0
A-11: 1
A-12: 2
A-13: 1
A-14: 1
A-15: (a)
2
f 1 ( x ) = 15x4 ´ 30x2 + 15 = 15 x4 ´ 2x2 + 1 = 15 x2 ´ 1 ě 0

The derivative is nonnegative everywhere. The only values of x for which f 1 ( x ) = 0 are 1
and ´1, so f 1 ( x ) ą 0 for every x in (´1, 1).
(b) If f ( x ) has two roots a and b in [´1, 1], then by Rolle’s Theorem, f 1 (c) = 0 for some x
strictly between a and b. But since a and b are in [´1, 1], and c is between a and b, that
means c is in (´1, 1); however, we know for every c in (´1, 1), f 1 (c) ą 0, so this can’t
happen. Therefore, f ( x ) does not have two roots a and b in [´1, 1]. This means f ( x ) has at
most one root in [´1, 1].
T
e ´1
A-16: log
T
π
A-17: C = 2. See the solution for the justification.
A-18: Since e´ f ( x) is always positive (regardless of the value of f ( x )),
1 1
f 1 (x) = f ( x )
ă =1
1+e ´ 1+0
for every x.
Since f 1 ( x ) exists for every x, we see that f is differentiable, so the Mean Value Theorem
applies. If f (100) is greater than or equal to 100, then by the Mean Value Theorem, there
would have to be some c between 0 and 100 such that
f (100) ´ f (0) 100
f 1 (c) = ě =1
100 100
Since f 1 ( x ) ď 1 for every x, there is no value of c as described. Therefore, it is not possible

that f (100) ě 100. So, f (100) ă 100.
A-19: Domain of f ´1 ( x ): (´8, 8)
Interval where f is one–to–one, and range of f ´1 ( x ): (´8, 8)
´1 : ´ 2π , 2π

A-20: One–to–one interval, and range of f 3 3
h ? ? i
´1 π 3 π 3
Domain of f : ´ 3 + 2 , 3 + 2
176
A-21: Define h( x ) = f ( x ) ´ g( x ), and notice h( a) = f ( a) ´ g( a) ă 0 and
h(b) = f (b) ´ g(b) ą 0. Since h is the difference of two functions that are continuous over
[ a, b] and differentiable over ( a, b), also h is continuous over [ a, b] and differentiable over
( a, b). So, by the Mean Value Theorem, there exists some c P ( a, b) with
h(b) ´ h( a)
h1 ( c ) =
bá
Since ( a, b) is an interval, b ą a, so the denominator of the above expression is positive;
since h(b) ą 0 ą h( a), also the numerator of the above expression is positive. So,
h1 (c) ą 0 for some c P ( a, b). Since h1 (c) = f 1 (c) ´ g1 (c), we conclude f 1 (c) ą g1 (c) for
some c P ( a, b).
A-22: 3
A-23: 2

A-1: e x
A-2: (ii), (iv)
3
A-3:
15!
dy 11 11
A-4: The derivative is only at the point (1, 3): it is not constantly , so it is wrong
dx 4 4
11 d2 y
to differentiate the constant to find . Below is a correct solution.
4 dx2
´28x + 2y + 2xy1 + 2yy1 = 0
´28 + 6 + 2y1 + 6y1 = 0

11
y1 = at the point (1, 3)
4
Differentiating the equation ´28x + 2y + 2xy1 + 2yy1 = 0:
´28 + 2y1 + 2y1 + 2xy2 + 2y1 y1 + 2yy2 = 0

4y1 + 2(y1 )2 + 2xy2 + 2yy2 = 28
11
At the point (1, 3), y1 = . Plugging in:
4
2
11 11
4 +2 + 2(1)y2 + 2(3)y2 = 28
4 4
15
y2 =
64
177
1
A-5: f 2 ( x ) =
x
d2 ´2x
A-6: tarctan xu =
dx2 (1 + x 2 )2
d2 y ´1
A-7: 2
= 3
dx y
A-8: 0
d3 2 100x (5x2 + 36)
A-9: tlog ( 5x ´ 12 ) u =
dx3 (5x2 ´ 12)3
A-10: speeding up
A-11: slower
A-12: ´4
A-13: (a) true (b) true (c) false
A-14: (ii)
A-15: f (n) = 2x (log 2)n
A-16: n = 4
2 2
A-17: (a) f 1 ( x ) = (1 + 2x )e x+ x f 2 ( x ) = (4x2 + 4x + 3)e x+ x h1 ( x ) = 1 + 3x
h2 ( x ) = 3
(b) f (0) = h(0) = 1; f 1 (0) = h1 (0) = 1; f 2 (0) = h2 (0) = 3
(c) f and h “start at the same place,” since f (0) = h(0). Also f 1 (0) = h1 (0), and
2 2
f 2 ( x ) = (4x2 + 4x + 3)e x+ x ą 3e x+ x ą 3 = h2 ( x ) when x ą 0. Since f 1 (0) = h1 (0), and
since f 1 grows faster than h1 for positive x, we conclude f 1 ( x ) ą h1 ( x ) for all positive x.
Now we can conclude that (since f (0) = h(0) and f grows faster than h when x ą 0) also
f ( x ) ą h( x ) for all positive x.
A-18:
4
(a) y1 (1) =
13
(b)
y
tangent
2
curve
1 x
A-19: (a) g2 ( x ) = [ f ( x ) + 2 f 1 ( x ) + f 2 ( x )]e x
(b) g3 ( x ) = [ f ( x ) + 3 f 1 ( x ) + 3 f 2 ( x ) + f 3 ( x )]e x
(c) g(4) ( x ) = [ f ( x ) + 4 f 1 ( x ) + 6 f 2 ( x ) + 4 f 3 ( x ) + f (4) ( x )]e x
A-20: m + n
178
A-21: 2
(a) In order to make f ( x ) a little more tractable, let’s change the format. Since
A-22: "
x xě0
|x| = , then:
´x x ă 0
´x2 x ă 0
"
f (x) =
x2 x ě 0.
Now, we turn to the definition of the derivative to figure out whether f 1 (0) exists.
f (0 + h ) ´ f (0) f (h) ´ 0 f (h)

f 1 (0) = lim = lim = lim if it exists.
hÑ0 h hÑ0 h hÑ0 h
Since f looks different to the left and right of 0, in order to evaluate this limit, we look at
the corresponding one-sided limits. Note that when h approaches 0 from the right, h ą 0
so f (h) = h2 . By contrast, when h approaches 0 from the left, h ă 0 so f (h) = ´h2 .
f (h) h2
lim = lim = lim h = 0
hÑ0+ h hÑ0+ h hÑ0+
f (h) ´h 2
lim = lim = lim ´h = 0
hÑ0´ h hÑ0´ h hÑ0´
Since both one-sided limits exist and are equal to 0,
f (0 + h ) ´ f (0)
lim =0
hÑ0 h
and so f is differentiable at x = 0 and f 1 (0) = 0.

(b) From (a), f 1 (0) = 0 and
´x2 x ă 0
"
f (x) =
x2 x ě 0.
So, "
1 ´2x x ă 0
f (x) =
2x x ě 0.
Then, we know the second derivative of f everywhere except at x = 0:

$
& ´2 x ă 0
2
f (x) = ?? x = 0
2 x ą 0.
%
So, whenever x ‰ 0, f 2 ( x ) exists. To investigate the differentiability of f 1 ( x ) when x = 0,

again we turn to the definition of a derivative. If
f 1 (0 + h ) ´ f 1 (0)
lim
hÑ0 h
179
exists, then f 2 (0) exists.
f 1 (0 + h ) ´ f 1 (0) f 1 (h) ´ 0 f 1 (h)

lim = lim = lim
Since f (h) behaves differently when h is greater than or less than zero, we look at the
one-sided limits.
f 1 (h) 2h
lim = lim =2
hÑ0+ h hÑ0+ h
f 1 (h) ´2h
lim = lim = ´2
hÑ0´ h hÑ0´ h
Since the one-sided limits do not agree,
f 1 (0 + h ) ´ f 1 (0)
lim = DNE
hÑ0 h
So, f 2 (0) does not exist. Now we have a complete picture of f 2 ( x ):
$
& ´2 xă0
2
f (x) = DNE x = 0
2 x ą 0.
%

A-1: False (but its velocity is zero)
m m
A-2: It takes 10 seconds to accelerate from 2 s to 3 s, and 100 seconds to accelerate from
3 ms to 13 ms .
A-4: True
A-5: The pot is falling at 14 metres per second, just as it hits the ground.
A-6: (a) 4.9x2 metres (b) 4.9x2 + x metres
?
A-7: 8 ´ 4 3 « 1 sec
A-8: 7.2 sec
A-9: 72 000 kph per hour
100
A-10: « 14 kph
7
A-11: about 1240 miles
A-12: 49 metres per second
A-13: About 416 metres
?
A-14: v0 = 1960 « 44 metres per second
180
A-15: « 74.2 kph
1 1
A-16: Time elapsed: + « 0.3 seconds
4.9 9.8
A-17: The acceleration is given by 2t v0 log 2, where v0 is the velocity of the object at time
t = 0.

A-1: ii and iv
3
A-2: ´
2
A-3: 6%
F1
A-4: (a) 0 (b) 100 = 15%, or F1 = 0.15F
F
17
A-5: ´ units per second
5
4
A-6: units per second
5
A-7: increasing at 7 mph
A-8: 8 cm per minute
13
A-9: ´ metres per second
6
3 cm
A-10: The height of the water is decreasing at = 0.1875 min .
16
1
A-11: metres per second (or about 1 centimetre every five minutes)
29200
 

2  6175 rad
A-12:  « 13.8 hour « 0.0038 rad

2 
3 sec
1235
72 + 4
24
A-13: (a) « 1.85 km/min (b) about .592 radians/min
13
?
55 21π
A-14: « 19 centimetres per hour.
42
dA cm2
A-15: = ´2π
dt s
A-16: 288π cubic units per unit time
A-17: 0 square centimetres per minute
7π cm3
A-18: ´ « ´1.8
12 sec2
181
?
7 m3
A-19: The flow is decreasing at a rate of .
1000 sec2
´15
A-20: « ´0.097 cm per minute
49π
dD 1
A-21: (a) = ? metres per hour
dt 2 2
(b) The river is higher than 2 metres.
(c) The river’s flow has reversed direction. (This can happen near an ocean at high tide.)
1
A-22: (a) 2 units per second (b) Its y-coordinate is decreasing at unit per second.
? 2
5
The point is moving at units per second.
2
h a i
A-23: (a) 10π = π 3( a + b) ´ ( a + 3b)(3a + b) or equivalently,
a
10 = 3( a + b) ´ ( a + 3b)(3a + b)
(b) 20πab
(c) The water is spilling out at about 375.4 cubic centimetres per second. The exact
?
cm3

200π 3 35 ´ 11
amount is ´ ? 1´2 ? .
9 ´ 35 3 35 ´ 13 sec
A-24: B(10) = 0
Answers to Exercises 3.3.1 — Jump to TABLE OF CONTENTS

A-1: (a), (b)
A-2: (a), (d)
A-3: If C = 0, then there was none to start out with, and Q(t) = 0 for all values of t.
If C ‰ 0, then Q(t) will never be 0 (but as t gets bigger and bigger, Q(t) gets closer and
closer to 0).
1
A-4: A = 5, k = ¨ log (π/5)
7
A-5: y(t) = 2e´3(t´1) , or equivalently, y(t) = 2e3 e´3t
10000
A-6: 5 ¨ 2´ 5730 « 1.5 µg
A-7: Radium-226 has a half life of about 1600 years.
log 2
A-8: = log6 (2) years, which is about 139 days
log 6
log 10
A-9: 120 ¨ seconds, or about six and a half minutes.
log 2
A-10:
About 0.5%of the sample decays in a day. The exact amount is
1
100 1 ´ 2´ 138 %.
182
A-11: After ten years, the sample contains between 6.2 and 6.8 µg of Uranium-232.

A-1: (a), (c), (d)
A-2: The temperature of the room is -10 degrees, and the room is colder than the object.
A-3: K is a negative number. It cannot be positive or zero.
A-4: If the object has a different initial temperature than its surroundings, then T (t) is
never equal to A. (But as time goes on, it gets closer and closer.)
If the object starts out with the same temperature as its surrounding, then T (t) = A for
all values of t.
´10 log(750)
A-5: 2
« 32.9 seconds
log 15
log(10)
A-6: 10 « 14.3 minutes
log(5)
A-7: If Newton adds his cream just before drinking, the coffee ends up cooler by 0.85˝ C.
A-8:

dT 1 4
(a) = log ( T ´ 30)
dt 5 5
5 log(2/5)
(b) « 20.53 min
log(4/5)
A-9: positive

A-1: If P(0) = 0, yes. If P(0) ‰ 0, no: it does not take into account external constraints on
population growth.
A-2: The Malthusian model predicts the herd will number 217 individuals in 2020.
log(3)
A-3: « 1.6 hours
log(2)
A-4: 1912 or 1913
106
A-5: « 1603
54 ´ 1
A-6: (a) At t = 0, there are 100 units of the radioactive isotope in the sample. k is negative.
(b) At t = 0, there are 100 individuals in the population. k is positive.
(c) The ambient temperature is 0 degrees. k is negative.
183
A-1: f (2) = 2e2π
A-2: Solutions to the differential equation have the form

9 7t 9
T ( t ) = T (0) + e ´
7 7
for some constant T (0).
8 log(0.4)
A-3: « 32.85 days
log(0.8)
A-4: 25˝ C
A-5:
(a) A(t) = 90,000 ¨ e0.05t ´ 40,000
When the graduate is 65, they will have $625,015.05 in the account.
(b) $49,437.96
A-6: (a) A(t) = 150,000 ´ 30,000 e0.06t (b) after 26.8 yrs
9 log 2
A-7: « 5.68 hr
log 3
g g
A-8: (a) v(t) = v0 + k e´kt ´ k
g
(b) lim v(t) = ´
tÑ8 k

A-1: Since f (0) is closer to g(0) than it is to h(0), you would probably want to estimate
f (0) « g(0) = 1 + 2 sin(1) if you had the means to efficiently figure out what sin(1) is,
and if you were concerned with accuracy. If you had a calculator, you could use this
estimation. Also, later in this chapter we will learn methods of approximating sin(1) that
do not require a calculator, but they do require time.
Without a calculator, or without a lot of time, using f (0) « h(0) = 0.7 probably makes
the most sense. It isn’t as accurate as f (0) « g(0), but you get an estimate very quickly,
without worrying about figuring out what sin(1) is.
A-2: log(0.93) « log(1) = 0
y
0.93 y = f (x)
x
1 y=0
184
A-3: arcsin(0.1) « 0
?
A-4: 3 tan(1) « 3
A-5: 10.13 « 103 = 1000

A-1: (a) f (5) = 6 (b) f 1 (5) = 3 (c) not enough information to know
A-2:
y
y = f (x)
x
2
The linear approximation is shown in red.

A-3: f ( x ) = 2x + 5
A-4: log(0.93) « ´0.07
y
y = x´1
y = f (x)
x
1
? 9
A-5: 5«
4
? 79
A-6: 5 30 «
40
A-7: 10.13 « 1030, 10.13 = 1030.301
A-8: There are many possible answers. One is f ( x ) = sin x, a = 0, and b = π.
?
A-9: a = 3
185
A-1: f (3) = 9, f 1 (3) = 0, f 2 (3) = ´2; there is not enough information to know f 3 (3).
A-2: f ( x ) « 2x + 5
A-3: log(0.93) « ´0.07245

1 449
A-4: cos «
15 450
A-5: e2x « 1 + 2x + 2x2
2754
A-6: One approximation: e 3 «
32
10 75
A-7: (a) 26 (b) 16 (c) (d)
11 64
A-8: For each of these, there are many solutions. We provide some below.
5
ÿ
(a) 1 + 2 + 3 + 4 + 5 = n
n =1
4
ÿ
(b) 2 + 4 + 6 + 8 = 2n
n =1
5
ÿ
(c) 3 + 5 + 7 + 9 + 11 = (2n + 1)
n =1
7
ÿ
(d) 9 + 16 + 25 + 36 + 49 = n2
n =3
7
ÿ
(e) 9 + 4 + 16 + 5 + 25 + 6 + 36 + 7 + 49 + 8 = ( n2 + n + 1)
n =3
7
ÿ
(f) 8 + 15 + 24 + 35 + 48 = ( n2 ´ 1)
n =3
6
ÿ
(g) 3 ´ 6 + 9 ´ 12 + 15 ´ 18 = (´1)n+1 3n
n =1
A-9: f (1) « 2, f (1) = π

A-10: e « 2.5
A-11: t(a), (d), (e)u, t(b), (g)u, t(c), (f)u

A-1: f 2 (1) = ´4
186
A-2: f (10) (5) = 10!
A-3: T3 ( x ) = ´x3 + x2 ´ x + 1
A-4: T3 ( x ) = ´7 + 7( x ´ 1) + 9( x ´ 1)2 + 5( x ´ 1)3 , or equivalently,
T3 ( x ) = 5x3 ´ 6x2 + 4x ´ 10
11 ¨ 10!
A-5: f (10) (5) =
6
?
A-6: a = e

A-1:
1 1 1 1 1 1 1 1 1 10 1 11
T16 ( x ) = 1+ x´ x2 ´ x3 + x4 + x5 ´ x6 ´ x7 + x8 + x9 ´ x ´ x
2 3! 4! 5! 6! 7! 8! 9! 10! 11!
1 1 1 1 1
+ x12 + x13 ´ x14 ´ x15 + x16
12! 13! 14! 15! 16!
A-2: T100 (t) = 127.5 + 48(t ´ 5) + 4.9(t ´ 5)2 = 4.9t2 ´ t + 10

n
ÿ 2(log 2)k
A-3: Tn ( x ) = ( x ´ 1) k
k!
k =0
A-4: T6 ( x ) = 7 + 5( x ´ 1) + 27 ( x ´ 1)2 + 13 ( x ´ 1)3 ´ 12

1
( x ´ 1)4 + 1
30 ( x ´ 1)
5 1
´ 60 ( x ´ 1)6
n
ÿ
A-5: Tn ( x ) = xk
k =0
A-6: T3 ( x ) = 1 + ( x ´ 1) + ( x ´ 1)2 + 21 ( x ´ 1)3

1 82 ?
A-7: π = 6 arctan ? « 3 « 3.156
3 45
100
ÿ (´1)k
A-8: T100 ( x ) = ´1 + ( x ´ 1) k
k ( k ´ 1)
k =2
řn (´1?)` π 2`
řn´1 (´1)`? π 2`+1

A-9: T2n ( x ) = `=0 (2`)! 2 x´ 4 + `=0 (2`+1)! 2 x´ 4
A-10:
1 1 1 1
1+
+ + +¨¨¨+ « e´1
2 3! 4! 157!
1
A-11: We estimate that the sum is close to ´ ? .
2

A-1:
187
y
f ( x + ∆x )
∆y y = f (x)
f (x)
∆x
x
x x + ∆x
A-2: Let f ( x ) be the number of problems finished after x minutes of work. The question
tells us that 5∆y « ∆x. So, if ∆x = 15, ∆y « 3. That is, in 15 minutes more, you will finish
about 3 more problems.
1 51
A-3: (a) ∆y « « 0.003846 (b) ∆y « « 0.003772
260 13520
A-4: (a) ∆y « 1.1 metres per second
(b) The increase from the first to the second jump is bigger.

A-1: False.
A-2: Absolute error: 0.17; percentage error: 2.92%
A-3: The linear approximates estimates the error in f ( x ) to be about 60, while the
quadratic approximates estimates the error in f ( x ) to be about 63.
A-4: 1%

9 d 9
A-5: (a) θ (b) θ = 2 arcsin (c) ? ¨ 0.02 « 0.03
2 6 36 ´ 0.682
A-6: We estimate that the volume decreased by about 0.00245 cubic metres, or about 2450
cubic centimetres.
A-7: Correct to within about 10.4 years (or about 53%)

A-1: (a) False (b) True (c) True (d) True
A-2: Equation 3.4.33 gives us the bound | f (2) ´ T3 (2)| ă 6. A calculator tells us actually
| f (2) ´ T3 (2)| « 1.056.
A-3: | f (37) ´ T (37)| = 0
A-4: You do, you clever goose!
188
9
A-5: | f (11.5) ´ T5 (11.5)| ă ă 0.02
7 ¨ 26
1
A-6: | f (0.1) ´ T2 (0.1)| ă
1125
ˇ ˇ
ˇ 1 1 ˇ 1
A-7: ˇˇ f ´ ´ T5 ´ ˇˇ ă ă 0.00004
4 4 6 ¨ 46
A-8: Your answer may vary. One reasonable answer is
14
| f (30) ´ T3 (30)| ă 7 ă 0.000002. Another reasonable answer is
5 ¨ 9 ¨ 15
14
| f (30) ´ T3 (30)| ă 7 ă 0.00002.
5 ¨9
2
2 100
A-9: Equation 3.4.33 gives the bound | f (0.01) ´ Tn (0.01)| ď 100 π ´1 .
A more reasonable bound on the error is that it is less than 5.

A-10: Using Equation 3.4.33,
ˇ ˇ
ˇ f 1 ´ T2 1 ˇ ă 1 .
ˇ ˇ
ˇ 2 2 ˇ 10
The actual error is

ˇ ˇ
ˇ f 1 ´ T2 1 ˇ = π ´ 1
ˇ ˇ
ˇ 2 2 ˇ 6 2
which is about 0.02.

A-11: Any n greater than or equal to 3.
?7 13
A-12: 2200 « 3 + « 3.00255
7 ¨ 36
A-13: If we’re going to use Equation 3.4.33, then we’ll probably be taking a Taylor
polynomial. Using Example 3.4.16, the 6th-degree Maclaurin polynomial for sin x is
x3 x5
T6 ( x ) = T5 ( x ) = x ´ +
3! 5!
so let’s play with this a bit. Equation 3.4.33 tells us that the error will depend on the
seventh derivative of f ( x ), which is ´ cos x:
17
f (1) ´ T6 (1) = f (7) (c)
7!
1 1 ´ cos c
sin(1) ´ 1 ´ + =
3! 5! 7!
101 ´ cos c
sin(1) ´ =
5! 7!
4242 ´ cos c
sin(1) =
7!
189
for some c between 0 and 1. Since ´1 ď cos c ď 1,
4242 ´ 1 4242 + 1
ď sin(1) ď
7! 7!
4241 4243
ď sin(1) ď
7! 7!
4241 4243
ď sin(1) ď
5040 5040
Remark: there are lots of ways to play with this idea to get better estimates. One way is
to take a higher-degree Maclaurin polynomial. Another is to note that, since
π 1
0 ă c ă 1 ă , then ă cos c ă 1, so
3 2
4242 ´ 1 4242 ´ 21
ă sin(1) ă
7! 7!
4241 8483 4243
ă sin(1) ă ă
5040 10080 5040
If you got tighter bounds than asked for in the problem, congratulations!
n
A-14: (a) T4 ( x ) = 4n=0 xn! (b) T4 (1) = 65
ř
24 (c) See the solution.

A-1: f 1 (0) = 0 and f 2 (0) = 6.
A-2: 4
1
A-3: h1 (2) = , h2 (2) = 0
2
A-4: (a) 1.92 (b) 1.918
13
A-5: 101/3 « ; this approximation is too big.
6
? 3
A-6: 2«
2
? 80
A-7: 3 26 «
27
A-8: (10.1)5 « 105, 000

101π π
A-9: sin «´
100 100

π 1
A-10: arctan(1.1) « +
4 20
8012
A-11:
1000
190
402 201
A-12: (8.06)2/3 « =
100 50
14
A-13: 1 + x + 2x2 + x3
3
A-14:
• By Equation 3.4.33, the absolute value of the error is
ˇ 3 ˇ ˇ ˇ
ˇ f (c) 3ˇ
ˇ ˇ c ˇ
ˇ
ˇ 3! ¨ ( 2 ´ 1 ) = ˇ ˇ
ˇ ˇ 6(22 ´ c2 ) ˇ
for some c P (1, 2).

• When 1 ď c ď 2, we know that 18 ď 22 ´ c2 ď 21, and that numerator and
denominator are non-negative, so
ˇ ˇ
ˇ c ˇ c 2 2
ˇ 6(22 ´ c2 ) ˇ = 6(22 ´ c2 ) ď 6(22 ´ c2 ) ď 6 ¨ 18
ˇ ˇ
1 1
= ď
54 50
as required.
• Alternatively, notice that c is an increasing function of c, while 22 ´ c2 is a
decreasing function of c. Hence the fraction is an increasing function of c and takes
its largest value at c = 2. Hence
ˇ ˇ
ˇ c ˇ 2 1 1
ˇ 6(22 ´ c2 ) ˇ ď 6 ˆ 18 = 54 ď 50 .
ˇ ˇ
A-15:
• By Equation 3.4.33, there is c P (0, 0.5) such that the error is
f (4) ( c )
R4 = (0.5 ´ 0)4
4!
1 cos(c2 )
= ¨
24 ¨ 16 3 ´ c
• For any c we have | cos(c2 )| ď 1, and for c ă 0.5 we have 3 ´ c ą 2.5, so that
ˇ cos(c2 ) ˇ
ˇ ˇ
1
ˇ 3 ´ c ˇ ď 2.5 .
ˇ ˇ
• We conclude that
1 1 1 1 1
|R4 | ď = ă = ă
2.5 ¨ 24 ¨ 16 60 ¨ 16 60 ¨ 10 600 500
A-16:
191
• By Equation 3.4.33, there is c P (0, 1) such that the error is
ˇ 3 ˇ ˇ ˇ
ˇ f (c) 3
ˇ ˇ eć ˇ
ˇ 3! ¨ (1 ´ 0) ˇ = ˇ 6(8 + c2 ) ˇ .
ˇ ˇ ˇ ˇ
• When 0 ă c ă 1, we know that 1 ą eć ą e´1 and 8 ď 8 + c2 ă 9, so

ˇ ˇ
ˇ eć ˇ eć
ˇ ˇ=
ˇ 6(8 + c2 ) ˇ 6(8 + c2 )
1
ă
6|8 + c2 |
1 1 1
ă = ă
6ˆ8 48 40
as required.
A-17: (a) 2.9259 (b) 2.9241 (c) 94 25´5/3
A-18: T3 ( x ) = 5x2 ´ 9
A-19: (a) 1.05 (b) 1.0483
(c)
y
y = 1 + 12 (x − 2)
y = y(x) ≈ 1 + 21 (x − 2) − 61 (x − 2)2
1
2 x
A-20: (a) 0.9 (b) 0.8867 (c)

y
y = −1 − (x + 1)
y = y(x) ≈ −1 − (x + 1) − 43 (x + 1)2 1
x −1
A-21: log 10.3 « 2.33259
The error is between ´0.00045 and ´0.00042.
A-22: (a) L( x ) = e + ex (b) Q( x ) = e + ex + ex2
(c) Since ex2 ą 0 for all x ą 0, L( x ) ă Q( x ) for all x ą 0.
From the error formula, we know that
1
f ( x ) = f (0) + f 1 (0) x + 12 f 2 (0) x2 + f 3 (c) x3
3!
1 c c
= Q( x ) + e + 3e2c + e3c ee x3
6
192
c
for some c between 0 and x. Since 61 ec + 3e2c + e3c ee is positive for any c, for all x ą 0,
1 c 2c 3c eec x3 ą 0, so Q ( x ) ă f ( x ).

6 e + 3e + e
(d) 1.105 ă e0.1 ă 1.115

A-1:
y
y = f (x)
There is a critical point at x = 0. The x-value of the red dot is a singular point, and a local
maximum occurs there.
A-2:
y
y = f (x)
x
a b
The x-coordinate corresponding to the blue dot (let’s call it a) is a critical point, and f ( x )
193
has a local and global minimum at x = a. The x-coordinate corresponding to the
discontinuity (let’s call it b) is a singular point, but there is not a global or local extremum
at x = b.
A-3: One possible answer is shown below.
x
2
A-4: The critical points are x = 3 and x = ´1. These two points are the only places where
local extrema might exist. There are no singular points.
A-5:
y y y y
x x x x
2 2 2 2
local max neither neither local max
A-6: There are many possible answers. Every answer must have x = 2 as a singular point
strictly inside the domain of f ( x ). Two possibilities are shown below.
y y
x x
2 2
A-7: x = ´7, x = ´1, and x = 5
194
A-8: Every real number c is a critical point of f ( x ), and f ( x ) has a local and global
maximum and minimum at x = c. There are no singular points.

A-1: Two examples are given below, but many are possible.
y
x
y = f (x)
y = f (x)
A-2: Two examples are given below, but many are possible.
y
y = ex
y = arctan x + 2
A-3: One possible answer:
195
y
y = f (x)
x
´5 5
A-4: The global maximum is 45 at x = 5 and the global minimum is ´19 at x = ´3.
A-5: The global maximum over the interval is 61 at x = ´3, and the global minimum is 7
at x = 0.

A-1: The global maximum is f (´1) = 6, the global minimum is f (´2) = ´20.
A-2: Global maximum is f (2) = 12, global minimum is f (1) = ´14.
A-3: Global maximum is f (4) = 30, global minimum is f (2) = ´10.
A-4: Local max at (´2, 20), local min at (2, ´12).
A-5: (´2, 33) max, and (2, ´31) min
?
A-6: Q should be 4 3 kilometres from A
A-7: 10 ˆ 30 ˆ 15
A-8: 2 ˆ 2 ˆ 6
?
A-9: X = Y = 2
?
A-10: The largest possible perimeter is 2 5R and the smallest possible perimeter is 2R.
A3/2
A-11: ?
3 6π
P2
A-12:
2( π + 4)
d c a
A Ap Ap
A-13: (a) x = ,y= , and z = ?
3p 3 3(1 + p )
(b) p = 1
196
c c
A 1 A
(The dimensions of the resulting baking pan are x = y = and z = .)
3 2 3
1
A-14: (a) x x (1 + log x ) (b) x = (c) local minimum
e
A-15: Maximum area: do not cut, make a circle and no square.

4
Minimum area: make a square out of a piece that is of the total length of the wire.
4+π
A-2: f ( x ) = A( x ) g( x ) = C ( x ) h( x ) = B( x ) k( x) = D( x)
A-3: (a) p = e2 (b) b = é2 1 ´ e2
2
A-4: vertical asymptote at x = 3; horizontal asymptotes lim f ( x ) =
xÑ˘8 3
A-5: horizontal asymptote y = 0 as x Ñ ´8; no other asymptotes
A-1: A1 ( x ) = l ( x ) B1 ( x ) = p ( x ) C1 ( x ) = n( x ) D1 ( x ) = o ( x ) E1 ( x ) = m ( x )
A-2: (´2, 8)
A-3: (1, 4)
A-4: (´8, 1)
A-1:
197
y
concave down
concave up concave down concave up
A-2:
y
x
´5 5

A-4: x = 1, y = 11
A-5: Let
g( x ) = f 2 ( x ) = x3 + 5x ´ 20.
Then g1 ( x ) = 3x2 + 5, which is always positive. That means g( x ) is strictly increasing for
all x. So, g( x ) can change signs once, from negative to positive, but it can never change
back to negative. An inflection point of f ( x ) occurs when g( x ) changes signs. So, f ( x )
has at most one inflection point.
Since g( x ) is continuous, we can apply the Intermediate Value Theorem to it. Notice
g(3) ą 0 while g(0) ă 0. By the IVT, g( x ) = 0 for at least one x P (0, 3). Since g( x ) is
strictly increasing, at the point where g( x ) = 0, g( x ) changes from negative to positive.
So, the concavity of f ( x ) changes. Therefore, f ( x ) has at least one inflection point.
Now that we’ve shown that f ( x ) has at most one inflection point, and at least one
inflection point, we conclude it has exactly one inflection point.
198
A-6: (a) Let
g( x ) = f 1 ( x )
Then f 2 ( x ) is the derivative of g( x ). Since f 2 ( x ) ą 0 for all x, g( x ) = f 1 ( x ) is strictly
increasing for all x. In other words, if y ą x then g(y) ą g( x ).
Suppose g( x ) = 0. Then for every y that is larger than x, g(y) ą g( x ), so g(y) ‰ 0.
Similarly, for every y that is smaller than x, g(y) ă g( x ), so g(y) ‰ 0. Therefore, g( x ) can
only be zero for at most one value of x. Since g( x ) = f 1 ( x ), that means f ( x ) can have at
most one critical point.
Suppose f 1 (c) = 0. Since f 1 ( x ) is a strictly increasing function, f 1 ( x ) ă 0 for all x ă c and
f 1 ( x ) ą 0 for all x ą c.
x ă c, so x ą c, so
f 1 (x)
ă f 1 (c) = 0 f 1 (x) ą f 1 (c) = 0
Then f ( x ) is decreasing for x ă c and increasing for x ą c. So f ( x ) ą f (c) for all x ‰ c.
y = f (x)
xăc c xąc
f ( x ) decreasing, so f ( x ) increasing, so
f ( x ) ą f (c) f ( x ) ą f (c)
Since f ( x ) ą f (c) for all x ‰ c, so c is an absolute minimum for f ( x ).

(b) We know that the maximum over an interval occurs at an endpoint, a critical point, or
a singular point.
• Since f 1 ( x ) exists everywhere, there are no singular points.
• If the maximum were achieved at a critical point, that critical point would have to
provide both the absolute maximum and the absolute minimum (by part (a)). So,
the function would have to be a constant and consequently could not have a
nonzero second derivative. So the maximum is not at a critical point.
That leaves only the endpoints of the interval.
A-7: If x = 3 is an inflection point, then the concavity of f ( x ) changes at x = 3. That is,
there is some interval strictly containing 3, with endpoints a and b, such that
199
Since f 2 ( a) ă 0 and f 2 (b) ą 0, and since f 2 ( x ) is continuous, the Intermediate Value

Theorem tells us that there exists some x strictly between a and b with f 2 ( x ) = 0. So, we
know f 2 ( x ) = 0 somewhere between a and b. The question is, where exactly could that
be?
• f 2 ( x ) ă 0 (and hence f 2 ( x ) ‰ 0) for all x between a and 3
• f 2 ( x ) ą 0 (and hence f 2 ( x ) ‰ 0) for all x between b and 3
• So, any number between a and b that is not 3 has f 2 ( x ) ‰ 0.
So, x = 3 is the only possible place between a and b where f 2 ( x ) could be zero.
Therefore, f 2 (3) = 0.
A-1: even
A-2: odd, periodic
A-3:
A-4:
200
y
A-5: A function is even if f (´x ) = f ( x ).
(´x )4 ´ (´x )6
f (´x ) = 2
e(´x)
x4 ´ x6
= 2
ex
= f (x)
So, f ( x ) is even.
A-6: For any real number x, we will show that f ( x ) = f ( x + 4π ).

x + 4π
f ( x + 4π ) = sin( x + 4π ) + cos
2
x
= sin( x + 4π ) + cos + 2π
x 2
= sin( x ) + cos
2
= f (x)
So, f ( x ) is periodic.
A-7: even
A-8: none
A-9: 1
A-10: π
201
A-1: (a) (´8, 3]

(b) f ( x ) in increasing on (´8, 2) and decreasing on (2, 3). There is a local maximum at
x = 2 and a local minimum at the endpoint x = 3.
(c) f ( x ) is always concave down and has no inflection points.
(d) (3, 0)
(e)
(2.2)
x
(3, 0)
A-2: The open dot is the inflection point, and the closed dot is the local and global
maximum.
y 2, 83

x
? ?
3
2 2 3 20
A-3: The open dot marks the inflection point.
202
y
x
? ?
´34 ´1 3
2
A-4:
?
3 3
2
x
? ?
´ 3 ´1 1 3
?
´3 2 3
A-5: (a) One branch of the function, the exponential function e x , is continuous
x2 + 3
everywhere. So f ( x ) is continuous for x ă 0. When x ě 0, f ( x ) = , which is
3( x + 1)
continuous whenever x ‰ ´1 (so it’s continuous for all x ą 0). So, f ( x ) is continuous for
203
x ą 0. To see that f ( x ) is continuous at x = 0, we see:
lim f ( x ) = lim e x = 1
xÑ0´ xÑ0´
x2 + 3
lim f ( x ) = lim =1
xÑ0+ xÑ0+ 3( x + 1)
So, lim f ( x ) = 1 = f (0)
xÑ0
Hence f ( x ) is continuous at x = 0, so f ( x ) is continuous everywhere.

(b) i.
f ( x ) is increasing for x ă 0and x ą 1, decreasing for 0 ă x ă 1, has a local max at (0, 1),
and has a local min at 1, 32 .
ii.
f ( x ) is concave upwards for all x ‰ 0.
iii.
The x–axis is a horizontal asymptote as x Ñ ´8.
(c)
(0, 1)
(1, 23 )
A-6:
y
x
−1 0.5
? (´1, 1?)
A-7: (a) Increasing: decreasing: (´8, ´1) Y (1, 8?) ?
concave up: (´ 3, 0) Y ?( 3, 8) concave down: (´8, ´ 3) Y (0, 3)
inflection points: x = ˘ 3, 0
(b) The local and global minimum of f ( x ) is at (´1, ´1

? ), and the local and global
e
maximum of f ( x ) is at (1, ?1e ).
(c) In the graph below, open dots are inflection points, and solid dots are extrema.
204
y
1
?
e
? x
´ 3 ?
´1 1 3
´1
?
e
A-8: Local maxima occur at x = 2π3 + 2πn for all integers n, and local minima occur at
2π
x = ´ 3 + 2πn for all integers n. Inflection points occur at every integer multiple of π.
x
´14π ´10π´8π ´4π ´2π 2π 4π 8π 10π 14π
3 3 3 3 3 3 3 3 3 3
A-9: Below is the graph y = f ( x ) over the interval [´π, π ]. The sketch of the curve over a
larger domain is simply a repetition of this figure.
205
y
´π ´ b b a πá
x
´π ´ π2 π π
2
´ 5π
6
´ π6
´2
´3
On the interval [0, π ], the maximum value of f ( x ) is 6 and the minimum value is ´2.
?
´1 ´1 + 33
Let a = sin « 0.635 « 0.2π and
? 8

´1 ´1 ´ 33
b = sin « ´1.003 « ´0.3π. The points ´π ´ b, b, a, and π ´ a are
8
inflection points.
A-10: The closed dot is?the local minimum, and the open dots are inflection points at
x = ´1 and x = ´2 ˘ 1.5. The graph has horizontal asymptotes y = 0 as x goes to ˘8.
206
y
x
´2 ´1
´1
A-11: (a) decreasing for x ă 0 and x ą 2, increasing for 0 ă x ă 2, minimum at (0, 0),
maximum at (2, 2). ? ? ? ?
(b) concave up for x ă 2 ´ ?2 and x ą 2 + 2, concave down for 2 ´ 2 ă x ă 2 + 2,
inflection points at x = 2 ˘ 2.
(c)8
(d) Open dots indicate inflection points, and closed dots indicate local extrema.
x
? ?
2´ 2 2 2+ 2
A-12: (a)
207
y
1
y = f (x)
x
1
There are no inflection points or extrema, except the endpoint (0, 1).
(b)
1 y = g( x )
x
1
(c) The domain of g is (0, 1]. The range of g is [0, 8).
(d) g1 ( 21 ) = ´2
A-13: (a)
208
y
4
? 1
?
545 4
5
x
´1
? ´4
4 ?
5 545
1 1
Local maximum at x = ´ ? 4 ; local minimum at x = 4 ;
? inflection point at the origin;
5 5
concave down for x ă 0 ; concave up for x ą 0.
(b) The number of distinct real roots of x5 ´ x + k is:
4
• 1 when |k| ą ?
545
4
• 2 when |k| = ?
545
4
• 3 when |k| ă ?
545
A-14: (a)
y y
y = sinh x y = cosh x
1
x x
(b) For any real x, define sinh´1 ( x ) to be the unique solution of sinh(y) = x. For every
x P [1, 8), define cosh´1 ( x ) to be the unique y P [0, 8) that obeys cosh(y) = x.
209
y y
y = sinh´1 x y = cosh´1 x
x x
d 1
(c) tcosh´1 ( x )u = ?
dx x2 ´ 1

A-1: There are many possible answers. Here is one: f ( x ) = 5x, g( x ) = 2x.
A-2: There are many possible answers. Here is one: f ( x ) = x, g( x ) = x2 .
A-3: There are many possible answers. Here is one: f ( x ) = 1 + 1x , g( x ) = x log 5 (recall
we use log to mean logarithm base e).
2
A-4: ´
π
A-5: ´8
A-6: 0
A-7: 0
A-8: 3
A-9: 2
A-10: 0
1
A-11: 2
A-12: 0
A-13: 5
A-14: 8
A-15: 3
3
A-16: 2
A-17: 0
1
A-18: 3
A-19: 0
A-20: ?1
e
210
A-21: 1
A-22: 1
A-23: c = 0
$
2
ek sin( x ) ´ (1 + 2x2 ) & ´8 k ă 2
A-24: lim = 2 k=2
xÑ0 x4 %
8 ką2
A-25:
• We want to find the limit as n goes to infinity of the percentage error,
|S(n) ´ A(n)|
lim 100 . Since A(n) is a nicer function than S(n), let’s simplify:
nÑ8 |S(n)| ˇ ˇ
|S(n) ´ A(n)| ˇ A ( n ) ˇ
lim 100 = 100 ˇˇ1 ´ lim ˇ.
nÑ8 |S(n)| nÑ8 S ( n ) ˇ
We figure out this limit the natural way:

ˇ ˇ
ˇ ˇ
ˇ ˇ
4
ˇ ˇ ˇ ˇ
ˇ A(n) ˇ ˇ ˇ 5n ˇ
100 ˇ1 ´ lim
ˇ = 100 ˇ 1 ´ lim 4 ´ 13n3 ´ 4n + log( n ) ˇ
ˇ
nÑ8 S ( n ) ˇ ˇ
ˇ nÑ8 5n
ˇ loooooooooooooooomoooooooooooooooonˇˇ
ˇ numÑ8 ˇ
denÑ8
ˇ ˇ
ˇ 20n3 ˇ
= 100 ˇ1 ´ lim
ˇ ˇ
ˇ 3 2
nÑ8 20n ´ 39n ´ 4 + ˇ 1 ˇ
n
ˇ ˇ
ˇ n 3 20 ˇ
= 100 ˇ1 ´ lim 3 ¨
ˇ ˇ
nÑ8 n 39 4 1 ˇ
ˇ 20 ´ ´ 3 + 4 ˇ
n n n
= 100|1 ´ 1| = 0
So, as n gets larger and larger, the relative error in the approximation gets closer
and closer to 0.
• Now, let’s look at the absolute error.
lim |S(n) ´ A(n)| = lim | ´ 13n3 ´ 4n + log n| = 8

nÑ8 nÑ8
So although the error gets small relative to the giant numbers we’re talking about, the
absolute error grows without bound.

A-1: F ( x ) = f ( x ) + C
A-2: C ( x )
A-3: F ( x ) = x3 + x5 + 5x2 ´ 9x + C
3 8 18 5 1 2
A-4: F ( x ) = x ´ x + x +C
40 5 2
211
4 45
A-5: F ( x ) = 3x 3 + +C
17x1.7
2?
A-6: F ( x ) = x+C
7
1
A-7: F ( x ) = e5x+11 + C
5
3 7
A-8: F ( x ) = ´ cos(5x ) + sin(13x ) + C
5 13
A-9: F ( x ) = tan( x + 1) + C
A-10: F ( x ) = log |x + 2| + C
7
A-11: F ( x ) = ? arcsin( x ) + C
3
1
A-12: F ( x ) = arctan(5x ) + C
5
9 19
A-13: f ( x ) = x3 ´ x2 + 4x +
2 2
1
A-14: f ( x ) = sin(2x ) + π
2
A-15: f ( x ) = log |x|
A-16: f ( x ) = arcsin x + x ´ π ´ 1
A-17: It takes 12 log 7 hours (about 58 minutes) for the initial colony to increase by 300
individuals.
t
A-18: At time t, the amount of money in your bank account is 75000e 50 + C dollars, for
some constant C.
24
A-19: + 6 « 13.6 kWh
π
1
A-20: f 1 ( x ) = g1 ( x ) = ? ; f and g differ only by a constant.
x ´ x2
A-21: F ( x ) = sin(2x ) cos(3x ) + C
ex
A-22: F ( x ) = +C
x2 + 1
3
A-23: F ( x ) = e x + C
5
A-24: F ( x ) = ´ cos( x2 ) + C
2
1
A-25: F ( x ) = x2 + C.
2

x
A-26: F ( x ) = 7 arcsin ? +C
3
212

1 5 2 3
A-27: V ( H ) = 2π H + H +H
5 3
213
214
Part IV
S OLUTIONS TO PROBLEMS
215
Solutions to Exercises 1.1 — Jump to TABLE OF CONTENTS
S-1:
b
y
y = f (x)
c
Q
P
x
The tangent line to y = f ( x ) at a point should go through the point, and be “in the same
direction” as f at that point. The secant line through P and Q is simply the straight line
passing through P and Q.
S-2:
(a) True: since y = 2x + 3 is the tangent line to y = f ( x ) at the point x = 2, this means the
function and the tangent line have the same value at x = 2. So f (2) = 2(2) + 3 = 7.
(b) In general, this is false. We are only guaranteed that the curve y = f ( x ) and its
tangent line y = 2x + 3 agree at x = 2. The functions f ( x ) and 2x + 3 may or may not
take the same values when x ‰ 2. For example, if f ( x ) = 2x + 3, then of course f ( x )
and 2x + 3 agree for all values of x. But if f ( x ) = 2x + 3 + ( x ´ 2)2 , then f ( x ) and
2x + 3 agree only for x = 2.
S-3: Since the tangent line to the curve at point P passes through point P, the curve and
the tangent line touch at point P. So, they must intersect at least once. By drawing
various examples, we can see that different curves may touch their tangent lines exactly
once, exactly twice, exactly three times, etc.
y y = f (x)
x
P
Blue tangent line touches curve only once, at P
217
y
P
y
y = f (x) x
P
x y = f (x)
Blue tangent line touches curve twice Blue tangent line touches curve many times
S-1: Speed is nonnegative; velocity has a sign (positive or negative) that indicates
direction.
S-2: Yes–an object that is not moving has speed 0.
S-3: Since you started and ended in the same place, your difference in position was 0,
0
and your difference in time was 24 hours. So, your average velocity was = 0 kph.
24
S-4: Objects accelerate as they fall — their speed gets bigger and bigger. So in the entire
first second of falling, the object is at its fastest at the one-second mark. We have defined
the average speed over a time interval to be distance moved
time taken . We do not yet know how to
compute the distanced travelled by an object whose speed is not constant. But our
intuition certainly says that the distance travelled is no more than the maximum speed
times the time taken, so that the average speed is no larger that the maximum speed. In
the next CLP-1 chapter we will verify that intuition mathematically. See Example 2.13.6
in the CLP-1 text. So the average speed will be a smaller number then the speed at the
one-second mark, which is the maximum speed.
change in vertical component

S-5: The slope of a curve is given by . The change in the
change in horizontal component
vertical component is exactly s(b) ´ s( a), and the change in the horizontal component is
exactly b ´ a.
S-6: The velocity is positive when the object is going in the increasing direction; it is
going “up” on the graph when t is between 0 and 2, and when t is between 6 and 7. So,
the velocity is positive when t is in (0, 2) Y (6, 7).
218
S-7:
change in position s (5) ´ s (3) ( 3 ¨ 52 + 5 ) ´ ( 3 ¨ 32 + 5 )
(a) Average velocity: = = = 24
change in time 5´3 5´3
units per second.
(b) From the notes, we know the velocity of an object at time a is
s( a + h) ´ s( a)
v( a) = lim
hÑ0 h
So, in our case:
s (1 + h ) ´ s (1) [3(1 + h )2 + 5] ´ [3(1)2 + 5] 6h + 3h2

v(1) = lim = lim = lim = lim 6 + 3h = 6
hÑ0 h hÑ0 h hÑ0 h hÑ0
So the velocity when t = 1 is 6 units per second.
S-8:
change in position s (9) ´ s (1) 3´1 1
(a) Average velocity: = = = units per second.
change in time 9´1 9´1 4
(b) From the notes, we know the velocity of an object at time a is
s( a + h) ´ s( a)
v( a) = lim
hÑ0 h
So, in our case:
?
s (1 + h ) ´ s (1) 1+h´1
v(1) = lim = lim
hÑ0 h hÑ0 h
? ?
1+h´1 1+h+1
= lim ¨ ?
hÑ0 h 1+h+1
(1 + h ) ´ 1
= lim ?
hÑ0 h ( 1 + h + 1)
1 1
= lim ? = units per second
hÑ0 1 + h + 1 2
(c)
?
s (9 + h ) ´ s (9) 9+h´3
v(9) = lim = lim
hÑ0 h hÑ0 h
? ?
9+h´3 9+h+3
= lim ¨ ?
hÑ0 h 9+h+3
(9 + h ) ´ 9
= lim ?
hÑ0 h ( 9 + h + 3)
1 1
= lim ? = units per second
hÑ0 9 + h + 3 6
219
Remark: the average velocity is not the average of the two instantaneous velocities.
S-1:
(a) lim f ( x ) = 1: as x gets very close to ´2, y gets very close to 1.

xÑ´2
(b) lim f ( x ) = 0: as x gets very close to 0, y also gets very close to 0.

xÑ0
(c) lim f ( x ) = 2: as x gets very close to 2, y gets very close to 2. We ignore the value of
xÑ2
the function where x is exactly 2.
S-2: The limit does not exist. As x approaches 0 from the left, y approaches -1; as x
approaches 0 from the right, y approaches 1. This tells us lim f ( x ) = ´1 and
xÑ0´
lim f ( x ) = 1, but neither of these are what the question asked. Since the limits from left
xÑ0+
and right do not agree, the limit does not exist. Put another way, there is no single
number y approaches as x approaches 0, so the limit lim f ( x ) does not exist.
xÑ0
S-3:
(a) lim f ( x ) = 2: as x approaches ´1 from the left, y approaches 2. It doesn’t matter

xÑ´1´
that the function isn’t defined at x = ´1, and it doesn’t matter what happens to the
right of x = ´1.
(b) lim f ( x ) = ´2: as x approaches ´1 from the right, y approaches -2. It doesn’t
xÑ´1+
matter that the function isn’t defined at ´1, and it doesn’t matter what happens to
the left of ´1.
(c) lim f ( x ) = DNE: since the limits from the left and right don’t agree, the limit does
xÑ´1
not exist.
(d) lim f ( x ) = 0: as x approaches ´2 from the right, y approaches 0. It doesn’t matter

xÑ´2+
that the function isn’t defined at 2, or to the left of 2.
(e) lim f ( x ) = 0: as x approaches 2 from the left, y approaches 0. It doesn’t matter that
xÑ2´
the function isn’t defined at 2, or to the right of 2.
S-4: Many answers are possible; here is one.
220
y
y = f (x)
10
x
3
As x gets closer and closer to 3, y gets closer and closer to 10: this shows lim f ( x ) = 10.
xÑ3
Also, at 3 itself, the function takes the value 10; this shows f (3) = 10.
y
y = f (x)
10
x
3
Note that, as x gets closer and closer to 3 except at 3 itself, y gets closer and closer to 10:
this shows lim f ( x ) = 10. Then, when x = 3, the function has value 0: this shows
xÑ3
f (3) = 0.
S-6: In general, this is false. The limit as x goes to 3 does not take into account the value
of the function at 3: f (3) can be anything.
S-7: False. The limit as x goes to 3 does not take into account the value of the function at
3: f (3) tells us nothing about lim f ( x ).
xÑ3
S-8: lim f ( x ) = 16: in order for the limit lim f ( x ) to exist and be equal to 16, both one
xÑ´2´ xÑ2
221
sided limits must exist and be equal to 16.
S-9: Not enough information to say. If lim f ( x ) = 16, then lim f ( x ) = 16. If
xÑ´2+ xÑ´2
lim f ( x ) ‰ 16, then lim f ( x ) does not exist.
xÑ´2+ xÑ´2
S-10: lim sin t = 0: as t approaches 0, sin t approaches 0 as well.

tÑ0
S-11: lim log x = ´8: as x approaches 0 from the right, log x is negative and
xÑ0+
increasingly large, growing without bound.
S-12: lim y2 = 9: as y gets closer and closer to 3, y2 gets closer and closer to 32 .
yÑ3
1 1
S-13: lim = ´8: as x gets closer and closer to 0 from the left, becomes a larger and
xÑ0´ x x
larger negative number.
1 1
S-14: lim = DNE: as x gets closer and closer to 0 from the left, becomes a larger and
xÑ0 x x
1
larger negative number; but as x gets closer and closer to 0 from the right, becomes a
x
larger and larger positive number. So the limit from the left is not the same as the limit
1
from the right, and so lim = DNE. Contrast this with Question 15.
xÑ0 x
1 1
S-15: lim = 8: as x gets closer and closer to 0 from the either side, becomes a larger
xÑ0 x2 x2
and larger positive number, growing without bound. Contrast this with Question 14.
1 1 1 1
S-16: lim = : no matter what x is, is always . In particular, as x approaches 3,
xÑ3 10 10 10 10
1 1
stays put at .
10 10
S-17: When x is very close to 3, f ( x ) looks like the function x2 . So: lim f ( x ) = lim x2 = 9
xÑ3 xÑ3
S-1: Zeroes cause a problem when they show up in the denominator, so we can only
compute (a) and (d). (Both these limits are zero.) Be careful: there is no such rule as “zero
divided by zero is one,” or “zero divided by zero is zero.”
222
f (x)
S-2: The statement lim = 10 tells us that, as x gets very close to 3, f ( x ) is 10 times as
xÑ3 g ( x )
f (x) f (x)
large as g( x ). We notice that if f ( x ) = 10g( x ), then = 10, so lim = 10 wherever
g( x ) xÑ g ( x )
f and g exist. So it’s enough to find a function g( x ) that has limit 0 at 3. Such a function is
(for example) g( x ) = x ´ 3. So, we take f ( x ) = 10( x ´ 3) and g( x ) = x ´ 3. It is easy now
f (x) 10( x ´ 3)
to check that lim f ( x ) = lim g( x ) = 0 and lim = lim = lim 10 = 10.
xÑ3 xÑ3 xÑ3 g ( x ) xÑ3 x´3 xÑ3
S-3:
• As we saw in Question 2, x ´ 3 is a function with limit 0 at x = 3. So one way of
f (x)
thinking about this question is to try choosing f ( x ) so that g( x) = g( x ) = x ´ 3 too,
which leads us to the solution f ( x ) = ( x ´ 3)2 and g( x ) = x ´ 3. This is one of
many, many possible answers.
• Another way of thinking about this problem is that f ( x ) should go to 0 “more
strongly” than g( x ) when x approaches 3. One way of a function going to 0 really
strongly is to make that function identically zero. So we can set f ( x ) = 0 and
f (x)
g( x ) = x ´ 3. Now is equal to 0 whenever x ‰ 3, and is undefined at x = 3.
g( x )
Since the limit as x goes to three does not take into account the value of the function
f (x)
at 3, we have lim = 0.
xÑ3 g ( x )
There are many more possible answers.
1 1
S-4: One way to start this problem is to remember lim 2
= 8. (Using 2 as opposed to
xÑ0 x x
1 1
is important, since lim does not exist.) Then by “shifting” by three, we find
x xÑ0 x
1 f (x) 1
lim 2
= 8. So it is enough to arrange that = . We can achieve this
xÑ3 ( x ´ 3) g( x ) ( x ´ 3)2
with f ( x ) = x ´ 3 and g( x ) = ( x ´ 3)3 , and maintain lim f ( x ) = lim g( x ) = 0. Again,
xÑ3 xÑ3
this is one of many possible solutions.
S-5: Any real number; positive infinity; negative infinity; does not exist.
0
This is an important thing to remember: often, people see limits that look like and
0
think that the limit must be 1, or 0, or infinite. In fact, this limit could be anything–it
depends on the relationship between f and g.
Questions 2 and 3 show us examples where the limit is 10 and 0; they can easily be
modified to make the limit any real number.
Question 4 show us an example where the limit is 8; it can easily be modified to make
the limit ´8 or DNE.
223
S-6: Since we’re not trying to divide by 0, or multiply by
2(t ´ 10)2 2¨0
infinity: lim = =0
tÑ10 t 10
S-7: Since we’re not doing anything dodgy like putting 0 in the denominator,
(y + 1)(y + 2)(y + 3) (0 + 1)(0 + 2)(0 + 3) 6
lim = = = 6.
yÑ0 cos y cos 0 1
S-8: Since the limits of the numerator and denominator exist, and since the limit of the
4x ´ 2 4 4(3) ´ 2 4

denominator is nonzero: lim = = 16
xÑ3 x+2 3+2
S-9:
lim (1 ´ t)
1´t tÑ´3
lim = = 4/ cos(´3) = 4/ cos(3)
tÑ´3 cos(t) lim cos(t)
tÑ´3
S-10: If try naively then we get 0/0, so we expand and then simplify:
(2 + h )2 ´ 4 h2 + 4h + 4 ´ 4 h
= = +2
2h 2h 2

h
Hence the limit is lim + 2 = 2.
hÑ0 2
S-11:

t´5 limtÑ´2 (t ´ 5)
lim = = ´7/2.
tÑ´2 t+4 limtÑ´2 (t + 4)
S-12: a c b ?
lim 5x3 + 4 = lim 5x3 + 4 = 5 lim( x3 ) + 4 = 9 = 3.
tÑ1 tÑ1 tÑ1
S-13:
lim (t ´ 2)
t´2 tÑ´1
lim = = ´3/2.
tÑ´1 t+3 lim (t + 3)
tÑ´1

log(1 + x ) ´ x
S-14: We simply plug in x = 1: lim = log(2) ´ 1.
xÑ1 x2
224
S-15: If we try naively then we get 0/0, so we simplify first:
x´2 x´2 1
2
= =
x ´4 ( x ´ 2)( x + 2) x+2
1
Hence the limit is lim = 1/4.
xÑ2 x + 2
S-16: If we try to plug in x = 4, we find the denominator is zero. So to get a better idea of
what’s happening, we factor the numerator and denominator:
x2 ´ 4x x ( x ´ 4)
lim 2 = lim
xÑ4 x ´ 16 xÑ4 ( x + 4)( x ´ 4)
x
= lim
xÑ4 x + 4
4 1
= =
8 2
S-17: If we try to plug in x = 2, we find the denominator is zero. So to get a better idea of
what’s happening, we factor the numerator:
x2 + x ´ 6 ( x + 3)( x ´ 2)
lim = lim
xÑ2 x´2 xÑ2 x´2
= lim ( x + 3) = 5
xÑ2
S-18: If we try naively then we get 0/0, so we simplify first:
x2 ´ 9 ( x ´ 3)( x + 3)
= = x´3
x+3 ( x + 3)
Hence the limit is lim ( x ´ 3) = ´6.

xÑ´3
S-19: To calculate the limit of a polynomial, we simply evaluate the polynomial:

1 1
lim t4 ´ 3t3 + t = ¨ 24 ´ 3 ¨ 23 + 2 = ´14
tÑ2 2 2
225
S-20:
? ? ?
x2 + 8 ´ 3
x2 + 8 ´ 3 x2 + 8 + 3
= ¨?
x+1 x+1 x2 + 8 + 3
( x2 + 8) ´ 32
= ?
( x + 1)( x2 + 8 + 3)
x2 ´ 1
= ?
( x + 1)( x2 + 8 + 3)
( x ´ 1)( x + 1)
= ?
( x + 1)( x2 + 8 + 3)
( x ´ 1)
=? 2
? x +8+3
2
x +8´3 ( x ´ 1)
lim = lim ?
xÑ´1 x+1 xÑ´1 x2 + 8 + 3
´2
=?
9+3
2 1
=´ =´ .
6 3
S-21: If we try to do the limit naively we get 0/0. Hence we must simplify.
? ? ? ? ? ?
x + 7 ´ 11 ´ x x + 7 ´ 11 ´ x x + 7 + 11 ´ x
= ¨ ? ?
2x ´ 4 2x ´ 4 x + 7 + 11 ´ x
( x + 7) ´ (11 ´ x )
= ? ?
(2x ´ 4)( x + 7 + 11 ´ x )
2x ´ 4
= ? ?
(2x ´ 4)( x + 7 + 11 ´ x )
1
=? ?
x + 7 + 11 ´ x
? ?
x + 7 ´ 11 ´ x 1
So, lim = lim ? ?
xÑ2 2x ´ 4 xÑ2 x + 7 + 11 ´ x
1
=? ?
9+ 9
1
=
6
226
? ? ? ? ? ?
x+2´ 4´x x+2´ 4´x x+2+ 4´x
= ¨? ?
x´1 x´1 x+2+ 4´x
( x + 2) ´ (4 ´ x )
= ? ?
( x ´ 1)( x + 2 + 4 ´ x )
2x ´ 2
= ? ?
( x ´ 1)( x + 2 + 4 ´ x )
2
=? ?
x+2+ 4´x
So the limit is
? ?
x+2´ 4´x 2
lim = lim ? ?
xÑ1 x´1 xÑ1 x + 2 + 4 ´ x
2
=? ?
3+ 3
1
=?
3
? ? ? ? ? ?
x´2´ 4´x x´2´ 4´x x´2+ 4´x
= ¨? ?
x´3 x´3 x´2+ 4´x
( x ´ 2) ´ (4 ´ x )
= ? ?
( x ´ 3)( x ´ 2 + 4 ´ x )
2x ´ 6
= ? ?
( x ´ 3)( x ´ 2 + 4 ´ x )
2
=? ?
x´2+ 4´x
? ?
x´2´ 4´x 2
So, lim = lim ? ?
xÑ3 x´3 xÑ3 x ´ 2 + 4 ´ x
2
=
1+1
= 1.
227
S-24: Here we get 0/0 if we try the naive approach. Hence we must simplify.
?
3t ´ 3 3t ´ 3 2+ 5´t
? = ? ˆ ?
2´ 5´t 2´ 5´t 2+ 5´t
? 3t ´ 3
= 2+ 5´t 2
2 ´ (5 ´ t )
? 3t ´ 3
= 2+ 5´t
t´1
? 3( t ´ 1)
= 2+ 5´t
t´1
So there is a cancelation. Hence the limit is
3t ´ 3 ?
lim ? = lim 2 + 5 ´ t ¨ 3
tÑ1 2 ´ 5 ´ t tÑ1
= 12
S-25: This is a classic example of the Squeeze Theorem.

It is tempting to try to use
3
arithmetic of limits: lim ´x2 = 0, and lim cos = something, and zero times
xÑ0 xÑ0 x
something is 0. However, this is invalid reasoning,
because we can only use arithmetic of
3
limits when those limits exist, and lim cos does not exist. So, we need the Squeeze
xÑ0 x
Theorem.
Since ´1 ď cos 3x ď 1, we can bound our function of interest from above and below

(being careful of the sign!):

2 2 3
´x (1) ď ´x cos ď ´x2 (´1)
x
So our function of interest is between ´x2 and x2 . Since lim ´x2 = lim x2 = 0, by the
xÑ0 xÑ0
2 3
Squeeze Theorem, also lim ´x cos = 0.
xÑ0 x
Advice about writing these up: whenever we use the Squeeze Theorem, we need to
explicitly write that two things are true: that the function we’re interested is bounded
above and below by two other functions, and that both of those functions have the same
limit. Then we can conclude (and we need to write this down as well!) that our original
function also shares that limit.
S-26: Recall that sine and cosine, no matter what (real-number) input we feed them, spit
out numbers between ´1 and 1. So we can bound our horrible numerator, rather than
trying to deal with it directly.

4 2 4 1 2 1
x (´1) + 5x (´1) + 2 ď x sin + 5x cos + 2 ď x4 (1) + 5x2 (1) + 2
x x
228
Further, notice that our bounded functions tend to the same value as x goes to 0:
lim x4 (´1) + 5x2 (´1) + 2 = lim x4 + 5x2 + 2 = 2.

xÑ0 xÑ0
So, by the Squeeze Theorem, also

4 1 2 1
lim x sin + 5x cos + 2 = 2.
xÑ0 x x
Now, we evaluate our original limit:

x sin x + 5x cos 1x + 2
4 1 2
2 1
lim 2
= 2
= .
xÑ0 ( x ´ 2) (´2) 2

2 1
S-27: lim sin = DNE, so we think about using the Squeeze Theorem. We’ll need to
xÑ0 x
2 1
bound the expression x sin , but the bounding is a little delicate. For any non-zero
x
1
value we plug in for x, sin2 is a number in the interval [0, 1]. If a is a number in the
x
interval [0, 1], then:
0 ăxa ă x when x is positive

x ăxa ă 0 when x is negative
We’ll show you two ways to use this information to create bound that will allow you to
apply the Squeeze Theorem.
• Solution 1: We will evaluate separately the limit from the right and from the left.
When x ą 0,
1
2
0 ď x sin ďx
x

1
2
because 0 ď sin ď 1. Since
x
lim 0 = 0 and lim x = 0

xÑ0+ xÑ0+
then by the Squeeze Theorem, also

21
lim x sin = 0.
xÑ0 + x
Similarly, When x ă 0,
1
2
x ď x sin ď0
x
229

1
2
because 0 ď sin ď 1. Since
x
lim x = 0 and lim 0 = 0

xÑ0´ xÑ0´

12
lim x sin = 0.
xÑ0´ x
Since the one-sided limits are both equal to zero,

2 1
lim x sin = 0.
xÑ0 x
Remark: this is a perfectly fine proof, but it seems to repeat itself. Since the cases
x ă 0 and x ą 0 are so similar, we would like to take care of them together. This can
be done as shown below.

2 1
• Solution 2: If x ‰ 0, then 0 ď sin ď 1, so
x

2 1
´|x| ď x sin ď |x|.
x
Since
lim ´|x| = 0 and lim |x| = 0
xÑ0 xÑ0

2 1
lim x sin = 0.
xÑ0 x
S-28: When we plug w = 5 in to the numerator and denominator, we find that each
becomes zero. Since we can’t divide by zero, we have to dig a little deeper. When a
polynomial has a root, that also means it has a factor: we can factor (w ´ 5) out of the
top. That lets us cancel:
2w2 ´ 50 2(w ´ 5)(w + 5) 2( w + 5)

lim = lim = lim .
wÑ5 ( w ´ 5)( w ´ 1) wÑ5 ( w ´ 5)( w ´ 1) wÑ5 ( w ´ 1)
2w2 ´ 50
Note that the function is NOT defined at w = 5, while the function
(w ´ 5)(w ´ 1)
2( w + 5)
IS defined at w = 5; so strictly speaking, these two functions are not equal.
( w ´ 1)
However, for every value of w that is not 5, the functions are the same, so their limits are
230
equal. Furthermore, the limit of the second function is quite easy to calculate, since we’ve
2( w + 5) 2(5 + 5)
eliminated the zero in the denominator: lim = = 5.
wÑ5 ( w ´ 1) 5´1
2w2 ´ 50 2( w + 5)
So lim = lim = 5.
wÑ5 ( w ´ 5)( w ´ 1) wÑ5 ( w ´ 1)
S-29: When we plug in r = ´5 to the denominator, we find that it becomes 0, so we need

to dig deeper. The numerator is not zero, so cancelling is out. Notice that the
denominator is factorable: r2 + 10r + 25 = (r + 5)2 . As r approaches ´5 from either side,
the denominator gets very close to zero, but stays positive. The numerator gets very
close to ´5. So, as r gets closer to ´5, we have something close to ´5 divided by a very
small, positive number. Since the denominator is small, the fraction will have a large
magnitude; since the numerator is negative and the denominator is positive, the fraction
r
will be negative. So, lim 2 = ´8
rÑ´5 r + 10r + 25
x3 + x2 + x + 1
S-30: First, we find lim . When we plug in x = ´1 to the top and the
xÑ´1 3x + 3
bottom, both become zero. In a polynomial, where there is a root, there is a factor, so this
tells us we can factor out ( x + 1) from both the top and the bottom. It’s pretty easy to see
how to do this in the bottom. For the top, if you’re having a hard time, one factoring
method (of many) to try is long division of polynomials; another is to factor out ( x + 1)
from the first two terms and the last two terms. (Detailed examples of long division are
given in Appendix A.16 and Examples 1.10.2 and 1.10.3 of the CLP-2 text.)
x3 + x2 + x + 1 x 2 ( x + 1) + ( x + 1) ( x + 1)( x2 + 1)
lim = lim = lim
xÑ´1 3x + 3 xÑ´1 3x + 3 xÑ´1 3( x + 1)
2
x +1 2
(´1) + 1 2
= lim = = .
xÑ´1 3 3 3
x3 + x2 + x + 1
One thing to note here is that the function is not defined at x = ´1
3x + 3
x2 + 1
(because we can’t divide by zero). So we replaced it with the function , which IS
3
defined at x = ´1. These functions only differ at x = ´1; they are the same at every other
point. That is why we can use the second function to find the limit of the first function.
Now we’re ready to find the actual limit asked in the problem:
d c
x3 + x2 + x + 1 2
lim = .
xÑ´1 3x + 3 3
231
S-31: When we plug x = 0 into the denominator, we get 0, which means we need to look
harder. The numerator is not zero, so we won’t be able to cancel our problems away.
Let’s factor to make things clearer.
x2 + 2x + 1 ( x + 1)2
=
3x5 ´ 5x3 x3 (3x2 ´ 5)
As x gets close to 0, the numerator is close to 1; the term (3x2 ´ 5) is negative; and the
sign of x3 depends on the direction we’re approaching 0 from. Since we’re dividing a
numerator that is very close to 1 by something that’s getting very close to 0, the
magnitude of the fraction is getting bigger and bigger without bound. Since the sign of
the fraction flips depending on whether we are using numbers slightly bigger than 0, or
slightly smaller than 0, that means the one-sided limits are 8 and ´8, respectively. (In
x2 + 2x + 1 x2 + 2x + 1
particular, lim = 8 and lim = ´8.) Since the one-sided limits
xÑ0´ 3x5 ´ 5x3 xÑ0+ 3x5 ´ 5x3
don’t agree, the limit does not exist.
S-32: As usual, we first try plugging in t = 7, but the denominator is 0, so we need to

think harder. The top and bottom are both squares, so let’s go ahead and factor:
t2 x2 + 2tx + 1 (tx + 1)2
= . Since x is positive, the numerator is nonzero. Also, the
t2 ´ 14t + 49 ( t ´ 7)2
numerator is positive near t = 7. So, we have something positive and nonzero on the top,
and we divide it by the bottom, which is positive and getting closer and closer to zero.
The quotient is always positive near t = 7, and it is growing in magnitude without
t2 x2 + 2tx + 1
bound, so lim 2 = 8.
tÑ7 t ´ 14t + 49
Remark: there is an important reason we specified that x must be a positive constant.

Suppose x were ´ 17 (which is negative and so was not allowed in the question posed). In
this case, we would have
t2 x2 + 2tx + 1 (tx + 1)2

lim = lim
tÑ7 t2 ´ 14t + 49 tÑ7 ( t ´ 7)2
(´t/7 + 1)2
= lim
tÑ7 ( t ´ 7)2
(´1/7)2 (t ´ 7)2
= lim
tÑ7 ( t ´ 7)2
= lim(´1/7)2
tÑ7
1
=
49
‰8
S-33: The function whose limit we are taking does not depend on d. Since x is a constant,
232
x5 ´ 32x + 15 is also a constant–it’s just some number, that doesn’t change, regardless of
what d does. So lim x5 ´ 32x + 15 = x5 ´ 32x + 15.
dÑ0
S-34: There’s a lot going on inside that sine function... and we don’t have to care about
any of it. No matter what horrible thing we put inside a sine function, the sine function
will spit out a number between ´1 and 1. So that means we can bound our horrible
function like this:
2 ´ 3x + 2 2
" #
x
( x ´ 1)2 ¨ (´1) ď ( x ´ 1)2 sin + 15 ď ( x ´ 1)2 ¨ (1)
x2 ´ 2x + 1
Since lim ( x ´ 1)2 ¨ (´1) = lim ( x ´ 1)2 ¨ (1) = 0, the Squeeze Theorem tells us that
xÑ1 xÑ1
" 2 #
x2 ´ 3x + 2
lim ( x ´ 1)2 sin + 15 = 0
xÑ1 x2 ´ 2x + 1
as well.
S-35: Since ´1 ď sin x ď 1 for all values of x,
´1 ď sin x´100 ď 1

(´1) x1/101 ď x1/101 sin x´100 ď (1) x1/101 when x ą 0, and

(1) x1/101 ď x1/101 sin x´100 ď (´1) x1/101 when x ă 0. Also,

lim x1/101 = lim ´x1/101 = 0 So, by the Squeeze Theorem,

xÑ0 xÑ0
1/101 ´100
= 0 = lim x1/101 sin x´100 and so

lim x sin x
xÑ0´ xÑ0+
0 = lim x1/101 sin x´100 .

xÑ0
Remark: there is a technical point here. When x is a positive number, ´x is negative, so

(´1) x ă x. But, when x is negative, ´x is positive, so (´1) x ą x. This is why we take the
one-sided limits of our function, and apply the Squeeze Theorem
1/101 1/101 ´100
to them separately. It is
not true to say that, for instance, (´1) x ďx sin x when x is near zero,
because this does not hold when x is less than zero.
S-36:
x2 ´ 4 ( x ´ 2)( x + 2) x+2
lim 2 = lim = lim =2
xÑ2 x ´ 2x xÑ2 x ( x ´ 2) xÑ2 x
S-37: When we plug in x = 5 to the top and the bottom, both limits exist, and the bottom
( x ´ 5)2 0
is nonzero. So lim = = 0.
xÑ5 x + 5 10
233
S-38: Since we can’t plug in t = 12 , we’ll simplify. One way to start is to add the fractions
in the numerator. We’ll need a common demoninator, such as 3t2 (t2 ´ 1).
t2 ´1 2
1
+ 1
3t (t2 ´1)
2 + 3t2 (3tt2 ´1)
3t2 t2 ´1
lim = lim
tÑ 12 2t ´ 1 tÑ 21 2t ´ 1
4t2 ´1
3t2 (t2 ´1)
= lim
tÑ 21 2t ´ 1
4t2 ´ 1
= lim
tÑ 21 3t2 (t2 ´ 1)(2t ´ 1)
(2t + 1)(2t ´ 1)
= lim
tÑ 21 3t2 (t2 ´ 1)(2t ´ 1)
2t + 1
= lim
tÑ 21 3t2 (t2´ 1)
Since we cancelled out the term that was causing the numerator and denominator to be
zero when t = 21 , now t = 12 is in the domain of our function, so we simply plug it in:
1+1
=
3 1
4 4 ´ 1
2
= 3
´ 43

4
32
=´
9
S-39: We recall that

"
x , xě0
|x| =
´x , x ă 0
So,
x
"
|x| x , xą0
= ´x
x x , xă0
"
1 , xą0
=
´1 , x ă 0
Therefore,
"
|x| 4 , xą0
3+ =
x 2 , xă0
Since our function gives a 4 when x is to the right of zero, and a value of 2 when
value of
|x|
x is to the left of zero, lim 3 + does not exist.
xÑ0 x
To further clarify the situation, the graph of y = f ( x ) is sketched below:
234
y
|d + 4|
S-40: If we factor out 3 from the numerator, our function becomes 3 . We recall that
d+4
"
X , Xě0
|X| =
´X , X ă 0
So, with X = d + 4,
$
d +4
& 3 d +4
’ , d+4 ą 0
|d + 4|
3 =
d+4 % 3 ´( d +4) , d + 4 ă 0
’
d +4
"
3 , d ą ´4
=
´3 , d ă ´4
Since our function gives a value of 3 when d ą ´4, and a value of ´3 when d ă ´4,
|3d + 12|
lim does not exist.
dÑ´4 d + 4

y
x
´4
´3
S-41: Note that x = 0 is in the domain of our function, and nothing “weird” is happening
there: we aren’t dividing by zero, or taking the square root of a negative number, or
joining two pieces of a piecewise-defined function. So, as x gets extremely close to zero,
5x ´ 9 0´9 ´9
is getting extremely close to = .
|x| + 2 0+2 2
5x ´ 9 9
That is, lim =´ .
xÑ0 |x| + 2 2
235
S-42: Since we aren’t dividing by zero, and all these limits exist:
x f (x) + 3 (´1)(´1) + 3
lim = = ´4.
xÑ´1 2 f ( x ) + 1 2(´1) + 1
S-43: As x Ñ ´2, the denominator goes to 0, and the numerator goes to ´2a + 7. For the
7
ratio to have a limit, the numerator must also converge to 0, so we need a = . Then,
2
x2 + ax + 3 x2 + 27 x + 3
lim = lim
xÑ´2 x2 + x ´ 2 xÑ´2 ( x + 2)( x ´ 1)
( x + 2)( x + 23 )
= lim
xÑ´2 ( x + 2)( x ´ 1)
x + 23
= lim
xÑ´2 x ´ 1
1
=
6
7
so the limit exists when a = .
2
S-44:
(a) lim f ( x ) = 0: as x approaches 0, so does 2x.

xÑ0
(b) lim g( x ) = DNE: the left and right limits do not agree, so the limit does not exist. In
xÑ0
particular: lim g( x ) = ´8 and lim g( x ) = 8.
xÑ0´ xÑ0+
1
(c) lim f ( x ) g( x ) = lim 2x ¨
= lim 2 = 2.
xÑ0 xÑ0 x xÑ0
Remark: although the limit of g( x ) does not exist here, the limit of f ( x ) g( x ) does.
f (x) 2x
(d) lim = lim 1 = lim 2x2 = 0
xÑ0 g( x ) xÑ0
x
xÑ0
1 1 9
(e) lim f ( x ) + g( x ) = lim 2x + = 4+ =
xÑ2 xÑ2 x 2 2
f (x) + 1 2x + 1 1
(f) lim = lim 1 = =1
xÑ0 g ( x + 1) xÑ0
x +1
1
236
S-45: We can begin by plotting the points that are easy to read off the diagram.
1
x f (x) f (x)
´1
´3 ´3 3
´2 0 UND
1
´1 3 3
1
0 3 3
3 2
1 2 3
2 0 UND
3 1 1
1 1
Note that f (x)
is undefined when f ( x ) = 0. So f (x)
is undefined at x = ´2 and x = 2. We
1
shall look more closely at the behaviour of f (x)
for x near ˘2 shortly.
Plotting the above points, we get the following picture:

y
x
1
1
Since f ( x ) is constant when x is between -1 and 0, then also f (x)
is constant between -1
and 0, so we update our picture:
y
x
1
1
The big question that remains is the behaviour of f (x)
when x is near -2 and 2. We can
237
answer this question with limits. As x approaches ´2 from the left, f ( x ) gets closer to
zero, and is negative. So f (1x) will be negative, and will increase in magnitude without
1
bound; that is, lim = ´8. Similarly, as x approaches ´2 from the right, f ( x ) gets
xÑ´2´ f ( x )
closer to zero, and is positive. So f (1x) will be positive, and will increase in magnitude
1
without bound; that is, lim = 8. We add this behaviour to our graph:
xÑ´2+ f ( x )
x
1
Now, we consider the behaviour at x = 2. Since f ( x ) gets closer and closer to 0 AND is
1
positive as x approaches 2, we conclude lim = 8. Adding to our picture:
xÑ2 f ( x )
x
1
Now the only remaining blank space is between x = 0 and x = 1. Since f ( x ) is a smooth
curve that stays away from 0, we can draw some kind of smooth curve here, and call it
good enough. (Later on we’ll go into more details about drawing graphs. The purpose of
this exercise was to utilize what we’ve learned about limits.)
238
y
x
1
S-46: We can start by examining points.
f (x)
x f (x) g( x )
g( x )
´3 ´3 ´1.5 2
´2 0 0 UND
´1 3 1.5 2
´0 3 1.5 2
1 1.5 .75 2
2 0 0 UND
3 1 .5 2
f (x)
We cannot divide by zero, so is not defined when x = ˘2. But for every other value
g( x )
f (x)
of x that we plotted, f ( x ) is twice as large as g( x ), = 2. With this in mind, we see
g( x )
that the graph of f ( x ) is exactly the graph of 2g( x ).
This gives us the graph below.
y
f (x)
y=
g( x )
x
1
239
f (2) f (x)
Remark: f (2) = g(2) = 0, so does not exist, but lim = 2. Although we are
g (2) xÑ2 g ( x )
trying to “divide by zero” at x = ˘2, it would be a mistake here to interpret this as a
vertical asymptote.
s (1 + h ) ´ s (1)
S-47: Velocity of white ball when t = 1 is lim , so the given information
hÑ0 h
s (1 + h ) ´ s (1)
tells us lim = 5. Then the velocity of the red ball when t = 1 is
hÑ0 h
2s(1 + h) ´ 2s(1) s (1 + h ) ´ s (1)
lim = lim 2 ¨ = 2 ¨ 5 = 10.
hÑ0 h hÑ0 h
S-48: (a) Neither limit exists. When x gets close to 0, these limits go to positive infinity
from one side, and negative infinity from the other.
1 1
(b) lim [ f ( x ) + g( x )] = lim ´ = lim 0 = 0.
xÑ0 xÑ0 x x xÑ0
(c) No: this is an example of a time when the two individual functions have limits that
don’t exist, but the limit of their sum does exist. This “sum rule” is only true when both
lim f ( x ) and lim g( x ) exist.
xÑa xÑa
S-49: (a) When we evaluate the limit from the left, we only consider values of x that are
less than zero. For these values of x, our function is x2 ´ 3. So,
lim f ( x ) = lim ( x2 ´ 3) = ´3.
xÑ0´ xÑ0´
(b) When we evaluate the limit from the right, we only consider values of x that are
greater than zero. For these values of x, our function is x2 + 3. So,
lim f ( x ) = lim ( x2 + 3) = 3.
xÑ0+ xÑ0+
(c) Since the limits from the left and right do not agree, lim f ( x ) = DNE.
xÑ0

y
3
x
´3
S-50: (a) When we evaluate lim f ( x ), we only consider values of x that are less than
xÑ´4´
240
´4. For these values, f ( x ) = x3 + 8x2 + 16x. So,
lim f ( x ) = lim ( x3 + 8x2 + 16x ) = (´4)3 + 8(´4)2 + 16(´4) = 0

xÑ´4´ xÑ´4´
Note that, because x3 + 8x2 + 16x is a polynomial, we can evaluate the limit by directly
substituting in x = ´4.
(b) When we evaluate lim f ( x ), we only consider values of x that are greater than ´4.
xÑ´4+
For these values,
x2 + 8x + 16
f (x) =
x2 + 30x ´ 4
So,
x2 + 8x + 16
lim f ( x ) = lim 2
xÑ´4+ xÑ´4+ x + 30x ´ 4
This is a rational function, and x = ´4 is in its domain (we aren’t doing anything suspect,
like dividing by 0), so again we can directly substitute x = ´4 to evaluate the limit:
(´4)2 + 8(´4) + 16 0
= 2
= =0
(´4) + 30(´4) ´ 4 ´108
(c) Since lim f ( x ) = lim f ( x ) = 0, we conclude lim f ( x ) = 0.

xÑ´4´ xÑ´4+ xÑ´4
S-1: Any polynomial of degree one or higher will go to 8 or ´8 as x goes to 8. So, we

need a polynomial of degree 0–that is, f ( x ) is a constant. One possible answer is
f ( x ) = 1.
S-2: This will be the case for any polynomial of odd degree. For instance, f ( x ) = x.
Many answers are possible: also f ( x ) = x15 ´ 32x2 + 9 satisfies lim f ( x ) = 8 and
xÑ8
lim f ( x ) = ´8.
xÑ´8
1
S-3: lim 2´x = lim =0
xÑ8 xÑ8 2x
S-4: As x gets larger and larger, 2x grows without bound. (For integer values of x, you
can imagine multiplying 2 by itself more and more times.) So, lim 2x = 8.
xÑ8
S-5: Write X = ´x. As x becomes more and more negative, X becomes more and more
positive. From Question 4, we know that 2X grows without bound as X gets larger and
241
larger. Since 2x = 2´(´x) = 2´X = 21X , as we let x become a huge negative number, we
are in effect dividing by a huge positive number; hence lim 2x = 0.
xÑ´8
1
A more formulaic way to describe the above is this: lim 2x = lim 2´X = lim X = 0.
xÑ´8 XÑ8 XÑ8 2
S-6: There is no single number that cos x approaches as x becomes more and more
strongly negative: as x grows in the negative direction, the function oscillates between
´1 and +1, never settling close to one particular number. So, this limit does not exist.
S-7: The highest-order term in this polynomial is ´3x5 , so this dominates the function’s
behaviour as x goes to infinity. More formally:

5 2
5 1 100
lim x ´ 3x + 100x = lim ´3x 1 ´ 4 ´ 3
xÑ8 xÑ8 3x 3x
= lim ´3x5 = ´8
xÑ8
because

1 100
lim 1´ 4 ´ 3 = 1 ´ 0 ´ 0 = 1.
xÑ8 3x 3x
S-8: Our standard trick is to factor out the highest power of x in the denominator: x4 . We
just have to be a little careful with the square root.?Since we are?taking the limit as x goes
to positive infinity, we have positive x-values, so x2 = x and x8 = x4 .
b
?
3x8 + 7x4 + 10 x8 (3 + x74 ) + 10
lim = lim
xÑ8 x4 ´ 2x2 + 1 xÑ8 x 4 (1 ´ 2 + 1 )
x2 x4
? b
x8 3 + x74 + 10
= lim 4
xÑ8 x (1 ´ 2 + 1 )
x2 x4
b
x4 3 + x74 + 10
= lim 4
xÑ8 x (1 ´ 2 + 1 )
2 x4
b x
4 7 10
x 3 + x4 + x4
= lim
xÑ8 x 4 (1 ´ 2 + 1 )
x2 x4
b
3 + x74 + 10 x4
= lim 2 1
xÑ8 1 ´ + x4
x2
?
3+0+0 ?
= = 3
1´0+0
242
S-9: We have two terms, each getting extremely large. It’s unclear at first what happens
when we subtract?them. To get?this equation into another form, we multiply and divide
by the conjugate, x + 5x + x2 ´ x.
2
" ? ? ? ? #
ha a i ( x 2 + 5x ´ x2 ´ x )( x 2 + 5x + x 2 ´ x )
lim x2 + 5x ´ x2 ´ x = lim ? ?
xÑ8 xÑ8 x2 + 5x + x2 ´ x
( x2 + 5x ) ´ ( x2 ´ x )
= lim ? ?
xÑ8 x 2 + 5x + x 2 ´ x
6x
= lim ? ?
xÑ8 x 2 + 5x + x 2 ´ x
? numerator and denominator by x. In the case of the denominator,

Now we divide the
since x ą 0, x = x2 .
6( x )
= lim ? b ? b
xÑ8 5 1
x 1 + x + x2 1 ´
2
x
6( x )
= lim b b
xÑ8 5 1
(x) 1 + + (x) 1 ´
x x
6
= lim b b
xÑ8
1 + 5x + 1 ´ 1
x
6
=? ? =3
1+0+ 1´0
S-10:
? ? for large negative x, the first term in the denominator
Note that
2 2
?4x + x « 4x = |2x| = ´2x not +2x. A good way to avoid incorrectly computing
2
x when x is negative is to define y = ´x and express everything in terms of y. That’s
what we’ll do.
3x ´3y
lim ? = lim a
xÑ´8 2
4x + x ´ 2x yÑ + 8 2
4y ´ y + 2y
´3y
= lim b
yÑ+8
y 4 ´ 1y + 2y
´3
= lim b
yÑ+8
4 ´ y1 + 2
´3
=? since 1/y Ñ 0 as y Ñ +8
4´0+2
3
=´
4
243
S-11: The highest power of x in the denominator is x2 , so we divide the numerator and
denominator by x2 :
1 ´ x ´ x2 1/x2 ´ 1/x ´ 1
lim = lim
xÑ´8 2x2 ´ 7 xÑ´8 2 ´ 7/x2
0´0´1 1
= =´
2´0 2
S-12:
? ?
2+x´x 2+x+x
a x x ( x2 + x ) ´ x2
lim x2 + x ´ x = lim ? = lim ?
xÑ8 xÑ8 x2 + x + x xÑ8 x2 + x + x
x 1 1
= lim ? = lim b =
xÑ8 x2 + x + x xÑ8
1+ 1 +1 2
x
S-13: We have, after dividing both numerator and denominator by x2 (which is the
highest power of the denominator) that
5x2 ´ 3x + 1 5 ´ 3x + x12
= .
3x2 + x + 7 3 + 1x + x72
Since 1/x Ñ 0 and also 1/x2 Ñ 0 as x Ñ +8, we conclude that

5x2 ´ 3x + 1 5
lim 2
= .
xÑ+8 3x + x + 7 3
S-14: We have, after dividing both numerator and denominator by x (which is the
highest power of the denominator) that
b
? 4 2
4x+2 x + x2
= .
3x + 4 3 + 4x
Since 1/x Ñ 0 and also 1/x2 Ñ 0 as x Ñ +8, we conclude that

?
4x+2 0
lim = = 0.
xÑ+8 3 x + 4 3
S-15: The dominant terms in the numerator and denominator have order x3 . Taking out
that common factor we get
4x3 + x 4 + x12
= .
7x3 + x2 ´ 2 7 + 1x ´ x23
244
Since 1/x a Ñ 0 as x Ñ +8 (for a ą 0), we conclude that
4x3 + x 4
lim 3 2
= .
xÑ+8 7x + x ´ 2 7
S-16:
• Solution 1
We want to factor out x, the highest power in the denominator.
? Since our limit only
4 4
sees
? negative values of x, we must remember that x = |x| = ´x, although
3 3
x = x.
b b
? ? 3 3 ( 1 + 1 ) ´ 4 x 4 (1 + 5 )
3 2
x +x´ x +54 4 x x x2 x4
lim = lim
xÑ´8 x+1 xÑ´8 x (1 + 1x )
? b
3 3 3 1
?
4 4 4
b
1
x x + x2
´ x 1 + x54
= lim
xÑ´8 x (1 + 1x )
b b
3 1 1
x x + x2 ´ (´x ) 4 1 + x54
= lim
xÑ´8 x (1 + 1x )
b b
3 1 1
x + x2 +
4
1 + x54
= lim
xÑ´8 1 + 1x
? ?
3
0+0+ 41+0
= =1
1+0
• Solution 2
Alternately, we can use the transformation lim f ( x ) = lim f (´x ). Then we only
xÑ´8 ? xÑ8
4 4
look at positive values of x, so roots behave nicely: x = |x| = x.
? ? a a
x + x ´ 4 x4 + 5
3 2 3
(´x )2 ´ x ´ 4 (´x )4 + 5
lim = lim
xÑ´8 x+1 xÑ8
? ´x
? +1
3 2 4 4
x ´x´ x +5
= lim
b´x + 1 ? b
xÑ8
?3 3 3 1 4 4 4
1
x x ´ x2 ´ x 1 + x54
= lim
xÑ8 x (´1 + 1x )
b b
x x ´ x2 ´ x 4 1 + x54
3 1 1
= lim
xÑ8 x (´1 + 1x )
b b
3 1 1
x ´ x2
´ 4
1 + x54
= lim
xÑ8 ´1 + 1x
? ?
3
0´0´ 4 1+0 ´1
= = =1
´1 + 0 ´1
245
S-17: We have, after dividing both numerator and denominator by x3 (which is the
highest power of the denominator) that:
5 10
5x2 + 10 + 3 0
lim 3 2
= lim x 2 x 1 = = 0.
xÑ8 3x + 2x + x xÑ8 3 + + 3
x x2
?
S-18: Since we only consider negative values of x, x2 = |x| = ´x.
x+1 x+1
lim ? = lim
xÑ´8 x2 xÑ´8 ´x
x 1
= lim +
xÑ´8 ´x ´x
1
= lim ´1 ´
xÑ´8 x
= ´1
?
S-19: Since we only consider positive values of x, x2 = |x| = x.
x+1 x+1
lim ? = lim
xÑ8 x2 xÑ8 x
1
= lim 1 + = 1 + 0 = 1
xÑ8 x

π |x| 1
S-20: When x ă 0, |x| = ´x and so lim sin ¨ + = sin(´π/2) = ´1.
xÑ8 2 x x
S-21: We divide both the numerator and the denominator

? by the highest power of x in
2
the denominator, which is x. Since x ă 0, we have x = |x| = ´x, so that
? c c
x2 + 5 x2 + 5 5
=´ 2
= ´ 1 + 2.
x x x
Since 1/x Ñ 0 and also 1/x2 Ñ 0 as x Ñ ´8, we conclude that
3x + 5 3 + 5x 3 3
lim ? = lim b = =´ .
xÑ´8 x2 + 5 ´ x xÑ´8
´ 1 + x52 ´ 1 ´1 ´ 1 2
246
S-22: We divide both the numerator and the denominator
? by the highest power of x in
2
the denominator, which is x. Since x ă 0, we have x = |x| = ´x, so that
? ? c c
4x2 + 15 4x2 + 15 4x2 + 15 15
= ? =´ 2
= ´ 4 + 2.
x ´ x 2 x x
Since 1/x Ñ 0 and also 1/x2 Ñ 0 as x Ñ ´8, we conclude that
5x + 7 5 + 7x 5 5
lim ? = lim b = =´ .
2
xÑ´8 4x + 15 ´ x xÑ´8
´ 4 + 15 ´1 ´2 ´ 1 3
x2
S-23:
3x7 + x5 ´ 15 x2 (3x5 + x3 ´ 15
x2
)
lim = lim
xÑ´8 4x 2 + 32x xÑ´8 x2 (4 + 32
x)
3x5 + x3 ´ 15
x2
= lim 32
xÑ´8 4+ x
15
3(´x )5 + (´x )3 ´ (´x )2
= lim 32
xÑ+8 4+´x
´3x5 ´ x3 ´ 15
x2
= lim 32
xÑ+8 4´ x
= ´8
?
S-24: We multiply and divide the expression by its conjugate, n2 + 5n + n .
? !
a a n2 + 5n + n
lim n2 + 5n ´ n = lim n2 + 5n ´ n ?
nÑ8 nÑ8 n2 + 5n + n
(n2 + 5n) ´ n2
= lim ? 2
nÑ8 n + 5n + n
5n
= lim ? 2
nÑ8 n + 5n + n
5¨n
= lim ? b
nÑ8
n2 1 + n5 + n
247
?
Since n ą 0, we can simplify n2 = n.
5¨n
= lim b
nÑ8 5
n 1+ n +n
5
= lim b
nÑ8 5
1+ n +1
5 5
=? =
1+0+1 2
S-25:
• Solution 1:
When a approaches 0 from the right, the numerator approaches negative infinity,
a2 ´ 1a
and the denominator approaches ´1. So, lim = 8.
aÑ0+ a ´ 1
More precisely, using Theorem 1.5.9:
1
= +8lim
aÑ0+ a
Also, lim a2 = 0
aÑ0+
1
So, using Theorem 1.5.9, lim a2 ´ = ´8
aÑ0+ a
Furthermore, lim a ´ 1 = ´1
aÑ0+
a2 ´ 1a
So, using our theorem, lim =8
aÑ0+ a ´ 1
• Solution 2:
Since a = 0 is not in the domain of our function, a reasonable impulse is to simplify.
a2 ´ 1a a a3 ´ 1 ( a ´ 1)( a2 + a + 1)
= =
a´1 a a ( a ´ 1) a ( a ´ 1)
So,
a2 ´ 1a ( a ´ 1)( a2 + a + 1)
lim = lim
aÑ0+ a´1 aÑ0+ a ( a ´ 1)
a2 + a + 1
= lim
aÑ0+ a
1
= lim a + 1 + = 8
aÑ0+ a
248
S-26:
Since x = 3 is not in the domain of the function, we simplify, hoping we can cancel a
problematic term.
2x + 8 2x + 8
lim = lim x +3
xÑ3 1 + x21´9 xÑ3 + x21´9
x´3 x2 ´9
2x + 8
= lim x +4
xÑ3
x2 ´9
(2x + 8)( x2 ´ 9)
= lim =0
xÑ3 x+4
S-27: First, we need a rational function whose limit at infinity is a real number. This
means that the degree of the bottom is greater than or equal to the degree of the top.
There are two cases: the denominator has higher degree than the numerator, or the
denominator has the same degree as the numerator.
If the denominator has higher degree than the numerator, then
lim f ( x ) = lim f ( x ) = 0, so the limits are equal–not what we’re looking for.
xÑ8 xÑ´8
If the denominator has the same degree as the numerator, then the limit as x goes to ˘8
is the ratio of the leading terms: again, the limits are equal. So no rational function exists
as described.
S-28: The amount of the substance that will linger long-term is some positive
number–the substance will stick around. One example of a substance that does this is the
ink in a tattoo. (If the injection was of medicine, probably it will be metabolized, and
lim c(t) = 0.)
tÑ8
Remark: it actually doesn’t make much sense to let t go to infinity: after a few million
hours, you won’t even have a body, so what is c(t) measuring? Often when we use
formulas in the real world, there is an understanding that they are only good for some
fixed range. We often use the limit as t goes to infinity as a stand-in for the function’s
long-term behaviour.
S-1: Many answers are possible; the tangent function behaves like this.
S-2: If we let f be my height, the time of my birth is a, and now is b, then we know that
f ( a) ď 1 ď f (b). It is reasonable to assume that my height is a continuous function. So by
the IVT, there is some value c between a and b where f (c) = 1. That is, there is some time
(we called it c) between my birth and today when I was exactly one meter tall.
249
Notice the IVT does not say precisely what day I was one meter tall; it only guarantees
that such a day occurred between my birth and today.
"
0 when 0 ď x ď 1
S-3: One example is f ( x ) = . The IVT only guarantees f (c) = 1 for
2 when 1 ă x ď 2
some c in [0, 2] when f is continuous over [0, 2]. If f is not continuous, the IVT says
nothing.
S-4: Yes. This is a straightforward application of IVT.
S-5: No. IVT says that f ( x ) = 0 for some x between 10 and 20, but it doesn’t have to be
exactly half way between.
S-6: No. IVT says nothing about functions that are not guaranteed to be continuous at the
outset. It’s quite easy to construct a function that is as described, but not continuous. For
example, the function pictured below, " whose equation format is somewhat less
26
´ 5 x + 65, 10 ď x ă 15
enlightening than its graph: f ( x ) = .
´ 26
5 x + 91, 15 ď x ď 20
13
x
10 15 20
´13
S-7: True. Since f (t) is continuous at t = 5, that means lim f (t) = f (5). For that to be
tÑ5
true, f (5) must exist — that is, 5 is in the domain of f ( x ).
250
S-8: True. Using the definition of continuity, lim f (t) = f (5) = 17.
tÑ5
S-9: In general, false. If f (t) is continuous at t = 5, then f (5) = 17; if f (t) is

discontinuous at t = 5, then f (5) either does not exist, or is a number other than 17.
"
17 , t ‰ 5
An example of a function with lim f (t) = 17 ‰ f (5) is f (t) = , shown
tÑ5 0 , t=5
below.
y
x
5
S-10: Since f ( x ) and g( x ) are continuous at zero, and since g2 ( x ) + 1 must be nonzero,
then h( x ) is continuous at 0 as well. According to the definition of continuity, then
0 f (0)
lim h( x ) exists and is equal to h(0) = g2 (0)+1 = 0.
xÑ0
Since the limit lim h( x ) exists and is equal to zero, also the one-sided limit lim h( x )
xÑ0 xÑ0+
exists and is equal to zero.
S-11: Using the definition of continuity, we need k = lim f ( x ). Since the limit is blind to
xÑ0
1
what actually happens to f ( x ) at x = 0, this is equivalent to k = lim x sin . So if we
xÑ0 x
find the limit, we solve the problem.

As x gets small, sin 1x goes a little crazy (see example 1.3.5), so let’s get rid of it by
using the Squeeze Theorem. We can bound the function above and below, but we should
be a little careful about whether we’re going from the left or the right. The reason we
need to worry about direction is illustrated with the following observation: If a ď b and
x ą 0, then xa ď xb. (For example, plug in x = 1, a = 2, b = 3.) But if a ď b and x ă 0,
1
then xa ě xb. (For example, plug in x = ´1, a = 2, b = 3.) So first, let’s find lim sin x .
xÑ0´
When x ă 0,

1
1 ě sin ě ´1
x

1
so: x (1) ď x sin ď x (´1)
x

1
and lim x = lim ´x = 0, so by the Squeeze Theorem, also lim x sin = 0.
xÑ0 ´ xÑ0 ´ xÑ0 ´ x
251

1
Now, let’s find lim x sin . When x ą 0,
xÑ0 + x

1
´1 ď sin ď1
x

1
so: x (´1) ď xsin ď x (1)
x

1
and lim x = lim ´x = 0, so by the Squeeze Theorem, also lim x sin = 0.
xÑ0+ xÑ0+ xÑ0+ x

1
Since the limits from the left and right agree, we conclude lim x sin = 0, so when
xÑ0 x
k = 0, the function is continuous at x = 0.
S-12: Since f is a polynomial, it is continuous over all real numbers. f (0) = 1 ă 12345
and f (12345) = 123453 + 123452 + 12345 + 1 ą 12345 (since all terms are positive). So by
the IVT, f (c) = 12345 for some c between 0 and 12345.
S-13: f ( x ) is a rational function and so is continuous except when its denominator is

zero. That is, except when x = 1 and x = ´1.
S-14: The function is continuous when x2 ´ 1 ą 0, i.e. ( x ´ 1)( x + 1) ą 0, which yields the
intervals (´8, ´1) Y (1, +8).
?
S-15: The function 1/ x is continuous on (0, +8) and the function cos( x ) + 1 is
continuous everywhere.
a
So 1/ cos( x ) + 1 is continuous except when cos x = ´1. This happens when x is an odd
multiple of π. Hence the function is continuous except at x = ˘π, ˘3π, ˘5π, . . . .
S-16: The function is continuous when sin( x ) ‰ 0. That is, when x is not an integer
multiple of π.
S-17: The function is continuous for x ‰ c since each of those two branches are
polynomials. So, the only question is whether the function is continuous at x = c; for this
we need
lim f ( x ) = f (c) = lim f ( x ).
xÑc´ xÑc+
We compute
lim f ( x ) = lim 8 ´ cx = 8 ´ c2 ;
xÑc´ xÑc´
f (c) = 8 ´ c ¨ c = 8 ´ c2 and
lim f ( x ) = lim x2 = c2 .
xÑc+ xÑc+
So, we need 8 ´ c2 = c2 , which yields c2 = 4, i.e. c = ´2 or c = 2.
252
S-18: The function is continuous for x ‰ 0 since x2 + c and cos cx are continuous
everywhere. It remains to check continuity at x = 0. To do this we must check that the
following three are equal.
lim f ( x ) = lim x2 + c = c
xÑ0+ xÑ0+
f (0) = c
lim f ( x ) = lim cos cx = cos 0 = 1
xÑ0´ xÑ0´
Hence when c = 1 we have the limits agree.
S-19: The function is continuous for x ‰ c since each of those two branches are defined
by polynomials. Thus, the only question is whether the function is continuous at x = c.
Furthermore,
lim f ( x ) = c2 ´ 4
xÑc´
and
lim f ( x ) = f (c) = 3c .
xÑc+
For continunity we need both limits and the value to agree, so f is continuous if and only
if c2 ´ 4 = 3c, that is if and only if
c2 ´ 3c ´ 4 = 0 .
Factoring this as (c ´ 4)(c + 1) = 0 yields c = ´1 or c = +4.
S-20: The function is continuous for x ‰ c since each of those two branches are
polynomials. So, the only question is whether the function is continuous at x = c; for this
we need
lim f ( x ) = f (2c) = lim f ( x ).
xÑ2c´ xÑ2c+
We compute
lim f ( x ) = lim 6 ´ cx = 6 ´ 2c2 ;
xÑ2c´ xÑ2c´
f (2c) = 6 ´ c ¨ 2c = 6 ´ 2c2 and

lim f ( x ) = lim x2 = 4c2 .
xÑ2c+ xÑ2c+
So, we need 6 ´ 2c2 = 4c2 , which yields c2 = 1, i.e. c = ´1 or c = 1.
S-21: This isn’t the kind of equality that we can just solve; we’ll need a trick, and that
trick is the IVT. The general idea is to show that sin x is somewhere bigger, and
somewhere smaller, than x ´ 1. However, since the IVT can only show us that a function
is equal to a constant, we need to slightly adjust our language. Showing sin x = x ´ 1 is
equivalent to showing sin x ´ x + 1 = 0, so let f ( x ) = sin x ´ x + 1, and let’s show that it
has a real root.
First, we need to note that f ( x ) is continuous (otherwise we can’t use the IVT). Now, we
need to find a value of x for which it is positive, and for which it’s negative. By checking
253
a few values, we find f (0) is positive, and f (100) is negative. So, by the IVT, there exists
a value of x (between 0 and 100) for which f ( x ) = 0. Therefore, there exists a value of x
for which sin x = x ´ 1.
S-22: We let f ( x ) = 3x ´ x2 . Then f ( x ) is a continuous function, since both 3x and x2 are

continuous for all real numbers.
We find a value a such that f ( a) ą 0. We observe immediately that a = 0 works since
f (0) = 30 ´ 0 = 1 ą 0.
We find a value b such that f (b) ă 0. We see that b = ´1 works since
1
f (´1) = ´ 1 ă 0.
3
So, because f ( x ) is continuous on (´8, 8) and f (0) ą 0 while f (´1) ă 0, then the
Intermediate Value Theorem guarantees the existence of a real number c in the interval
(´1, 0) such that f (c) = 0.
S-23: We let f ( x ) = 2 tan( x ) ´ x ´ 1. Then f ( x ) is a continuous function on the interval

(´π/2, π/2) since tan( x ) = sin( x )/ cos( x ) is continuous on this interval, while x + 1 is a
polynomial and therefore continuous for all real numbers.
We find a value a P (´π/2, π/2) such that f ( a) ă 0. We observe immediately that a = 0
works since
f (0) = 2 tan(0) ´ 0 ´ 1 = 0 ´ 1 = ´1 ă 0.
We find a value b P (´π/2, π/2) such that f (b) ą 0. We see that b = π/4 works since
f (π/4) = 2 tan(π/4) ´ π/4 ´ 1 = 2 ´ π/4 ´ 1 = 1 ´ π/4 = (4 ´ π )/4 ą 0,
because 3 ă π ă 4.
So, because f ( x ) is continuous on [0, π/4] and f (0) ă 0 while f (π/4) ą 0, then the
Intermediate Value Theorem guarantees the existence of a real number c P (0, π/4) such
that f (c) = 0.
a
S-24: Let f ( x ) = cos(πx ) ´ sin(2πx ) ´ 1/2. This function is continuous provided
cos(πx ) ě 0. This is true for 0 ď x ď 12 .
Now f takes positive values on [0, 1/2]:
b ?
f (0) = cos(0) ´ sin(0) ´ 1/2 = 1 ´ 1/2 = 1/2.
And f takes negative values on [0, 1/2]:

b
f (1/2) = cos(π/2) ´ sin(π ) ´ 1/2 = 0 ´ 0 ´ 1/2 = ´1/2
254
? ?
(Notice that f (1/3) = ( 2 ´ 3)/2 ´ 1/2 also works)
So, because f ( x ) is continuous on [0, 1/2) and f (0) ą 0 while f (1/2) ă 0, then the
Intermediate Value Theorem guarantees the existence of a real number c P (0, 1/2) such
that f (c) = 0.
1 3
S-25: We let f ( x ) = ´ x ´ . Then f ( x ) is a continuous function on the interval
cos2 (πx ) 2
(´1/2, 1/2) since cos x is continuous everywhere and non-zero on that interval.
The function f takes negative values. For example, when x = 0:
1 3 3 1
f (0) = 2
´ 0 ´ = 1 ´ = ´ ă 0.
cos (0) 2 2 2
It also takes positive values, for instance when x = 1/4:
1 1 3
f (1/4) = 2
´ ´
(cos π/4) 4 2
1 1+6
= ´
1/2 4
= 2 ´ 7/4 = 1/4 ą 0.
By the IVT there is c, 0 ă c ă 1/4 such that f (c) = 0, in which case

1 3
2
= c+ .
(cos πc) 2
S-26: f ( x ) is a polynomial, so it’s continuous everywhere. If we can find values a and b

so that f ( a) ą 0 and f (b) ă 0, then by the IVT, there will exist some c in ( a, b) where
f (c) = 0; that is, there is a root in the interval [ a, b]. Let’s start plugging in easy values of
x.
f (0) = 15, and f (1) = 1 ´ 15 + 9 ´ 18 + 15 = ´8. Since 0 is between f (0) and f (1), and
since f is continuous, by IVT there must be some x in [0, 1] for which f ( x ) = 0: that is,
there is some root in [0, 1].
That was the easiest interval to find. If you keep playing around, you find two more.
f (´1) = 26 (positive) and f (´2) = ´1001 (negative), so there is a root in [´2, ´1].
The arithmetic is nasty, but there is also a root in [14, 15].
This is an important trick. For high-degree polynomials, it is often impossible to get the
exact values of the roots. Using the IVT, we can at least give a range where a root must
be.
S-27: Let f ( x ) = x3 . Since

? f is a polynomial, it is continuous everywhere. If
3
f ( a) ă 7 ă f (b), then 7 is somewhere between a and b. If we can find a and b that
satisfy these inequalities,
? and are very close together, that will give us a good
3
approximation for 7.
255
?
• Let’s start with integers. 13 ă 7 ă 23 , so 3
7 is in the interval (1, 2).
?
• Let’s narrow this down, say by testing f (1.5). (1.5)3 = 3.375 ă 7, so 3 7 is in the
interval (1.5, 2).
?
• Let’s narrow further, say by testing f (1.75). (1.75)3 « 5.34 ă 7, so 3 7 is in the
interval (1.75, 2).
?
• Testing various points, we find f (1.9) ă 7 ă f (2), so 3 7 is between 1.9 and 2.
?
• By testing more, we find f (1.91) ă 7 ă f (1.92), so 3 7 is in (1.91, 1.92).
?3
• In order to get an approximation
? for 7 that is rounded to two decimal places, we
3
have
? to know whether 7 is greater or less than 1.915;
? indeed f (1.915) « 7.02 ą 7,
3 3
so 7 ă 1.915; then rounded to two decimal places, 7 « 1.91.
?
If this seems like the obvious way to approximate 3 7, that’s good. The IVT is a formal
statement of a very intuitive principle.
S-28:
• If f ( a) = g( a), or f (b) = g(b), then we simply take c = a or c = b.
• Suppose f ( a) ‰ g( a) and f (b) ‰ g(b). Then f ( a) ă g( a) and g(b) ă f (b), so if we
define h( x ) = f ( x ) ´ g( x ), then h( a) ă 0 and h(b) ą 0. Since h is the difference of
two functions that are continuous over [ a, b], also h is continuous over [ a, b]. So, by
the Intermediate Value Theorem, there exists some c P ( a, b) with h(c) = 0; that is,
f ( c ) = g ( c ).
S-1: If Q is to the left of the y axis, the line through Q and P is increasing, so the secant
line has positive slope. If Q is to the right of the y axis, the line through Q and P is
decreasing, so the secant line has negative slope.
S-2: (a) By drawing a few pictures, it’s easy to see that sliding Q closer to P, the slope of
the secant line increases.
(b) Since the slope of the secant line increases the closer Q gets to P, that means the
tangent line (which is the limit as Q approaches P) has a larger slope than the secant line
between Q and P (using the location where Q is right now).
Alternately, by simply sketching the tangent line at P, we can see that has a steeper slope
than the secant line between P and Q.
f (2) ´ f (´2) f (2) ´ f (´2)

S-3: The slope of the secant line will be = , in every part. So,
2 ´ (´2) 4
if two lines have the same slope, that means their differences f (2) ´ f (´2) will be the
same.
256
The graphs in (a),(c), and (e) all have f (2) ´ f (´2) = 1, so they all have the same secant
line slope. The graphs in (b) and (f) both have f (2) ´ f (´2) = ´1, so they both have the
same secant line slope. The graph in (d) has f (2) ´ f (´2) = 0, and it is the only graph
with this property, so it does not share its secant line slope with any of the other graphs.
S-4: A good approximation from the graph is f (5) = 0.5. We want to find a secant line
whose endpoints are both very close to x = 5, but that also give us clear y-values. It looks
like f (5.25) « 1, and f (4.75) « 81 . The secant line from x = 5 to x = 5.25 has approximate
f (5.25) ´ f (5) 1 ´ .5
slope « = 2. The secant line from x = 5 to x = 4.75 has
5.25 ´ 5 .25
0.5 ´ 18 3
approximate slope = .
5 ´ 4.75 2
The graph increases more and more quickly (gets steeper and steeper), so it makes sense
that the secant line to the left of x = 5 has a smaller slope than the secant line to the right
of x = 5. Also, if you’re taking secant lines that have endpoints farther out from x = 5,
you’ll notice that the slopes of the secant lines change quite dramatically. You have to be
very, very close to x = 5 to get any kind of accuracy.
If we split the difference, we might approximate the slope of the secant line to be the
average of 32 and 2, which is 74 .
Another way to try to figure out the tangent line is by carefully drawing it in with a ruler.
This is shown here in blue:
y
x
1 5
It’s much easier to take the slope of a line than a curve, and this one looks like it has slope
about 1.5. However, we drew this with a computer: by hand it’s much harder to draw an
accurate tangent line. (That’s why we need calculus!)
The actual slope of the tangent line to the function at x = 5 is about 1.484. This is
extremely hard to figure out just from the graph–by hand, a guess between 1.25 and 1.75
would be very accurate.
S-5: There is only one tangent line to f ( x ) at P (shown in blue), but there are infinitely
many choices of Q and R (one possibility shown in red). One easy way to sketch the
257
secant line on paper is to draw any line parallel to the tangent line, and choose two
intercepts with y = f ( x ).
y
y = f (x)
x
R
Q
P
S-6: Any place the graph looks flat (if you imagine zooming in) is where the tangent line
has slope 0. This occurs three times.
y
y = f (x)
Notice that two of the indicated points are at a low point and a high point, respectively.
Later, we’ll use these places where the tangent line has slope zero to find where a graph
achieves its biggest and smallest values.
S-1: The function shown is a line, so it has a constant slope–(a) . Since the function is
always increasing, f 1 is always positive, so also (d) holds. Remark: it does not matter that
the function itself is sometimes negative; the slope is always positive because the
function is always increasing. Also, since the slope is constant, f 1 is neither increasing
nor decreasing: it is the function that is increasing, not its derivative.
258
S-2: The function is always decreasing, so f 1 is always negative, option (e). However, the
function alternates between being more and less steep, so f 1 alternates between
increasing and decreasing several times, and no other options hold.
Remark: f is always positive, but (d) does not hold!
S-3: At the left end of the graph, f is decreasing rapidly, so f 1 is a strongly negative
number. Then as we move towards x = 0, f decreases less rapidly, so f 1 is a less strongly
negative number. As we pass 0, f increases, so f 1 is a positive number. As we move to
the right, f increases more and more rapidly, so f 1 is an increasing positive number. This
description tells us that f 1 increases for the entire range shown. So (b) holds, but not (a)
or (c). Since f 1 is negative to the left of the y axis, and positive to the right of it, also (d)
and (e) do not hold.
S-4: By definition, f ( x ) = x3 is differentiable at x = 0 if the limit

f ( h ) ´ f (0) h3 ´ 0
lim = lim
hÑ0 h hÑ0 h
exists.
S-5: f 1 (´1) does not exist, because to the left of x = ´1 the slope is a pretty big positive
number (looks like around +1) and to the right the slope is ´1/4. Since the derivative
involves a limit, that limit needs to match the limit from the left and the limit from the
right. The sharp angle made by the graph at x = ´1 indicates that the left and right
limits do not match, so the derivative does not exist.
f 1 (3) also does not exist. One way to see this is to notice that the function is
discontinuous here. More viscerally, note that f (3) = 1, so as we take secant lines with
one endpoint (3, 1), and the other endpoint just to the right of x = 3, we get slopes that
are more and more strongly negative, as shown in the picture below. If we take the limit
of the slopes of these secant lines as x goes to 3 from the right, we get ´8. (This certainly
doesn’t match the slope from the left, which is ´ 14 .)
y
1
x
1
At x = ´3, there is some kind of “change” in the graph; however, it is a smooth curve, so
the derivative exists here.
S-6: True. The definition of the derivative tells us that

f ( a + h) ´ f ( a)
f 1 ( a) = lim ,
hÑ0 h
259
if it exists. We know from our work with limits that if both one-sided limits
f ( a + h) ´ f ( a) f ( a + h) ´ f ( a)
lim and lim exist and are equal to each other, then
hÑ0´ h hÑ0+ h
f ( a + h) ´ f ( a)
lim exists and has the same value as the one-sided limits. So, since the
hÑ0 h
one-sided limits exist and are equal to one, we conclude f 1 ( a) exists and is equal to one.
S-7: In general, this is false. The key problem that can arise is that f ( x ) might not be
continuous at x = 1. One example is the function
"
x xă0
f (x) =
x´1 x ě 0
where f 1 ( x ) = 1 whenever x ‰ 0 (so in particular, lim f 1 ( x ) = lim f 1 ( x ) = 1) but f 1 (0)
xÑ0´ xÑ0+
does not exist.
There are two ways to see that f 1 (0) does not exist. One is to notice that f is not
continuous at x = 0.
y
y = f (x)
Another way to see that f 1 (0) does not exist is to use the definition of the derivative.
Remember, in order for a limit to exist, both one-sided limits must exist. Let’s consider
the limit from the left. If h Ñ 0´ , then h ă 0, so f (h) is equal to h (not h ´ 1).
f (0 + h ) ´ f (0) (h) ´ (´1)

lim = lim
hÑ0´ h hÑ0´ h
h+1
= lim
hÑ0´ h
1
= lim 1 +
hÑ0´ h
= ´8
In particular, this limit does not exist. Since the one-sided limit does not exist,
f (0 + h ) ´ f (0)
lim = DNE
hÑ0 h
260
and so f 1 (0) does not exist.
S-8: Using the definition of the derivative,
s(t + h) ´ s(t)
s1 (t) = lim
hÑ0 h
The units of the numerator are meters, and the units of the denominator are seconds
(since the denominator comes from the change in the input of the function). So, the units
of s1 (t) are metres per second.
Remark: we learned that the derivative of a position function gives velocity. In this
example, the position is given in metres, and the velocity is measured in metres per
second.
S-9: We can use point-slope form to get the equation of the line, if we have a point and its
slope. The point is given: (1, 6). The slope is the derivative:
y (1 + h ) ´ y (1)
y1 (1) = lim
hÑ0 h
[(1 + h)3 + 5] ´ [13 + 5]
= lim
hÑ0 h
[1 + 3h + 3h2 + h3 + 5] ´ [1 + 5]
= lim
hÑ0 h
2
3h + 3h + h 3
= lim
hÑ0 h
= lim 3 + 3h + h2
hÑ0
=3
So our slope is 3, which gives a line of equation y ´ 6 = 3( x ´ 1).
261
S-10: We set up the definition of the derivative.
f ( x + h) ´ f ( x )
f 1 ( x ) = lim
hÑ0 h
1 1
´
= lim x+h x
hÑ0 h
x
x ( x +h)
´ x(xx++hh)
= lim
hÑ0 h
x´( x +h)
x ( x +h)
= lim
hÑ0 h
´h
x ( x +h)
= lim
hÑ0 h
´1
= lim
hÑ0 x ( x + h )
´1
= 2
x
S-11: By definition
f ( h ) ´ f (0) h|h|
f 1 (0) = lim = lim = lim |h| = 0
hÑ0 h hÑ0 h hÑ0
In particular, the limit exists, so the derivative exists (and is equal to zero).
S-12: We set up the definition of the derivative.
1 f ( x + h) ´ f ( x ) 1 2 2 2 ( x + 1) ´ ( x + h + 1)
f ( x ) = lim = lim ´ = lim
hÑ0 h hÑ0 h x + h + 1 x+1 hÑ0 h ( x + h + 1)( x + 1)
2 ´h ´2 ´2
= lim = lim =
hÑ0 h ( x + h + 1)( x + 1) hÑ0 ( x + h + 1)( x + 1) ( x + 1)2
S-13:
f ( x + h) ´ f ( x ) 1 1 1 1 x 2 ´ ( x + h )2
f 1 ( x ) = lim = lim ´ = lim
hÑ0 h hÑ0 h ( x + h )2 + 3 x2 + 3 hÑ0 h [( x + h )2 + 3][ x 2 + 3]
1 ´2xh ´ h2 ´2x ´ h ´2x

= lim 2 2
= lim 2 2
= 2
hÑ0 h [( x + h ) + 3][ x + 3] hÑ0 [( x + h ) + 3][ x + 3] [ x + 3]2
S-14: The slope of the tangent line is the derivative. We set this up using the same
definition of the derivative that we always do. This limit is hard to take for general x, but
262
easy when x = 0.
f (0 + h ) ´ f (0)
f 1 (0) = lim
hÑ0 h
h log10 (2h + 10) ´ 0
= lim
hÑ0 h
= lim log10 (2h + 10) = log10 (10) = 1
hÑ0
So, the slope of the tangent line is 1.
S-15:
1 1
f ( x + h) ´ f ( x ) ( x + h )2
´ x2 x 2 ´ ( x + h )2 ´2xh ´ h2
1
f ( x ) = lim = lim = lim = lim
hÑ0 h hÑ0 h hÑ0 ( x + h )2 x 2 h hÑ0 ( x + h )2 x2 h
´2x ´ h ´2x 2
= lim 2 2
= 4 =´ 3
hÑ0 ( x + h ) x x x
S-16: When x is not equal to 2, then the function is differentiable– the only place we have
to worry about is when x is exactly 2.
In ˇorder for
2
differentiable at x = 2, it must also be continuous at x = 2. This forces
f to be
x x=2 = ax + b x=2 or
ˇ
2a + b = 4.
In order for a limit to exist, the left- and right-hand limits must exist and be equal to each
other. Since a derivative is a limit, in order for f to be differentiable at x = 2, the left hand
derivative of ax + b at x = 2 must be the same as the right hand derivative of x2 at x = 2.
Since ax + b is a line, its derivative is a everywhere. We’ve already seen the derivative of
x2 is 2x, so we need
ˇ
a = 2xˇx=2 = 4.
So, the values of a and b that makes f differentiable everywhere are a = 4 and b = ´4.
S-17: We plug in f ( x ) to the definition of a derivative. To evaluate the limit, we multiply
263
and divide by the conjugate of the numerator, then simplify.
? ?
1 f ( x + h) ´ f ( x ) 1+x+h´ 1+x
f ( x ) = lim = lim
hÑ0 h hÑ0 h
? ? ? ?
1+x+h´ 1+x 1+x+h+ 1+x
= lim ? ?
hÑ0 h 1+x+h+ 1+x
(1 + x + h ) ´ (1 + x )
= lim ? ?
hÑ0 h ( 1 + x + h + 1 + x )
h
= lim ? ?
hÑ0 h ( 1 + x + h + 1 + x )
1
= lim ? ?
hÑ0 1 + x + h + 1 + x
1 1
=? ? = ?
1+x+0+ 1+x 2 1+x
The domain of the function is [´1, 8). In particular, f ( x ) is defined when x = ´1.
However, f 1 ( x ) is not defined when x = ´1, so f 1 ( x ) only exists over (´1, 8).
Remark: lim f 1 ( x ) = 8, so the tangent line to f ( x ) at the point x = ´1 has a vertical
xÑ´1+
slope.
S-18: From Section 1.2, we see that the velocity is exactly the derivative.
s(t + h) ´ s(t)
v(t) = lim
hÑ0 h
( t + h )4 ´ ( t + h )2 ´ t4 + t2
= lim
hÑ0 h
(t + 4t h + 6t2 h2 + 4th3 + h4 ) ´ (t2 + 2th + h2 ) ´ t4 + t2
4 3
= lim
hÑ0 h
4t h + 6t h + 4th + h ´ 2th ´ h2
3 2 2 3 4
= lim
hÑ0 h
= lim 4t + 6t h + 4th + h3 ´ 2t ´ h
3 2 2
hÑ0
3
= 4t ´ 2t
So, the velocity is given by v(t) = 4t3 ´ 2t.
S-19: The function is differentiable at x = 0 if the following limit:

f ( x ) ´ f (0) f (x) ´ 0 f (x)
lim = lim = lim
xÑ0 x´0 xÑ0 x xÑ0 x
exists (note that we used the fact that f (0) = 0 as per the definition of the first branch
which includes the point x = 0). We start
? by computing the left limit. For this
computation, recall that if x ă 0 then x2 = |x| = ´x.
? ? ? ?
f (x) x2 + x4 x2 1 + x2 ´x 1 + x2
lim = lim = lim = lim = ´1
xÑ0´ x xÑ0´ x xÑ0´ x xÑ0 x
264
Now, from the right:
x cos x
lim = lim cos x = 1.
xÑ0+ x xÑ0+
Since the limit from the left does not equal the limit from the right, the derivative does
not exist at x = 0.
f ( x ) ´ f (0) f (x) ´ 0 f (x)

lim = lim = lim
which includes the point x = 0).
We start by computing the left limit.
f (x) x cos x
lim = lim = lim cos x = 1.
xÑ0´ x xÑ0´ x xÑ0´
Now, from the right:

? ? ?
1+x´1 1+x´1 1+x+1
lim = lim ¨?
xÑ0+ x xÑ0+ x 1+x+1
1+x´1 1 1
= lim ? = lim ? =
xÑ0+ x ( 1 + x + 1) xÑ0+ 1+x+1 2
Since the limit from the left does not equal the limit from the right, the derivative does
not exist at x = 0.
f ( x ) ´ f (0) f (x) ´ 0 f (x)

lim = lim = lim
which includes the point x = 0). We compute left and right limits; so
f (x) x3 ´ 7x2
lim = lim = lim x2 ´ 7x = 0
xÑ0´ x xÑ0´ x xÑ0´
and

1
x3 cos x

1
2
lim = lim x ¨ cos .
xÑ0+ x xÑ0 + x
This last limit equals 0 by the Squeeze Theorem since

1
´1 ď cos ď1
x
265
and so,

2 12
´x ď x ¨ cos ď x2 ,
x
where in these inequalities we used the fact that x2 ě 0. Finally, since

limxÑ0+ ´x2 =lim 2
xÑ0+ x = 0, the Squeeze Theorem yields that also
1
limxÑ0+ x2 cos x = 0, as claimed.
Since the left and right limits match (they’re both equal to 0), we conclude that indeed
f ( x ) is differentiable at x = 0 (and its derivative at x = 0 is actually equal to 0).
f ( x ) ´ f (1) f (x) ´ 0 f (x)

lim = lim = lim
xÑ1 x´1 xÑ1 x ´ 1 xÑ1 x ´ 1
which includes the point x = 0). We compute left and right limits; so
f (x) 4x2 ´ 8x + 4 4( x ´ 1)2

lim = lim = lim = lim 4( x ´ 1) = 0
xÑ1´ x´1 xÑ1´ x´1 xÑ1´ x´1 xÑ1´
and

1
( x ´ 1)2 sin x´1

1

lim = lim ( x ´ 1) ¨ sin .
xÑ1+ x´1 xÑ1+ x´1
This last limit equals 0 by the Squeeze Theorem since

1
´1 ď sin ď1
x´1
and so,

1
´( x ´ 1) ď ( x ´ 1) ¨ sin ď x ´ 1,
x´1
where in these inequalities we used the fact that x Ñ 1+ yields positive values for x ´ 1.
+ ´x + 1 = limxÑ1+ x ´ 1 = 0, the Squeeze Theorem yields that also
Finally, since limxÑ1
1
limxÑ1+ ( x ´ 1) sin x´1 = 0, as claimed.
Since the left and right limits match (they’re both equal to 0), we conclude that indeed
f ( x ) is differentiable at x = 1 (and its derivative at x = 1 is actually equal to 0).
266
y
1
x
1
The key is to realize that the few points you’re given suggest a pattern, but don’t
guarantee it. You only know nine points; anything can happen in between.
S-24:
p( x + h) ´ p( x )
p1 ( x ) = lim
hÑ0 h
f ( x + h) + g( x + h) ´ f ( x ) ´ g( x )
= lim
hÑ0 h
f ( x + h) ´ f ( x ) + g( x + h) ´ g( x )
= lim
hÑ0 h

f ( x + h) ´ f ( x ) g( x + h) ´ g( x )
= lim +
hÑ0 h h

f ( x + h) ´ f ( x ) g( x + h) ´ g( x )
(˚) = lim + lim
hÑ0 h hÑ0 h
1 1
= f (x) + g (x)
At step (˚), we use the limit law that lim [ F ( x ) + G ( x )] = lim F ( x ) + lim G ( x ), as long as
xÑa xÑa xÑa
lim F ( x ) and lim G ( x ) exist. Because the problem states that f 1 ( x ) and g1 ( x ) exist, we
xÑa xÑa
f ( x + h) ´ f ( x ) g( x + h) ´ g( x )
know that lim and lim exist, so our work is valid.
hÑ0 h hÑ0 h
S-25: (a) Since y = f ( x ) = 2x and y = g( x ) = x are straight lines, we don’t need the
definition of the derivative (although you can use it if you like). f 1 ( x ) = 2 and g1 ( x ) = 1.
267
(b) p( x ) = 2x2 , so p( x ) is not a line: we use the definition of a derivative to find p1 ( x ).
p( x + h) ´ p( x )
p1 ( x ) = lim
hÑ0 h
2( x + h)2 ´ 2x2
= lim
hÑ0 h
2x + 4xh + 2h2 ´ 2x2
2
= lim
hÑ0 h
4xh + 2h 2
= lim
hÑ0 h
= lim 4x + 2h
hÑ0
= 4x
(c) No, p1 ( x ) = 4x ‰ 2 ¨ 1 = f 1 ( x ) ¨ g1 ( x ). In general, the derivative of a product is not the

same as the derivative of the functions being multiplied.
S-26: We know that y1 = 2x. So, if we choose a point (α, α2 ) on the curve y = x2 , then the
tangent line to the curve at that point has slope 2α. That is, the tangent line has equation
(y ´ α2 ) = 2α( x ´ α)
simplified, y = (2α) x ´ α2
So, if (1, ´3) is on the tangent line, then
´3 = (2α)(1) ´ α2
ðñ 0 = α2 ´ 2α ´ 3
ðñ 0 = (α ´ 3)(α + 1)
ðñ α = 3, or α = ´1.
So, the tangent lines y = (2α) x ´ α2 are

y = 6x ´ 9 and y = ´2x ´ 1.
S-27: Using the definition of the derivative, f is differentiable at 0 if and only if

f ( h ) ´ f (0)
lim
hÑ0 h
exists. In particular, this means f is differentiable at 0 if and only if both one-sided limits
exist and are equal to each other.
When h ă 0, f (h) = 0, so
f ( h ) ´ f (0) 0´0
lim = lim =0
hÑ0´ h hÑ0´ h
268
So, f is differentiable at x = 0 if and only if
f ( h ) ´ f (0)
lim = 0.
hÑ0+ h

1
To evaluate the limit above, we note f (0) = 0 and, when h ą 0, f (h) = h a sin h , so

1
f ( h ) ´ f (0) h a sin h
lim = lim
hÑ0 + h hÑ0+ h

a´1 1
= lim h sin
hÑ0+ h
We will spend the rest of this solution evaluating the limit above for different values of a,
to find when it is equal to zero and when it is not. Let’s consider the different values that
could be taken by h a´1 .
• If a = 1, then a ´ 1 = 0, so h a´1 = h0 = 1 for all values of h. Then

a´1 1 1
lim h sin = lim sin = DNE
hÑ0 + h hÑ0 + h

(Recall that the function sin 1x oscillates faster and faster as x goes to 0. We first
saw this behaviour in Example 1.3.5.)
• If a ă 1, then a ´ 1 ă 0, so lim h a´1 = 8. (Since we have a negative exponent, we
hÑ0+
are in effect dividing by a smaller and smaller positive number. For example, if
1 a´1 ´ 12 1
a = 2 , then lim h = lim h = lim ? = 8.) Since sin 1x goes back and
hÑ0+ hÑ0+ hÑ0+ h
forth between one and negative one,

a´1 1
lim h sin = DNE
hÑ0 + x
since as h goes to 0, the function oscillates between positive and negative numbers
of ever-increasing magnitude.

• If a ą 1, then a ´ 1 ą 0, so lim h a´1 = 0. Although sin 1x oscillates wildly near
hÑ0+
x = 0, it is bounded by ´1 and 1. So,

a´1 a´1 1
(´1)h ďh sin ď h a´1
h
Since both lim (´1)h a´1 = 0 and lim h a´1 = 0, by the Squeeze Theorem,
hÑ0+ hÑ0+

a´1 1
lim h sin =0
hÑ0+ x
as well.
269
In the above cases, we learned
f ( h ) ´ f (0) 1
lim = lim h a´1 sin = 0 when a ą 1, and
hÑ0+ h hÑ0+ x

f ( h ) ´ f (0) a´1 1
lim = lim h sin ‰ 0 when a ď 1.
hÑ0 + h hÑ0 + x
So, f is differentiable at x = 0 if and only if a ą 1.
h(24) ´ h(0) m
S-1: (a) The slope of the secant line is ; this is the change in height over
24 ´ 0 hr
the first day divided by the number of hours in the first day. So, it is the average rate of
change of the height over the first day, measured in meters per hour.
(b) Consider (a). The secant line gives the average rate of change of the height of the dam;
as we let the second point of the secant line get closer and closer to (0, h(0)), its slope
approximates the instantaneous rate of change of the height of the water. So the slope of
the tangent line is the instantaneous rate of change of the height of the water at the time
m
t = 0, measured in hr .
p(t + h) ´ p(t) p ( t + 1) ´ p ( t )
S-2: p1 (t) = lim « = p(t + 1) ´ p(t), or the difference in
hÑ0 h 1
profit caused by the sale of the (t + 1)st widget. So, p1 (t) is the profit from the (t + 1)st
widget. That is, p1 (t) is the profit per additional widget sold, when t widgets are being
sold. This is called the marginal profit per widget, when t widgets are being sold.
S-3: How quickly the temperature is changing per unit change of depth, measured in
degrees per metre. In an ordinary body of water, the temperature near the surface (d = 0)
is pretty variable, depending on the sun, but deep down it is more stable (unless there
are heat sources). So, one might reasonably expect that |T 1 (d)| is larger when d is near 0.
C (w + h) ´ C (w) C ( w + 1) ´ C ( w )
S-4: C1 (w) = lim « = C (w + 1) ´ C (w), which is the
hÑ0 h 1
number of calories in C (w + 1) grams minus the number of calories in C (w) grams. This
is the number of calories per additional gram, when there are w grams.
S-5: The rate of change of velocity is acceleration. (If your velocity is increasing, you’re
accelerating; if your velocity is decreasing, you have negative acceleration.)
S-6: The rate of change in this case will be the relationship between the heat added and
T ( j + h) ´ T ( j) T ( j + 1) ´ T ( j )
the temperature change. lim « = T ( j + 1) ´ T ( j), or the
hÑ0 h 1
change in temperature after the application of one joule. (This is closely related to heat
capacity and to specific heat — there’s a nice explanation of this on Wikipedia.)
270
S-7: As usual, it is instructive to think about the definition of the derivative:
P( T + h) = P( T ) P ( T + 1) ´ P ( t )
P1 ( T ) = lim « = P ( T + 1) ´ P ( T ).
hÑ0 h 1
This is the difference in population between two hypothetical populations, raised one
degree in temperature apart. So, it is the number of extra individuals that exist in the
hotter experiment (with the understanding that this number could be negative, as one
would expect in conditions that are hotter than the bacteria prefer). So P1 ( T ) is the
number of bacteria added to the colony per degree.
S-8: R1 (t) is the rate at which the wheel is rotating measured in rotations per second. To
convert to degrees, we multiply by 360: 360R1 (t) .
S-9: If P1 (t) is positive, your sample is below the ideal temperature, because adding heat
increases the population. If P1 (t) is negative, your sample is above the ideal temperature,
because adding heat decreases the population. If P1 (t) = 0, then adding a little bit of heat
doesn’t change the population, but it’s unclear why this is. Perhaps your sample is
deeply frozen, and adding heat doesn’t change the fact that your population is 0. Perhaps
your sample is boiling, and again, changing the heat a little will keep the population
constant at “none.” But also, at the ideal temperature, you would expect P1 (t) = 0. This
is best seen by noting in the curve below, the tangent line is horizontal at the peak.
0
y
to
)ą
o
ho
(1 t
t,
,P
P
1
ld
(t
co
)ă
o
to
frozen, P1 (t) = 0 ideal, P1 (t) = 0 boiled, P1 (t) = 0
S-1: True: this is exactly what the Sum Rule states.
d
S-2: False, in general. The product rule tells us dx t f ( x ) g ( x )u = f 1 ( x ) g ( x ) + f ( x ) g1 ( x ).
An easy example of why we can’t do it the other way is to take f ( x ) = g( x ) = x. Then
d
the equation becomes dx tx2 u = (1)(1), which is false.
271
S-3: True: the quotient rule tells us
" *
d f (x) g ( x ) f 1 ( x ) ´ f ( x ) g1 ( x ) g( x ) f 1 ( x ) f ( x ) g1 ( x ) f 1 (x) f ( x ) g1 ( x )
= = ´ = ´ .
dx g( x ) g2 ( x ) g2 ( x ) g2 ( x ) g( x ) g2 ( x )
S-4: If you’re creative, you can find lots of ways to differentiate!

Constant multiple: g1 ( x ) = 3 f 1 ( x ).
d
Product rule: g1 ( x ) = dx t3u f ( x ) + 3 f 1 ( x ) = 0 f ( x ) + 3 f 1 ( x ) = 3 f 1 ( x ).
d
Sum rule: g1 ( x ) = dx t f ( x )"+ f (* x ) + f ( x )u = f 1 ( x ) + f 1 ( x ) + f 1 ( x ) = 3 f 1 ( x ).
1 1 1 1
d f (x) 3 f ( x )´ f ( x )(0) 3 f (x) 1
Quotient rule: g1 ( x ) = dx 1 = 1 = 1 = 9 1 1
3 f ( x ) = 3 f ( x ).
3 9 9
All rules give g1 ( x ) = 3 f 1 ( x ).
d 2
S-5: We know, from Examples 2.2.5 and 2.2.9 in the CLP-1 text, that dx x = 2x and
d 1/2
dx x = 2?1 x . So, by linearity,
1 2
f 1 ( x ) = 3 ¨ 2x + 4 ¨ ? = 6x + ?
2 x x
d ? 1
S-6: We have already seen dx t xu = ?
2 x
in Example 2.2.9 of the CLP-1 text. Now:

1
? 8
f ( x ) = (2)(8 x ´ 9x ) + (2x + 5) ? ´9
2 x

? 4
= 16 x ´ 18x + (2x + 5) ? ´ 9
x
? 20
= ´36x + 24 x + ? ´ 45
x
d d 2
S-7: We already know that dx x = 1 and dx x = 2x, so we can compute the derivative of
3 3 2
x by writing x = ( x )( x ),
d 3 d
x = ( x )( x2 ) = (1)( x2 ) + ( x )(2x ) = 3x2
dx dx
3
When this is evaluated at x = 2 we get 4 . Since we also compute 12 = 81 , the equation
1 3
of the tangent line is

1 3 1
y´ = ¨ x´ .
8 4 2
272
d 3
S-8: Let f (t) = t3 ´ 4t2 + 1. We saw in Question 7 that dt t = 3t2 . So
f 1 (t) = 3t2 ´ 8t f 1 (2) = 3 ˆ 4 ´ 8 ˆ 2 = ´4

f 2 (t) = 6t ´ 8 f 2 (2) = 6 ˆ 2 ´ 8 = 4
Hence at t = 2, (a) the particle has speed of magnitude 4, and (b) is moving towards the
left. At t = 2, f 2 (2) ą 0, so f 1 is increasing, i.e. becoming less negative. Since f 1 is getting
closer to zero, (c) the magnitude of the speed is decreasing.
S-9: We can use the quotient rule here.

" *
d 2x ´ 1 (2x + 1)(2) ´ (2x ´ 1)(2) 4 1
= = =
dx 2x + 1 (2x + 1)2 (2x + 1)2 ( x + 1/2)2
S-10: First, we find the y1 for general x. Using the corollary to Theorem 2.4.3 and the
quotient rule:
" *
1 3x + 1 d 3x + 1
y =2 ¨
3x ´ 2 dx 3x ´ 2

3x + 1 (3x ´ 2)(3) ´ (3x + 1)(3)
=2
3x ´ 2 (3x ´ 2)2

3x + 1 ´9
=2
3x ´ 2 (3x ´ 2)2
´18(3x + 1)
=
(3x ´ 2)3
So, plugging in x = 1:
´18(3 + 1)
y1 (1) = = ´72
(3 ´ 2)3
?
S-11: We need f 1 (1), so first we must find f 1 ( x ). Since f ( x ) is the reciprocal of x + 1, we
can use the Corollary 2.4.6:
d ? ´ 2?1 x
1 ´ dx t x + 1u ´1
f (x) = ? 2
= ? 2
= ? ? ,
( x + 1) ( x + 1) 2 x ( x + 1)2
´1
so f 1 (1) = ? ?
2
= ´1
8 .
2 1( 1 + 1)

Now, using the point 1, 12 and the slope ´1
8 , our tangent line has equation
y ´ 21 = ´ 18 ( x ´ 1).
273
S-12: Population growth is rate of change of population. Population in year 2000 + t is
given by P(t) = P0 + b(t) ´ d(t), where P0 is the initial population of the town. Then
P1 (t) is the expression we’re looking for, and P1 (t) = b1 (t) ´ d1 (t).
It is interesting to note that the initial population does not obviously show up in this
calculation. It would probably affect b(t) and d(t), but if we know these we do not need
to know P0 to answer our question.
d 2
S-13: We already know that dx x = 2x. So the slope of y = 3x2 at x = a is 6a. The tangent
line to y = 3x2 at x = a, y = 3a2 is y ´ 3a2 = 6a( x ´ a). This tangent line passes through
(2, 9) if
9 ´ 3a2 = 6a(2 ´ a)
3a2 ´ 12a + 9 = 0
a2 ´ 4a + 3 = 0
( a ´ 3)( a ´ 1) = 0
ùñ a = 1, 3
The points are (1, 3), (3, 27).
S-14: This limit represents the derivative computed at x = 100180 of the function
? 1
f ( x ) = x. Since the derivative of f ( x ) is ? , then its value at x = 100180 is exactly
2 x
1
? .
2 100180
S-15: Let w(t) and l (t) be the width and length of the rectangle. Given in the problem is
that w1 (t) = 2 and l 1 (t) = 5. Since both functions have constant slopes, both must be
lines. Their slopes are given, and their intercepts are w(0) = l (0) = 1. So, w(t) = 2t + 1
and l (t) = 5t + 1.
The area of the rectangle is A(t) = w(t) ¨ l (t), so using the product rule, the rate at which
the area is increasing is A1 (t) = w1 (t)l (t) + w(t)l 1 (t) = 2(5t + 1) + 5(2t + 1) = 20t + 7
square metres per second.
S-16: Using the product rule, f 1 ( x ) = (2x ) g( x ) + x2 g1 ( x ), so

f 1 (0) = 0 ¨ g( x ) + 0 ¨ g1 ( x ) = 0. (Since g is differentiable, g1 exists.)
S-17:
g( x )
First expression, f ( x ) = :
h( x )
h ( x ) g1 ( x ) ´ g ( x ) h1 ( x )
f 1 (x) =
h2 ( x )
274
g( x ) k( x )
Second expresson, f ( x ) = ¨ :
k( x ) h( x )
k ( x ) g1 ( x ) ´ g ( x ) k 1 ( x ) h ( x ) k 1 ( x ) ´ k ( x ) h1 ( x )

k( x) g( x )
f 1 (x) = +
k2 ( x ) h( x ) k( x) h2 ( x )
1 1 1
k( x ) g ( x ) ´ g( x )k ( x ) g( x )h( x )k ( x ) ´ g( x )k( x )h ( x ) 1
= +
k( x )h( x ) k ( x ) h2 ( x )
h ( x ) k ( x ) g1 ( x ) ´ h ( x ) g ( x ) k 1 ( x ) g ( x ) h ( x ) k 1 ( x ) ´ g ( x ) k ( x ) h1 ( x )
= +
k ( x ) h2 ( x ) k ( x ) h2 ( x )
h ( x ) k ( x ) g1 ( x ) ´ h ( x ) g ( x ) k 1 ( x ) + g ( x ) h ( x ) k 1 ( x ) ´ g ( x ) k ( x ) h1 ( x )
=
k ( x ) h2 ( x )
h ( x ) k ( x ) g1 ( x ) ´ g ( x ) k ( x ) h1 ( x )
=
k ( x ) h2 ( x )
h ( x ) g1 ( x ) ´ g ( x ) h1 ( x )
=
h2 ( x )
and this is exactly what we got from differentiating the first expression.
S-1: In the quotient rule, there is a minus, not a plus. Also, 2( x + 1) + 2x is not the same
as 2( x + 1).
The correct version is:
2x
f (x) =
x+1
2 ( x + 1)´2x
f 1 (x) =
( x + 1)2
2
=
( x + 1)2
d
S-2: False: Lemma 2.6.9 tells us that, for a constant n, tx n u = nx n´1 . Note that the base
dx
x is the variable and the exponent n is a constant. In the equation given in the question,
the base 2 is a constant, and the exponent x is the variable: this is the opposite of the
situation where Lemma 2.6.9 applies.
We do not yet know how to differentiate 2x . We’ll learn about it in Section 2.7.
275
S-3: f ( x ) = 32 x6 + 5x4 + 12x2 + 9 is a polynomial:
f 1 ( x ) = 4x5 + 20x3 + 24x

= 4x ( x4 + 5x2 + 6)
= 4x (( x2 )2 + 5( x2 ) + 6)
= 4x ( x2 + 2)( x2 + 3)
S-4: We can rewrite slightly to make every term into a power of t:
s(t) = 3t4 + 5t3 ´ t´1

s1 (t) = 4 ¨ 3t3 + 3 ¨ 5t2 ´ (´1) ¨ t´2
1
= 12t3 + 15t2 + 2
t
S-5: We could use the product rule here, but it’s easier to simplify first. Don’t be confused
by the role reversal of x and y: x is the name of the function, and y is the variable.

1
x (y) = 2y + ¨ y3
y
= 2y4 + y2
x1 (y) = 8y3 + 2y
d ?
S-6: We’ve already seen that dx t xu = 2?1 x , but if you forget this formula it is easy to
? d ?
figure out: x = x1/2 , so dx t xu = 12 x´1/2 = 2?1 x .
Using the quotient rule:

?
x+1
T (x) =
x2 +3
2 1
?
( x + 3) 2 x ´ ( x + 1)(2x )
?
T1 (x) =
( x 2 + 3)2
S-7: We use quotient rule:
( x2 + 3) ¨ 7 ´ 2x ¨ (7x + 2) 21 ´ 4x ´ 7x2
=
( x 2 + 3)2 ( x 2 + 3)2
276
S-8: Instead of multiplying to get our usual form of this polynomial, we can use the
product rule. If f 1 ( x ) = 3x3 + 4x2 + x + 1 and f 2 ( x ) = 2x + 5, then
f 11 ( x ) = 9x2 + 8x + 1 and f 21 ( x ) = 2. Then
f 1 (0) = f 11 (0) f 2 (0) + f 1 (0) f 21 (0)

= (1)(5) + (1)(2) = 7
S-9: Using the quotient rule,
( x2 + 5x )(9x2 ) ´ (3x3 + 1)(2x + 5) 3x4 + 30x3 ´ 2x ´ 5

f 1 (x) = =
( x2 + 5x )2 ( x2 + 5x )2
(2 ´ x )(6x ) ´ (3x2 + 5)(´1) ´3x2 + 12x + 5

=
(2 ´ x )2 ( x ´ 2)2
(3x2 + 5)(´2x ) ´ (2 ´ x2 )(6x ) ´22x

2 2
=
(3x + 5) (3x2 + 5)2
6x2 ¨ ( x + 2) ´ (2x3 + 1) ¨ 1 4x3 + 12x2 ´ 1

=
( x + 2)2 ( x + 2)2
S-13: The derivative of the function is

?
(1 ´ x2 ) ¨ 2?1 x ´ x ¨ (´2x ) (1 ´ x2 ) ´ 2x ¨ (´2x )
= ?
(1 ´ x 2 )2 2 x (1 ´ x 2 )2
The derivative is undefined if either x ă 0 or x = 0, ˘1 (since the square-root is

undefined for x ă 0 and the denominator is zero when x = 0, 1, ´1. Putting this together
— the derivative exists for x ą 0, x ‰ 1.
277
S-14: Using the product rule seems faster than expanding.
d ? ? ( ? ? d ! 2 )
f 1 (x) = 3 5 x + 15 3 x + 8 3x2 + 8x ´ 5 + 3 5 x + 15 3 x + 8 3x + 8x ´ 5
dx dx
d ! 1 1
) ? ? d ! )
3x 5 + 15x 3 + 8 3x2 + 8x ´ 5 + 3 5 x + 15 3 x + 8 3x2 + 8x ´ 5

=
dx
dx
3 ´4 ´2
? ?
3x2 + 8x ´ 5 + 3 5 x + 15 3 x + 8 (6x + 8)

= x 5 + 5x 3
5
S-15: To avoid the quotient rule, we can divide through the denominator:
? ? ? ?
( x2 + 5x + 1)( x + 3 x ) 2 ( x + 3 x)
f (x) = = ( x + 5x + 1)
x x
2 ´1/2 ´2/3
= ( x + 5x + 1)( x +x )
Now, product rule:

1 ´1/2 ´2/3 2 ´1 ´3/2 2 ´5/3
f ( x ) = (2x + 5)( x +x ) + ( x + 5x + 1) x ´ x
2 3
(If you simplified differently, or used the quotient rule, you probably came up with a
different-looking answer. There is only one derivative, though, so all correct answers will
look the same after sufficient algebraic manipulation.)
S-16: The question asks us to find where f 1 ( x ) = 0 and f ( x ) exists. We can use the
formula for the derivative of a reciprocal, Corollary 2.4.6:
! )
d 1 5
´ dx 5x + x4 ´ 53 x3
f 1 (x) = 2
1 5 5 3
5x + x4 ´ 3x
´ x4 + 4x3 ´ 5x2

= 2
1 5 4 ´ 5 x3
5 x + x 3
´x2 x2 + 4x ´ 5

= 2
1 5 4 5 3
5x + x ´ 3x
´x2 ( x + 5)( x ´ 1)
= 2
1 5 4 ´ 5 x3
5 x + x 3
So our candidates for x-values where f 1 ( x ) = 0 are x = 0, x = ´5, and x = 1. However,

we need to check that f exists at these places: f (0) is undefined (and f 1 (0) doesn’t exist).
So f 1 ( x ) = 0 only when x = ´5 and x = 1.
278
S-17: Denote by m the slope of the common tangent, by ( x1 , y1 ) the point of tangency
with y = x2 , and by ( x2 , y2 ) the point of tangency with y = x2 ´ 2x + 2. Then we must
have
y2 ´ y1
y1 = x12 y2 = x22 ´ 2x2 + 2 m = 2x1 = 2x2 ´ 2 =
x2 ´ x1
m m
From the “m” equations we get x1 = 2 , x2 = 2 + 1 and
y2 ´ y1
m=
x2 ´ x1
= y2 ´ y1
= x22 ´ 2x2 + 2 ´ x12
= ( x2 ´ x1 )( x2 + x1 ) ´ 2( x2 ´ 1)
m m m m m
= +1´ +1+ ´2 +1´1
2 2 2 2 2
m
= (1)(m + 1) ´ 2
2
=1
1 3 9 5
So, m = 1, x1 = 12 , y1 = , x2 = , y2 = ´3+2 =
4 2 4 4
An equation of the common tangent is y = x ´ 14 .
S-18: The line y = mx + b is tangent to y = x2 at x = α if

2α = m and α2 = mα + b ðñ m = 2α and b = ´α2
The same line y = mx + b is tangent to y = ´x2 + 2x ´ 5 at x = β if
´2β + 2 = m and ´ β2 + 2β ´ 5 = mβ + b
ðñ m = 2 ´ 2β and b = ´β2 + 2β ´ 5 ´ (2 ´ 2β) β = β2 ´ 5
For the line to be simultaneously tangent to the two parabolas we need
m = 2α = 2 ´ 2β and b = ´α2 = β2 ´ 5
Substituting α = 1 ´ β into ´α2 = β2 ´ 5 gives ´(1 ´ β)2 = β2 ´ 5 or 2β2 ´ 2β ´ 4 = 0 or
β = ´1, 2. The corresponding values of the other parameters are α = 2, ´1, m = 4, ´2
and b = ´4, ´1. The two lines are y = 4x ´ 4 and y = ´2x ´ 1.
y
x
−1 2
279
S-19: This limit represents the derivative computed at x = 2 of the function f ( x ) = x2015 .
To see this, simply use the definition of the derivative at a = 2 with f ( x ) = x2015 :
ˇ
d ˇ f ( x ) ´ f ( a)
t f ( x )uˇˇ = lim
dx xÑa xá
ˇa
d x2015 ´ 22015
tx2015 uˇˇ = lim
ˇ
dx 2 xÑ2 x´2
Since the derivative of f ( x ) is 2015 ¨ x2014 , then its value at x = 2 is exactly 2015 ¨ 22014 .
S-1: Since 1x = 1 for any x, we see that (b) is just the constant function y = 1, so D
matches to (b).
x
Since 2´x = 21x = 12 , functions ( a) and (d) are the same. This is the only function out
of the lot that grows as x Ñ ´8 and shrinks as x Ñ 8, so A matches to ( a) and (d).
This leaves B and C to match to (c) and (e). Since 3 ą 2, when x ą 0, 3x ą 2x . So, (e)
matches to the function that grows more quickly to the right of the x-axis: B matches to
(e), and C matches to (c).
S-2: First, let’s consider the behaviour of exponential

" functions a x based on"whether a is
8 aą1 0 aą1
greater or less than 1. As we know, lim a x = and lim a x = .
xÑ8 0 aă1 xÑ´8 8 aă1
Our function has lim f ( x ) = 8 and lim f ( x ) = 0, so we conclude a ą 1: thus (d) and
xÑ8 xÑ´8
also (b) hold. (We could have also seen that (b) holds because a x is defined for all real
numbers.)
It remains to decide whether a is greater or less than e. (If a were equal to e, then f 1 ( x )
d
would be the same as f ( x ).) We saw in the text that dx ta x u = C ( a) a x for the function
ah ´ 1
C ( a) = lim . We know that C (e) = 1. (Actually, we chose e to be the number that
hÑ0 h
has this property.) From our graph, we see that f 1 ( x ) ă f ( x ), so C ( a) ă 1 = C (e). In
ah ´ 1 eh ´ 1
other words, lim ă lim ; so, a ă e. Thus (e) holds.
hÑ0 h hÑ0 h
d
S-3: The power rule tells us that dx tx n u = nx n´1 . In this equation, the variable is the
base, and the exponent is a constant. In the function e x , it’s reversed: the variable is the
exponent, and the base it a constant. So, the power rule does not apply.
S-4: P(t) is an increasing function over its domain, so the population is increasing.
280
There are a few ways to see that P(t) is increasing.
What we really care about is whether e0.2t is increasing or decreasing, since an increasing
function multiplied by 100 is still an increasing function, and a decreasing function
multiplied by 100 is still a decreasing function. Since f (t) = et is an increasing function,
we can use what we know about graphing functions to see that f (0.2t) = e0.2t is also
increasing.
2xe x ´ 2e x e x (2x ´ 2) ( x ´ 1) e x
f 1 (x) = = =
4x2 4x2 2x2
S-6:
d 2x d d
f 1 (x) = te u = t(e x )2 u = 2 te x ue x = 2e x e x = 2(e x )2 = 2e2x
dx dx dx
S-7:
e a+ x = e a e x
Since e a is just a constant,

d a x d
te e u = e a te x u = e a e x = e a+ x
dx dx
So, f 1 ( x ) = f ( x ) = e a+ x .
S-8: If the derivative is positive, the function is increasing, so let’s start by finding the
derivative. We use the product rule (although Question 12 gives a shortcut).
f 1 ( x ) = 1 ¨ e x + xe x = (1 + x )e x
Since e x is always positive, f 1 ( x ) ą 0 when 1 + x ą 0. So, f ( x ) is increasing when x ą ´1.
S-9:
1
e´x =
ex
Using the rule for differentiating the reciprocal:
d ´x é x ´1
te u = x 2 = x = é´x
dx (e ) e
281
S-10: Using the product rule,
f 1 ( x ) = (e x )(e x ´ 1) + (e x + 1)(e x ) = e x (e x ´ 1 + e x + 1) = 2(e x )2 = 2e2x
Alternate solution: using Question 6:
f ( x ) = e2x ´ 1 ùñ f 1 ( x ) = 2e2x .
S-11: The question asks when s1 (t) is negative. So, we start by differentiating. Using the
product rule:
s1 (t) = et (t2 + 2t)

= e t ¨ t ( t + 2)
et is always positive, so s1 (t) is negative when t and 2 + t have opposite signs. This
occurs when ´2 ă t ă 0.
S-12: Using the product rule, g1 ( x ) = f 1 ( x )e x + f ( x )e x = [ f ( x ) + f 1 ( x )]e x
S-13: We simplify the functions to get a better idea of what’s going on.
3
(a): y = e3 log x + 1 = elog x + 1 = x3 + 1. This is not a line.
(b): 2y + 5 = e3+log x = e3 elog x = e3 x. Since e3 is a constant, 2y + 5 = e3 x is a line.
(c): There isn’t a fancy simplification here–this isn’t a line. If that isn’t a satisfactory
answer, we can check: a line is a function( with a constant slope. For our function,
d d d
y = dx te + 4u = dx te u = dx (e ) = 2e e x = 2e2x . Since the derivative isn’t
1 2x 2x x 2 x
constant, the function isn’t a line.
(d): y = elog x 3e + log 2 = 3e x + log 2. Since 3e and log 2 are constants, this is a line.
S-14: When we say a function is differentiable without specifying a range, we mean that
it is differentiable over its domain. The function f ( x ) is differentiable when x ‰ 1 for any
values of a and b; it is up to us to figure out which constants make it differentiable when
x = 1.
In order to be differentiable, a function must be continuous. The definition of continuity
tells us that, for f to be continuous at x = 1, we need lim f ( x ) = f (1). From the
xÑ1
definition of f , we see f (1) = a + b = lim f ( x ), so we need lim f ( x ) = a + b. Since
xÑ1´ xÑ1+
1
lim f ( x ) = e = e, we specifically need
xÑ1+
e = a + b.
282
Now, let’s consider differentiability of f at x = 1. We need the following limit to exist:
f (1 + h ) ´ f (1)
lim
hÑ0 h
In particular, we need the one-sided limits to exist and be equal:
f (1 + h ) ´ f (1) f (1 + h ) ´ f (1)
lim = lim
hÑ0´ h hÑ0+ h
If h ă 0, then 1 + h ă 1, so f (1 + h) = a(1 + h)2 + b. If h ą 0, then 1 + h ą 1, so

f (1 + h) = e1+h . With this in mind, we begin to evaluate the one-sided limits:
f (1 + h ) ´ f (1) [ a (1 + h )2 + b ] ´ [ a + b ]
lim = lim
hÑ0´ h hÑ0´ h
2
ah + 2ah
= lim = 2a
hÑ0´ h
f (1 + h ) ´ f (1) e 1+ h ´ ( a + b )
lim = lim
hÑ0+ h hÑ0+ h
Since we take a + b to be equal to e (to ensure continuity):
e 1+ h ´ e 1
= lim
hÑ0+ h
ˇ
d x ˇ
= e1 = e
ˇ
= te uˇ
dx x =1
So, we also need

2a = e
e
Therefore, the values of a and b that make f differentiable are a = b = .
2
S-1:
y y = cos x
y = sin x
x
´π π
The graph f ( x ) = sin x has horizontal tangent lines precisely at those points where
d
cos x = 0. This must be true, since dx tsin xu = cos x: where the derivative of sine is zero,
cosine itself is zero.
283
S-2:
y y = cos x
y = sin x
x
´π π
The graph f ( x ) = sin x has maximum slope at those points where cos x has a maximum.
This makes sense, because f 1 ( x ) = cos x: the maximum values of the slope of sine
correspond to the maximum values of cosine.
S-3: You should memorize the derivatives of sine, cosine, and tangent.
f 1 ( x ) = cos x ´ sin x + sec2 x
S-4: f 1 ( x ) = cos x ´ sin x, so f 1 ( x ) = 0 precisely when sin x = cos x. This happens at π/4,
but it also happens at 5π/4. By looking at the unit circle, it is clear that sin x = cos x
whenever x = π4 + πn for some integer n.
y
sin(π/4) = cos(π/4)
sin(5π/4) = cos(5π/4)
S-5:
• Solution 1: f ( x ) = sin2 x + cos2 x = 1, so f 1 ( x ) = d
dx t1u = 0.
• Solution 2: Using the formula for the derivative of a squared function,
f 1 ( x ) = 2 sin x cos x + 2 cos x (´ sin x ) = 2 sin x cos x ´ 2 sin x cos x = 0.
S-6: It is true that 2 sin x cos x = sin(2x ), but we don’t know the derivative of sin(2x ). So,
we use the product rule:
f 1 ( x ) = 2 cos x cos x + 2 sin x (´ sin x ) = 2(cos2 x ´ sin2 x ).
284
S-7:
• Solution 1: using the product rule,
f 1 ( x ) = e x cot x + e x (´ csc2 x ) = e x (cot x ´ csc2 x ).
• Solution 2: using the formula from Question 12, Section 2.7,

f 1 ( x ) = e x (cot x ´ csc2 x ).
S-8: We use the quotient rule.

(cos x + tan x )(2 cos x + 3 sec2 x ) ´ (2 sin x + 3 tan x )(´ sin x + sec2 x )
f 1 (x) =
(cos x + tan x )2
2 cos2 x + 3 cos x sec2 x + 2 cos x tan x + 3 tan x sec2 x
=
(cos x + tan x )2
2 sin2 x ´ 2 sin x sec2 x + 3 sin x tan x ´ 3 tan x sec2 x
+
(cos x + tan x )2
2 + 3 sec x + 2 sin x ´ 2 tan x sec x + 3 sin x tan x
=
(cos x + tan x )2
S-9: We use the quotient rule.

e x (5 sec x tan x ) ´ (5 sec x + 1)e x
f 1 (x) =
( e x )2
5 sec x tan x ´ 5 sec x ´ 1
=
ex
S-10: We use the product rule:

f 1 ( x ) = (e x + cot x )(30x5 + csc x cot x ) + (e x ´ csc2 x )(5x6 ´ csc x )
S-11: We don’t know how to differentiate this function as it is written, but an identity
d
helps us. Since sin π2 ´ θ = cos θ, we see f 1 (θ ) = dθ tcos θu = ´ sin(θ ).
S-12: We know the derivative of sin x, but not of sin(´x ). So we re-write f ( x ) using
identities:
f ( x ) = sin(´x ) + cos(´x )
= ´ sin x + cos x
1
f ( x ) = ´ cos x ´ sin x
285
S-13: We apply the quotient rule.
(cos θ ´ sin θ )(´ sin θ + cos θ ) ´ (cos θ + sin θ )(´ sin θ ´ cos θ )
s1 ( θ ) =
(cos θ ´ sin θ )2
(cos θ ´ sin θ )2 + (cos θ + sin θ )2
=
(cos θ ´ sin θ )2
cos θ + sin θ 2

= 1+
cos θ ´ sin θ
S-14: In order for f to be differentiable at x = 0, it must also be continuous at x = 0. This

forces
lim f ( x ) = lim f ( x ) = f (0) or lim cos( x ) = lim ( ax + b) = 1

xÑ0´ xÑ0+ xÑ0´ xÑ0+
or b = 1. In order for f to be differentiable at x = 0, we need the limit
f (0 + h ) ´ f (0)
lim
hÑ0 h
to exist. This is the case if and only if the two one–sided limits
f (0 + h ) ´ f (0) cos(h) ´ cos(0)

lim = lim
hÑ0´ h hÑ0´ h
and
f (0 + h ) ´ f (0) ( ah + b) ´ cos(0)
lim = lim =a since b = 1
hÑ0+ h hÑ0+ h
exist and are equal. Because cos( x ) is differentiable at x = 0 we have

ˇ
cos(h) ´ cos(0) d ˇ ˇ
lim = cos( x )ˇˇ = ´ sin( x )ˇ =0
ˇ
hÑ0 ´ h dx x =0 x =0
So, we need a = 0 and b = 1.
S-15: We compute the derivative of cos( x ) + 2x as being ´ sin( x ) + 2, which evaluated at

x = π2 yields ´1 + 2 = 1. Since we also compute cos(π/2) + 2(π/2) = 0 + π, then the
equation of the tangent line is
y ´ π = 1 ¨ ( x ´ π/2).
286
f ( x ) = cos( x ). To see this, simply use the definition of the derivative at a = 2015 with
f ( x ) = cos x:
ˇ
d ˇ f ( x ) ´ f ( a)
dx xÑa xá
ˇ a
d ˇ cos( x ) ´ cos(2015)
tcos xuˇˇ = lim
dx 2015 xÑ2015 x ´ 2015
Since the derivative of f ( x ) is ´ sin( x ), its value at x = 2015 is exactly ´ sin(2015).
S-17: This limit represents the derivative computed at x = π/3 of the function
f ( x ) = cos x. To see this, simply use the definition of the derivative at a = π/3 with
f ( x ) = cos x:
ˇ
d ˇ f ( x ) ´ f ( a)
dx xÑa xá
ˇ a
d ˇ cos( x ) ´ cos(π/3)
tcos xuˇˇ = lim
dx π/3 xÑπ/3 x ´ π/3
cos( x ) ´ 1/2
= lim
xÑπ/3 x ´ π/3
? of f ( x ) is ´ sin x, then its value at x = π/3 is exactly
Since the derivative
´ sin(π/3) = ´ 3/2.
S-18: This limit represents the derivative computed at x = π of the function

f ( x ) = sin( x ). To see this, simply use the definition of the derivative at a = π with
f ( x ) = sin x:
ˇ
d ˇ f ( x ) ´ f ( a)
dx xÑa xá
ˇa
d ˇ sin( x ) ´ sin(π )
tsin xuˇˇ = lim
dx π
xÑπ x´π
sin( x )
= lim
xÑπ x ´ π
Since the derivative of f ( x ) is cos( x ), then its value at x = π is exactly

cos(π ) = ´1.
S-19:
sin θ
tan θ =
cos θ
So, using the quotient rule,
d cos θ cos θ ´ sin θ (´ sin θ ) cos2 θ + sin2 θ
ttan θu = =
dθ cos2 θ cos2 θ
2
1
= = sec2 θ
cos θ
287
S-20: In order for the function f ( x ) to be continuous at x = 0, the left half formula ax + b
6 cos x
and the right half formula must match up at x = 0. This forces
2 + sin x + cos x
6 cos 0 6
aˆ0+b = = ùñ b = 2
2 + sin 0 + cos 0 3
f ( h ) ´ f (0)
In order for the derivative f 1 ( x ) to exist at x = 0, the limit lim must exist. In
hÑ0 h
f ( h ) ´ f (0) f ( h ) ´ f (0)
particular, the limits lim and lim must exist and be equal to
hÑ0 ´ h hÑ0 + h
each other.
When h Ñ 0´ , this means h ă 0, so f (h) = ah + b = ah + 2. So:
ˇ
f ( h ) ´ f (0) ( ah + 2) ´ 2 d ˇ
lim = lim = tax + 2uˇˇ = a.
hÑ0´ h hÑ0´ h dx x =0
6 cos h
Similarly, when h Ñ 0+ , then h ą 0, so f (h) = and
1 + sin h + cos h
" *ˇ
f ( h ) ´ f (0) d 6 cos x ˇ
lim = ˇ
hÑ0+ h dx 2 + sin x + cos x ˇx=0
ˇ
´6 sin x (2 + sin x + cos x ) ´ 6 cos x (cos x ´ sin x ) ˇˇ
= .
(2 + sin x + cos x )2 ˇ
x =0
Since the limits from the left and right must be equal, this forces
´6 sin 0(2 + sin 0 + cos 0) ´ 6 cos 0(cos 0 ´ sin 0) ´6 2

a= 2
= 2
ùñ a = ´
(2 + sin 0 + cos 0) (2 + 1) 3
S-21: In order for f 1 ( x ) to exist, f ( x ) has to exist. We already know that tan x does not
exist whenever x = π2 + nπ for any integer n. If we look a little deeper, since
sin x
tan x = cos x , the points where tangent does not exist correspond exactly to the points
where cosine is zero.
From its graph, tangent looks like a smooth curve over its domain, so we might guess
that everywhere tangent is defined, its derivative is defined. We can check this:
2
f 1 ( x ) = sec2 x = cos1 x . Indeed, wherever cos x is nonzero, f 1 exists.
So, f 1 ( x ) exists for all values of x except when x = π

2 + nπ for some integer n.
288
S-22: The function is differentiable whenever x2 + x ´ 6 ‰ 0 since the derivative equals
10 cos( x ) ¨ ( x2 + x ´ 6) ´ 10 sin( x ) ¨ (2x + 1)
,
( x 2 + x ´ 6)2
which is well-defined unless x2 + x ´ 6 = 0. We solve x2 + x ´ 6 = ( x ´ 2)( x + 3) = 0,
and get x = 2 and x = ´3. So, the function is differentiable for all real values x except for
x = 2 and for x = ´3.
S-23: The function is differentiable whenever sin( x ) ‰ 0 since the derivative equals
sin( x ) ¨ (2x + 6) ´ cos( x ) ¨ ( x2 + 6x + 5)
,
(sin x )2
which is well-defined unless sin x = 0. This happens when x is an integer multiple of π.
So, the function is differentiable for all real values x except x = nπ,, where n is any
integer.
S-24: We compute the derivative of tan( x ) as being sec2 ( x ), which evaluated at x = π4

yields 2. Since we also compute tan(π/4) = 1, then the equation of the tangent line is
y ´ 1 = 2 ¨ ( x ´ π/4).
S-25: We compute the derivative y1 = cos( x ) ´ sin( x ) + e x , which evaluated at x = 0

yields 1 ´ 0 + 1 = 2. Since we also compute y(0) = 0 + 1 + 1 = 2, the equation of the
tangent line is
y ´ 2 = 2( x ´ 0)
ie y = 2x + 2.
S-26: We are asked to solve f 1 ( x ) = 0. That is, e x [sin x + cos x ] = 0. Since e x is always
positive, that means we need to find all points where sin x + cos x = 0. That is, we need
to find all values of x where sin x = ´ cos x. Looking at the unit circle, we see this
happens whenever x = 3π 4 + nπ for any integer n.
sin(3π/4) = ´ cos(3π/4)
sin(´π/4) = ´ cos(´π/4)
289
S-27: First, we note that our function is continuous, because
sin x
lim f ( x ) = lim = 1 = f (0)
xÑ0 xÑ0 x
This is a handy thing to check: if the function were discontinuous at x = 0, then we

would automatically know that it was not differentiable there.
Now, on to the derivative. We can use the limit definition:
f (0 + h ) ´ f (0)
f 1 (0) = lim if it exists
hÑ0 h
f (h) ´ 1
= lim
hÑ0 h
sin h
´1
= lim h
hÑ0 h
As h approaches 0, both the numerator and the denominator approach 0. So, to evaluate
the limit, we need to do more work. The key insight we can use is a result that was
shown in the text while evaluating the derivative of sine. When h is close to 0,
cos h ď sinh h ď 1. We use this to bound our limit, and then apply the squeeze theorem.
sin h
cos h ď ď1
h
sin h
So, cos h ´ 1 ď ´ 1 ď1 ´ 1 = 0
h

and 1h sinh h ´ 1 is between 0 and cos hh´1 . We can evaluate the limit of cos h´1
h , by noticing
its similarity to the definition of the derivative.
cos h ´ 1 cos(0 + h) ´ cos(0)
lim = lim
hÑ0 h hÑ0 h
d
= tcos xu|x=0 = 0
dx
sin h
h ´1
So, by the Squeeze Theorem, f 1 (0) = lim = 0.
hÑ0 h
S-28: As usual, when dealing with the absolute value function, we can make things a
little clearer by splitting it up into two pieces.
"
x xě0
|x| =
´x x ă 0
So,
" "
sin x x ě 0 sin x x ě 0
sin |x| = =
sin(´x ) x ă 0 ´ sin x x ă 0
290
where we used the identity sin(´x ) = ´ sin x. From here, it’s easy to see h1 ( x ) when x is
anything other than zero.
$
d & cos x x ą 0
tsin |x|u = ?? x = 0
dx
´ cos x x ă 0
%
To decide whether h( x ) is differentiable at x = 0, we use the definition of the derivative.

One word of explanation: usually in the definition of the derivative, h is the tiny “change
in x” that is going to zero. Since h is the name of our function, we need another letter to
stand for the tiny change in x, the size of which is tending to zero. We chose t.
h ( t + 0) ´ h (0) sin |t|
lim = lim
tÑ0 t tÑ0 t
We consider the behaviour of this function to the left and right of t = 0:
#
sin t " sin t
sin |t| t tě0 t tě0
= sin(´t) = sin t
t t tă0 ´ t tă0
sin t
Since we’re evaluating the limit as t goes to zero, we need the fact that lim = 1. We
tÑ0 t
saw this in Section 2.7, but also we know enough now to evaluate it another way. Using
the definition of the derivative:
ˇ
sin t sin(t + 0) ´ sin(0) d ˇ
lim = lim = tsin xuˇˇ = cos 0 = 1
tÑ0 t tÑ0 t dx t =0
sin t
At any rate, since we know lim = 1, then:
tÑ0 t
h ( t + 0) ´ h (0) sin t h ( t + 0) ´ h (0) ´ sin t

lim = lim =1 lim = lim = ´1
tÑ0+ t tÑ0+ t tÑ0´ t tÑ0´ t
So, since the one-sided limits disagree,
h ( t + 0) ´ h (0)
lim = DNE
tÑ0 t
so h( x ) is not differentiable at x = 0. Therefore,
"
1 cos x x ą 0
h (x) =
´ cos x x ă 0
S-29: Statement i is false, since f (0) = 0. Statement iv cannot hold, since a function that is
differentiable is also continuous.
sin x
Since lim = 1 (we saw this in Section 2.8),
xÑ0+ x
sin x
lim f ( x ) = lim ?
xÑ0+ xÑ0+ x
? sin x
= lim x
xÑ0+ x
= 0¨1 = 0
291
So f is continuous at x = 0, and so Statement ii does not hold. Now, let’s consider f 1 ( x ).
?x
sin
´0
f ( x ) ´ f (0) x
lim = lim
xÑ0+ x xÑ0+ x
1 sin x
= lim ? = +8
xÑ0+ x x
Therefore, using the definition of the derivative,

f ( x ) ´ f (0)
f 1 (0) = lim if it exists, but
xÑ0 x
f ( x ) ´ f (0)
lim = DNE
xÑ0 x
since one of the one-sided limits does not exist. So f is continuous but not differentiable
at x = 0. The correct statement is iii.
sin x
S-30: Recall that lim = 1. In order to take advantage of this knowledge, we divide
xÑ0 x
the numerator and denominator by x5 (because 5 is the power of sine in the denominator,
and a denominator that goes to zero generally makes a limit harder).
sin x27 99
sin x27 + 2x5 e x
99
5
+ 2e x
lim = lim x
sin5 x sin x 5

xÑ0 xÑ0
x
Now the denominator goes to 1, which is nice, but we need to take care of the fraction
sin x27
in the numerator. This fraction isn’t very familiar, but we know that, as x goes to
x5
sin x27
zero, x27 also goes to zero, so that goes to 1. Consequently,
x27
sin x27 99
sin x27 + 2x5 e x
99 x22 27
+ 2e x 0 ˆ 1 + 2 ˆ e0
lim = lim x = =2
sin5 x sin x 5 15

xÑ0 xÑ0
S-1: (a) More urchins means less kelp, and fewer urchins means more kelp. This means
dK
kelp and urchins are negatively correlated, so dU ă 0.
If you aren’t sure why that is, we give a more detailed explanation here, using the
definition of the derivative. When h is a positive number, U + h is greater than U, so
K (U + h) is less than U, hence K (U + h) ´ K (U ) ă 0. Therefore:
K (U + h ) ´ K (U ) negative
lim = ă 0.
hÑ0+ h positive
292
Similarly, when h is negative, U + h is less than U, so K (U + h) ´ K (U ) ą 0, and
K (U + h ) ´ K (U ) positive
lim = ă 0.
hÑ0´ h negative
Therefore:
dK K (U + h ) ´ K (U )
= lim ă 0.
dU hÑ0 h
(b) More otters means fewer urchins, and fewer otters means more urchins. So, otters
and urchins are negatively correlated: dUdO ă 0.
dK dK dU
(c) Using the chain rule, dO = dU ¨ dO . Parts (a) and (b) tell us both these derivatives are
dK
negative, so their product is positive: dO ą 0.
dK
We can also see that dO ą 0 by thinking about the relationships as described. When the
otter population increases, the urchin population decreases, so the kelp population
increases. That means when the otter population increases, the kelp population also
increases, so kelp and otters are positively correlated. The chain rule is a formal version
of this kind of reasoning.
S-2:
dA dA dB dC dD
= ¨ ¨ ¨ ă0
dE dB dC dD dE
since we multiply three positive quantities and one negative.
S-3: Applying the chain rule:
d d
tcos(5x + 3)u = ´ sin(5x + 3) ¨ t5x + 3u
dx dx
= ´ sin(5x + 3) ¨ 5
S-4: Using the chain rule,

" 5 *
1 d 2
f (x) = x +2
dx
4 d
= 5 x2 + 2 ¨ tx2 + 2u
dx
2 4
= 5( x + 2) ¨ 2x
= 10x ( x2 + 2)4
293
" 17 *
1 d 4 2
T (k) = 4k + 2k + 1
dk
d
= 17(4k4 + 2k2 + 1)16 ¨ t4k4 + 2k2 + 1u
dk
= 17(4k + 2k + 1) ¨ (16k3 + 4k)
4 2 16
S-6: Using the chain rule:

$d ,
2
" 2 *
d & x + 1. 1 d x +1
= b ¨
dx % x2 ´ 1 - 2 x2 +1 dx x2 ´ 1
x2 ´1
d " 2 *
1 x2 ´ 1 d x +1
= ¨
2 x2 + 1 dx x2 ´ 1
And now, the quotient rule:

d
x2 ´ 1 ( x2 ´ 1)(2x ) ´ ( x2 + 1)2x

1
= ¨
2 x2 + 1 ( x 2 ´ 1)2
d
1 x2 ´ 1

´4x
= ¨
2 x2 + 1 ( x 2 ´ 1)2
d
x2 ´ 1

´2x
= ¨
x2 + 1 ( x 2 ´ 1)2
´2x
= ?
( x 2 ´ 1) x 4 ´ 1
S-7: If we let g( x ) = e x and h( x ) = cos( x2 ), then f ( x ) = g(h( x )), so

f 1 ( x ) = g1 (h( x )) ¨ h1 ( x ).
2 d
f 1 ( x ) = ecos( x ) ¨ tcos( x2 )u
dx
d 2
In order to evaluate dx tcos( x )u, we’ll need the chain rule again.
2 d
= ecos(x ) ¨ [´ sin( x2 )] ¨ tx2 u
dx
2
= écos(x ) ¨ sin( x2 ) ¨ 2x
294
S-8: We use the chain rule, followed by the quotient rule:
" *
1 1 x d x
f (x) = g ¨
h( x ) dx h( x )
h( x ) ´ xh1 ( x )

x
= g1 ¨
h( x ) h ( x )2
When x = 2:
h(2) ´ 2h1 (2)

1 1 2
f (2) = g
h (2) h (2)2
2´2ˆ3
=4 = ´4
22
S-9: Using the chain rule, followed by the product rule:
d ! x cos( x) ) d
e = e x cos x tx cos xu
dx dx
= [cos x ´ x sin x ]e x cos(x)
S-10: Using the chain rule:
d ! x2 +cos( x) ) 2 d ! 2 )
e = e x +cos x x + cos x
dx dx
2
= [2x ´ sin x ]e x +cos(x)
S-11: Using the chain rule, followed by the quotient rule:

#c + " *
d x´1 1 d x´1
= c
dx x+2 x ´ 1 dx x + 2
2
x+2
?
x + 2 ( x + 2) ´ ( x ´ 1)
= ? ¨
2 x´1 ( x + 2)2
3
= ? ? 3
2 x´1 x+2
S-12: First, we manipulate our function to make it easier to differentiate:
f ( x ) = x´2 + ( x2 ´ 1)1/2
295
1
Now, we can use the power rule to differentiate 2 . This will be easier than
x
1
differentiating 2 using quotient rule, but if you prefer, quotient rule will also work.
x
1 d
f 1 ( x ) = ´2x´3 + ( x2 ´ 1)´1/2 ¨ tx2 ´ 1u
2 dx
1
= ´2x´3 + ( x2 ´ 1)´1/2 (2x )
2
´2 x
= 3 +?
x x2 ´ 1
The function f ( x ) is only defined when x ‰ 0 and when x2 ´ 1 ě 0. That is, when x is in
(´8, ´1] Y [1, 8). We have an added restriction on the domain of f 1 ( x ): x2 ´ 1 must not
be zero. So, the domain of f 1 ( x ) is (´8, ´1) Y (1, 8).
d
S-13: We use the quotient rule, noting that tsin 5xu = 5 cos 5x:
dx
(1 + x2 )(5 cos 5x ) ´ (sin 5x )(2x )
f 1 (x) =
(1 + x 2 )2
S-14: If we let g( x ) = sec x and h( x ) = e2x+7 , then f ( x ) = g(h( x )), so by the chain rule,
f 1 ( x ) = g1 (h( x )) ¨ h1 ( x ). Since g1 ( x ) = sec x tan x:
f 1 ( x ) = g1 (h( x )) ¨ h1 ( x )
= sec(h( x )) tan(h( x )) ¨ h1 ( x )
d ! 2x+7 )
= sec(e2x+7 ) tan(e2x+7 ) ¨ e
dx
Here, we need the chain rule again:

2x +7 2x +7 d
2x +7
= sec(e ) tan(e )¨ e ¨ t2x + 7u
dx
h i
2x +7 2x +7 2x +7
= sec(e ) tan(e )¨ e ¨2
= 2e2x+7 sec(e2x+7 ) tan(e2x+7 )
S-15: It is possible to start in on this problem with the product rule and then the chain
rule, but it’s easier if we simplify first. Since tan2 x + 1 = sec2 x = cos12 x , we see
cos2 x
f (x) = =1
cos2 x
for all values of x for which cos x is nonzero. That is, f ( x ) = 1 for every x that is not an
integer multiple of π/2 (and f ( x ) is not defined when x is an integer multiple of π/2).
296
Therefore, f 1 ( x ) = 0 for every x on which f exists, and in particular f 1 (π/4) = 0. Also,
f (π/4) = 1, so the tangent line to f at x = π/4 is the line with slope 0, passing through
the point (π/4, 1):
y=1
S-16: Velocity is the derivative of position with respect to time. So, the velocity of the
particle is given by s1 (t). We need to find s1 (t), and determine when it is zero.
To differentiate, we us the chain rule.
d 3
3 ´7t2 +8t
s1 ( t ) = e t tt ´ 7t2 + 8tu
¨
dt
t3 ´7t2 +8t
=e ¨ (3t2 ´ 14t + 8)
To determine where this function is zero, we factor:
3 ´7t2 +8t
= et ¨ (3t ´ 2)(t ´ 4)
3 2
So, the velocity is zero when et ´7t +8t = 0, when 3t ´ 2 = 0, and when t ´ 4 = 0. Since
3 2
et ´7t +8t is never zero, this tells us that the velocity is zero precisely when t = 32 or t = 4.
2
S-17: The slope of the tangent line is the derivative. If we let f ( x ) = tan x and g( x ) = e x ,
2
then f ( g( x )) = tan(e x ), so y1 = f 1 ( g( x )) ¨ g1 ( x ):
d x2 2
y1 = sec2 (e x ) ¨
te u
dx
We find ourselves once more in need of the chain rule:
2 2 d
= sec2 (e x ) ¨ e x tx2 u
dx
2 2
= sec2 (e x ) ¨ e x ¨ 2x
Finally, we evaluate this derivative at the point x = 1:
y1 (1) = sec2 (e) ¨ e ¨ 2
= 2e sec2 e
S-18: Using the Product rule,

d 4x
y1 = te u tan x + e4x sec2 x
dx
and the chain rule:
d
= e4x ¨ t4xu ¨ tan x + e4x sec2 x
dx
= 4e4x tan x + e4x sec2 x
297
d
1 (3x2 )(1 + e3x ) ´ ( x3 ) ¨ dx t1 + e3x u
f (x) =
(1 + e3x )2
Now, the chain rule:
(3x2 )(1 + e3x ) ´ ( x3 )(3e3x )

=
(1 + e3x )2
So, when x = 1:
3(1 + e3 ) ´ 3e3 3
f 1 (1) = =
(1 + e3 )2 (1 + e3 )2
S-20: This requires us to apply the chain rule twice.
d ! sin2 ( x) ) sin2 ( x ) d
!
2
)
e =e ¨ sin ( x )
dx dx
2 d
= esin (x) (2 sin( x )) ¨ sin( x )
dx
2
= esin (x) (2 sin( x )) ¨ cos( x )
d ! ) d ! )
sin(e5x ) = cos e5x ¨ e5x
dx dx
d
= cos(e5x )(e5x ) ¨ t5xu
dx
= cos(e5x )(e5x ) ¨ 5
S-22: We’ll use the chain rule twice.

d ! cos( x2 ) ) 2 d
e = ecos(x ) ¨ tcos( x2 )u
dx dx
2 d
= ecos(x ) ¨ (´ sin( x2 )) ¨ tx2 u
dx
2
= écos(x ) ¨ sin( x2 ) ¨ 2x
298
S-23: We start with the chain rule:
a d ! 2 a )
y1 = ´ sin x2 + x2 + 1 ¨ x + x2 + 1
dx

a d !a 2 )
2

2
= ´ sin x + x + 1 ¨ 2x + x +1
dx
and find ourselves in need of chain rule a second time:
)
2
a
2
1 d ! 2
= ´ sin x + x + 1 ¨ 2x + ? 2 ¨ x +1
2 x + 1 dx

2
a
2
2x
= ´ sin x + x + 1 ¨ 2x + ?
2 x2 + 1
S-24:
y = (1 + x2 ) cos2 x
Using the product rule,

d
y1 = (2x ) cos2 x + (1 + x2 ) tcos2 xu
dx
Here, we’ll need to use the chain rule. Remember cos2 x = [cos x ]2 .
d
= 2x cos2 x + (1 + x2 )2cos x ¨ tcos xu
dx
= 2x cos2 x + (1 + x2 )2 cos x ¨ (´ sin x )
= 2x cos2 x ´ 2(1 + x2 ) sin x cos x
d 3x
S-25: We use the quotient rule, noting by the chain rule that dx te u = 3e3x :
(1 + x2 ) ¨ 3e3x ´ e3x (2x )

y1 =
(1 + x 2 )2
e3x (3x2 ´ 2x + 3)
=
(1 + x 2 )2
S-26: By the chain rule,

d ! 2 )
h x = h1 ( x2 ) ¨ 2x
dx
Using the product rules and the result above,
g1 ( x ) = 3x2 h( x2 ) + x3 h1 ( x2 )2x
299
Plugging in x = 2:
g1 (2) = 3(22 )h(22 ) + 23 h1 (22 )2 ˆ 2

= 12h(4) + 32h1 (4) = 12 ˆ 2 ´ 32 ˆ 2
= ´40
2 ´1) /2 2 ) /2
S-27: Let f ( x ) = xe´( x = xe(1´x . Then, using the product rule,
1 (1´x2 )/2 d ! (1´x2 )/2 )

f (x) = e +x¨ e
dx
Here, we need the chain rule:
" *
(1´x2 )/2 (1´x2 )/2 d 1
=e +xë (1 ´ x 2 )
dx 2
2 ) /2 2 ) /2
= e(1´x + x ¨ e(1´x ¨ (´x )
(1´x2 )/2
= (1 ´ x 2 ) e
There is no power of e that is equal to zero; so if the product above is zero, it must be that
1 ´ x2 = 0. This happens for x = ˘1. On the curve, when x = 1, y = 1, and when
x = ´1, y = ´1. So the points are (1, 1) and (´1, ´1).
S-28: The question asks when s1 (t) is negative. So, we start by differentiating. Using the
chain rule:
" *
1 1 d 1
s (t) = cos ¨
t dt t

1 ´1
= cos ¨ 2
t t
1
When t ě 1, isbetween
t 0 and 1. Since cos θ is positive for 0 ď θ ă π/2, and π/2 ą 1,
we see that cos 1t is positive for the entire domain of s(t). Also, ´1t2
is negative for the
entire domain of the function. We conclude that s1 (t) is negative for the entire domain of
s(t), so the particle is always moving in the negative direction.
S-29: We present two solutions: one where we dive right in and use the quotient rule,
and another where we simplify first and use the product rule.
• Solution 1: We begin with the quotient rule:
d d
cos3 (5x ´ 7) dx te x u ´ e x dx tcos3 (5x ´ 7)u
f 1 (x) =
cos6 (5x ´ 7)
d
cos3 (5x ´ 7)e x ´ e x dx tcos3 (5x ´ 7)u
=
cos6 (5x ´ 7)
300
Now, we use the chain rule. Since cos3 (5x ´ 7) = [cos(5x ´ 7)]3 , our “outside”
function is g( x ) = x3 , and our “inside” function is h( x ) = cos(5x ´ 1).
d
cos3 (5x ´ 7)e x ´ e x ¨ 3cos2 (5x ´ 7) ¨ dx tcos(5x ´ 7)u
= 6
cos (5x ´ 7)
We need the chain rule again!
d
cos3 (5x ´ 7)e x ´ e x ¨ 3cos2 (5x ´ 7) ¨ [´ sin(5x ´ 7) ¨ dx t5x ´ 7u]
= 6
cos (5x ´ 7)
3 x x 2
cos (5x ´ 7)e ´ e ¨ 3cos (5x ´ 7) ¨ [´ sin(5x ´ 7) ¨ 5]
=
cos6 (5x ´ 7)
We finish by simplifying:
e x cos2 (5x ´ 7) (cos(5x ´ 7) + 15 sin(5x ´ 7))

=
cos6 (5x ´ 7)
cos(5x ´ 7) + 15 sin(5x ´ 7)
= ex
cos4 (5x ´ 7)
= e x (sec3 (5x ´ 7) + 15 tan(5x ´ 7) sec3 (5x ´ 7))
= e x sec3 (5x ´ 7)(1 + 15 tan(5x ´ 7))
• Solution 2: We simplify to avoid the quotient rule:

ex
f (x) =
cos3 (5x ´ 7)
= e x sec3 (5x ´ 7)
Now we use the product rule to differentiate:
d
f 1 ( x ) = e x sec3 (5x ´ 7) + e x tsec3 (5x ´ 7)u
dx
Here, we’ll need the chain rule. Since sec3 (5x ´ 7) = [sec(5x ´ 7)]3 , our “outside”
function is g( x ) = x3 and our “inside” function is h( x ) = sec(5x ´ 7), so that
g(h( x )) = [sec(5x ´ 7)]3 = sec3 (5x ´ 7).
d
= e x sec3 (5x ´ 7) + e x ¨ 3 sec2 (5x ´ 7) ¨ tsec(5x ´ 7)u
dx
d
We need the chain rule again! Recall dx tsec xu = sec x tan x.
d
= e x sec3 (5x ´ 7) + e x ¨ 3 sec2 (5x ´ 7) ¨ sec(5x ´ 7) tan(5x ´ 7) ¨ t5x ´ 7u
dx
= e x sec3 (5x ´ 7) + e x ¨ 3 sec2 (5x ´ 7) ¨ sec(5x ´ 7) tan(5x ´ 7) ¨ 5
We finish by simplifying:
= e x sec3 (5x ´ 7)(1 + 15 tan(5x ´ 7))
301
S-30:
• Solution 1: In Example 2.6.6, we generalized the product rule to three factors:
d
t f ( x ) g ( x ) h ( x )u = f 1 ( x ) g ( x ) h ( x ) + f ( x ) g1 ( x ) h ( x ) + f ( x ) g ( x ) h1 ( x )
dx
Using this rule:
d !
2x
) d 2x d ! 2x ) d
(x) e (cos 4x ) = txu ¨ e cos 4x + x ¨ e ¨ cos 4x + xe2x ¨ tcos 4xu
dx dx dx dx
= e2x cos 4x + x 2e2x cos 4x + xe2x (´4 sin 4x )
= e2x cos 4x + 2xe2x cos 4x ´ 4xe2x sin 4x
• Solution 2: We can use the product rule twice. In the first step, we split the function
xe2x cos 4x into the product of two functions.
d ! 2x ) d ! 2x ) d
xe (cos 4x ) = xe ¨ cos 4x + xe2x ¨ tcos 4xu
dx dx
dx

d 2x d ! 2x ) d
= txu ¨ e + x ¨ e ¨ cos 4x + xe2x ¨ tcos 4xu
dx dx dx

= e2x + x 2e2x ¨ cos 4x + xe2x (´4 sin 4x )
= e2x cos 4x + 2xe2x cos 4x ´ 4xe2x sin 4x

S-31: At time t, the particle is at the point x (t), y(t) , with x (t) = cos t and y(t) = sin t.
Over time, the particle traces out a curve; let’s call that curve y = f ( x ). Then

y(t) = f x (t) , so the slope of the curve at the point x (t), y(t) is f 1 x (t) . You are to

determine the values of t for which f 1 x (t) = ´1.
By the chain rule
y1 ( t ) = f 1 x ( t ) ¨ x 1 ( t )

Substituting in x (t) = cos t and y(t) = sin t gives
cos t = f 1 x (t) ¨ ´ sin t

so that
cos t
f 1 x (t) = ´

sin t
π
is ´1 precisely when sin t = cos t. This happens whenever t = 4.
Remark: the path traced by the particle is a semicircle. You can think about the point on
the unit circle with angle t, or you can notice that x2 + y2 = sin2 t + cos2 t = 1.
302
2
S-32: Let f ( x ) = e x+ x and g( x ) = 1 + x. Then f (0) = g(0) = 1.
2
f 1 ( x ) = (1 + 2x )e x+ x and g1 ( x ) = 1. When x ą 0,
2 2 2 2
f 1 ( x ) = (1 + 2x )e x+ x ą 1 ¨ e x+ x = e x+ x ą e0+0 = 1 = g1 ( x ).
Since f (0) = g(0), and f 1 ( x ) ą g1 ( x ) for all x ą 0, that means f and g start at the same
place, but f always grows faster. Therefore, f ( x ) ą g( x ) for all x ą 0.
S-33: Since sin 2x and 2 sin x cos x are the same function, they have the same derivative.
sin 2x = 2 sin x cos x

d d
ñ tsin 2xu = t2 sin x cos xu
dx dx
2 cos 2x = 2[cos2 x ´ sin2 x ]
cos 2x = cos2 x ´ sin2 x
We conclude cos 2x = cos2 x ´ sin2 x, which is another common trig identity.

Remark: if we differentiate both sides of this equation, we get the original identity back.
S-34:
d
2
3 ecsc x
f (x) = ?
x3 ´ 9 tan x
2
!1
3
ecsc x
= ?
x3 ´ 9 tan x
1
To begin the differentiation, we can choose our “outside” function to be g( x ) = x 3 , and
2
ecsc x
our “inside” function to be h( x ) = ? . Then f ( x ) = g(h( x )), so
x3 ´ 9 tan x
2
f 1 ( x ) = g1 (h( x )) ¨ h1 ( x ) = 31 (h( x ))´ 3 h1 ( x ):
2
! ´2 # 2
+
3
1 ecsc x d ecsc x
f 1 (x) = ? ¨ ?
3 x3 ´ 9 tan x dx x3 ´ 9 tan x
? !2 # 2
+
3
1 x3 ´ 9 tan x d ecsc x
= 2 ¨ ?
3 ecsc x dx x3 ´ 9 tan x
This leads us to use the quotient rule:
? !2 ? 3 d
! )
csc x2 ´ ecsc x2 d
!?
3 ´ 9 tan x
)
1 x3´ 9 tan x
3 x ´ 9 tan x dx e dx x
=  
3 ecsc x
2
(tan2 x )( x3 ´ 9)
303
Let’s figure out those two derivatives on their own, then plug them in. Using the chain
rule twice:
d ! csc x2 ) csc x2 d
! ) 2 d
e =e csc x = ecsc x ¨ (´ csc( x2 ) cot( x2 )) ¨
2
tx2 u
dx dx dx
2
= ´2xecsc x csc( x2 ) cot( x2 )
For the other derivative, we start with the product rule, then chain:
d !a 3 ) d !a 3 ) a
x ´ 9 tan x = x ´ 9 ¨ tan x + x3 ´ 9 sec2 x
dx dx
1 d ! 3 ) a
= ? x ´ 9 ¨ tan x + x3 ´ 9 sec2 x
2 x3 ´ 9 dx
3x2 tan x a 3
= ? + x ´ 9 sec2 x
2 x3 ´ 9
Now, we plug these into our equation for f 1 ( x ):
? !2 ? 3 d
! )
csc x2 ´ ecsc x2 d
!?
3 ´ 9 tan x
)
1 x3´ 9 tan x
3 x ´ 9 tan x dx e dx x
f 1 (x) = 2

2

3 ecsc x (tan x )( x3 ´ 9)
? !2
3
1 x3 ´ 9 tan x
= 2 ¨
3 ecsc x
? 
2 2 3x 2 tan x ?
3 csc x 2 2 csc x 2
+ x3 ´ 9 sec x 
 x ´ 9 tan x (´2x )e csc( x ) cot( x ) ´ e ?
2 x3 ´9
 
 (tan2 x )( x3 ´ 9) 
S-35:
(a) The table below gives us a number of points on our graph, and the times they occur.
t (sin t, cos2 t)
0 (0, 1)
π/4 ( ?12 , 12 )
π/2 (1, 0)
3π/4 ( ?12 , 12 )
π (0, 1)
5π/4 (´ ?12 , 12 )
3π/2 (´1, 0)
7π/4 (´ ?12 , 12 )
2π (0, 1)
These points will repeat with a period of 2π. With this information, we have a pretty
good idea of the particle’s motion:
304
y
(0,1)
t=0,π,2π

´ ?1 , 12 ?1 , 1
2 2 2
t=5π/4,7π/4 t=π/4,3π/4
x
(´1,0) (1,0)
t=3π/2 t=π/2
The particle traces out an arc, pointing down. It starts at t = 0 at the top part of the graph
at (1, 0), then is moves to the right until it hits (1, 0) at time t = π/2. From there it
reverses direction and moves along the curve to the left, hitting the top at time t = π and
reaching (´1, 0) at time t = 3π/2. Then it returns to the top at t = 2π and starts again.
So, it starts at the top of the curve, then moves back for forth along the length of the
curve. If goes right first, and repeats its cycle every 2π units of time.
(b) Let y = f ( x ) be the curve the particle traces in the xy-plane.

Since x is a function of t,
df 10π df
y(t) = f ( x (t)). What we want to find is when t = . Since is a function of
dx 3 ?dx

10π 10π 4π 3
x, we note that when t = , x = sin = sin = ´ . So, the quantity
3 3 3 2
we want to find (the slope of the tangent line to the curve y = f ( x ) traced by the particle
?
10π df 3
at the time t = is given by ´ .
3 dx 2
Using the chain rule:
y(t) = f ( x (t))
dy d d f dx
= t f ( x (t))u = ¨
dt dt dx dt
df dy dx
so, = ˜
dx dt dt
Using y(t) = cos2 t and x (t) = sin t:
df
= (´2 cos t sin t) ˜ (cos t) = ´2 sin t = ´2x
dx
?
10π 3
So, when t = and x = ´ ,
3 2
? ?
df ´ 3 ´ 3 ?
= ´2 ¨ = 3.
dx 2 2
305
Remark: The standard way to write this problem is to omit the notation f ( x ), and let the
variable y stand for two functions. When t is the variable, y(t) = cos2 t gives the
y-coordinate of the particle at time t. When x is the variable, y( x ) gives the y-coordinate
of the particle given its position along the x-axis. This is an abuse of notation, because if
we write y(1), it is not clear whether we are referring to the y-coordinate of the particle
when t = 1 (in this case, y = cos2 (1) « 0.3), or the y-coordinate of the particle when x = 1
(in this case, looking at our table of values, y = 0). Although this notation is not strictly
“correct,” it is very commonly used. So, you might see a solution that looks like this:
dy dy
The slope of the curve is . To find , we use the chain rule:
dx dx
dy dy dx
= ¨
dt dx dt
d ! 2 ) dy d
cos t = ¨ tsin tu
dt dx dt
dy
´2 cos t sin t = ¨ cos t
dx
dy
= ´2 sin t
dx
10π
So, when t = ,
3
?
?

dy 10π 3
= ´2 sin = ´2 ´ = 3.
dx 3 2
In this case, it is up to the reader to understand when y is used as a function of t, and

when it is used as a function of x. This notation (using y to be two functions, y(t) and
y( x )) is actually the accepted standard, so you should be able to understand it.
S-1: We are given that one speaker produces 3dB. So if P is the power of one speaker,

P
3 = V ( P) = 10 log10 .
S
So, for ten speakers:

10P P
V (10P) = 10 log10 = 10 log10 + 10 log10 (10)
S S
= 3 + 10(1) = 13dB
and for one hundred speakers:

100P P
V (100P) = 10 log10 = 10 log10 + 10 log10 (100)
S S
= 3 + 10(2) = 23dB
306
S-2: The investment doubles when it hits $2000. So, we find the value of t that gives
A(t) = 2000:
2000 = A(t)
2000 = 1000et/20
2 = et/20
t
log 2 =
20
20 log 2 = t
S-3: From our logarithm rules, we know that when y is positive, log(y2 ) = 2 log y.
However, the expression cos x does not always take on positive values, so (a) is not
correct. (For instance, when x = π,
log(cos2 x ) = log(cos2 π ) = log (´1)2 = log(1) = 0, while 2 log(cos π ) = 2 log(´1),

which does not exist.)

Because cos2 x is never negative, we notice that cos2 x = | cos x|2 . When cos x is nonzero,
| cos x| is positive, so our logarithm rules tell us log | cos x|2 = 2 log | cos x|. When cos x
is exactly zero, then both log(cos2 x ) and 2 log | cos x| do not exist. So,
log(cos2 x ) = 2 log | cos x|.
S-4:
d 1 1
• Solution 1: Using the chain rule, tlog(10x )u = ¨ 10 = .
dx 10x x
d d 1 1
• Solution 2: Simplifying, tlog(10x )u = tlog(10) + log xu = 0 + = .
dx dx x x
S-5:
d ! 2
) 1 2
• Solution 1: Using the chain rule, log( x ) = 2 ¨ 2x = .
dx x x
d ! ) d 2
• Solution 2: Simplifying, log( x2 ) = t2 log( x )u = .
dx dx x
S-6: Don’t be fooled by a common mistake: log( x2 + x ) is not the same as log( x2 ) + log x.
d ! ) 1 2x + 1
We differentiate using the chain rule: log( x2 + x ) = 2 ¨ (2x + 1) = 2 .
dx x +x x +x
S-7: We know the derivative of the natural logarithm (base e), so we use the base-change
formula:
log x
f ( x ) = log10 x =
log 10
307
Since log 10 is a constant:
1
f 1 (x) = .
x log 10
S-8:
• Solution 1: Using the quotient rule,
x3 1x ´ (log x ) ¨ 3x2 x2 ´ 3x2 log x 1 ´ 3 log x

y1 = 6
= 6
= .
x x x4
• Solution 2: Using the product rule with y = log x ¨ x´3 ,
1 ´3
y1 = x + log x ¨ (´3) x´4 = x´4 (1 ´ 3 log x )
x
d 1
log(sec θ ) = ¨ (sec θ ¨ tan θ )
dθ sec θ
= tan θ
function log(sec θ ) is
Remark: the domain of the those
values of θ for which sec θ is
positive: so, the intervals 2n ´ 12 π, 2n + 1
2 π where n is any integer. Certainly the
tangent function has a larger domain than this, but outside the domain of log(sec θ ), tan θ
is not the derivative of log(sec θ ).
S-10: Let’s start in with the chain rule.

d
f 1 ( x ) = ecos(log x) ¨ tcos (log x )u
dx
We’ll need the chain rule again:
d
= ecos(log x) (´ sin(log x )) ¨ tlog xu
dx
1
= ecos(log x) (´ sin(log x )) ¨
x
é cos ( log x ) sin(log x )
=
x
Remark: Although we have a logarithm in the exponent, we can’t cancel. The expression
ecos(log x) is not the same as the expression xcos x , or cos x.
308
S-11:
a
y = log( x2 + x 4 + 1)
So, we’ll need the chain rule:
d
! ? )
dx x2 + x4 + 1
y1 = ?
x2 + ! x4 + 1 )
d
?
2x + dx x4 + 1
= ?
x2 + x4 + 1
We need the chain rule again:
d
tx4 +1u
2x + dx? 4
2 x +1
= ?
x2 + x4 + 1
3
2x + ?4x4
2 x +1
= ? .
x + x4 + 1
2

"b *
d 1 d
´ log(cos x ) = a ¨ t´ log (cos x )u
dx 2 ´ log(cos x ) dx
1 1 d
=´ a ¨ tcos xu
2 ´ log(cos x ) cos x dx
1 1
=´ a ¨ ¨ (´ sin x )
2 ´ log(cos x ) cos x
tan x
= a
2 ´ log(cos x )
Remark: it looks strange to see a negative sign in the argument of a square root. Since the
a values that are at most 1, log(cos x ) is always negative or
cosine function always gives
zero over its domain. So, log(cos x ) is only defined for the points where cos x = 1 (and
so log(cos x ) = 0–this isn’t a very interesting function! In contrast, ´ log(cos x ) is always
positive or zero over its domain – and therefore we can always take its square root.
d
S-13: Under the chain rule, dx log f ( x ) = f (1x) f 1 ( x ). So
d ! a ) 1 d ! a )
log x + x2 + 4 = ? ¨ x + x2 + 4
dx x + x2 + 4 dx

1 2x
= ? ¨ 1+ ?
x + x2 + 4 2 x2 + 4
? !
1 2 x2 + 4 + 2x
= ? ¨ ?
x + x2 + 4 2 x2 + 4
1
=? 2
x +4
309
?
d x2 4
dx te +? 1 + x u
g1 ( x ) = 2
e x + 1 + x4
2 3
2xe x + ?4x 4 ? !
2 1+ x 1 + x4
= 2 ? ?
e x + 1 + x4 1 + x4
2?
2xe x 1 + x4 + 2x3
= 2?
e x 1 + x4 + 1 + x4
S-15: Using logarithm rules makes this an easier problem:

g( x ) = log(2x ´ 1) ´ log(2x + 1)
2 2
So, g1 ( x ) = ´
2x ´ 1 2x + 1
2 2 4
and g1 (1) = ´ =
1 3 3
S-16: We begin by simplifying:

d 
( x2 + 5)3
f ( x ) = log  
x4 + 10
1/2 !
( x 2 + 5)3

= log
x4 + 10
2
( x + 5)3

1
= log
2 x4 + 10
1 h i
= log ( x2 + 5)3 ´ log( x4 + 10)
2
1h i
= 3 log ( x2 + 5) ´ log( x4 + 10)
2
Now, we differentiate using the chain rule:
4x3

1 1 2x
f (x) = 3 ´
2 x2 + 5 x4 + 10
3x 2x3
= 2 ´ 4
x + 5 x + 10
Remark: it is a common mistake to write log( x2 + 4) as log( x2 ) + log(4). These
expressions are not equivalent!
310
S-17: We use the chain rule twice, followed by the product rule:
1 d
f 1 (x) = ¨ tg( xh( x ))u
g( xh( x )) dx
1 d
= ¨ g1 ( xh( x )) ¨ txh( x )u
g( xh( x )) dx
1
¨ g1 xh( x ) ¨ h( x ) + xh1 ( x )

=
g xh( x )
In particular, when x = 2:
1
f 1 (2) = ¨ g1 2h(2) ¨ h(2) + 2h1 (2)

g 2h(2)
g1 (4) 5
= 2+2ˆ3 = 2+2ˆ3
g (4) 3
40
=
3
d
S-18: In the text, we saw that ta x u = a x log a for any constant a. So,
dx
d
tπ x u = π x log π.
dx
d
By the power rule, tx π u = πx π´1 .
dx
Therefore, g1 ( x ) = π x log π + πx π´1 .
Remark: we had to use two different rules for the two different terms in g( x ). Although
the functions π x and x π look superficially the same, they behave differently, as do their
derivatives. A function of the form (constant) x is an exponential function and not eligible
for the power rule, while a function of the form xconstant is exactly the class of function
the power rule applies to.
S-19: We have the power rule to tell us the derivative of functions of the form x n , where n
is a constant. However, here our exponent is not a constant. Similarly, in this section we
learned the derivative of functions of the form a x , where a is a constant, but again, our
d x
base is not a constant! Although the result a = a x log a is not what we need, the
dx
method used to differentiate a x will tell us the derivative of x x .
We’ll set g( x ) = log( x x ), because now we can use logarithm rules to simplify:
g( x ) = log( f ( x )) = x log x
Now, we can use the product rule to differentiate the right side, and the chain rule to
differentiate log( f ( x )):
f 1 (x) 1
g1 ( x ) = = log x + x = log x + 1
f (x) x
311
Finally, we solve for f 1 ( x ):
f 1 ( x ) = f ( x )(log x + 1) = x x (log x + 1)
d
S-20: In Question 19, we saw tx x u = x x (log x + 1). Using the base-change formula,
dx
log x
log10 ( x ) = . Since log10 is a constant,
log 10
" *
1 d x log x
f (x) = x +
dx log 10
1
= x x (log x + 1) +
x log 10
S-21: Rather than set in with a terrible chain rule problem, we’ll use logarithmic
differentiation. Instead of differentiating f ( x ), we differentiate a new function log( f ( x )),
after simplifying.
d
( x4 + 12)( x4 ´ x2 + 2)
4
log( f ( x )) = log
x3
4
( x + 12)( x4 ´ x2 + 2)

1
= log
4 x3
1
= log( x4 + 12) + log( x4 ´ x2 + 2) ´ 3 log x
4
Now that we’ve simplified, we can efficiently differentiate both sides. It is important to
remember that we aren’t differentiating f ( x ) directly–we’re differentiating log( f ( x )).
f 1 (x) 4x3 4x3 ´ 2x

1 3
= + ´
f (x) 4 x4 + 12 x4 ´ x2 + 2 x
Our final step is to solve for f 1 ( x ):
4x3 4x3 ´ 2x

1 1 3
f (x) = f (x) + ´
4 x4 + 12 x4 ´ x2 + 2 x
d 
4 4 2
1  4 ( x + 12)( x ´ x + 2) 

4x3 4x3 ´ 2x 3

= + ´
4 x3 x4 + 12 x4 ´ x2 + 2 x
It was possible to differentiate this function without logarithms, but the logarithms make
it more efficient.
312
S-22: It’s possible to do this using the product rule a number of times, but it’s easier to
use logarithmic differentiation. Set
h i
g( x ) = log( f ( x )) = log ( x + 1)( x2 + 1)2 ( x3 + 1)3 ( x4 + 1)4 ( x5 + 1)5
Now we can use logarithm rules to change g( x ) into a form that is friendlier to
differentiate:
= log( x + 1) + log( x2 + 1)2 + log( x3 + 1)3 + log( x4 + 1)4 + log( x5 + 1)5

= log( x + 1) + 2 log( x2 + 1) + 3 log( x3 + 1) + 4 log( x4 + 1) + 5 log( x5 + 1)
Now, we differentiate g( x ) using the chain rule:
f 1 (x) 1 4x 9x2 16x3 25x4

g1 ( x ) = = + 2 + 3 + 4 + 5
f (x) x+1 x +1 x +1 x +1 x +1
9x2 16x3 25x4

1 1 4x
f (x) = f (x) + + + +
x + 1 x2 + 1 x3 + 1 x4 + 1 x5 + 1
= ( x + 1)( x2 + 1)2 ( x3 + 1)3 ( x4 + 1)4 ( x5 + 1)5
9x2 16x3 25x4

1 4x
¨ + + + +
x + 1 x2 + 1 x3 + 1 x4 + 1 x5 + 1
S-23: We could do this with quotient and product rules, but it would be pretty painful.
Insteady, let’s use a logarithm.
2 2
5x + 10x + 15 1 x + 2x + 3 1
f (x) = =
3x4 + 4x3 + 5 10( x + 1) 3x4 + 4x3 + 5 2( x + 1)
2
x + 2x + 3 1
log( f ( x )) = log
3x4 + 4x3 + 5 2( x + 1)
2
x + 2x + 3 1
= log + log
3x4 + 4x3 + 5 2( x + 1)

= log x2 + 2x + 3 ´ log 3x4 + 4x3 + 5 ´ log( x + 1) ´ log(2)
Now we have a function that we can differentiate more cleanly than our original
function.
d d ! )
tlog( f ( x ))u = log x2 + 2x + 3 ´ log 3x4 + 4x3 + 5 ´ log ( x + 1) ´ log (2)
dx dx
1
f (x) 2x + 2 12x3 + 12x2 1
= 2 ´ 4 3
´
f (x) x + 2x + 3 3x + 4x + 5 x + 1
2( x + 1) 12x2 ( x + 1) 1
= 2 ´ 4 3
´
x + 2x + 3 3x + 4x + 5 x + 1
313
Finally, we solve for f ( x ):
12x2 ( x + 1)

1 2( x + 1) 1
f (x) = f (x) ´ ´
x2 + 2x + 3 3x4 + 4x3 + 5 x + 1
2
12x2 ( x + 1)

x + 2x + 3 1 2( x + 1) 1
= ´ ´
3x4 + 4x3 + 5 2( x + 1) x2 + 2x + 3 3x4 + 4x3 + 5 x + 1
2
6x2

x + 2x + 3 1 1
= ´ ´
3x4 + 4x3 + 5 x2 + 2x + 3 3x4 + 4x3 + 5 2( x + 1)2
S-24: Since f ( x ) has the form of a function raised to a functional power, we will use
logarithmic differentiation.

log( f ( x )) = log (cos x )sin x = sin x ¨ log(cos x )
Logarithm rules allowed us to simplify. Now, we differentiate both sides of this equation:
f 1 (x) ´ sin x
= (cos x ) log(cos x ) + sin x ¨
f (x) cos x
= (cos x ) log(cos x ) ´ sin x tan x
f 1 ( x ) = f ( x ) [(cos x ) log(cos x ) ´ sin x tan x ]

= (cos x )sin x [(cos x ) log(cos x ) ´ sin x tan x ]
Remark: negative numbers behave in a complicated manner when they are the base of an
exponential expression. For example, the expression (´1) x is defined when x is the
reciprocal of an odd number (like x = 51 or x = 17 ), but not when x is the reciprocal of an
even number (like x = 12 ). Since the domain of f ( x ) was restricted to (0, π2 ), cos x is
always positive, and we avoid these complications.
S-25: Since f ( x ) has the form of a function raised to a functional power, we will use
logarithmic differentiation. We take the logarithm of the function, and make use of
logarithm rules:
log ((tan x ) x ) = x log(tan x )
Now, we can differentiate:

d
dx t(tan x ) x u sec2 x
= log(tan x ) + x ¨
(tan x ) x tan x
x
= log(tan x ) +
sin x cos x
314
d
Finally, we solve for the derivative we want, t(tan x ) x u:
dx
d x
t(tan x ) x u = (tan x ) x log(tan x ) +
dx sin x cos x
Remark: the restricted domain (0, π/2) ensures that tan x is a positive number, so we
avoid the problems that arise by raising a negative number to a variety of powers.
S-26: We use logarithmic differentiation.
log f ( x ) = log( x2 + 1) ¨ ( x2 + 1)
We differentiate both sides to obtain:

f 1 (x) d ! )
= log( x2 + 1) ¨ ( x2 + 1)
f (x) dx
2x
= 2 ( x2 + 1) + 2x log( x2 + 1)
x +1
= 2x (1 + log( x2 + 1))
Now, we solve for f 1 ( x ):
f 1 ( x ) = f ( x ) ¨ 2x (1 + log( x2 + 1))
2 +1
= ( x 2 + 1) x ¨ 2x (1 + log( x2 + 1))
S-27: We use logarithmic differentiation: we modify our function to consider
log f ( x ) = log( x2 + 1) ¨ sin x
We differentiate using the product and chain rules:
f 1 (x) d ! ) 2x sin x
= log( x2 + 1) ¨ sin x = cos x ¨ log( x2 + 1) + 2
f (x) dx x +1
Finally, we solve for f 1 ( x )

1 2 2x sin x
f ( x ) = f ( x ) ¨ cos x ¨ log( x + 1) + 2
x +1

2 sin( x ) 2 2x sin x
= ( x + 1) ¨ cos x ¨ log( x + 1) + 2
x +1
S-28: We use logarithmic differentiation; so we modify our function to consider
log f ( x ) = log( x ) ¨ cos3 ( x )
315
Differentiating, we find:
f 1 (x) d ! ) cos3 ( x )
= log( x ) ¨ cos3 ( x ) = 3 cos2 ( x ) ¨ (´ sin( x )) ¨ log( x ) +
f (x) dx x
cos3 ( x )

1 2
f ( x ) = f ( x ) ¨ ´3 cos ( x ) sin( x ) log( x ) +
x
cos3 ( x )

cos3 ( x ) 2
=x ¨ ´3 cos ( x ) sin( x ) log( x ) +
x
Remark: negative numbers behave in a complicated manner when they are the base of an
exponential expression. For example, the expression (´1) x is defined when x is the
reciprocal of an odd number (like x = 51 or x = 17 ), but not when x is the reciprocal of an
even number (like x = 12 ). Since the domain of f ( x ) was restricted so that x is always
positive, we avoid these complications.
S-29: We use logarithmic differentiation. So, we modify our function and consider
log f ( x ) = ( x2 ´ 3) ¨ log(3 + sin( x )) .
We differentiate:
f 1 (x) d ! 2 )
= ( x ´ 3) ¨ log(3 + sin( x ))
f (x) dx
cos( x )
= 2x log(3 + sin( x )) + ( x2 ´ 3)
3 + sin( x )
( x2 ´ 3) cos( x )

1
f ( x ) = f ( x ) ¨ 2x log(3 + sin( x )) +
3 + sin( x )
( x2 ´ 3) cos( x )

x2 ´3
= (3 + sin( x )) ¨ 2x log(3 + sin( x )) +
3 + sin( x )
S-30: We will use logarithmic differentiation. First, we take the logarithm of our function,
so we can use logarithm rules.

log [ f ( x )] g( x) = g( x ) log( f ( x ))
Now, we differentiate. On the left side we use the chain rule, and on the right side we use
product and chain rules.
d !
g( x )
) d
log [ f ( x )] = tg( x ) log( f ( x ))u
dx dx
d g( x ) u
dx t[ f ( x )] 1 f 1 (x)
= g ( x ) log ( f ( x )) + g ( x ) ¨
[ f ( x )] g(x) f (x)
316
Finally, we solve for the derivative of our original function.
f 1 (x)

d
t[ f ( x )] g( x) u = [ f ( x )] g( x) 1
g ( x ) log( f ( x )) + g( x ) ¨
dx f (x)
Remark: in this section, we have differentiated problems of this type several times–for
example, Questions 24 through 29.
f 1 (x)
S-31: Let g( x ) := log( f ( x )). Notice g1 ( x ) = f ( x) .
In order to show that the two curves have horizontal tangent lines at the same values of
x, we will show two things: first, that if f ( x ) has a horizontal tangent line at some value
of x, then also g( x ) has a horizontal tangent line at that value of x. Second, we will show
that if g( x ) has a horizontal tangent line at some value of x, then also f ( x ) has a
horizontal tangent line at that value of x.
Suppose f ( x ) has a horizontal tangent line where x = x0 for some point x0 . This means
f 1 (x ) f 1 (x )
f 1 ( x0 ) = 0. Then g1 ( x0 ) = f ( x 0) . Since f ( x0 ) ‰ 0, f ( x 0) = f (0x ) = 0, so g( x ) also has a
0 0 0
horizontal tangent line when x = x0 . This shows that whenever f has a horizontal
tangent line, g has one too.
Now suppose g( x ) has a horizontal tangent line where x = x0 for some point x0 . This
f 1 (x )
means g1 ( x0 ) = 0. Then g1 ( x0 ) = f ( x 0) = 0, so f 1 ( x0 ) exists and is equal to zero.
0
Therefore, f ( x ) also has a horizontal tangent line when x = x0 . This shows that
whenever g has a horizontal tangent line, f has one too.
Remark: if we were not told that f ( x ) gives only positive numbers, it would not
necessarily be true that f ( x ) and log( f ( x )) have horizontal tangent lines at the same
values of x. If f ( x ) had a horizontal tangent line at an x-value where f ( x ) were negative,
then log( f ( x )) would not exist there, let alone have a horizontal tangent line.
S-1: We use the power rule (a) and the chain rule (b): the power rule tells us to “bring
down the 2”, and the chain rule tells us to multiply by y1 .
There is no need for the quotient rule here, as there are no quotients. Exponential
functions have the form (constant)function , but our function has the form
(function)constant , so we did not use (d).
S-2: At (0, 4) and (0, ´4), the curve looks to be horizontal, if you zoom in: a tangent line
here would have derivative zero. At the origin, the curve looks like its tangent line is
dy
vertical, so does not exist.
dx
317
y
S-3: (a) No. A function must pass the vertical line test: one input cannot result in two (or
more) outputs. Since one value of ?x sometimes corresponds to two values of y (for
example, when x = π/4, y is ˘1/ 2), there is no function f ( x ) so that y = f ( x ) captures
every point on the circle.
?
Remark: y = ˘ 1 ´ x2 does capture every point on the unit circle. However, since one
input x sometimes results in two outputs y, this expression is not a function.
(b) No, for the same reasons as (a). If f 1 ( x ) is a function, then it can give at most one
slope corresponding to one value of x. Since one value of x can correspond to two points
on the circle with different slopes, f 1 ( x ) cannot give the slope of every point on the circle.
For example, fix any 0 ă a ă 1. There are two points on the circle with x-coordinate equal
to a. At the upper one, the slope is strictly negative. At the lower one, the slope is strictly
positive.
(c) We differentiate:
dy
2x + 2y =0
dx
dy
and solve for
dx
dy x
=´
dx y
But there is a y in the right-hand side of this equation, and it’s not clear how to get it out.
Our answer in (b) tells us that, actually, we can’t get it out, if we want the right-hand side
to be a function of x. The derivative cannot be expressed as a function of x, because one
value of x corresponds to multiple points on the circle.
318
?
Remark: since y = ˘ 1 ´ x2 , we could try writing
dy x x
= ´ = ˘?
dx y 1 ´ x2
but this is not a function of x. Again, in a function, one input leads to at most one output,
dy
but here one value of x will usually lead to two values of dx .
S-4: Remember that y is a function of x. We begin with implicit differentiation.

xy + e x + ey = 1
dy dy
y+x + e x + ey =0
dx dx
dy
Now, we solve for .
dx
dy dy
x + ey = ´( e x + y )
dx dx
dy
( x + ey ) = ´( e x + y )
dx
dy ex + y
=´ y
dx e +x
S-5: Differentiate both sides of the equation with respect to x:

dy dy
ey = x ¨ 2y + y2 + 1
dx dx
Now, get the derivative on one side and solve
dy dy
ey ´ 2xy = y2 + 1
dx dx
dy y
(e ´ 2xy) = y2 + 1
dx
dy y2 + 1
= y
dx e ´ 2xy
S-6:
• First we find the x-coordinates where y = 1.
π
x2 tan + 2x log(1) = 16
4
x2 ¨ 1 + 2x ¨ 0 = 16
x2 = 16
So x = ˘4.
319
• Now we use implicit differentiation to get y1 in terms of x, y:
x2 tan(πy/4) + 2x log(y) = 16
π 2x 1
2x tan(πy/4) + x2 sec2 (πy/4) ¨ y1 + 2 log(y) + ÿ = 0.
4 y
?
• Now set y = 1 and use tan(π/4) = 1 , sec(π/4) = 2 to get
π
2x tan(π/4) + x2 sec2 (π/4)y1 + 2 log(1) + 2x ¨ y1 = 0
4
π
2x + x2 y1 + 2xy1 = 0
2
2x 4
y1 = ´ 2 =´
x π/2 + 2x πx + 4
4 1
• So at ( x, y) = (4, 1) we have y1 = ´ =´
4π + 4 π+1
1
• and at ( x, y) = (´4, 1) we have y1 =
π´1
S-7: Differentiate the equation and solve:

2 3 dy 2 dy
3x + 4y = ´ sin( x + y) ¨ 2x +
dx dx
dy 2
2x sin( x + y) + 3x 2
=´
dx 4y3 + sin( x2 + y)
S-8:
x2 e0 + 4x cos(0) = 5
x2 + 4x ´ 5 = 0
( x + 5)( x ´ 1) = 0
So x = 1, ´5.
x2 ey + 4x cos(y) = 5 differentiate both sides

x2 ¨ ey ¨ y1 + 2xey + 4x (´ sin(y)) ¨ y1 + 4 cos(y) = 0
320
• Now set y = 0 to get
x2 ¨ e0 ¨ y1 + 2xe0 + 4x (´ sin(0)) ¨ y1 + 4 cos(0) = 0

x2 y1 + 2x + 4 = 0
4 + 2x
y1 = ´ .
x2
• So at ( x, y) = (1, 0) we have y1 = ´6,

6
• and at ( x, y) = (´5, 0) we have y1 = 25 .
S-9: Differentiate the equation and solve:

dy dy
2x + 2y = cos( x + y) ¨ 1 +
dx dx
dy cos( x + y) ´ 2x
=
dx 2y ´ cos( x + y)
S-10:
x2 cos(0) + 2xe0 = 8
x2 + 2x ´ 8 = 0
( x + 4)( x ´ 2) = 0
So x = 2, ´4.
x2 cos(y) + 2xey = 8 differentiate both sides

2 1 y 1 y
x ¨ (´ sin y) ¨ y + 2x cos y + 2xe ¨ y + 2e = 0
x2 ¨ (´ sin 0) ¨ y1 + 2x cos 0 + 2xe0 ¨ y1 + 2e0 = 0

0 + 2x + 2xy1 + 2 = 0
2 + 2x 1+x
y1 = ´ =´
2x x
• So at ( x, y) = (2, 0) we have y1 = ´ 32 ,
• and at ( x, y) = (´4, 0) we have y1 = ´ 34 .
321
dy
S-11: The question asks at which points on the ellipse = 1. So, we begin by
dx
differentiating, implicitly:
dy
2x + 6y =0
dx
dy dy
We could solve for at this point, but it’s not necessary. We want to know when is
dx dx
equal to one:
2x + 6y(1) = 0
x = ´3y
dy
That is, = 1 at those points along the ellipse where x = ´3y. We plug this into the
dx
equation of the ellipse to find the coordinates of these points.
(´3y)2 + 3y2 = 1
12y2 = 1
1 1
y = ˘? = ˘ ?
12 2 3
So, the points along the ellipse where the tangent line is parallel to the line y = x occur
1 ´1
when y = ? and x = ´3y, and when y = ? and x = ´3y. That is, the points
? 2 3 ? 2 3
´ 3 1 3 ´1
, ? and , ? .
2 2 3 2 2 3
S-12: First, we differentiate implicitly with respect to x.

?
xy = x2 y ´ 2
1 d dy
? ¨ txyu = (2x )y + x2
2 xy dx dx
dy
y + x dx dy
? = 2xy + x2
2 xy dx
dy
Now, we plug in x = 1, y = 4, and solve for :
dx
dy
4 + dx dy
= 8+
4 dx
dy 28
=´
dx 3
322
S-13: Implicitly differentiating x2 y( x )2 + x sin(y( x )) = 4 with respect to x gives
2xy2 + 2x2 yy1 + sin y + xy1 cos y = 0
Then we gather the terms containing y1 on one side, so we can solve for y1 :
2x2 yy1 + xy1 cos y = ´2xy2 ´ sin y

y1 (2x2 y + x cos y) = ´2xy2 ´ sin y
2xy2 + sin y
y1 = ´
2x2 y + x cos y
S-14:
• First we find the x-ordinates where y = 0.
x 2 + (1) e0 = 5
x2 + 1 = 5
x2 = 4
So x = 2, ´2.
dy dy
2x + (y + 1)ey + ey =0
dx dx

dy dy
2x + (0 + 1)e0 + e0 =0
dx dx
dy dy
2x + + =0
dx dx
dy
2x = ´2
dx
dy
x=´
dx
• So at ( x, y) = (2, 0) we have y1 = ´2,

• and at ( x, y) = (´2, 0) we have y1 = 2.
S-15: The slope of the tangent line is, of course, given by the derivative, so let’s start by
dy
finding dx of both shapes.
For the circle, we differentiate implicitly

dy
2x + 2y =0
dx
323
dy
and solve for
dx
dy x
=´
dx y
For the ellipse, we also differentiate implicitly:
dy
2x + 6y =0
dx
dy
and solve for
dx
dy x
=´
dx 3y
What we want is a value of x where both derivatives are equal. However, they might
have different values of y, so let’s let y1 be the y-values associated with x on the circle,
and let y2 be the y-values associated with x on the ellipse. That is, x2 + y21 = 1 and
x2 + 3y22 = 1. For the slopes at ( x, y1 ) on the circle and ( x, y2 ) on the ellipse to be equal,
we need:
x x
´ =´
y1 3y2

1 1
x ´ =0
y1 3y2
So x = 0 or y1 = 3y2 . Let’s think about which x-values will have a y-coordinate of the
circle be three times as large as a y-coordinate of the ellipse. If y1 = 3y2 , ( x, y1 ) is on the
circle, and ( x, y2 ) is on the ellipse, then x2 + y21 = x2 + (3y2 )2 = 1 and x2 + 3y22 = 1. In
this case:
x2 + 9y22 = x2 + 3y22
9y22 = 3y22
y2 = 0
x = ˘1
We need to be a tiny bit careful here: when y = 0, y1 is not defined for either curve. For
both curves, when y = 0, the tangent lines are vertical (and so have no real-valued
slope!). Two vertical lines are indeed parallel.
So, for x = 0 and for x = ˘1, the two curves have parallel tangent lines.
324
y
x2 + 3y2 = 1
x 2 + y2 = 1
S-1: (a) We can plug any number into the cosine function, and it will return a number in
[´1, 1]. The domain of arcsin x is [´1, 1], so any number we plug into cosine will give us
a valid number to plug into arcsine. So, the domain of f ( x ) is all real numbers.
(b) We can plug any number into the cosine function, and it will return a number in
[´1, 1]. The domain of arccsc x is (´8, ´1] Y [1, 8), so in order to have a valid number to
plug into arccosecant, we need cos x = ˘1. That is, the domain of g( x ) is all values
x = nπ for some integer n.
(c) The domain of arccosine is [´1, 1]. The domain of sine is all real numbers, so no
matter what number arccosine spits out, we can safely plug it into sine. So, the domain of
h( x ) is [´1, 1].
S-2: False: cos t = 1 for infinitely many values of t; arccosine gives only the single value
t = 0 for which cos t = 1 and 0 ď t ď π. The particle does not start moving until t = 10,
so t = 0 is not in the domain of the function describing its motion.
The particle will have height 1 at time 2πn, for any integer n ě 2.
S-3: First, we restrict the domain of f to force it to be one–to–one. There are many
intervals we could choose over which f is one–to–one, but the question asks us to contain
x = 0 and be as large as possible; this leaves us with the following restricted function:
325
y
The inverse of a function swaps the role of the input and output; so if the graph of
y = f ( x ) contains the point ( a, b), then the graph of Y = f ´1 ( X ) contains the point (b, a).
That is, the graph of Y = f ´1 ( X ) is the graph of y = f ( x ) with the x-coordinates and
y-coordinates swapped. (So, since y = f ( x ) crosses the y-axis at y = 1, then Y = f ´1 ( X )
crosses the X-axis at X = 1.) This swapping is equivalent to reflecting the curve y = f ( x )
over the line y = x.
y
y=x
y = f (x)
1
y = f ´1 ( x )
x
1
Remark: while you’re getting accustomed to inverse functions, it is sometimes clearer to

consider y = f ( x ) and Y = f ´1 ( X ): using slightly different notations for x (the input of
f , hence the output of f ´1 ) and X (the input of f ´1 , which comes from the output of f ).
However, the convention is to use x for the inputs of both functions, and y as the outputs
of both functions, as is written on the graph above.
S-4: The tangent line is horizontal when 0 = y1 = a ´ sin x. That is, when a = sin x.
• If |a| ą 1, then there is no value of x for which a = sin x, so the curve has no
horizontal tangent lines.
• If |a| = 1, then there are infinitely many solutions to a = sin x, but only one solution
in the interval [´π, π ]: x = arcsin( a) = arcsin(˘1) = ˘ π2 . Then the values of x for
326
which a = sin x are x = 2πn + a π2 for any integer n.
• If |a| ă 1, then there

are infinitely many solutions to a = sin x. The solution in the
π π
interval ´ 2 , 2 is given by x = arcsin( a). The other solution in the interval
(´π, π ) is given by x = π ´ arcsin( a), as shown in the unit circles below.
y y
a)
in (
a a
arcs
π´
in ( a)
arcs arcsin( a)
x x
So, the values of x for which x = sin a are x = 2πn + arcsin( a) and
x = 2πn + π ´ arcsin( a) for any integer n.
Remark: when a = 1, then
π π
2πn + arcsin( a) = 2πn + = 2πn + π ´ = 2πn + π ´ arcsin( a).
2 2
Similarly, when a = ´1,
π π
2πn + arcsin( a) = 2πn ´ = 2π (n ´ 1) + π ´ ´ = 2π (n ´ 1) + π ´ arcsin( a).
2 2
So, if we try to use the descriptions in the third bullet point to describe points where the
tangent line is horizontal when |a| = 1, we get the correct points but each point is listed
twice. This is why we separated the case |a| = 1 from the case |a| ă 1.
S-5: The function arcsin x is only defined for |x| ď 1, and the function arccsc x is only
defined for |x| ě 1, so f ( x ) has domain |x| = 1. That is, x = ˘1.
In order for f ( x ) to be differentiable at a point, it must exist in an open interval around

that point. (See Definition 2.2.1.) Since our function does not exist over any open interval,
f ( x ) is not differentiable anywhere.
So, actually, f ( x ) is a pretty boring function, which we can entirely describe as:
f (´1) = ´π and f (1) = π.
327
d ! x ) 1 1
arcsin =b ¨
dx 3 x 2 3

1´ 3
1
= b
x2
3 1´ 9
1
=?
9 ´ x2
x
Since the domain of arcsine is [´1, 1], and we are plugging in to arcsine, the values of x
3
x
that we can plug in are those that satisfy ´1 ď ď 1, or ´3 ď x ď 3. So the domain of f
3
is [´3, 3].

" * ( t2 ´ 1) ?´1 ´ (arccos t)(2t)
d arccos t 1´t2
=
dt t2 ´ 1 ( t2 ´ 1)2
The domain of arccosine is [´1, 1], and since t2 ´ 1 is in the denominator, the domain of f
requires t2 ´ 1 ‰ 0, that is, t ‰ ˘1. So the domain of f (t) is (´1, 1).
S-8: The domain of arcsec x is |x| ě 1: that is, we can plug into arcsecant only values with
absolute value greater than or equal to one. Since ´x2 ´ 2 ď ´2, every real value of x
gives us an acceptable value to plug into arcsecant. So, the domain of f ( x ) is all real
numbers.
d 1
To differentiate, we use the chain rule. Remember tarcsec xu = ? .
dx |x| x2 ´ 1
d ! 2
) 1
arcsec(´x ´ 2) = a ¨ (´2x )
dx | ´ x ´ 2| (´x2 ´ 2)2 ´ 1
2
´2x
= ? .
( x2 + 2) x4 + 4x + 3
S-9: We use the chain rule, remembering that a is a constant.

" x * 1
d 1 1 1
arctan = ¨ 2 ¨
dx a a a 1+ x a
a
1
= 2
a + x2
The domain of arctangent is all real numbers, so the domain of f ( x ) is also all real
numbers.
328
S-10: We differentiate using the product and chain rules.
d ! a ) x ´2x
x arcsin x + 1 ´ x2 = arcsin x + ? + ?
dx 1 ´ x2 2 1 ´ x2
= arcsin x
?
The domain of arcsin x is [´1, 1], and the domain of 1 ´ x2 is all values of x so that
1 ´ x2 ě 0, so x in [´1, 1]. Therefore, the domain of f ( x ) is [´1, 1].
S-11: We differentiate using the chain rule:

d 2x
tarctan( x2 )u =
dx 1 + x4
This is zero exactly when x = 0.
S-12: Using formulas you should memorize from this section,

d 1 ´1
tarcsin x + arccos xu = ? +? =0
dx 1 ´ x2 1 ´ x2
Remark: the only functions with derivative equal to zero everywhere are constant
functions, so arcsin x + arccos x should be a constant. Since sin θ = cos π2 ´ θ , we can
set
π
sin θ = x cos ´θ = x
2
where x and θ are the same in both expressions, and ´ 2 ď θ ď π2 . Then
π
π
arcsin x = θ arccos x = ´θ
2

We note here that arcsine is the inverse of the sine function restricted to ´ π2 , π2 . So, since
we restricted θ to this domain, sin θ = x really does imply arcsin x = θ. (For an example
of why this matters, note sin(2π ) = 0, but arcsin(0) = 0 ‰ 2π.) Similarly, arccosine is the
inverse of the cosine function restricted
to [0, π ]. Since ´ π2 ď θ ď π2 , then
0 ď ( 2 ´ θ ) ď π, so cos 2 ´ θ = x really does imply arccos x = π2 ´ θ.
π π
So,
π π
arcsin x + arccos x = θ + ´θ =
2 2
d !π)
which means the derivative we were calculating was actually just = 0.
dx 2

´ x12 ´1
y1 = c 2 = 2 b .
1
1 ´ 1x x 1´ x2
329
1
´ x12 ´1
y = 2 = 2 .
x +1
1 + 1x
S-15: Using the product rule:
d ! ) 1
(1 + x ) arctan x = 2x arctan x + (1 + x2 )
2
dx 1 + x2
= 2x arctan x + 1
S-16: Let θ = arctan x. Then θ is the angle of a right triangle that gives tan θ = x. In
particular, the ratio of the opposite side to the adjacent side is x. So, we have a triangle
?
x2 + 1
x
θ
1
where the length of the hypotenuse came from the Pythagorean Theorem. Now,
opp x
sin (arctan x ) = sin θ = =?
hyp x2 + 1
" * ?x2 + 1 ´ x ?2x
d x 2 x 2 +1
? = 2
dx x2 + 1 x +1
? 2
 ?
x2 + 1 ´ ? x2 x2 + 1
x +1 
= ¨ ?
x2 + 1 x2 + 1
( x 2 + 1) ´ x 2
=
( x2 + 1)3/2
1
= 2 = ( x2 + 1)´3/2
( x + 1)3/2
Remark: another strategy is to differentiate first, using the chain rule, then draw a
d cos(arctan x )
triangle to simplify the resulting expression tsin (arctan x )u = .
dx 1 + x2
330
S-17:
Let θ = arcsin x. Then θ is the angle of a right triangle that gives sin θ = x. In particular,
the ratio of the opposite side to the hypotenuse is x. So, we have a triangle that looks like
this:
1
x
θ
?
1 ´ x2
where the length of the adjacent side came from the Pythagorean Theorem. Now,
?
adj 1 ´ x2
cot (arcsin x ) = cot θ = =
opp x
#? ?
2
+ x ?´2x 2 ´ 1 ´ x2
d 1´x 2 1´x
=
dx x x2
´x2 ´ (1 ´ x2 )
= ?
x2 1 ´ x2
´1
= ?
x2 1 ´ x2
Remark: another strategy is to differentiate first, using the chain rule, then draw a
d ´ csc2 (arcsin x )
triangle to simplify the resulting expression tcot (arcsin x )u = ? .
dx 1 ´ x2
S-18: The line y = 2x + 9 has slope 2, so we must find all values of x between ´1 and 1
d
(arcsin x is only defined for these values of x) for which dx tarcsin xu = 2. Evaluating the
331
derivative:
y = arcsin x
1
2 = y1 = ?
1 ´ x2
1
4=
1 ´ x2
1
= 1 ´ x2
4
3
x2 =
4?
3
x=˘
2?
3 π
( x, y) = ˘ ,
2 3
S-19: We differentiate using the chain rule:

d 1 d
tarctan(csc x )u = 2
¨ tcsc xu
dx 1 + csc x dx
´ csc x cot x
=
1 + csc2 x
´ 1 ¨ cos x
= sinx sinx2
1 + sin1 x
´ cos x
=
sin2 x + 1
(2n + 1)π
So if f 1 ( x ) = 0, then cos x = 0, and this happens when x = for any integer n.
2
We should check that these points are in the domain of f . Arctangent is defined for all
(2n + 1)π
real numbers, so we only need to check the domain of cosecant; when x = ,
2
1
then sin x = ˘1 ‰ 0, so csc x = exists.
sin x
S-20: Since g(y) = f ´1 (y),

f ( g(y)) = f f ´1 (y) = y
Now, we can differentiate with respect to y using the chain rule.

d d
t f ( g(y))u = tyu
dy dy
f 1 ( g(y)) ¨ g1 (y) = 1
1 1
g1 ( y ) = 1 =
f ( g(y)) 1 ´ sin g(y)
332
S-21: Write g(y) = f ´1 (y). Then g( f ( x )) = x, so differentiating both sides (using the
chain rule), we see
g1 ( f ( x )) ¨ f 1 ( x ) = 1
What we want is g1 (π ´ 1), so we need to figure out which value of x gives f ( x ) = π ´ 1.

A little trial and error leads us to x = π2 .
π
g1 ( π ´ 1) ¨ f 1 =1
2

Since f 1 ( x ) = 2 ´ cos( x ), f 1 π2 = 2 ´ 0 = 2:
g1 ( π ´ 1) ¨ 2 = 1
1
g1 ( π ´ 1) =
2
S-22: Write g(y) = f ´1 (y). Then g( f ( x )) = x, so differentiating both sides (using the
chain rule), we see
g1 ( f ( x )) f 1 ( x ) = 1
What we want is g1 (e + 1), so we need to figure out which value of x gives f ( x ) = e + 1.

A little trial and error leads us to x = 1.
g1 ( f (1)) f 1 (1) = 1
g1 ( e + 1) ¨ f 1 (1) = 1
1
g1 ( e + 1) =
f 1 (1)
It remains only to note that f 1 ( x ) = e x + 1, so f 1 (1) = e + 1
1
g1 ( e + 1) =
e+1
S-23: We use logarithmic differentiation, our standard method of differentiating an

expression of the form (function)function .
f ( x ) = [sin x + 2]arcsec x
log( f ( x )) = arcsec x ¨ log[sin x + 2]
f 1 (x) 1 cos x
= ? log[sin x + 2] + arcsec x ¨
f (x) |x| x2 ´ 1 sin x + 2

1 arcsec x log[sin x + 2] arcsec x ¨ cos x
f ( x ) = [sin x + 2] ? +
|x| x2 ´ 1 sin x + 2
333
The domain of arcsec x is |x| ě 1. For any x, sin x + 2 is positive, and a positive number
can be raised to any power.
? (Recall negative numbers cannot be raised to any power–for
example, (´1) 1/2 = ´1 is not a real number.) So, the domain of f ( x ) is |x| ě 1.
1
S-24: The function ? exists only for those values of x with x2 ´ 1 ą 0: that is, the
x2 ´1
1
domain of ? is |x| ą 1. However, the domain of arcsine is |x| ď 1. So, there is not
x2 ´ 1
1
one single value of x where arcsin x and ? are both defined.
x2 ´ 1
1
If the derivative of arcsin( x ) were given by ? , then the derivative of arcsin( x )
x2 ´ 1
would not exist anywhere, so we would probably just write “derivative does not exist,”
instead of making up a function with a mismatched domain. Also, the function
f ( x ) = arcsin( x ) is a smooth curve–its derivative exists at every point strictly inside its
domain. (Remember not all curves are like this: for instance, g( x ) = |x| does not have a
derivative at x = 0, but x = 0 is strictly inside its domain.) So, it’s a pretty good bet that
1
the derivative of arcsine is not ? .
x2 ´ 1

f ( x ) = arctan x. To see this, simply use the definition of the derivative at a = 1:
ˇ
d ˇ f ( x ) ´ f ( a)
dx xÑa xá
ˇa
d ˇ arctan x ´ arctan 1
tarctan xuˇˇ = lim
dx 1 xÑ1 x´1
arctan x ´ π4
= lim
xÑ1 x´1
π
= lim ( x ´ 1)´1 arctan x ´ .
xÑ1 4
1 1
Since the derivative of f ( x ) is 2
, its value at x = 1 is exactly .
1+x 2
S-26: First, let’s interpret the given information: when the input of our function is 2x + 1
5x ´ 9
for some x, then its output is , for that same x. We’re asked to evaluate f ´1 (7),
3x + 7
which is the number y with the property that f (y) = 7. If the output of our function is 7,
that means
5x ´ 9
7=
3x + 7
and so
7(3x + 7) = 5x ´ 9
29
x=´
8
334
29 5x ´ 9
So, when x = ´ , our equation f (2x + 1) = becomes:
8 3x + 7
5 ¨ ´29
8 ´9

´29
f 2¨ +1 =
8 3 ¨ ´29
8 +7
Or, equivalently:

25
f ´ =7
4
25
Therefore, f ´1 (7) = ´ .
4
S-27: If f ´1 (y) = 0, that means f (0) = y. So, we want to find out what we plug into f ´1
to get 0. Since we only know f ´1 in terms of a variable x, let’s figure out what x gives us
an output of 0:
2x + 3
=0
x+1
2x + 3 = 0
3
x=´
2
2x + 3 ´3
Now, the equation f ´1 (4x ´ 1) = with x = tells us:
x+1 2
2 ¨ ´3
2 +3

´1 ´3
f 4¨ ´1 = ´3
2 2 +1
Or, equivalently:
f ´1 (´7) = 0
Therefore, f (0) = ´7.
S-28:
• Solution 1: We begin by differentiating implicitly. Following the usual convention,
335
we use y1 to mean y1 ( x ).
arcsin( x + 2y) = x 2 + y2 Using the chain rule:

1 + 2y1
a = 2x + 2yy1
1 ´ ( x + 2y)2
1 2y1
a +a = 2x + 2yy1
1 ´ ( x + 2y)2 1 ´ ( x + 2y)2
2y1 1
a ´ 2yy1 = 2x ´ a
1 ´ ( x + 2y) 2 1 ´ ( x + 2y)2
!
2 1
y1 a ´ 2y = 2x ´ a
1 ´ ( x + 2y)2 1 ´ ( x + 2y)2
2x ´ ? 1 a
2
!
1´ ( x + 2y ) 2 1 ´ ( x + 2y )
y1 = a
? 2 ´ 2y 1 ´ ( x + 2y)2
1´( x +2y)2
a
2x 1 ´ ( x + 2y)2 ´ 1
y1 = a
2 ´ 2y 1 ´ ( x + 2y)2
• Solution 2: We begin by taking the sine of both sides of the equation.
x + 2y = sin( x2 + y2 )
Now, we differentiate implicitly.
1 + 2y1 = cos( x2 + y2 ) ¨ (2x + 2yy1 )

1 + 2y1 = 2x cos( x2 + y2 ) + 2yy1 cos( x2 + y2 )
2y1 ´ 2yy1 cos( x2 + y2 ) = 2x cos( x2 + y2 ) ´ 1

1 2 2
y 2 ´ 2y cos( x + y ) = 2x cos( x2 + y2 ) ´ 1
2x cos( x2 + y2 ) ´ 1
y1 =
2 ´ 2y cos( x2 + y2 )
• We used two different methods, and got two answers that look pretty different.
However, the answers ought to be equivalent. To see this, we remember that for all
values of x and y that we care about (those pairs ( x, y) in the domain of our curve),
the equality
holds. Drawing a triangle:
336
1
x + 2y
2
y
x +
2
a
1 ´ ( x + 2y)2
where the adjacent

a side (in red) come from the Pythagorean Theorem. Then,
2 2
cos( x + y ) = 1 ´ ( x + 2y)2 , so using our second solution:
2x cos( x2 + y2 ) ´ 1
y1 =
2 ´ 2y cos( x2 + y2 )
a
2x 1 ´ ( x + 2y)2 ´ 1
= a
2 ´ 2y 1 ´ ( x + 2y)2
which is exactly the answer from our first solution.
S-1: We know the top speed of the caribou, so we can use this to give the minimum
possible number of hours the caribou spent
travelling
during its migration. If the caribou
1hr
travels at 70kph, it will take (5000km) « 71.4hrs to travel 5000 kilometres.
70km
Probably the caribou wasn’t sprinting the whole time, so probably it took it longer than
that, but we can only say for sure that the caribou spent at least about 71.4 hours
migrating.
S-2: If f ( x ) is the position of the crane at time x, measured in hours, then (if we let x = 0
be the beginning of the day) we know that f (24) ´ f (0) = 240. Since f ( x ) is the position
of the bird, f ( x ) is continuous and differentiable. So, the MVT says there is a c in (0, 24)
f (24) ´ f (0) 240
such that f 1 ( x ) = = = 10. That is, at some point c during the day, the
24 ´ 0 24
speed of the crane was exactly 10 kph.
S-3: The MVT guarantees there is some point c strictly between a and b where the tangent
line to f ( x ) at x = c has the same slope as the secant line of f ( x ) from x = a to x = b. So,
let’s start by drawing in the secant line.
337
y
x
a b
What we’re looking for is a point on the curve where the tangent line is parallel to this
secant line. In fact, there are two.
x
a b
So, either of the two values c1 and c2 marked below can serve as the point guaranteed by
the MVT:
338
y
x
a c1 c2 b
S-4: Since f ( x ) is differentiable for all x P (0, 10), then f ( x ) is also continuous for all
x P (0, 10). If f ( x ) were continuous on the closed interval [0, 10], then the MVT would
f (10) ´ f (0)
guarantee f 1 ( x ) = = 1 for some c P (0, 10); however, this is not the case. So,
10 ´ 0
it must be that f ( x ) is continuous for all x P (0, 10), but not for all x P [0, 10].
Since f 1 (c) = 0 for c P (0, 10), that means f is constant on that interval. So, f ( x ) is a
function like this:
y
where the height of the constant function can be anything.

"
0 x ‰ 10
So, one possible answer is f ( x ) = .
10 x = 10
S-5: (a) No such function is possible: Rolle’s Theorem guarantees f 1 (c) = 0 for at least
one point c P (1, 2).
For the other functions, examples are below, but many answers are possible.
339
(b) (c) (d)
y y y
x x x
1 2 1 2 1 2
S-6: The function f ( x ) is continuous over all real numbers, but it is only differentiable
when x ‰ 0. So, if we want to apply the MVT, our interval must consist of only positive
numbers or only negative numbers: the interval (´4, 13) is not valid.
It is possible to use the mean value theorem to prove what we want: if a = 1 and b = 144,
then f ( x ) is differentiable over the interval (1, 144) (since 0 is not contained in that
interval), and f ( x ) is continuous
a everywhere,
a so by the mean value theorem there exists
|144| ´ |1| 11 1
some point c where f 1 ( x ) = = = .
144 ´ 1 143 13
That being said, an easier way to prove that a point exists is to simply
find
it–without
? 1 169 1
using the MVT. When x ą 0, f ( x ) = x, so f 1 ( x ) = ? . Then f 1 = .
2 x 4 13
S-7: We note that f (0) = f (2π ) = 0. Then using the Mean Value Theorem (note that the
function is differentiable for all real numbers), we conclude that there exists c in (0, 2π )
such that
f (2π ) ´ f (0)
f 1 (c) = = 0.
2π ´ 0
S-8: We note that f (0) = f (1) = 0. Then using the Mean Value Theorem (note that the
function is differentiable for all real numbers), we get that there exists c P (0, 1) such that
f (1) ´ f (0)
f 1 (c) = = 0.
1´0
?
S-9: We note that f (0) = f (2π ) = 3 + π 2 . Then using the Mean Value Theorem (note
that the function is differentiable for all real numbers since 3 + sin x ą 0), we get that
there exists c P (0, 2π ) such that
f (2π ) ´ f (0)
f 1 (c) = = 0.
2π ´ 0
340
S-10: We note that f (0) = 0 and f (π/4) = 0. Then using the Mean Value Theorem (note
that the function is differentiable for all real numbers), we get that there exists
c P (0, π/4) such that
f (π/4) ´ f (0)
f 1 (c) = = 0.
π/4 ´ 0
S-11: By inspection, we see x = 0 is a root of f ( x ). The question now is whether there

could possibly be other roots. Since f ( x ) is differentiable over all real numbers, if there is
another root a, then by Rolle’s Theorem, f 1 (c) = 0 for some c strictly between 0 and a.
However, f 1 ( x ) = 3 ´ cos x is never zero. So, there is no second root: f ( x ) has precisely
one root.
S-12: The function f ( x ) is continuous and differentiable over all real numbers. If a and b
are distinct roots of f ( x ), then f 1 (c) = 0 for some c strictly between a and b (Rolle’s
Theorem). So, let’s think about f 1 ( x ).
f 1 ( x ) = (4x + 1)3 + 1
This is simple enough that we can find its zero explicitly:
(4x + 1)3 + 1 = 0 ô (4x + 1)3 = ´1 ô 4x + 1 = ´1 ô 4x = ´2
´1
Hence f 1 (c) = 0 only when c = . That the derivative only has a single zero is very
2
useful. It means (via Rolle’s theorem) that if f ( x ) has distinct roots a and b with a ă b,
´1
then we must have a ă ă b. This also means that f ( x ) cannot have 3 distinct roots
2
a, b and (say) q with a ă b ă q, because then Rolle’s theorem would imply that f 1 ( x )
would have two zeros — one between a and b and another between b and q.
We’ve learned that f ( x ) has at most two roots, and we’ve learned something about
where those roots can exist, if there are indeed two of them. But that means f ( x ) could
have 0, 1, or 2 roots.
It’s not easy to find a root of f ( x ) by inspection. But we can get a good enough picture of
the graph of y = f ( x ) to tell exactly how many roots there are, just by exploiting the
following properties of f ( x ) and f 1 ( x ).
• As x tends to ˘8, f ( x ) tends to +8.
• The derivative f 1 ( x ) = (4x + 1)3 + 1 is negative for x ă ´ 12 and is positive for

x ą ´ 12 . That is, f ( x ) is decreasing for x ă ´ 12 and increasing for x ą ´ 12 .

• f ´ 12 = 16 1
´ 12 ă 0.
This means the function must look something like the graph below:
341
y
except we are unsure of the locations of the x-intercepts. At present we only know that
one is to the left of x = 1/2 and one is to the right.
Note what happens to f ( x ) as x increases from strongly negative values to strongly

positive values.
• When x is large and negative, f ( x ) ą 0.
• As x increases, f ( x ) decreases continuously until x = ´ 12 , where f ( x ) ă 0. In

65
ą 0 and f ´ 12 ă 0 and f ( x ) is continuous, the

particular, since f (´1) = 16
intermediate value theorem guarantees that f ( x ) takes the value zero for some x
between ´ 12 and ´1. More descriptively put, as x increases from hugely negative
numbers to ´ 21 , f ( x ) passes through zero exactly once.
• As x increases beyond ´ 12 , f ( x ) increases continuously, starting negative and

becoming very large and positive
1 1
when x becomes large and positive. In particular,
since f (0) = 16 ą 0 and f ´ 2 ă 0 and f ( x ) is continuous, the intermediate value
theorem guarantees that f ( x ) takes the value zero for some x between ´ 12 and 0.
So, as x increases from ´ 21 to near +8, f ( x ) again passes through zero exactly once.
So f ( x ) must have exactly two roots, one with x ă ´ 21 and one with x ą ´ 12 .
S-13:
• We can see by inspection that f (0) = 0, so there is at least one root. We have to
determine how many any other roots there are, if any.
• The function f ( x ) is the sum of the two terms x3 and sin( x5 ). A number x is a root
of f ( x ) if and only if the two terms cancel each other exactly for that value of x.
That is, x is a root of f if and only if x3 = ´ sin( x5 ). To develope some intuition in
our hunt for other roots, we sketch, in the same figure, the graphs y = x3 and
y = ´ sin( x5 ). Then the roots of f ( x ) are precisely the x’s where the two curves
y = x3 and y = ´ sin( x5 ) intersect.
342
y y = x3
x=−1
y = − sin(x5 ) y=1
x
y=−1
x=1
• Looking at the sketch, we see that the two curves cannot possibly intersect at any
point having |x| ą 1 — if |x| ą 1, then |x3 | ą 1 but | ´ sin( x5 )| ď 1 and we cannot
possibly have x3 = ´ sin( x5 ). That is, all roots of f ( x ) are in the interval [´1, 1].
• From the sketch, we would probably guess that y = x3 and y = ´ sin( x5 ) cross only
at x = 0. We can use Rolle’s Theorem to verify that that is indeed the case.
• If there is a root a ‰ 0, then by Rolle’s Theorem (since f ( x ) is continuous and

differentiable for all real numbers x) f 1 (c) = 0 for some c strictly between 0 and a.
In particular, since we already know any roots a will be between ´1 and 1, if f ( x )
has two roots then f 1 (c) = 0 for some c P (´1, 0) Y (0, 1).
• The derivative of f is

f 1 ( x ) = 3x2 + 5x4 cos x5 = x2 3 + 5x2 cos x5

So, if f 1 ( x ) = 0, then x = 0 or 3 + 5x2 cos x5

= 0. If x P (´1, 0) Y (0, 1), then
|x| ă 1 ùñ |x5 | ă 1 ă ùñ cos x5 ą 0

π

2
ùñ 3 + 5x2 cos x5 ą 3

ùñ f 1 ( x ) ‰ 0
That is, there is no c P (´1, 0) Y (0, 1) with f 1 (c) = 0. Therefore, following our last
bullet point, f ( x ) has only one root.
Note here that f 1 ( x ) has many zeroes — infinitely many, in fact. However, x = 0 is the
only root of f 1 ( x ) in the interval (´1, 1).
S-14:
We are to find the number of positive solutions to the equation e x = 4 cos(2x ). The figure
below contains the graphs of y = e x and y = 4 cos(2x ). The solutions to e x = 4 cos(2x )
are precisely the x’s where y = e x and y = 4 cos(2x ) cross.
343
y y = ex
x
1 log 4
y = 4 cos(2x)
−4
It sure looks like there is exactly one crossing with xě 0 and that one
crossing is
x
somewhere between x = 0 and x = 1. Indeed since e ´ 4 cos(2x ) x=0 = ´3 ă 0 and
x
e ´ 4 cos(2x ) x=1 ą e ą 0 and f ( x ) = e x ´ 4 cos(2x ) is continuous, the intermediate

value theorem guarantees that there is at least one root with 0 ă x ă 1.
We still have to show that there is no second root — even if our graphs are not accurate.
Recall that the range of the cosine function is [´1, 1]. If e x = 4 cos(2x ), then e x ď 4, so
x ď log(4) « 1.39. So, we only need to search for roots of f ( x ) on the interval (0, 1.4): we
are guaranteed there are no roots elsewhere. Over this interval, 2x P (0, 2.8), so
sin(2x ) ą 0, and thus f 1 ( x ) = e x + 8 sin(2x ) ą 0. Since f 1 ( x ) has no roots in (0, 1.4), we
conclude by Rolle’s Theorem that f ( x ) has at most one root in (0, 1.4) (and so at most one
positive root total). Since we’ve already found that a root of f ( x ) exists in (0, 1), we
conclude e x = 4 cos(2x ) has precisely one positive-valued solution.
S-15: (a)
2
f 1 ( x ) = 15x4 ´ 30x2 + 15 = 15 x4 ´ 2x2 + 1 = 15 x2 ´ 1 ě 0

The derivative is nonnegative everywhere. The only values of x for which f 1 ( x ) = 0 are 1
and ´1, so f 1 ( x ) ą 0 for every x in (´1, 1).
(b) If f ( x ) has two roots a and b in [´1, 1], then by Rolle’s Theorem, f 1 (c) = 0 for some x
strictly between a and b. But since a and b are in [´1, 1], and c is between a and b, that
means c is in (´1, 1); however, we know for every c in (´1, 1), f 1 (c) ą 0, so this can’t
happen. Therefore, f ( x ) does not have two roots a and b in [´1, 1]. This means f ( x ) has at
most one root in [´1, 1].
S-16: Write f ( x ) = e x . Since f ( x ) is continuous and differentiable, the Mean Value

Theorem asserts that there exists some c between 0 and T such that
f ( T ) ´ f (0)
f 1 (c) =
T´0
344
The problem asks us to find this value of c. Solving:
e T ´ e0
ec =
T
T
e ´1
ec =
T
eT ´ 1

c = log
T
S-17: The domains of arcsec x and C ´ arcsec x are the same: |x| ě 1. Define
f ( x ) = arcsec x + arccsc x, and note that the domain of f ( x ) is also |x| ě 1. Using
Theorem 2.12.7 in the CLP-1 text,
d d 1 ´1
f 1 (x) = arcsec x + arccsc x = ? + ? =0
dx dx |x| x2 ´ 1 |x| x2 ´ 1
By Corollary 2.13.11 in the CLP-1 text, this means that f ( x ) is constant on any interval in
|x| ě 1. So f ( x ) is a constant, call it C+ , on x ě 1, and f ( x ) is also a constant, call it C´ , on
x ď ´1.
In order to find C+ , we find f (1), because we know angles for which the secant and
cosecant are x = 1.
1
cos(0) = 1 ùñ sec(0) = 1 = 1 ùñ arcsec(1) = 0
1

sin π2 = 1 ùñ csc π2 = 1 = 1 ùñ arccsc(1) = π2
π
ùñ C+ = arcsec(1) + arccsc(1) = 2
In order to find C´ , we find f (´1), because we know angles for which the secant and
cosecant are x = ´1.
1
cos(π ) = ´1 ùñ sec(π ) = ´1 = ´1 ùñ arcsec(´1) = π
1

sin ´ 2 = ´1 ùñ csc ´ π2 =
π
´1 = ´1 ùñ arccsc(´1) = ´ π2
π
ùñ C´ = arcsec(´1) + arccsc(´1) = 2
π π
So f ( x ) = arcsec x + arccsc x = 2 for all |x| ě 1 and arcsec x = 2 ´ arccsc x for all |x| ě 1.
S-18: Since e´ f ( x) is always positive (regardless of the value of f ( x )),

1 1
f 1 (x) = ă =1
1 + e´ f ( x ) 1+0
for every x.
Since f 1 ( x ) exists for every x, we see that f is differentiable, so the Mean Value Theorem
applies. If f (100) is greater than or equal to 100, then by the Mean Value Theorem, there
would have to be some c between 0 and 100 such that
f (100) ´ f (0) 100
f 1 (c) = ě =1
100 100
345
Since f 1 ( x ) ă 1 for every x, there is no value of c as described. Therefore, it is not possible
that f (100) ě 100. So, f (100) ă 100.
S-19: If 2x + sin x is one–to–one over an interval, it never takes the same value for two
distinct numbers in that interval. By Rolle’s Theorem, if f ( a) = f (b) for distinct a and b,
then f 1 (c) = 0 for some c between a and b. However, f 1 ( x ) = 2 + cos x, which is never
zero. In fact, f 1 ( x ) ě 1 for all x, so f ( x ) is strictly increasing over its entire domain.
Therefore, our function f never takes the same value twice, so it is one–to–one over all the
real numbers, (´8, 8).
When we define the inverse function f ´1 ( x ), the domain of f is the range of f ´1 , and
vice-versa. In general, we might have to restrict the domain of f (and hence the range of
f ´1 ) to an interval where f is one–to–one, but in our case, this isn’t necessary. So, the
range of f ´1 is (´8, 8) and the domain of f ´1 is the range of f : (´8, 8).
x
S-20: If f ( x ) = + sin x is one–to–one over an interval, it never takes the same value for
2
two distinct numbers in that interval. By Rolle’s Theorem, if f ( a) = f (b) for distinct a
and b, then f 1 (c) = 0 for some c between a and b. Since f 1 ( x ) = 12 + cos x, f 1 ( x ) = 0 when
x = 2nπ ˘ 2π 3 forsome integer n. So, in particular, if a and b are distinct numbers in the
interval ´ 3 , 2π
2π 1
3 , then for every c strictly between a and b, f ( c ) ‰0, so by Rolle’s
Theorem f ( a) ‰ f (b). Therefore f ( x ) is one–to–one on the interval ´ 2π 2π
3 , 3 .
We should also show that the interval ´ 2π 2π

3 , 3 cannot be extended to a larger interval
over which f ( x ) is still one–to–one. Consider the derivative f 1 ( x ) = 12 + cos x. For all
´ 2π 2π 1
3 ă x ă 3 , we have cos x ą ´ 2 (sketch the graph of cos x yourself) so that f ( x ) ą 0
1
and f ( x ) is increasing. But at x = 3 , f 1 ( x ) = 0, and then for x a bit bigger than 2π

2π
3 we
1 1
have cos x ă ´ 2 so that f ( x ) ă 0 and f ( x ) is decreasing. So the graph “reverses
direction”, and f ( x ) repeats values. (See the graph of y = f ( x ) below.) The same is true
for x a little smaller than ´ 2π 3 .
x
y = f (x) = + sin x
2
x
´ 2π
3
2π
3
346
When we define the inverse function f ´1 ( x ), first we restrict f to ´ 2π , 2π

3 3 . Then the
´1
2π 2π ´1
h of f isalso ´ i
range
? 3 , 3 . The domain of f
?
is the range of f over this interval, so
π 3 π 3
´ 3 + 2 , 3 + 2 .
S-21: Define h( x ) = f ( x ) ´ g( x ), and notice h( a) = f ( a) ´ g( a) ă 0 and

h(b) = f (b) ´ g(b) ą 0. Since h is the difference of two functions that are continuous over
[ a, b] and differentiable over ( a, b), also h is continuous over [ a, b] and differentiable over
( a, b). So, by the Mean Value Theorem, there exists some c P ( a, b) with
h(b) ´ h( a)
h1 ( c ) =
bá
Since ( a, b) is an interval, b ą a, so the denominator of the above expression is positive;

since h(b) ą 0 ą h( a), also the numerator of the above expression is positive. So,
h1 (c) ą 0 for some c P ( a, b). Since h1 (c) = f 1 (c) ´ g1 (c), we conclude f 1 (c) ą g1 (c) for
some c P ( a, b).
S-22: Since f ( x ) is differentiable over all real numbers, it is also continuous over all real
numbers. We claim that f ( x ) cannot have four or more distinct roots. For every two
distinct roots a ă b, Rolle’s Theorem tells us there is a c P ( a, b) such that f 1 (c) = 0: that
is, c is a root of f 1 . Since f 1 has only two distinct roots, f can have at most three distinct
roots.
roots of f 1 ( x )
c1 c2
a1 a2 a3
roots of f ( x )
S-23: We are asked to find the number of solutions to the equation x2 + 5x + 1 = ´ sin x.
The figure below contains the graphs of y = x2 + 5x + 1 and y = ´ sin x. The solutions to
x2 + 5x + 1 = ´ sin x are precisely the x’s where y = x2 + 5x + 1 and y = ´ sin x cross.
347
y
y = x2 + 5x + 1
y=1
x
y = ´ sin x
y = ´1
From the figure, it sure looks like there are two crossings. Since the function ´ sin x has
range [´1, 1], if the two functions cross, then also ´1 ď x2 + 5x + 1 ď 1. This portion of
the quadratic function is highlighted in blue in the figure. The x coordinates of the end
points of the blue arcs are found?by solving x2 + 5x + 1 = ˘1, i.e. x = 0, ´5, and (using
the quadratic equation) x = ´5˘2 17 .
We are now in a position to exploit the intuition that we have built using the above figure
to write a concise argument showing that f ( x ) has exactly two roots. Remember that, in
general, if we want to show that a function has n roots, we need to show that there exist
n distinct roots somewhere, and that there do not exist n + 1 distinct roots. This
argument is given below, in blue text.
• f ( x ) is continuous over all real numbers
• f ( x ) = 0 only when ´ sin x = x2 + 5x + 1, which only happens h when ? i
|x2 + 5x + 1| ď 1. Thus, f ( x ) only has roots in the intervals ´5, ´5´2 17 and
h ? i
´5+ 17
2 ,0 .
? ?
´5´ 17 ´5´ 17
• f (´5) = sin(´5) + 1 ą 0, and f 2 = sin 2 ´1 ă 0.
?
´5´ 17
So, by the IVT, f (c) = 0 for some c P ´5, 2 .
? ?
´5+ 17 ´5+ 17
• f (0) = 1 ą 0, and f 2 = sin ´1 ă 0.
2?
´5+ 17
So, by the IVT, f (c) = 0 for some c P 2 ,0 .
• f 1 ( x ) = cos x + 2x + 5. If f 1 ( x ) = 0, then 2x + 5 = ´ cos x, so |2x + 5| ď 1. So, the

only interval that can contain roots of f 1 ( x ) is [´3, ´2].
• Suppose f ( x ) has more than two roots. Then it has two roots in the interval
348
h ? i h ? i
´5´ 17 ´5+ 17
´5, 2 OR it has two roots in the interval 2 ,0 . Since f ( x ) is
1
differentiable
? ? Rolle’s Theorem tells us that f ( x ) has a root in
for allreal numbers,
´5, ´5´2 17 or in ´5+2 17 , 0 . However, since all roots of f 1 ( x ) are in the interval
? ?
[´3, ´2], and this interval shares no points with ´5, ´5´2 17 or ´5+2 17 , 0 , this
cannot be the case. Therefore f ( x ) does not have more than two roots.
roots of f 1 ( x )
´3 ´2
? ?
´5 ´5´ 17 ´5+ 17 0
2 2
roots of f ( x )
• Since f ( x ) has at least two roots, and not more than two roots, f ( x ) has exactly two
roots.
S-1: The derivative of e x is e x : taking derivatives leaves the function unchanged, even if
we do it 180 times. So f (180) = e x .
S-2: Since f 1 ( x ) ą 0 over ( a, b), we know from Corollary 2.13.11 that f ( x ) is increasing
over ( a, b), so (ii) holds. Since f 2 ( x ) ą 0, and f 2 ( x ) is the derivative of f 1 ( x ), by the same
reasoning we see that f 1 ( x ) is increasing. Since f 1 ( x ) is the rate at which f ( x ) is
increasing, that means that the rate at which f 1 ( x ) is increasing is itself increasing: this is,
(iv) holds (and not (iii)).
There is no reason to think (i) or (v) holds, but to be thorough we will give an example
showing that they do not need to be true. If f ( x ) = x2 ´ 10 and ( a, b) = (0, 1), then
f 1 ( x ) = 2x ą 0 over (0, 1), and f 2 ( x ) = 2 ą 0 everywhere, but f ( x ) ă 0 for all x P (0, 1),
so (i) does not hold. Also, f 3 ( x ) = 0 everywhere, so (v) does not hold either.
S-3: Every time we differentiate f ( x ), the constant out front gets multiplied by an
ever-decreasing constant, while the power decreases by one. As in Example 2.14.2,
d15 15 3
15
ax = a ¨ 15!. So, if a ¨ 15! = 3, then a = .
dx 15!
349
dy 11 11
S-4: The derivative is only at the point (1, 3): it is not constantly , so it is wrong
dx 4 4
11 d2 y
to differentiate the constant to find . Below is a correct solution.
4 dx2
´28x + 2y + 2xy1 + 2yy1 = 0
´28 + 6 + 2y1 + 6y1 = 0

11
y1 = at the point (1, 3)
4
Differentiating the equation ´28x + 2y + 2xy1 + 2yy1 = 0:
´28 + 2y1 + 2y1 + 2xy2 + 2y1 y1 + 2yy2 = 0

4y1 + 2(y1 )2 + 2xy2 + 2yy2 = 28
11
At the point (1, 3), y1 =
. Plugging in:
4
2
11 11
4 +2 + 2(1)y2 + 2(3)y2 = 28
4 4
15
y2 =
64
S-5:
f ( x ) = x log x ´ x
1
f 1 ( x ) = log x + x ¨ ´1
x
= log x
1
f 2 (x) =
x
S-6:
d 1
tarctan xu = 2
dx " * 1 +!x
d 1 d 2 ´1
)
= ( 1 + x )
dx 1 + x2 dx
= (´1)(1 + x2 )´2 (2x )
´2x
=
(1 + x 2 )2
350
S-7: We use implicit differentiation, twice.
2x + 2yy1 = 0
2 + (2y)y2 + (2y1 )y1 = 0
( y1 )2 + 1
y2 = ´
y
x
So, we need an expression for y1 . We use the equation 2x + 2yy1 = 0 to conclude y1 = ´ :
y
2
´ yx +1
y2 = ´
y
x2
y2
+1
=´
y
x 2 + y2
=´
y3
1
=´ 3
y
S-8: The question asks for s2 (1). We start our differentiation using the quotient rule:
1 et (t2 + 1) ´ et (2t)
s (t) =
( t2 + 1)2
et (t2 ´ 2t + 1)
=
( t2 + 1)2
Using the quotient rule again,

d d
2 (t2 + 1)2 dt tet (t2 ´ 2t + 1)u ´ et (t2 ´ 2t + 1) dt t( t2 + 1)2 u
s (t) =
( t2 + 1)4
(t2 + 1)2 ¨ et (2t ´ 2) + et (t2 ´ 2t + 1) ´ et (t2 ´ 2t + 1) ¨ 2(t2 + 1)(2t)

=
( t2 + 1)4
et (t2 + 1)2 (t2 ´ 1) ´ 4tet (t ´ 1)2 (t2 + 1)
=
( t2 + 1)4
s2 (1) = 0
S-9: We differentiate using the chain rule.
d 10x
tlog(5x2 ´ 12)u = 2
dx 5x ´ 12
351
Using the quotient rule:
d2
" *
2 d 10x
tlog(5x ´ 12)u =
dx2 dx 5x2 ´ 12
(5x2 ´ 12)(10) ´ 10x (10x )
=
(5x2 ´ 12)2
´10(5x2 + 12)
=
(5x2 ´ 12)2
Using the quotient rule one last time:
d3 d ´10(5x2 + 12)
" *
2
tlog(5x ´ 12)u =
dx3 dx (5x2 ´ 12)2
(5x2 ´ 12)2 (´10)(10x ) + 10(5x2 + 12)(2)(5x2 ´ 12)(10x )
=
(5x2 ´ 12)4
(5x2 ´ 12)(´100x ) + (200x )(5x2 + 12)
=
(5x2 ´ 12)3
100x (´5x2 + 12 + 10x2 + 24)
=
(5x2 ´ 12)3
100x (5x2 + 36)
=
(5x2 ´ 12)3
S-10: The velocity of the particle is given by h1 (t) = sin t. Note 0 ă 1 ă π, so

h1 (1) ą 0–the particle is rising (moving in the positive direction, in this case “up”). The
acceleration of the particle is h2 (t) = cos t. Since 0 ă 1 ă π2 , h2 (t) ą 0, so h1 (t) is
increasing: the particle is moving up, and it’s doing so at an increasing rate. So, the
particle is speeding up.
S-11: For this problem, remember that velocity has a sign indicating direction, while
speed does not.
The velocity of the particle is given by h1 (t) = 3t2 ´ 2t ´ 5. At t = 1, the velocity of the
particle is ´4, so the particle is moving downwards with a speed of 4 units per second.
The acceleration of the particle is h2 (t) = 6t ´ 2, so when t = 1, the acceleration is
(positive) 4 units per second per second. That means the velocity (currently ´4 units per
second) is becoming a bigger number–since the velocity is negative, a bigger number is
closer to zero, so the speed of the particle is getting smaller. (For instance, a velocity of
´3 represents a slower motion than a velocity of ´4.) So, the particle is slowing down at
t = 1.
S-12:
x2 + x + y = sin( xy)
352
dy
We differentiate implicitly. For ease of notation, we write y1 for .
dx
2x + 1 + y1 = cos( xy)(y + xy1 )
We’re interested in y2 , so we implicitly differentiate again.
2 + y2 = ´ sin( xy)(y + xy1 )2 + cos( xy)(2y1 + xy2 )
We want to know what y2 is when x = y = 0. Plugging these in yields the following:
2 + y2 = 2y1
So, we need to know what y1 is when x = y = 0. We can get this from the equation
2x + 1 + y1 = cos( xy)(y + xy1 ), which becomes 1 + y1 = 0 when x = y = 0. So, at the
origin, y1 = ´1, and
2 + y2 = 2(´1)
y2 = ´4
Remark: a common mistake is to stop at the equation 2x + 1 + y1 = cos( xy)(y + xy1 ),

d
plug in x = y = 0, find y1 = ´1, and decide y2 = t´1u = 0. This is due to a slight
dx
1
sloppiness in the usual notation. When we wrote ˇy = 1, what we meant is that at the
dy dy ˇˇ
point (0, 0), = ´1. More properly written: = ´1. This is not the same as
dx dx ˇx=0, y=0
saying y1 = 1 everywhere (in which case, indeed, y2 would be 0 everywhere).
S-13: For (a) and (b), notice the following:
d
sin x = cos x
dx
d
cos x = ´ sin x
dx
d
t´ sin xu = ´ cos x
dx
d
t´ cos xu = sin x
dx
d
sin x = cos x
dx
353
The fourth derivative is sin x is sin x, and the fourth derivative of cos x is cos x, so (a) and
(b) are true.
d
tan x = sec2 x
dx
d
sec2 x = 2 sec x (sec x tan x ) = 2 sec2 x tan x
dx
d
t2 sec2 x tan xu = (4 sec x ¨ sec x tan x ) tan x + 2 sec2 x sec2 x
dx
= 4 sec2 x tan2 x + 2 sec4 x
d
t4 sec2 x tan2 x + 2 sec4 xu = (8 sec x ¨ sec x tan x ) tan2 x + 4 sec2 x (2 tan x ¨ sec2 x )
dx
+ 8 sec3 x ¨ sec x tan x
= 8 sec2 x tan3 x + 16 sec4 x tan x
d4
So, tan x = 8 sec2 x tan3 x + 16 sec4 x tan x. It certainly seems like this is not the same
dx4
as tan x, but remember that sometimes trig identities can fool you: tan2 x + 1 = sec2 x,
and so on. So, to be absolutely sure that these are not equal, we need to find a value of x
so that the output of one is not the same as the output of the other. When x = π4 :
? 2 ? 4
8 sec2 x tan3 x + 16 sec4 x tan x = 8 2 (1)3 + 16 2 (1) = 80 ‰ 1 = tan x.
So, (c) is false.
S-14: Since f 1 ( x ) ă 0, we need a decreasing function. This only applies to (ii), (iii), and
(v). Since f 2 ( x ) ą 0, that means f 1 ( x ) is increasing, so the slope of the function must be
increasing. In (v), the slope is constant, so f 2 ( x ) = 0–therefore, it’s not (v). In (iii), the
slope is decreasing, because near a the curve is quite flat ( f 1 ( x ) near zero) but near b the
curve is very steeply decreasing ( f 1 ( x ) is a large negative number), so (iii) has a negative
second derivative. By contrast, in (ii), the line starts out as steeply decreasing ( f 1 ( x ) is a
strongly negative number) and becomes flatter and flatter ( f 1 ( x ) nears 0), so f 1 ( x ) is
increasing–in other words, f 2 ( x ) ą 0. So, (ii) is the only curve that has f 1 ( x ) ă 0 and
f 2 ( x ) ą 0.
S-15: We differentiate a few time to find the pattern.

d x
t2 u = 2x log 2
dx
d2 x
t2 u = 2x log 2 ¨ log 2 = 2x (log 2)2
dx2
d3 x
t2 u = 2x (log 2)2 ¨ log 2 = 2x (log 2)3
dx3
Every time we differentiate, we multiply the original function by another factor of log 2.
So, the nth derivative is given by:
dn x
t2 u = 2x (log 2)n
dx n
354
S-16: We differentiate using the power rule.
df
= 3ax2 + 2bx + c
dx
d2 f
= 6ax + 2b
dx2
d3 f
= 6a
dx3
d4 f
=0
dx4
In the above work, remember that a, b, c, and d are all constants. Since they are nonzero
d3 f
constants, = 6a ‰ 0. So, the fourth derivative is the first derivative to be identically
dx3
zero: n = 4.
S-17: (a) Using the chain rule for f ( x ):

2
f 1 ( x ) = (1 + 2x )e x+ x
2 2 2
f 2 ( x ) = (1 + 2x )(1 + 2x )e x+ x + (2)e x+ x = (4x2 + 4x + 3)e x+ x
h1 ( x ) = 1 + 3x
h2 ( x ) = 3
(b) f (0) = h(0) = 1; f 1 (0) = h1 (0) = 1; f 2 (0) = h2 (0) = 3

(c) f and h “start at the same place,” since f (0) = h(0). If it were clear that f 1 ( x ) were
greater than h1 ( x ) for x ą 0, then we would know that f grows faster than h, so we could
conclude that f ( x ) ą h( x ), as desired. Unfortunately, it is not obvious whether
2
(1 + 2x )e x+x is always greater than 1 + 3x for positive x. So, we look to the second
2 2
derivative. f 1 (0) = h1 (0), and f 2 ( x ) = (4x2 + 4x + 3)e x+ x ą 3e x+ x ą 3 = h2 ( x ) when
x ą 0. Since f 1 (0) = h1 (0), and since f 1 grows faster than h1 for positive x, we conclude
f 1 ( x ) ą h1 ( x ) for all positive x. Now we can conclude that (since f (0) = h(0) and f
grows faster than h when x ą 0) also f ( x ) ą h( x ) for all positive x.
S-18:
(a) We differentiate implicitly.
x3 y( x ) + y( x )3 = 10x
3x2 y( x ) + x3 y1 ( x ) + 3y( x )2 y1 ( x ) = 10
Subbing in x = 1 and y(1) = 2 gives
(3)(1)(2) + (1)y1 (1) + (3)(4)y1 (1) = 10
13y1 (1) = 4
4
y1 (1) =
13
355
4
(b) From part (a), the slope of the curve at x = 1, y = 2 is , so the curve is increasing,
13

4
but fairly slowly. The angle of the tangent line is tan´1 13 « 17˝ . We are also told
that y2 (1) ă 0. So the slope of the curve is decreasing as x passes through 1. That is,
the line is more steeply increasing to the left of x = 1, and its slope is decreasing
(getting less sleep, then possibly the slope even becomes negative) as we move past
x = 1.
y
tangent
2
curve
1 x
S-19: (a) Using the product rule,
g2 ( x ) = [ f 1 ( x ) + f 2 ( x )]e x + [ f ( x ) + f 1 ( x )]e x = [ f ( x ) + 2 f 1 ( x ) + f 2 ( x )]e x
(b) Using the product rule and our answer from (a),
g3 ( x ) = [ f 1 ( x ) + 2 f 2 ( x ) + f 3 ( x )]e x + [ f ( x ) + 2 f 1 ( x ) + f 2 ( x )]e x
= [ f ( x ) + 3 f 1 ( x ) + 3 f 2 ( x ) + f 3 ( x )]e x
(c) We notice that the coefficients of the derivatives of f correspond to the entries in the
rows of Pascal’s Triangle.
1
1 1
1 2 1
1 3 3 1
1 4 6 4 1
Pascal’s Triangle
• In the first derivative of g, the coefficients of f and f 1 correspond to the entries in

the second row of Pascal’s Triangle.
• In the second derivative of g, the coefficients of f , f 1 , and f 2 correspond to the
entries in the third row of Pascal’s Triangle.
• In the third derivative of g, the coefficients of f , f 1 , f 2 , and f 3 correspond to the
entries in the fourth row of Pascal’s Triangle.
• We guess that, in the fourth derivative of g, the coefficients of f , f 1 , f 2 , f 3 , and f (4)
will correspond to the entries in the fifth row of Pascal’s Triangle.
That is, we guess
g(4) ( x ) = [ f ( x ) + 4 f 1 ( x ) + 6 f 2 ( x ) + 4 f 3 ( x ) + f (4) ( x )]e x
356
This is verified by differentiating our answer from (a) using the product rule:
g3 ( x ) = [ f ( x ) + 3 f 1 ( x ) + 3 f 2 ( x ) + f 3 ( x )]e x
g(4) ( x ) = [ f 1 ( x ) + 3 f 2 ( x ) + 3 f 3 ( x ) + f (4) ( x )]e x + [ f ( x ) + 3 f 1 ( x ) + 3 f 2 ( x ) + f 3 ( x )]e x
= [ f ( x ) + 4 f 1 ( x ) + 6 f 2 ( x ) + 4 f 3 ( x ) + f (4) ( x )]e x .
S-20: Since f ( x ) is differentiable over all real numbers, it is also continuous over all real
numbers. Similarly, f 1 ( x ) is differentiable over all real numbers, so it is also continuous
over all real numbers, and so on for the first n derivatives of f ( x ).
Rolle’s Theorem tells us that if a and b are distinct roots of a function g, then g1 ( x ) = 0 for
some c in ( a, b). That is, g1 has a root strictly between a and b. Expanding this idea, if g
has m + 1 distinct roots, then g1 must have at least m distinct roots, as in the sketch below.
roots of g1 ( x )
c1 c2 cm
a1 a2 a3 a m +1
roots of g( x )
So, if f (n) ( x ) has only m roots, then f (n´1) ( x ) has at most m + 1 roots. Similarly, since
f (n´1) ( x ) has at most m + 1 roots, f (n´2) ( x ) has at most m + 2 roots. Continuing in this
way, we see f ( x ) = f (nń) ( x ) has at most m + n distinct roots.
S-21:
• Let’s begin by noticing that the domain of f ( x ) is (´1, 8).
• By inspection, f (0) = 0, so f ( x ) has at least one root.
• If x P (´1, 0), then ( x + 1) is positive, log( x + 1) is negative, sin( x ) is negative, and
´x2 is negative. Therefore, if x ă 0 is in the domain of f , then f ( x ) ă 0. So, f ( x ) has
no negative roots. We focus our attention on the case x ą 0.
• f 1 ( x ) = 1 ´ 2x + log( x + 1) + cos x. We would like to know how many positive
roots f 1 ( x ) has, but it isn’t obvious. So, let’s differentiate again.
1 1
• f 2 ( x ) = ´2 + x+ 2
1 ´ sin x. When x ą 0, x +1 ă 1, so f ( x ) ă ´1 ´ sin( x ) ď 0, so
f 2 ( x ) has no positive roots. Since f 1 ( x ) is continuous and differentiable over (0, 8),
357
and since f 2 ( x ) ‰ 0 for all x P (0, 8), by Rolle’s Theorem, f 1 ( x ) has at most one root
in [0, 8).
• Since f ( x ) is continuous and differentiable over [0, 8), and f 1 ( x ) has at most one
root in (0, 8), by Rolle’s Theorem f ( x ) has at most two distinct roots in [0, 8).
(Otherwise, f ( a) = f (b) = f (c) = 0 for some values 0 ď a ă b ă c, so
f 1 (d) = f 1 (e) = 0 for some d P ( a, b) and some e P (b, c), but since f 1 ( x ) has at most
one root, this is impossible.)
• We know f (0) = 0, so the remaining question is whether or not f ( x ) has a second
root (which would have to be positive). As usual, we can show another root exists
using the intermediate value theorem. We see that for large values of x, f ( x ) is
negative, for example:
f (4) = 5 log 5 + sin(4) ´ (4)2 ă 5 log(e2 ) + 1 ´ 16 = 11 ´ 16 ă 0
For positive values of x closer to zero, we hope to find a positive value of f ( x ).
However, it’s quite difficult to get a number c that obviously gives f (c) ą 0. It
suffices to observe that f (0) = 0 and f 1 (0) = 2 ą 0. From the definition of the
derivative, we can conclude f ( x ) ą 0 for some x ą 0. (If it is not true that f ( x ) ą 0
for some x ą 0, then f ( x ) ď 0 for all x ą 0. The definition of the derivative tells us
f ( h ) ´ f (0) f (h)
that [since f 1 (0) exists] f 1 (0) = lim = lim ; the denominator is
hÑ0+ h hÑ0+ h
positive, so if the numerator were always less than or equal to zero, the limit would
be less than or equal to zero as well. However, the derivative is positive, so f ( x ) ą 0
for some x ą 0.) Therefore, f ( x ) has a second root, so f ( x ) has precisely two roots.
" In order to make f ( x ) a little more tractable, let’s change the format. Since
S-22: (a)
x xě0
|x| = , then:
´x x ă 0
´x2 x ă 0
"
f (x) =
x2 x ě 0.
Now, we turn to the definition of the derivative to figure out whether f 1 (0) exists.
f (0 + h ) ´ f (0) f (h) ´ 0 f (h)
f 1 (0) = lim = lim = lim if it exists.
Since f looks different to the left and right of 0, in order to evaluate this limit, we look at
the corresponding one-sided limits. Note that when h approaches 0 from the right, h ą 0
so f (h) = h2 . By contrast, when h approaches 0 from the left, h ă 0 so f (h) = ´h2 .
f (h) h2
lim = lim = lim h = 0
hÑ0+ h hÑ0+ h hÑ0+
f (h) ´h 2
lim = lim = lim ´h = 0
hÑ0´ h hÑ0´ h hÑ0´
Since both one-sided limits exist and are equal to 0,

f (0 + h ) ´ f (0)
lim =0
hÑ0 h
358
and so f is differentiable at x = 0 and f 1 (0) = 0.
(b) From (a), f 1 (0) = 0 and
´x2 x ă 0
"
f (x) =
x2 x ě 0.
So, "
1 ´2x x ă 0
f (x) =
2x x ě 0.
Then, we know the second derivative of f everywhere except at x = 0:

$
& ´2 x ă 0
2
f (x) = ?? x = 0
2 x ą 0.
%
So, whenever x ‰ 0, f 2 ( x ) exists. To investigate the differentiability of f 1 ( x ) when x = 0,

again we turn to the definition of a derivative. If
f 1 (0 + h ) ´ f 1 (0)
lim
hÑ0 h
exists, then f 2 (0) exists.
f 1 (0 + h ) ´ f 1 (0) f 1 (h) ´ 0 f 1 (h)

lim = lim = lim
Since f (h) behaves differently when h is greater than or less than zero, we look at the
one-sided limits.
f 1 (h) 2h
lim = lim =2
hÑ0 + h hÑ0 h
+
f 1 (h) ´2h
lim = lim = ´2
hÑ0 ´ h hÑ0´ h
Since the one-sided limits do not agree,
f 1 (0 + h ) ´ f 1 (0)
lim = DNE
hÑ0 h
So, f 2 (0) does not exist. Now we have a complete picture of f 2 ( x ):

$
& ´2 xă0
2
f (x) = DNE x = 0
2 x ą 0.
%
359
S-1: False. The acceleration of the ball is given by h2 (t) = ´9.8. This is constant
throughout its trajectory (and is due to gravity).
Remark: the velocity of the ball at t = 2 is zero, since h1 (2) = ´9.8(2) + 19.6 = 0, but the
velocity is only zero for an instant. Since the velocity is changing, the acceleration is
nonzero.
S-2: The acceleration is constant, which means the rate of change of the velocity is
constant. So, since it took 10 seconds for the velocity to increase by 1 metre per second
(from 1 ms to 2 ms ), then it always takes 10 seconds for the velocity to increase by 1 metre
per second.
So, it takes 10 seconds to accelerate from 2 ms to 3 ms . To accelerate from 3 ms to 13 ms (that

is, to change its velocity by 10 metres per second), it takes 10 ˆ 10 = 100 seconds.
S-3: Let v( a) = s1 ( a) be the velocity of the particle. If s2 ( a) ą 0, then v1 ( a) ą 0 — so the

velocity of the particle is increasing. However, that does not mean that its speed (the
absolute value of velocity) is increasing as well. For example, if a velocity is increasing
from ´4 kph to ´3 kph, the speed is decreasing from 4 kph to 3 kph. So, the statement is
false in general.
Contrast this to Question 4.
S-4: Since s1 ( a) ą 0, |s1 ( a)| = s1 ( a): that is, the speed and velocity of the particle are the
same. (This means the particle is moving in the positive direction.) If s2 ( a) ą 0, then the
velocity (and hence speed) of the particle is increasing. So, the statement is true.
S-5: From Example 3.1.2, we know that an object falling from rest on the Earth is subject
to the acceleration due to gravity, 9.8 m
s2
. So, if h(t) is the height of the flower pot t
seconds after it rolls out the window, then h2 (t) = ´9.8. (We make the acceleration
negative, since the measure “height” has “up” as the positive direction, while gravity
pulls the pot in the negative direction, “down.”)
Then h1 (t) is a function whose derivative is the constant ´9.8 and with h1 (0) = 0 (since
the object fell, instead of being thrown up or down), so h1 (t) = ´9.8t.
What we want to know is h1 (t) at the time t when the pot hits the ground. We don’t
know yet exactly what time that happens, so we go a little farther and find an expression
for h(t). The function h(t) has derivative ´9.8t and h(0) = 10, so (again following the
ideas in Example 3.1.2)
´9.8 2
h(t) = t + 10
2
Now, we can find the time when the pot hits the ground: it is the time when h(t) = 0
360
(and t ą 0).
´9.8 2
0= t + 10
2
9.8 2
t = 10
2
20
t2 =
9.8
c
20
t=+ « 1.4 sec
9.8
The velocity of the pot at this time is
c ! c !
1 20 20 ? m
h = ´9.8 = ´ 20 ¨ 9.8 = ´14
9.8 9.8 s
So, the pot is falling at 14 metres per second, just as it hits the ground.
S-6: (a)
• Let s(t) be the distance the stone has fallen t seconds after dropping it. Since the
acceleration due to gravity is 9.8 m
s2
, s2 (t) = 9.8. (We don’t make this negative,
because s(t) measures how far the stone has fallen, which means the positive
direction in our coordinate system is “down,” which is exactly the way gravity is
pulling.)
• Then s1 (t) has a constant derivative of 9.8, so s1 (t) = 9.8t + c for some constant c.
Notice s1 (0) = c, so c is the velocity of the stone at the very instant you dropped it,
which is zero. Therefore, s1 (t) = 9.8t.
• So, s(t) is a function with derivative 9.8t. It’s not too hard to figure out by guessing
and checking that s(t) = 9.8 2
2 t + d for some constant d. Notice s (0) = d, so d is the
distance the rock has travelled at the instant you dropped it, which is zero. So,
s(t) = 9.8 2 2
2 t = 4.9t .
Remark: this is exactly the formula found in Example 3.1.2. You may, in general,
use that formula without proof, but you need to know where it comes from and be
able to apply it in other circumstances where it might be slightly different–like part
(b) below.
• The rock falls for x seconds, so the distance fallen is
4.9x2
Remark: this is a decent (if imperfect) way to figure out how deep a well is, or how
tall a cliff is, when you’re out and about. Drop a rock, square the time, multiply by
5.
(b) We’ll go through a similar process as before.
361
Again, let s(t) be the distance the rock has fallen t seconds after it is let go. Then
s2 (t) = 9.8, so s1 (t) = 9.8t + c. In this case, since the initial speed of the rock is 1 metre
per second, 1 = s1 (0) = c, so s1 (t) = 9.8t + 1.
9.8 2
Then, s(t) is a function whose derivative is 9.8t + 1, so s(t) = 2 t + t + d for some
constant d. Since 0 = s(0) = d, we see s(t) = 4.9t2 + t.
So, if the rock falls for x seconds, the distance fallen is
4.9x2 + x
Remark: This means there is an error of x metres in your estimation of the depth of the
well.
S-7: Let s(t) be the distance your keys have travelled since they left your hand. The rate
at which they are travelling, s1 (t), is decreasing by 0.25 metres per second. That is,
s2 (t) = ´0.25. Therefore, s1 (t) = ´0.25t + c for some constant c. Since c = s1 (0) = 2, we
see
s1 (t) = 2 ´ 0.25t
Then s(t) has 2 ´ 0.25t as its derivative, so s(t) = 2t ´ 18 t2 + d for some constant d. At
time t = 0, the keys have not yet gone anywhere, so 0 = s(0) = d. Therefore,
1
s(t) = 2t ´ t2
8
The keys reach your friend when s(t) = 2 and t ą 0. That is:
1
2 = 2t ´ t2
8
1 2
0 = t ´ 2t + 2
8 ?
t = 8˘4 3
We need to figure out which of these values of t is really the time when the keys reach
your friend. The keys travel this way from t = 0 to the time they reach your friend.
(Then s(t) no longer describes their
? motion.) So, we need to find the first value of t that is
positive with s(t) = 2. Since 8 ´ 4 3 ą 0, this is the first time s(t) = 2 and t ą 0. So, the
keys take ?
8 ´ 4 3 « 1 second
to reach your friend.
S-8: We proceed with the technique of Example 3.1.3 in mind.

Let v(t) be the velocity (in kph) of the car at time t, where t is measured in hours and
t = 0 is the instant the brakes are applied. Then v(0) = 100 and v1 (t) = ´50 000. Since
v1 (t) is constant, v(t) is a line with slope ´50 000 and intercept (0, 100), so
v(t) = 100 ´ 50 000t
362
100 1
The car comes to a complete stop when v(t) = 0, which occurs at t = 50 000 = 500 hours.
This is a confusing measure, so we convert it to seconds:

1 3600 sec
hrs = 7.2 sec
500 1hr
S-9: Suppose the deceleration provided by the brakes is d km2 . Then if v(t) is the velocity
hr
of the car, v(t) = 120 ´ dt (at t = 0, the velocity is 120, and it decreases by d kph per
hour). The car stops when 0 = v(t), so t = 120 d hours.
Let s(t) be the distance the car has travelled t hours after applying the brakes. Then
s1 (t) = v(t), so s(t) = 120t ´ 2d t2 + c for some constant c. Since 0 = s(0) = c,
d
s(t) = 120t ´ t2
2
1
The car needs to stop in 100 metres, which is 10 kilometres. We already found that the
stopping time is t = 120
d . So:

1 120
=s
10 d
d 120 2

1 120
= 120 ´
10 d 2 d
Multiplying both sides by d:
d 1202
= 1202 ´
10 2
2
d = 5 ¨ 120
So, the brakes need to apply 72 000 kph per hour of deceleration.
S-10: Since your deceleration is constant, your speed decreases smoothly from 100 kph to
0 kph. So, one second before your stop, you only have 17 of our speed left: you’re going
100
7 kph.
A less direct way to solve this problem is to note that v(t) = 100 ´ dt is the velocity of car
7
t hours after braking, if d is its deceleration. Since it stops in 7 seconds (or 3600 hours),
7 7 360000

0 = v 3600 = 100 ´ 3600 d, so d = 7 . Then

6 360000 6 6 100
v = 100 ´ = 100 ´ ¨ 100 = kph
3600 7 3600 7 7
363
S-11: If the acceleration was constant, then it was
17500 mph miles
8.5
« 123500
60 hr hr2
So, the velocity t hours from liftoff is
v(t) = 123500t
Therefore, the position of the shuttle t hours from liftoff (taking s(0) = 0 to be its initial
position) is
123500 2
s(t) = t = 61750t2
2
8.5
So, after 60 hours, the shuttle has travelled
2
8.5 8.5
s = (61750) « 1240 miles
60 60
or a little less than 2000 kilometres.
S-12: We know that the acceleration of the ball will be constant. If the height of the ball is
given by h(t) while it is in the air, h2 (t) = ´9.8. (The negative indicates that the velocity is
decreasing: the ball starts at its largest velocity, moving in the positive direction, then the
velocity decreases to zero and then to a negative number as the ball falls.) As in
Example 3.1.2, we need a function h(t) with h2 (t) = ´9.8. Since this is a constant, h1 (t) is
a line with slope ´9.8, so it has the form
h1 (t) = ´9.8t + a
for some constant a. Notice when t = 0, h1 (0) = a, so in fact a is the initial velocity of the
ball–the quantity we want to solve for.
Again, as in Example 3.1.2, we need a function h(t) with h1 (t) = ´9.8t + a. Such a
function must have the form
h(t) = ´4.9t2 + at + b
for some constant b. You can find this by guessing and checking, or simply remember it
from the text. (In Section 4.1, you’ll learn more about figuring out which functions have a
particular derivative.) Notice when t = 0, h(0) = b, so b is the initial height of the
baseball, which is 0.
So, h(t) = ´4.9t2 + at = t(´4.9t + a). The baseball is at height zero when it is pitched
(t = 0) and when it hits the ground (which we want to be t = 10). So, we want
(´4.9)(10) + a = 0. That is, a = 49. So, the initial pitch should be at 49 metres per second.
Incidentally, this is on par with the fastest pitch in baseball, as recorded by Guiness
World Records:
https://www.guinnessworldrecords.com/world-records/
fastest-baseball-pitch-(male)
364
S-13: The acceleration of a falling object due to gravity is 9.8 metres per second squared. So, the
object’s velocity t seconds after being dropped is
m
v(t) = 9.8t
s
We want v(t) to be the speed of the peregrine’s dive, so we should convert that to metres per
second:
km 1000 m 1 hr 1625 m
325 ¨ =
hr 1 km 3600 sec 18 s
The stone will reach this velocity when
1625 1625
9.8t = ñ t=
18 18(9.8)
What is left to figure out is how far the stone will fall in this time. The position of the stone s(t)
has derivative 9.8t, so
s(t) = 4.9t2
1625
if we take s(0) = 0. So, if the stone falls for 18(9.8)
seconds, in that time it travels
2
1625 1625
s = 4.9 « 416 m
18(9.8) 18(9.8)
So, you would have to drop a stone from about 416 metres for it to fall as fast as the falcon.
m
S-14: Since gravity alone brings your cannon ball down, its acceleration is a constant ´9.8 s2
. So,
v(t) = v0 ´ 9.8t and thus its height is given by s(t) = v0 t ´ 4.9t2 (if we set s(0) = 0).
We want to know what value of v0 makes the maximum height 100 metres. The maximum height
v0
is reached when v(t) = 0, which is at time t = 9.8 . So, we solve:
v
0
100 = s
9.8
v v 2
0 0
100 = v0 ´ 4.9
9.8 9.8
1 4.9
100 = ´ v20
9.8 9.82
1
100 = v2
2 ¨ 9.8 0
v20 = 1960
? m
v0 = 1960 « 44
s
where we choose the positive square root because v0 must be positive for the cannon ball to get
off the ground.
S-15: The derivative of acceleration

3
is constant, so the acceleration a(t) has the form
3
mt + b. We
know a(0) = ´50 000 and a 3600 = ´60 000 (where we note that 3 seconds is 3600 hours). So, the
slope of a(t) is ´60 0003+50 000 = ´12 000 000, which leads us to
3600
a(t) = ´50 000 ´ (12 000 000)t
365
where t is measured in hours.
Since v1 (t) = a(t) = ´50 000 ´ (12 000 000)t, we see
´12 000 000 2

v(t) = t ´ 50 000t + c = ´6 000 000t2 ´ 50 000t + c
2
for some constant c. Since 120 = v(0) = c:
v(t) = ´6 000 000t2 ´ 50 000t + 120
Then after three seconds of braking,

2
3 3 3
v = ´6 000 000 ´ 50 000 + 120
3600 3600 3600
25 125
=´ ´ + 120
6 3
« 74.2 kph
Remark: When acceleration is constant, the position function is a quadratic function, but we don’t
want you to get the idea that position functions are always quadratic functions–in the example
you just did, it was the velocity function that was quadratic. Position, velocity, and acceleration
functions don’t have to be polynomial at all–it’s only in this section, where we’re dealing with the
simplest cases, that they seem that way.
S-16: Different forces are acting on you (1) after you jump but before you land on the trampoline,
and (2) while you are falling into the trampoline. In both instances, the acceleration is constant, so
both height functions are quadratic, of the form 2a t2 + vt + h, where a is the acceleration, v is the
velocity when t = 0, and h is the initial height.
• Let’s consider (1) first, the time during your jump before your feet touch the trampoline.
Let t = 0 be the moment you jump, and let the rim of the trampoline be height 0. Then,
since your initial velocity was (positive) 1 meter per second, your height is given by

´9.8 2 ´9.8
h1 ( t ) = t +t = t t+1
2 2
Notice that, because your acceleration is working against your positive velocity, it has a
negative sign.
• We’ll need to know your velocity when your feet first touch the trampoline on your fall.
The time your feet first first touch the trampoline after your jump is precisely when
2
h1 (t) = 0 and t ą 0. That is, when t = 9.8 . Now, since h1 (t) = ´9.8t + 1,
2 2

h1 9.8 = ´9.8 9.8 + 1 = ´1. So, you are descending at a rate of 1 metre per second at the
instant your feet touch the trampoline.
Remark: it is not only coincidence that this was your initial speed. Think about the
symmetries of parabolas, and conservation of energy.
• Now we need to think about your height as the trampoline is slowing your fall. One thing
to remember about our general equation 2a t2 + vt + h is that v is the velocity when t = 0.
2
But, you don’t hit the trampoline at t = 0, you hit it at t = 9.8 . In order to keep things
366
simple, let’s use a different time scale for this second part of your journey. Let’s let h2 ( T ) be
your height at time T, from the moment your feet touch the trampoline skin (T = 0) to the
bottom of your fall. Now, we can use the fact that your initial velocity is ´1 metres per
second (negative, since your height is decreasing) and your acceleration is 4.9 metres per
second per second (positive, since your velocity is increasing from a negative number to
zero):
4.9 2
h2 ( T ) = T ´T
2
where still the height of the rim of the trampoline is taken to be zero.
Remark: if it seems very confusing that your free-falling acceleration is negative, while
your acceleration in the trampoline is positive, remember that gravity is pushing you
down, but the trampoline is pushing you up.
• How long were you falling in the trampoline? The equation h2 ( T ) tells you your height
only as long as the trampoline is slowing your fall. You reach the bottom of your fall when
your velocity is zero.
h12 ( T ) = 4.9T ´ 1
1 1
so you reach the bottom of your fall at T = 4.9 . Be careful: this is 4.9 seconds after you entered
the trampoline, not after the peak of your fall, or after you jumped.
• The last piece of the puzzle is how long it took you to fall from the peak of your jump to the
surface of the trampoline. We know the equation of your motion during that time:
h1 (t) = ´9.8 2
2 t + t. You reached the peak when your velocity was zero:
1
h11 (t) = ´9.8t + 1 = 0 ñ t=
9.8
1 2
So, you fell from your peak at t = 9.8 and reached the level of the trampoline rim at t = 1.98 ,
1
which means the fall took 9.8 seconds.
Remark: by the symmetry mentioned early, the time it took to fall from the peak of your
jump to the surface of the trampoline is the same as one-half the time from the moment you
jumped off the rim to the moment you’re back on the surface of the trampoline.
1 1
• So, your time falling from the peak of your jump to its bottom was 9.8 + 4.9 « 0.3 seconds.
S-17: Let v(t) be the velocity of the object. From the given information:
• v(0) is some value, call it v0 ,
• v(1) = 2v0 (since the speed doubled in the first second),
• v(2) = 2(2)v0 (since the speed doubled in the second second),
• v(3) = 2(2)(2)v0 , and so on.
So, for general t:

v ( t ) = 2t v (0 )
d x
To find its acceleration, we simply differentiate. Recall t2 u = 2x log 2, where log denotes
dx
logarithm base e.
a(t) = 2t v0 log 2
367
Remark: we can also write a(t) = v(t) log 2. The acceleration doubles every second as well.
S-1: We have an equation relating P and Q:
P = Q3
We differentiate implicitly with respect to a third variable, t:
dP dQ
= 3Q2 ¨
dt dt
dP dQ
If we know two of the three quantities , Q, and , then we can find the third. Therefore, ii is
dt dt
a question we can solve. If we know P, then we also know Q (it’s just the cube root of P), so also
dP dQ
we can solve iv. However, if we know neither P nor Q, then we can’t find based only off ,
dt dt
dQ dP
and we can’t find based only off . So we can’t solve i or iii.
dt dt
S-2: Suppose that at time t, the point is at ( x (t), y(t)). Then x (t)2 + y(t)?
2 = 1 so that
?
1 1
2x (t) x (t) + 2y(t)y (t) = 0. We are told that at some time t0 , x (t0 ) = 2/ 5, y(t0 ) = 1/ 5 and
y1 (t0 ) = 3. Then
2x (t0 ) x1 (t0 ) + 2y(t0 )y1 (t0 ) = 0 ñ

2 1 1
2 ? x (t) + 2 ? (3) = 0 ñ
5 5
3
x 1 ( t0 ) = ´
2
S-3: The instantaneous percentage rate of change for R is
R1 ( PQ)1
100 = 100 R=PQ
R PQ
P1 Q + PQ1
= 100 product rule
PQ
1
Q1

P
= 100 + simplify
P Q
= 100[0.08 ´ 0.02] = 6%
P1 Q ´ PQ1 5 ˆ 5 ´ 25 ˆ 1
S-4: (a) By the quotient rule, F1 = 2
. At the moment in question, F1 = = 0.
Q 52
(b) We are told that, at the second moment in time, P1 = 0.1P and Q1 = ´0.05Q
1 1
(or equivalently 100 PP = 10 and 100 QQ = ´5). Substituting in these values:
368
P1 Q ´ PQ1
F1 =
Q2
0.1PQ ´ P(´0.05Q)
=
Q2
0.15PQ
=
Q2
P
= 0.15
Q
= 0.15F ùñ
F1 = 0.15F
1
or 100 FF = 15%. That is, the instantaneous percentage rate of change of F is 15%.
S-5:
• The distance z(t) between the particles at any moment in time is
z2 ( t ) = x ( t )2 + y ( t )2 ,
where x (t) is the position on the x-axis of the particle A at time t (measured in seconds) and
y(t) is the position on the y-axis of the particle B at the same time t.
• We differentiate the above equation with respect to t and get
2z ¨ z1 = 2x ¨ x1 + 2y ¨ y1 ,
• We are told that x1 = ´2 and y1 = ´3. (The values are negative because x and y are
decreasing.) It will take 3 seconds for particle A to reach x = 4, and in this time particle B
will reach y = 3.
a ?
• At this point z = x2 + y2 = 32 + 42 = 5.
• Hence
10z1 = 8 ¨ (´2) + 6 ¨ (´3) = ´34

34 17
z1 = ´ = ´ units per second.
10 5
S-6:
• We compute the distance z(t) between the two particles after t seconds as
z2 (t) = 32 + (y A (t) ´ y B (t))2 ,
where y A (t) and y B (t) are the y-coordinates of particles A and B after t seconds, and the
horizontal distance between the two particles is always 3 units.
369
• We are told the distance between the particles is 5 units, this happens when
(y A ´ y B )2 = 52 ´ 32 = 16
y A ´ yB = 4
That is, when the difference in y-coordinates is 4. This happens when t = 4.
• We differentiate the distance equation (from the first bullet point) with respect to t and get
2z ¨ z1 = 2(y A 1 ´ y B 1 )(y A ´ y B ),
• We know that (y A ´ y B ) = 4, and we are told that z = 5, y1A = 3, and y1B = 2. Hence
10z1 (4) = 2 ˆ 1 ˆ 4 = 8
• Therefore
8 4
z1 (4) = = units per second.
10 5
S-7:
H x(t) B
15 mph
y(t) z(t)
20 mph
As in the above figure, let x (t) be the distance between H (Hawaii) and ship B, and y(t) be the
distance between H and ship A, and z(t) be the distance between ships A and B, all at time t. Then
x ( t )2 + y ( t )2 = z ( t )2
Differentiating with respect to t,
2x (t) x1 (t) + 2y(t)y1 (t) = 2z(t)z1 (t)

x ( t ) x 1 ( t ) + y ( t ) y1 ( t ) = z ( t ) z1 ( t )
At the specified time, x (t) is decreasing, so x1 (t) is negative, and y(t) is increasing, so y1 (t) is
positive.
a
(300)(´15) + (400)(20) = 3002 + 4002 z1 (t)
500z1 (t) = 3500
z1 (t) = 7 mph
S-8:
370
• We compute the distance d(t) between the two snails after t minutes as
d2 (t) = 302 + (y1 (t) ´ y2 (t))2 ,
where y1 (t) is the altitude of the first snail, and y2 (t) the altitude of the second snail after t
minutes.
2d ¨ d1 = 2(y1 1 ´ y2 1 )(y1 ´ y2 )
d ¨ d1 = (y1 1 ´ y2 1 )(y1 ´ y2 )
• We are told that y1 1 = 25 and y2 1 = 15. It will take 4 minutes for the first snail to reach
y1 = 100, and in this time the second snail will reach y2 = 60.
• At this point d2 = 302 + (100 ´ 60)2 = 900 + 1600 = 2500, hence d = 50.
• Therefore
50d1 = (25 ´ 15) ˆ (100 ´ 60)

400
d1 = = 8 cm per minute.
50
S-9:
• If we write z(t) for the length of the ladder at time t and y(t) for the height of the top end of
the ladder at time t we have
z(t)2 = 52 + y(t)2 .
2z ¨ z1 = 2y ¨ y1 ,
• We are told that z1 (t) = ´2, so z(3.5) = 20 ´ 3.5 ¨ 2 = 13.

? ? ?
• At this point y = z2 ´ 52 = 169 ´ 25 = 144 = 12.
• Hence
2 ¨ 13 ¨ (´2) = 2 ¨ 12y1
2 ¨ 13 13
y1 = ´ = ´ meters per second.
12 6
dV
S-10: What we’re given is (where V is volume of water in the trough, and t is time), and what
dt
dh
we are asked for is (where h is the height of the water). So, we need an equation relating V
dt
and h. First, let’s get everything in the same units: centimetres.
371
cm
200
100 cm
50 cm
w
h
60 cm
We can calculate the volume of water in the trough by multiplying the area of
its trapezoidal
cross
section by 200 cm. A trapezoid with height h and bases b1 and b2 has area h b1 +2 b2 . (To see why
this is so, draw the trapezoid as a rectangle flanked by two triangles.) So, using w as the width of
the top of the water (as in the diagram above), the area of the cross section of the water in the
trough is
60 + w
A=h
2
and therefore the volume of water in the trough is
V = 100h(60 + w) cm3 .
We need a formula for w in terms of h. If we draw lines straight up from the bottom corners of the
trapezoid, we break it into rectangles and triangles.
20
50 a
60
a 20 2
Using similar triangles, = , so a = h. Then
h 50 5
w = 60 + 2a

2 4
= 60 + 2 h = 60 + h
5 5
so
V = 100h(60 + w)
4
= 100h(120 + h)
5
= 80h2 + 12000h
This is the equation we need, relating V and h. Differentiating implicitly with respect to t:
dV dh dh
= 2 ¨ 80h ¨ + 12000
dt dt dt
dh
= (160h + 12000)
dt
372
dV
We are given that h = 25 and = 3 litres per minute. Converting to cubic centimetres,
dt
dV
= ´3000 cubic centimetres per minute. So:
dt
dh
´3000 = (160 ¨ 25 + 12000)
dt
dh 3 cm
= ´ = ´.1875
dt 16 min
3
So, the water level is dropping at centimetres per minute.
16
dV
S-11: If V is the volume of the water in the tank, and t is time, then we are given . What we
dt
dh
want to know is , where h is the height of the water in the tank. A reasonable plan is to find an
dt
equation relating V and h, and differentiate it implicitly with respect to t.
Let’s be a little careful about units. The volume of water in the tank is
(area of cross section of water)ˆ(length of tank)
If we measure these values in metres (area in square metres, length in metres), then the volume is
going to be in cubic metres. So, when we differentiate with respect to time, our units will be cubic
metres per second. The water is flowing in at one litre per second, or 1000 cubic centimetres per
second. So, we either have to measure our areas and distances in centimetres, or convert litres to
cubic metres. We’ll do the latter, but both are fine.
If we imagine one cubic metre as a cube, with each side of length 1 metre, then it’s easy to see the
volume inside is (100)3 = 106 cubic centimetres: it’s the volume of a cube with each side of length
100 cm. Since a litre is 103 cubic centimetres, and a cubic metre is 106 cubic centimetres, one litre is
dV 1
10´3 cubic metres. So, = 3 cubic metres per second.
dt 10
Let h be the height of the water (in metres). We can figure out the area of the cross section by
breaking it into three pieces: a triangle on the left, a rectangle in the middle, and a trapezoid on
the right.
1m
1.25 m a b
3m 3m
373
• The triangle on the left has height h metres. Let its base be a metres. It forms a similar
triangle with the triangle whose height is 1.25 metres and width is 1 metre, so:
a 1
=
h 1.25
4
a= h
5
So, the area of the triangle on the left is
1 2
ah = h2
2 5
• The rectangle in the middle has length 3 metres and height h metres, so its area is 3h square
metres.
• The trapezoid on the right is a portion of a triangle with base 3 metres and height 1.25
metres. So, its area is

1 1
(3)(1.25) ´ (b)(1.25 ´ h)
2
looooooomooooooon 2
loooooooooomoooooooooon
area of big triangle area of little triangle
The little triangle (of base b and height 1.25 ´ h) is formed by the air on the right side of the
tank. It is a similar triangle to the triangle of base 3 and height 1.25, so
b 3
=
1.25 ´ h 1.25
3
b= (1.25 ´ h)
1.25
So, the area of the trapezoid on the right is

1 1 3
(3)(1.25) ´ (1.25 ´ h) (1.25 ´ h)
2 2 1.25
6 2
= 3h ´ h
5
So, the area A of the cross section of the water is

2 2 6 2
A= h + loo3hmoon + 3h ´ h
loo5moon 5
looomooon
rectangle
triangle trapezoid
4
= 6h ´ h2
5
So, the volume of water is

4
V = 5 6h ´ h2 = 30h ´ 4h2
5
Differentiating with respect to time, t:
dV dh dh
= 30 ´ 8h
dt dt dt
374
1 dV 1
When h = metre, and = 3 cubic metres per second,
10 dt 10

1 dh 1 dh
3
= 30 ´ 8
10 dt 10 dt
dh 1
= metres per second
dt 29200
This is about 1 centimetre every five minutes. You might want a bigger hose.
S-12: Let θ be the angle of your head, where θ = 0 means you are looking straight ahead, and
π dθ
θ = means you are looking straight up. We are interested in , but we only have information
2 dt
about h. So, a reasonable plan is to find an equation relating h and θ, and differentiate with
respect to time.
The right triangle formed by you, the rocket, and the rocket’s original position has adjacent side
(to θ) length 2km, and opposite side (to θ) length h(t) kilometres, so
h
tan θ =
2
Differentiating with respect to t:
dθ 1 dh
sec2 θ ¨ =
dt 2 dt
dθ 1 dh
= cos2 θ ¨
dt 2 dt
We know tan θ = 2h . We draw a right triangle with angle θ (filling in the sides using
SOH CAH TOA and the Pythagorean theorem) to figure out cos θ:
4
? h2
+
h
θ
2
375
2
Using the triangle, cos θ = ? , so
h2
+4
2
dθ 1 2 dh
= ? ¨
dt 2 2
h +4 dt

2 dh
=
h2 + 4 dt

1 1
So, the quantities we need to know one minute after liftoff (that is, when t = ) are h and
60 60
dh 1
. Recall h(t) = 61750t2 .
dt 60

1 61750 1235
h = =
60 3600 72
dh
= 2(61750)t
dt
dh 1 2(61750) 6175
= =
dt 60 60 3

dθ 2 dh
Returning to the equation = :
dt h2 + 4 dt
!
dθ 1 2 6175 rad rad
= 2
« 13.8 « 0.0038
dt 60 1235

+4 3 hour sec
72
S-13: (a) Let x (t) be the distance of the train along the track at time t, measured from the point on
the track nearest the camera. Let z(t) be the distance from the camera to the train at time t.
x train
0.5 km θ z
camera
?
Then x1 (t) = 2 and at the time in question, z(t) = 1.3 km and x (t) = 1.32 ´ 0.52 = 1.2 km. So
z(t)2 = x (t)2 + 0.52
2z(t)z1 (t) = 2x (t) x1 (t)
2 ˆ 1.3z1 (t) = 2 ˆ 1.2 ˆ 2
2 ˆ 1.2
z1 (t)= « 1.85 km/min
1.3
(b) Let θ (t) be the angle shown at time t. Then

x (t)
sin (θ (t)) =
z(t)
x 1 ( t ) z ( t ) ´ x ( t ) z1 ( t )
θ 1 (t) cos (θ (t)) =
z ( t )2
x 1 ( t ) z ( t ) ´ x ( t ) z1 ( t )
θ 1 (t) =
z(t)2 cos (θ (t))
376
0.5
From our diagram, we see cos (θ (t)) = , so:
z(t)
x 1 ( t ) z ( t ) ´ x ( t ) z1 ( t )
=2
z(t)
2 ˆ 1.2
Substituting in x1 (t) = 2, z(t) = 1.3, x (t) = 1.2, and z1 (t) = :
1.3
2ˆ1.2
1 2 ˆ 1.3 ´ 1.2 ˆ 1.3
θ (t) = 2 « .592 radians/min
1.3
S-14: Let θ be the angle between the two hands.
10cm
D
θ
5cm
The Law of Cosines (Appendix B.4.1) tells us that
D2 = 52 + 102 ´ 2 ¨ 5 ¨ 10 ¨ cos θ
D2 = 125 ´ 100 cos θ
Differentiating with respect to time t,
dD dθ
2D = 100 sin θ ¨
dt dt
dθ
Our tasks now are to find D, θ and when the time is 4:00. At 4:00, the minute hand is straight
dt
4 1 1 2π
up, and the hour hand is = of the way around the clock, so θ = (2π ) = at 4:00. Then
12 3 ? ? 3 3
D2 = 125 ´ 100 cos 2π = 125 ´ 100 ´ 21 = 175, so D = 175 = 5 7 at 4:00.

3
dθ
To calculate , remember that both hands are moving. The hour hand makes a full rotation
dt
2π π
every 12 hours, so its rotational speed is = radians per hour. The hour hand is being chased
12 6
by the minute hand. The minute hand makes a full rotation every hour, so its rotational speed is
2π
= 2π radians per hour. Therefore, the angle θ between the two hands is changing at a rate of
1
dθ π ´11π rad
= ´ 2π ´ = .
dt 6 6 hr
377
dθ dD
Now, we plug in D, θ, and to find :
dt dt
dD dθ
2D = 100 sin θ ¨
dt dt
? dD 2π ´11π

2 5 7 = 100 sin
dt 3 6
?
? dD

3 ´11π 275π
10 7 = 100 =´ ?
dt 2 6 3
?
dD ´55 21π cm
=
dt 42 hr
?
55 21π
So D is decreasing at « 19 centimetres per hour.
42
S-15: The area at time t is the area of the outer circle minus the area of the inner circle:
A ( t ) = π R ( t )2 ´ r ( t )2

So, A1 (t) = 2π R(t) R1 (t) ´ r (t)r1 (t)

Plugging in the given data,
A1 = 2π 3 ¨ 2 ´ 1 ¨ 7 = ´2π

cm2
So the area is shrinking at a rate of 2π .
s
S-16: The volume between the spheres, while the little one is inside the big one, is
4 4
V= πR3 ´ πr3
3 3
Differentiating implicitly with respect to t:
dV dR dr
= 4πR2 ´ 4πr2
dt dt dt
dR dr
We differentiate R = 10 + 2t and r = 6t to find = 2 and = 6. When R = 2r,
dt dt
10 + 2t = 2(6t), so t = 1. When t = 1, R = 12 and r = 6. So:
dV
= 4π 122 (2) ´ 4π 62 (6) = 288π

dt
So the volume between the two spheres is increasing at 288π cubic units per unit time.
Remark: when the radius of the inner sphere increases, we are “subtracting” more area. Since the
radius of the inner sphere grows faster than the radius of the outer sphere, we might expect the
area between the spheres to be decreasing. Although the radius of the outer sphere grows more
slowly, a small increase in the radius of the outer sphere results in a larger change in volume than
the same increase in the radius of the inner sphere. So, a result showing that the volume between
the spheres is increasing is not unreasonable.
378
S-17: We know something about the rate of change of the height h of the triangle, and we want to
know something about the rate of change of its area, A. A reasonable plan is to find an equation
relating A and h, and differentiate implicitly with respect to t. The area of a triangle with height h
and base b is
1
A = bh
2
Note, b will change with time as well as h. So, differentiating with respect to time, t:

dA 1 db dh
= ¨h+b¨
dt 2 dt dt
We are given dh
dt and h, but those b’s are a mystery. We need to relate them to h. We can do this by
breaking our triangle into two right triangles and using the Pythagorean Theorem:
cm 200
0 cm
15 h
? ?
1502 ´ h2 2002 ´ h2
So, the base of the triangle is

a a
b= 1502 ´ h2 + 2002 ´ h2
db ´2h dh
dt ´2h dh
dt
= ? + ?
dt 2 1502 ´ h2 2 2002 ´ h2
´h dh
dt ´h dh
dt
=? +?
1502 ´ h2 2002 ´ h2
dh
Using = ´3 centimetres per minute:
dt
db 3h 3h
=? +?
dt 1502 ´ h2 2002 ´ h2
? ?
When h = 120, 1502 ´ h2 = 90 and 2002 ´ h2 = 160. So, at this moment in time:
b = 90 + 160 = 250
db 3(120) 3(120) 9 25
= + = 4+ =
dt 90 160 4 4
We return to our equation relating the derivatives of A, b, and h.

dA 1 db dh
= ¨h+b¨
dt 2 dt dt
dh db 25
When h = 120 cm, b = 250, = ´3, and = :
dt dt 4

dA 1 25
= (120) + 250(´3)
dt 2 4
=0
379
ˇ
dA ˇˇ
Remark: What does it mean that = 0? Certainly, as the height changes, the area changes
dt ˇh=120
as well. As the height sinks to 120 cm, the area is increasing, but after it sinks past 120 cm, the area
is decreasing. So, at the instant when the height is exactly 120 cm, the area is neither increasing nor
decreasing: it is at a local maximum. You’ll learn more about this kind of problem in Section 3.5.
dS
S-18: Let S be the flow of salt (in cubic centimetres per second). We want to know : how fast
dt
the flow is changing at time t. We are given an equation for S:
1
S= A
5
where A is the uncovered area of the cut-out. So,
dS 1 dA
=
dt 5 dt
dA dS
If we can find , then we can find . We are given information about how quickly the door is
dt dt
rotating. If we let θ be the angle made by the leading edge of the door and the far edge of the
dθ π
cut-out (shown below), then = ´ radians per second. (Since the door is covering more and
dt 6
dθ
more of the cut-out, θ is getting smaller, so is negative.)
dt
edge of cut-out
or
do
of
ge
ed
dθ dA dS
Since we know , and we want to know (in order to get ), it is reasonable to look for an
dt dt dt
equation relating A and θ, and differentiate it implicitly with respect to t to get an equation
dA dθ
relating and .
dt dt
The area of an annulus with outer radius 6 cm and inner radius 1 cm is π ¨ 62 ´ π ¨ 12 = 35π
θ θ
square centimetres. A sector of that same annulus with angle θ has area 2π (35π ), since 2π is
the ratio of the sector to the entire annulus. (For example, if θ = π, then the sector is half of the
entire annulus, so its area is (1/2)35π.)
π
So, when 0 ď θ ď 2, the area of the cutout that is open is
θ 35
A= (35π ) = θ
2π 2
380
This is the formula we wanted, relating A and θ. Differentiating with respect to t,
dA 35 dθ 35 π 35π
= = ´ =´
dt 2 dt 2 6 12
dS 1 dA
Since = ,
dt 5 dt
dS 1 35π 7π cm3
=´ =´ « ´1.8
dt 5 12 12 sec2
Remark: the change in flow of salt is constant while the door covers more and more of the
cut-out, so we never used the fact that precisely half of the cut-out was open. We also never used
the radius of the lid, which is immaterial to the flow of salt.
1 dF
S-19: Let F be the flow of water through the pipe, so F = A. We want to know , so
5 dt
differentiating implicitly with respect to t, we find
dF 1 dA
= .
dt 5 dt
dA dF
If we can find , then we can find . We know something about the shape of the uncovered
dt dt
area of the pipe; a reasonable plan is to find an equation relating the height of the door with the
uncovered area of the pipe. Let h be the distance from the top of the pipe to the bottom of the
door, measured in metres.
door
h
?
2h ´ h2
1 1´h
Since the radius of the pipe is 1 metre, the orange line has length 1 ´ h metres, and the blue line
has
a length 1 metre.?Using the Pythagorean Theorem, the green line has length
12 ´ (1 ´ h)2 = 2h ´ h2 metres.
381
? uncovered area of the pipe can be broken up into a triangle (of height 1 ´ h and base
The
2 2h ´ h2 ) and a sector of a circle (with angle 2π ´ 2θ). The area of the triangle is
a
(loomoon 2h ´ h2 .
1 ´ h) loooomoooon
height 12 base
The area of the sector is

2π ´ 2θ
(π ¨ 12 ) = π ´ θ.
2π loomoon
area
loooooomoooooon
fraction of circle
of circle
dA dh
Remember: what we want is to find , and what we know is = 0.01 metres per second. If
dt dt
we find θ in terms of h, we find A in terms of h, and then differentiate with respect to t.
Since θ is an angle in a right triangle with hypotenuse 1 and adjacent side length 1 ´ h,
cos θ = 1´h
1 = 1 ´ h. We want to conclude that θ = arccos(1 ´ h ), but let’s be a little careful:
remember that the range of the arccosine function is angles in [0, π ]. We must be confident that
0 ď θ ď π in order to conclude θ = arccos(1 ´ h)–but clearly, θ isin this range. (Remark: we
? ?
2h´h2
could also have said sin θ = 1 , and so θ = arcsin 2h ´ h2 . This would require
´ π2 ď θ ď π2 , which is true when h ă 1, but false for h ą 1. Since our problem asks about h = 0.25,
we could also use arcsine.)
Now, we know the area of the open pipe in terms of h.
A = (area of triangle) + (area of sector)

a
= (1 ´ h) 2h ´ h2 + (π ´ θ )
a
= (1 ´ h) 2h ´ h2 + π ´ arccos (1 ´ h)
We want to differentiate with respect to t. Using the chain rule:
dA dA dh
= ¨
dt dh dt !
dA 2 ´ 2h a ´1 dh
= (1 ´ h ) ? + (´1) 2h ´ h2 + a
dt 2 2h ´ h 2 1 ´ (1 ´ h ) 2 dt
(1 ´ h )2

a 1 dh
= ? ´ 2h ´ h2 ´ ?
2h ´ h 2 2h ´ h 2 dt
2

(1 ´ h ) ´ 1 a dh
= ? ´ 2h ´ h2
2h ´ h2 dt
2

´(2h ´ h ) a dh
= ? ´ 2h ´ h2
2h ´ h 2 dt
a a dh
= ´ 2h ´ h2 ´ 2h ´ h2
dt
a dh
= ´2 2h ´ h2
dt
We note here that the negative sign makes sense: as the door lowers, h increases and A decreases,
dh dA
so and should have opposite signs.
dt dt
382
1 dh 1
When h = metres, and = metres per second:
4 dt 100
c ?
7 cm2

dA 2 1 1
= ´2 ´ 2 =´
dt 4 4 100 200 s
dF 1 dA
Since = :
dt 5 dt
?
dF 7 m3
=´
dt 1000 sec2
?
7 m3
That is, the flow is decreasing at a rate of .
1000 sec2
S-20: We are given the rate of change of the volume of liquid, and are asked for the rate of change
of the height of the liquid. So, we need an equation relating volume and height.
dV
The volume V of a cone with height h and radius r is 13 πr2 h. Since we know , and want to
dt
dh
know , we need to find a way to deal with the unwanted variable r. We can find r in terms of h
dt
by using similar triangles. Viewed from the side, the conical glass is an?equilateral triangle, as is
the water in it. Using the Pythagorean Theorem, the cone has height 5 3.
?
5 3
10
r
r 5 h
Using similar triangles, = ? , so r = ? . (Remark: we could also use the fact that the water
h 5 3 3
forms a cone that looks like an equilateral triangle when viewed from the side to conclude
h
r = ? .)
3
Now, we can write the volume of water in the cone in terms of h, and no other variables.
1 2
V= πr h
3
h 2

1
= π ? h
3 3
π
= h3
9
dV π dh
= h2
dt 3 dt
383
dV
When h = 7 cm and = ´5 mL per minute,
dt
π dh
´5 = (49)
3 dt
dh ´15
= « ´0.097 cm per minute
dt 49π
S-21: As is so often the case, we use a right triangle in this problem to relate the quantities.
2
D
θ
D
sin θ =
2
D = 2 sin θ
Using the chain rule, we differentiate both sides with respect to time, t.
dD dθ
= 2 cos θ ¨
dt dt
dθ π
So, if = 0.25 radians per hour and θ = radians, then
dt 4

dD π 1 1 1
( a) = 2 cos ¨ 0.25 = 2 ? = ? metres per hour.
dt 4 2 4 2 2
dθ dD
Setting aside part (b) for a moment, let’s think about (c). If and have different signs, then
dt dt
dD dθ
because = 2 cos θ ¨ , that means cos θ ă 0. We have to have a nonnegative depth, so D ą 0
dt dt
and D = 2 sin θ implies sin θ ą 0. If sin θ ě 0 and cos θ ă 0, then θ P (π/2, π ]. On the diagram,
2
D θ
That is: the water has reversed direction. This happens, for instance, when a river empties into
the ocean and the tide is high. Skookumchuck Narrows provincial park, in the Sunshine Coast,
has reversing rapids.
Now, let’s return to (b). If the rope is only 2 metres long, and the river rises higher than 2 metres,
then our equation D = 2 sin θ doesn’t work any more: the buoy might be stationary underwater
while the water rises or falls (but stays at or above 2 metres deep).
384
S-22: (a) When the point is at (0, ´2), its y-coordinate is not changing, because it is moving along
dx
a horizontal line. So, the rate at which the particle moves is simply . Let θ be the angle an
dt
dθ
observer would be looking at, in order to watch the point. Since we know , a reasonable plan is
dt
to find an equation relating θ and x, and then differentiate implicitly with respect to t. To do this,
let’s return to our diagram.
x
θ
2
When the point is a little to the right of (0, ´2), then we can make a triangle with the origin, as
dθ
shown. If we let θ be the indicated angle, then = 1 radian per second. (It is given that the
dt
observer is turning one radian per second, so this is how fast θ is increasing.) From the right
triangle in the diagram, we see
x
tan θ =
2
Now, we have to take care of a subtle point. The diagram we drew only makes sense for the point
when it is at a position a little to the right of (0, ´2). So, right now, we’ve only made a set-up that
will find the derivative from the right. But, with a little more thought, we see that even when x is
negative (that is, when the point is a little to the left of (0, ´2)), our equation holds if we are
careful about how we define θ. Let θ be the angle between the line connecting the point and the
origin, and the y-axis, where θ is negative when the point is to the left of the y-axis.
385
y
|θ| 2
|x|
Since x and θ are both negative when the point is to the left of the y-axis,
2
tan |θ| =
|x|
´x
tan(´θ ) =
2
So, since tan(´θ ) = ´ tan(θ ):

x
tan θ =
2
x
So, we’ve shown that the relationship tan θ = holds when our point is at ( x, ´2), regardless of
2
the sign of x.
dθ dx
Moving on, since we are given and asked for , we differentiate with respect to t:
dt dt
dθ 1 dx
sec2 θ ¨ = ¨
dt 2 dt
dθ
When the point is at (0, ´2), since the observer is turning at one radian per second, also = 1.
dt
Also, looking at the diagram, θ = 0. Plugging in these values:
1 dx
sec2 (0) ¨ (1) = ¨
2 dt
1 dx
1= ¨
2 dt
dx
=2
dt
So, the particle is moving at 2 units per second.
(b) When the point is at (0, 2), it is moving along a line with slope ´ 21 and y-intercept 2. So, it is on
the line
1
y = 2´ x
2
386
That is, at time t, if the point is at ( x (t), y(t)), then x (t) and y(t) satisfy y(t) = 2 ´ 21 x (t).
Implicitly differentiating with respect to t:
dy 1 dx
=´ ¨
dt 2 dt
dx dy 1 1
So, when = 1, = ´ . That is, its y-coordinate is decreasing at unit per second.
dt dt 2 2
For the question “How fast is the point moving?”, remember that the velocity of an object can be
found by differentiating (with respect to time) the equation that gives the position of the object.
The complicating factors in this case are that (1) the position of our object is not given as a
function of time, and (2) the position of our object is given in two dimensions (an x coordinate
and a y coordinate), not one.
Remark: the solution below is actually pretty complicated. It is within your abilities to figure it
out, but later on in your mathematical career you will learn an easier way, using vectors. For now,
take this as a relatively tough exercise, and a motivation to keep learning: your intuition that there
must be an easier way is well founded!
The point is moving along a straight line. So, to take care of complication (2), we can give its
position as a point on the line. We can take the line as a sort of axis. We’ll need to choose a point
on the axis to be the “origin”: (2, 1) is a convenient point. Let D be the point’s (signed) distance
along the “axis” from (2, 1). When the point is a distance of one unit to the left of (2, 1), we’ll have
D = ´1, and when the point is a distance of one unit to the right of (2, 1), we’ll have D = 1. Then
dD dx dy
D changes with respect to time, and is the velocity of the point. Since we know and ,a
dt dt dt
reasonable plan is to find an equation relating x, y, and D, and differentiate implicitly with respect
to t. (This implicit differentiation takes care of complication (1).) Using the Pythagorean Theorem:
D 2 = ( x ´ 2)2 + ( y ´ 1)2
dD dx dy
2D ¨ = 2( x ´ 2) ¨ + 2( y ´ 1) ¨
dt dt dt
dy a ?
We plug in x = 0, y = 2, dx
dt = 1, dt = ´ 1
2 , and D = ´ ( 0 ´ 2 ) 2 + (2 ´ 1)2 = ´ 5 (negative
because the point is to the left of (2, 1)):

? dD

1
´2 5 ¨ = 2(´2)(1) + 2(1) ´
dt 2
?
dD 5
= units per second
dt 2
S-23: (a) Since the perimeter of the bottle is unchanged (you aren’t stretching the plastic), it is
always the same as the perimeter before it was smooshed, which is the circumference of a circle of
radius 5, or 2π (5) = 10π. So, using our approximation for the perimeter of an ellipse,
b
10π = π 3( a + b) ´ ( a + 3b)(3a + b)
b
10 = 3( a + b) ´ ( a + 3b)(3a + b)
387
(b) The area of the base of the bottle is πab (see Section A.10), and its height is 20 cm, so the
volume of the bottle is
V = 20πab
(c) As you smoosh the bottle, its volume decreases, so the water spills out. (If it turns out that the
volume is increasing, then no water is spilling out–but life experience suggests, and our
dV
calculations verify, that this is not the case.) The water will spill out at a rate of ´ cubic
dt
centimetres per second, where V is the volume inside the bottle. We know something about a and
da
, so a reasonable plan is to differentiate the equation from (b) (relating V and a) with respect to
dt
t.
Using the product rule, we differentiate the equation in (b) implicitly with respect to t and get

dV da db
= 20π b+a
dt dt dt
da db
So, we need to find the values of a, b, , and at the moment when a = 2b.
dt dt
a
The equation from (a) tells us 10 = 3( a + b) ´ ( a + 3b)(3a + b). So, when a = 2b,
b
10 = 3(2b + b) ´ (2b + 3b)(6b + b)
b ?
10 = 9b ´ (5b)(7b) = b 9 ´ 35
10
b= ?
9 ´ 35
?
where we use the fact that b is a positive number, so b2 = |b| = b.
Since a = 2b,
20
a= ?
9 ´ 35
da db
Now we know a and b at the moment when a = 2b. We still need to know and at that
dt dt
da
moment. Since a = 5 + t, always = 1. The equation from (a) relates a and b, so differentiating
dt
da db
both sides with respect to t will give us an equation relating and . When differentiating the
dt dt
portion with a square root, be careful not to forget the chain rule.

da db da db

da db

dt + 3 dt ( 3a + b ) + ( a + 3b ) 3 dt + dt
0=3 + ´ a
dt dt 2 ( a + 3b)(3a + b)
da
Since = 1:
dt

db
1 + 3 db
dt ( 3a + b ) + ( a + 3b ) 3+ db
dt
0 = 3 1+ ´ a
dt 2 ( a + 3b)(3a + b)
388
At this point, we could plug in the values we know for a and b at the moment when a = 2b.
However, the algebra goes a little smoother if we start by plugging in a = 2b:

db db

db
1 + 3 dt ( 7b ) + ( 5b ) 3 + dt
0 = 3 1+ ´ a
dt 2 (5b)(7b)

b 7 + 21 db + 15 + 5 db
db dt dt
0 = 3 1+ ´ ?
dt 2b 35
db
22 + 26 dt

db
0 = 3 1+ ´ ?
dt 2 35
db 11 13 db
0 = 3+3 ´? ´?
dt 35 35 dt

11 13 db
´3 + ? = 3 ´ ?
35 35 dt
11 ?
db ´3 + ?35 ´3 35 + 11
= = ?
dt 3 ´ ?1335 3 35 ´ 13
dV
Now, we can calculate at the moment when a = 2b. We already found
dt

dV da db
= 20π b+a
dt dt dt
da db
So, plugging in the values of a, b, , and at the moment when a = 2b:
dt dt
?
dV 10 20 ´3 35 + 11
= 20π (1) ? + ? ?
dt 9 ´ 35 9 ´ 35 3 35 ´ 13
?
200π 3 35 ´ 11
= ? 1´2 ?
9 ´ 35 3 35 ´ 13
cm3
« ´375.4
sec
So the water is spilling out of the cup at about 375.4 cubic centimetres per second.
Remark: the algebra in this problem got a little nasty, but the method behind its solution is no
more difficult than most of the problems in this section. One of the reasons why calculus is so
widely taught in universities is to give you lots of practice with problem-solving: taking a big
problem, breaking it into pieces you can manage, solving the pieces, and getting a solution.
A problem like this can sometimes derail people. Breaking it up into pieces isn’t so hard, but
when you actually do those pieces, you can get confused and forget why you are doing the
calculations you’re doing. If you find yourself in this situation, look back a few steps to remind
yourself why you started the calculation you just did. It can also be helpful to write notes, like
“We are trying to find dV dV da db
dt . We already know that dt = .... We still need to find a, b, dt and dt .”
S-24: Since A = 0, the equation relating the variables tells us:

0 = log C2 + D2 + 1

1 = C2 + D2 + 1
0 = C2 + D2
0=C=D
389
This will probably be useful information. Since we’re also given the value of a derivative, let’s
differentiate the equation relating the variables implicitly with respect to t. For ease of notation,
dA
we will write = A1 , etc.
dt
2CC1 + 2DD1
A1 B + AB1 =
C2 + D2 + 1
At t = 10, A = C = D = 0:
0+0
A1 B + 0 =
0+0+1
1
AB=0
at t = 10, A1 = 2 units per second:
2B = 0
B = 0.
Solutions to Exercises 3.3.1 — Jump to TABLE OF CONTENTS
S-1: In the beginning of this section, the text says “A differential equation is an equation for an
unknown function that involves the derivative of the unknown function.” Our unknown function
dy
is y, so a differential equation is an equation that relates y and . This applies to (a) and (b), but
dx
not (c), (d), or (e).
dx
Note that = 1: this is the derivative of x with respect to x.
dx
dQ
S-2: Theorem 3.3.2 tells us that a function is a solution to the differential equation = kQ(t) if
dt
and only if the function has the form Q(t) = Cekt for some constant C. In our case, we want
dQ dQ 1
Q(t) = 5 , so = Q(t). So, the theorem tells us that the solutions are the functions of the
dt dt 5
form Q(t) = Cet/5 . This applies to (a) (with C = 0) and (d) (with C = 1), but none of the other
functions.
We don’t actually need a theorem to answer this question, though: we can just test every option.
dQ dQ
(a) = 0, so Q(t) = 0 = 5 ¨ 0 = 5 , so (a) is a solution.
dt dt
dQ dQ dQ
(b) = 5et = Q(t), so Q(t) = ‰5 , so (b) is not a solution.
dt dt dt
dQ 1 dQ dQ
(c) = 5e5t = 5Q(t), so Q(t) = ‰5 , so (c) is not a solution.
dt 5 dt dt
dQ 1 1 dQ
(d) = et/5 = Q(t), so Q(t) = 5 , so (d) is a solution.
dt 5 5 dt
dQ 1 1 dQ
(e) = et/5 = ( Q(t) ´ 1), so Q(t) = 5 + 1, so (e) is not a solution.
dt 5 5 dt
390
S-3: What we’re asked to find is when
Q(t) = 0
That is,
Ce´kt = 0
If C = 0, then this is the case for all t. There was no isotope to begin with, and there will continue
not being any undecayed isotope forever.
If C ą 0, then since e´kt ą 0, also Q(t) ą 0: so Q(t) is never 0 for any value of t. (But as t gets
bigger and bigger, Q(t) gets closer and closer to 0.)
Remark: The last result is somewhat disturbing: surely at some point the last atom has decayed.
The differential equation we use is a model that assumes Q runs continuously. This is a good
approximation only when there is a very large number of atoms. In practice, that is almost always
the case.
S-4: The two pieces of information give us
f (0) = A = 5 f (7) = Ae7k = π
Thus we know that A = 5 and so π = f (7) = 5e7k . Hence

π
e7k =
5
7k = log(π/5)
1
k = ¨ log(π/5).
7
where we use log to mean natural logarithm, loge .
dy
S-5: In Theorem 3.3.2, we saw that if y is a function of t, and = ´ky, then y = Ce´kt for some
dt
constant C.
dy
Our equation y satisfies = ´3y, so the theorem tells us y = Ce´3t for some constant C.
dt
We are also told that y(1) = 2. So, 2 = Ce´3ˆ1 tells us C = 2e3 . Then:
y = 2e3 ¨ e´3t = 2e´3(t´1) .
S-6: The amount of Carbon-14 in the sample t years after the animal died will be
Q(t) = 5e´kt
for some constant k (where 5 is the amount of Carbon-14 in the sample at time t = 0). So, the
answer we’re looking for is Q(10000). We need to replace k with an actual number to evaluate
391
Q(10000), and the key to doing this is the half-life. The text tells us that the half-life of Carbon-14
is 5730 years, so we know:
5
Q(5730) =
2
5
5e´k¨5730 =
2
5730 1
e´k =
2c
5730 1 1
e´k = = 2´ 5730
2
So:
t
Q(t) = 5 e´k
t
= 5 ¨ 2´ 5730
Now, we can evaluate:
10000
Q(10000) = 5 ¨ 2´ 5730 « 1.5 µg
Remark: after 2(5730) = 11, 460 years, the sample will have been sitting for two half-lives, so its
remaining Carbon-14 will be a quarter of its original amount, or 1.25 µg. It makes sense that at
10,000 years, the sample will contain slightly more Carbon-14 than at 11,460 years. Indeed, 1.5 is
slightly larger than 1.25, so our answer seems plausible.
It’s a good habit to look for ways to quickly check whether your answer seems plausible, since a
small algebra error can easily turn into a big error in your solution.
S-7: Let 100 years ago be the time t = 0. Then if Q(t) is the amount of Radium-226 in the sample,
Q(0) = 1, and
Q(t) = e´kt
for some positive constant k. When t = 100, the amount of Radium-226 left is 0.9576 grams, so
100
0.9576 = Q(100) = e´k¨100 = e´k
1
e´k = 0.9576 100
This tells us
t
Q(t) = 0.9576 100
So, if half the original amount of Radium-226 is left,
1 t
= 0.9576 100
2
1 t
log = log 0.9576 100
2
t
´ log 2 = log(0.9576)
100
log 2
t = ´100 « 1600
log 0.9576
392
So, the half life of Radium-226 is about 1600 years.
dQ
S-8: Let Q(t) denote the mass at time t. Then is the rate at which the mass is changing. Since
dt
dQ
the rate the mass is decreasing is proportional to the mass remaining, we know = ´kQ(t),
dt
dQ
where k is a positive constant. (Remark: since Q is decreasing, is negative. Since we cannot
dt
have a negative mass, if we choose k to be positive, then k and Q are both positive–this is why we
added the negative sign.)
The information given in the question is:
dQ
Q (0) = 6 = ´kQ(t) Q (1) = 1
dt
for some constant k ą 0. By Theorem 3.3.2, we know
Q(t) = Ce´kt
for some constant C. Since Q(0) = Ce0 = C, the given information tells us 6 = C. (This is the
initial mass of our sample.) So, Q(t) = 6e´kt . To get the full picture of the behaviour of Q, we
should find k. We do this using the given information Q(1) = 1:
1 = Q(1) = 6e´k(1)
1
6´1 = = e´k
6
So, all together,
t t
Q(t) = 6 e´k = 6 ¨ 6´1 = 61´t
6
The question asks us to determine the time th which obeys Q(th ) = = 3. Now that we know the
2
equation for Q(t), we simply solve:
Q(t) = 61´t
3 = Q(th ) = 61´th

log 3 = log 61´th = (1 ´ th ) log 6
log 3
= 1 ´ th
log 6
log 3 log 6 ´ log 3 log 2
th = 1 ´ = =
log 6 log 6 log 6
log 2
The half-life of Polonium-210 is years, or about 141 days.
log 6
Remark: The actual half-life of Polonium-210 is closer to 138 days. The numbers in the question
are made to work out nicely, at the expense of some accuracy.
393
S-9: The amount of Radium-221 in a sample will be
Q(t) = Ce´kt
where C is the amount in the sample at time t = 0, and k is some positive constant. We know the
half-life of the isotope, so we can find e´k :
C
= Q(30) = Ce´k¨30
2
1 30
= e´k
2
1
2´ 30 = e´k
So,
t t
Q(t) = C e´k = C ¨ 2´ 30
When only 0.01% of the original sample is left, Q(t) = 0.0001C:

t
0.0001C = Q(t) = C ¨ 2´ 30
t
0.0001 = 2´ 30

t
´ 30
log(0.0001) = log 2
t
log 10´4 = ´ log 2
30
t
´4 log 10 = ´ log 2
30
log 10
t = 120 ¨ « 398.6
log 2
It takes about 398.6 seconds (that is, roughly 6 and a half minutes) for all but 0.01% of the sample
to decay.
Remark: we can do another reality check here. The half-life is 30 seconds. 6 and a half minutes
13
represents 13 half-lives. So, the sample is halved 13 times: 12 « 0.00012 = 0.012%. So these 13
half-lives should reduce the sample to about 0.01% of its original amount, as desired.
S-10: We know that the amount of Polonium-210 in a sample after t days is given by
Q(t) = Ce´kt
where C is the original amount of the sample, and k is some positive constant.
The question asks us what percentage of the sample decays in a day. Since t is measured in days,
the amount that decays in a day is Q(t) ´ Q(t + 1). The percentage of Q(t) that this represents is
Q ( t ) ´ Q ( t + 1)
100 . (For example, if there were two grams at time t, and one gram at time t + 1,
Q(t)
2´1
then 100 = 50: 50% of the sample decayed in a day.)
1
394
In order to simplify, we should figure out a better expression for Q(t). As usual, we make use of
the half-life.
C
Q(138) =
2
C
Ce´k¨138 =
2
138 1
e´k = = 2´1
2
1
e´k = 2´ 138
Now, we have a better formula for Q(t):

t
Q(t) = C e´k
t
Q(t) = C ¨ 2´ 138
Finally, we can evaluate what percentage of the sample decays in a day.
t t +1 !
1
Q ( t ) ´ Q ( t + 1) C ¨ 2´ 138 ´ C ¨ 2´ 138 C
100 = 100 t 1
Q(t)
C ¨ 2´ 138 C
t t +1
2´ 138 ´ 2´ 138
= 100 t
2´ 138
t t +1 t
´ 138 ´ 138
= 100 2 ´2 2 138

1
´ 138
= 100 1 ´ 2 « 0.5
About 0.5% of the sample decays in a day.
Remark: when we say that half a percent of the sample decays in a day, we don’t mean half a
percent of the original sample. If a day starts out with, say, 1 microgram, then what decays in the
next 24 hours is about half a percent of that 1 microgram, regardless of what the “original”
sample (at some time t = 0) held.
In particular, since the sample is getting smaller and smaller, that half of a percent that decays
every day represents fewer and fewer actual atoms decaying. That’s why we can’t say that half of
the sample (50%) will decay after about 100 days, even though 0.5% decays every day and
100 ˆ 0.5 = 50.
S-11: The amount of Uranium-232 in the sample of ore at time t will be
Q(t) = Q(0)e´kt
where 6.9 ď Q(0) ď 7.5. We don’t exactly know Q(0), and we don’t exactly know the half-life, so
we also won’t exactly know Q(10): we can only say that is it between two numbers. Our strategy
is to find the highest and lowest possible values of Q(10), given the information in the problem.
395
In order for the most possible Uranium-232 to be in the sample after 10 years, we should start
with the most and have the longest half-life (since this represents the slowest decay). So, we take
Q(0) = 7.5 and Q(70) = 12 (7.5).
Q(t) = 7.5e´kt
1
(7.5) = Q(70) = 7.5e´k(70)
2
1 70
= e´k
2
1
2´ 70 = e´k
So, in this secenario,

t
Q(t) = 7.5 ¨ 2´ 70
After ten years,

10
Q(10) = 7.5 ¨ 2´ 70 « 6.79
So after ten years, the sample contains at most 6.8 µg.
Now, let’s think about the least possible amount of Uranium-232 that could be left after 10 years.
We should start with as little as possible, so take Q(0) = 6.9, and the sample should decay
quickly, so take the half-life to be 68.8 years.
Q(t) = 6.9e´kt
1
6.9 = Q(68.8) = 6.9e´k(68.8)
2
1 68.8
= e´k
2
1
2´ 68.8 = e´k
In this scenario,
t
Q(t) = 6.9 ¨ 2´ 68.8
After ten years,

10
Q(10) = 6.9 ¨ 2´ 68.8 « 6.24
So after ten years, the sample contains at least 6.2 µg.
After ten years, the sample contains between 6.2 and 6.8 µg of Uranium-232.
396
S-1: Using Corollary 3.3.8 (with A = 20 and K = 5), solutions to the differential equation all have
the form
T (t) = [ T (0) ´ 20]e5t + 20
for some constant T (0). This fits (a) (with T (0) = 20), (c) (with T (0) = 21), and (d) (with
T (0) = 40), but not (b) (since the constant has the wrong sign).
Instead of using the corollary, we can also just check each function for ourselves.
dT
(a) = 0 = 5 ¨ 0 = 5[ T (t) ´ 20], so (a) gives a solution to the differential equation.
dt
dT
(b) = 5[20e5t ] = 5[ T + 20] ‰ 5[ T ´ 20], so (b) does not give a solution to the differential
dt
equation.
dT
(c) = 5[e5t ] = 5[ T ´ 20], so (c) gives a solution to the differential equation.
dt
dT
(d) = 5[20e5t ] = 5[ T ´ 20], so (d) gives a solution to the differential equation.
dt
S-2: From Newton’s Law of Cooling and Corollary 3.3.8, the temperature of the object will be
T (t) = [ T (0) ´ A]eKt + A
where A is the ambient temperature (the temperature of the room), T (0) is the initial temperature
of the copper, and K is some constant. So, the ambient temperature–the temperature of the room–
is ´10 degrees. Since the coefficient of the exponential part of the function is positive, the
temperature of the object is higher than the temperature of the room.
S-3: As t grows very large, T (t) approaches A. That is:

lim T (t) = A
tÑ8
Kt
lim [ T (0) ´ A]e +A= A
tÑ8
lim [ T (0) ´ A]eKt = 0
tÑ8
Since the object is warmer than the room, T (0) ´ A is a nonzero constant. So,
lim eKt = 0
tÑ8
This tells us that K is a negative number. So, K must be negative–not zero, and not positive.
Remark: in our work, we used the fact that the object and the room have different temperatures
(but it didn’t matter which one was hotter). If not, then T (0) = A, and T (t) = A: that is, the
temperature of the object is constant. In this case, our usual form for the temperature of the object
looks like this:
T (t) = 0eKt + A
Keeping the exponential piece in there is overkill: the temperature isn’t changing, the function is
simply constant. If the object and the room have the same temperature, K could be any real
number since we’re multiplying eKt by zero.
Remark: contrast this to Question 9.
397
S-4: We want to know when
[ T (0) ´ A]ekt + A = A
That is, when
[ T (0) ´ A]ekt = 0
Since ekt ą 0 for all values of k and t, this happens exactly when
T (0) ´ A = 0
So: if the initial temperature of the object is not the same as the ambient temperature, then
according to this model, it never will be! (However, as t gets larger and larger, T (t) gets closer
and closer to A–it just never exactly reaches there.)
If the initial temperature of the object starts out the same as the ambient temperature, then
T (t) = A for all values of t.
S-5: From Newton’s Law of Cooling and Corollary 3.3.8, we know the temperature of the copper
will be
T (t) = [ T (0) ´ A]eKt + A
where A is the ambient temperature (100˝ ), T (0) is the temperature of the copper at time 0 (let’s
make this the instant it was dumped in the water, so T (0) = 25˝ ), and K is some constant. That is:
T (t) = [25 ´ 100]eKt + 100

= ´75eKt + 100
The information given tells us that T (10) = 90, so
90 = ´75e10K + 100
10
75 eK = 10
10 2
eK =
15
1
K 2 10
e =
15
This lets us describe T (t) without any unknown constants.
t
T (t) = ´75 eK + 100
t
2 10
= ´75 + 100
15
398
The question asks what value of t gives T (t) = 99.9.
t
2 10
99.9 = ´75 + 100
15
t
2 10
75 = 0.1
15
t
2 10 1
=
15 750
t
 

2 10  1
log  = log
15 750

t 2
log = ´ log(750)
10 15
´10 log(750)
t= 2
« 32.9
log 15
It takes about 32.9 seconds.
S-6: The temperature of the stone t minutes after taking it from the bonfire is
T (t) = [ T (0) ´ A]eKt + A

= [500 ´ 0]eKt + 0
= 500eKt
for some constant K. We are given that T (10) = 100.
100 = T (10) = 500e10K

1
e10K =
5
1
eK = 5´ 10
This gives us the more complete picture for the temperature of the stone.
t t
T (t) = 500 eK = 500 ¨ 5´ 10
If T (t) = 50 :
t
50 = T (t) = 500 ¨ 5´ 10
1 t
= 10´1 = 5´ 10
10
t
10 = 5 10
t
log(10) = log(5)
10
log(10)
t = 10 « 14.3
log(5)
So the stone has been out of the fire for about 14.3 minutes.
399
S-7:
• First scenario: At time 0, Newton mixes 9 parts coffee at temperature 95˝ C with 1 part
cream at temperature 5˝ C. The resulting mixture has temperature
9 ˆ 95 + 1 ˆ 5
= 86˝
9+1
The mixture cools according to Newton’s Law of Cooling, with initial temperature 86˝ and
ambient temperature 22˝ :
T (t) = [86 ´ 22]e´kt + 22

T (t) = 64e´kt + 22
After 10 minutes,
54 = T (10) = 22 + 64e´10k
54 ´ 22 1
e´10k = =
64 2
We could compute k from this, but we don’t need it.
• Second scenario: At time 0, Newton gets hot coffee at temperature 95˝ C. It cools according
to Newton’s Law of Cooling
T (t) = [ T (0) ´ 22]e´kt + 22
In this second scenario, T (0) = 95, so
T (t) = [95 ´ 22]e´kt + 22 = 73e´kt + 22
The value of k is the same as in the first scenario, so after 10 minutes

1
T (10) = 22 + 73e´10k = 22 + 73 = 58.5
2
This cooled coffee is mixed with cold cream to yield a mixture of temperature
9 ˆ 58.5 + 1 ˆ 5
= 53.15
9+1
Under the second (add cream just before drinking) scenario, the coffee ends up cooler by 0.85˝ C .
S-8:
(a) By Newton’s law of cooling, the rate of change of temperature is proportional to the
difference between T (t) and the ambient temperature, which in this case is 30˝ . Thus
dT
= k[ T (t) ´ 30]
dt
for some constant of proportionality k. To answer part (a), all we have to do is find k.
400
Under Newton’s Law of Cooling, the temperature at time t will be given by
T (t) = [ T (0) ´ A]ekt + A

= [5 ´ 30]ekt + 30
= ´25ekt + 30
We are told T (5) = 10:
10 = ´25e5k + 30
25e5k = 20
4
e5k =
5
5k = log(4/5)
1
k= 5 log(4/5)
So, the differential equation is
dT 1
(t) = log(4/5)[ T (t) ´ 30]
dt 5
(b) Since T (t) = 30 ´ 25ekt , the temperature of the tea is 20˝ when
30 ´ 25ekt = 20
10
ekt =
25
10
kt = log
25
1 2
t = log
k 5
5 log(2/5)
=
log(4/5)
« 20.53 min
S-9: As time goes on, temperatures that follow Newton’s Law of Cooling get closer and closer to
the ambient temperature. So, lim T (t) exists. In particular, lim 0.8kt exists.
tÑ8 tÑ8
• If k ă 0, then lim 0.8kt = 8, since 0.8 ă 1. So, k ě 0.

tÑ8
• If k = 0, then T (t) = 16 for all values of t. But, in the statement of the question, the object is
changing temperature. So, k ą 0.
Therefore, k is positive.
Remark: contrast this to Question 3. The reason we get a different answer is that our base in this
question (0.8) is less than one, while the base in Question 3 (e) is greater than one.
401
Although the given equation T (t) does not exactly look like the Newton’s Law equations we’re
used to, it is equivalent. Remembering elog(0.8) = 0.8:
T (t) = 0.8kt + 15
kt
= elog 0.8 + 15
= e(k log 0.8)t + 15
= [16 ´ 15]e(k log 0.8)t + 15
= [16 ´ 15]eKt + 15
with K = k log 0.8. This is now the more familiar form for Newton’s Law of Cooling (with A = 15
and T (0) = 16).
Since 0.8 ă 1, log(0.8) is negative, so k and K have opposite signs.
S-1: Since b is a positive constant, lim ebt = 8. Therefore:

tÑ8
"
bt 0 if P(0) = 0
lim P(t) = lim P(0)e =
tÑ8 tÑ8 8 if P(0) ą 0
If P(0) = 0, then the model simply says “a population that starts with no individuals continues to
have no individuals indefinitely,” which certainly makes sense. If P(0) ‰ 0, then (since we can’t
have a negative population) P(0) ą 0, and the model says “a population that starts out with some
individuals will end up with any gigantically huge number you can think of, given enough time.”
This one doesn’t make so much sense. Populations only grow to a certain finite amount, due to
scarcity of resources and such. In the derivation of the Malthusian model, we assume a constant
net birth rate–that the birth and death rates (per individual) don’t depend on the population,
which is not a reasonable assumption long-term.
S-2: The assumption that the animals grow according to the Malthusian model tells us that their
population t years after 2015 is given by P(t) = 121ebt for some constant b, since 121 = P(0), the
population 0 years after 2015. The information about 2016 tells us
136 = P(1) = 121eb
136
= eb
121
This gives us a better idea of P(t):
t
bt 136
P(t) = 121e = 121
121
2020 is 5 years after 2015, so in 2020 (assuming the population keeps growing according to the
Malthusian model) the population will be
136 5

P(5) = 121 « 217
121
In 2020, the Malthusian model predicts the herd will number 217 individuals.
402
S-3: Since the initial population of bacteria is 1000 individuals, the Malthusian model says that the
population of bacteria t hours after being placed in the dish will be P(t) = 1000ebt for some
constant b. Since P(1) = 2000,
2000 = P(1) = 1000eb
2 = eb
So, the population at time t is
P(t) = 1000 ¨ 2t
We want to know at what time the population triples, to 3,000 individuals.
3000 = 1000 ¨ 2t
3 = 2t
log(3) = log 2t = t log(2)

log(3)
t= « 1.6
log(2)
The population triples in about 1.6 hours. Note that the tripling time depends only on the
constant b. In particular, it does not depend on the initial condition P(0).
S-4: According to the Malthusian Model, if the ship wrecked at year t = 0 and 2 rats washed up
on the island, then t years after the wreck, the population of rats will be
P(t) = 2ebt
for some constant b. We want to get rid of this extraneous variable b, so we use the given
information. If 1928 is a years after the wreck:
1000 = P( a) = 2eba
1500 = P( a + 1) = 2eb(a+1) = 2eba eb
So,

(1000) eb = 2eba eb = 1500
Which tells us
1500
eb = = 1.5
1000
Now, our model is complete:
t
P(t) = 2 eb = 2 ¨ 1.5t
Since P( a) = 1000, we can find a:

1000 = P( a) = 2 ¨ 1.5a
500 = 1.5a
log(500) = log (1.5a ) = a log(1.5)
log(500)
a= « 15.3
log(1.5)
So, the year 1928 was about 15.3 years after the shipwreck. Since we aren’t given a month when
the rats reached exactly 1000 in number, that puts the shipwreck at the year 1912 or 1913.
403
S-5: The Malthusian model suggests that, if we start with P(0) cochineals, their population after 3
months will be
P(t) = P(0)ebt
for some constant b. The constant b is the net birthrate per population member per unit time.
Assuming that the net birthrate for a larger population will be the same as the test population, we
can use the data from the test to find eb . Let Q(t) be the number of individuals in the test
population at time t.
Q(t) = Q(0)ebt = 200ebt

1000 = Q(3) = 200e3b
5 = e3b
eb = 51/3
Now that we have an idea of the birthrate, we predict

t t
P ( t ) = P (0) e b = P (0) ¨ 5 3
We want P(12) = 106 + P(0).

12
106 + P(0) = P(12) = P(0) ¨ 5 3 = P(0) ¨ 54
h i
106 = P(0) ¨ 54 ´ P(0) = P(0) 54 ´ 1
106
P (0) = « 1603
54 ´ 1
The farmer should use an initial population of (at least) about 1603 individuals.
Remark: if we hadn’t specified that we need to save P(0) individuals to start next year’s
population, the number of individual cochineals we would want to start with to get a million in a
year would be 1600–almost the same!
S-6:
(a) Since f (t) gives the amount of the radioactive isotope in the sample at time t, the amount of
the radioactive isotope in the sample when t = 0 is f (0) = 100e0 = 100 units. Since the
sample is decaying, f (t) is decreasing, so f 1 (t) is negative. Differentiating,
f 1 (t) = k (100ekt ). Since 100ekt is positive and f 1 (t) is negative, k is negative.
(b) Since f (t) gives the size of the population at time t, the number of individuals in the
population when t = 0 is f (0) = 100e0 = 100. Since the population is growing, f (t) is
increasing, so f 1 (t) is positive. Differentiating, f 1 (t) = k (100ekt ). Since 100ekt is positive and
f 1 (t) is positive, k is also positive.
(c) Newton’s Law of Cooling gives the temperature of an object at time t as

f (t) = [ f (0) ´ A]ekt + A, where A is the ambient temperature surrounding the object. In
our case, the ambient temperature is 0 degrees. In an object whose temperature is being
modelled by Newton’s Law of Cooling, it doesn’t matter whether the object is heating or
404
cooling, k is negative. We saw this in Question 3 of Section 3.3.2, but it bears repeating.
Since f (t) approaches the ambient temperature (in this case, 0) as t goes to infinty:
lim 100ekt = 0
tÑ8
so k is negative.
df
S-1: The first piece of information given tells us = π f ( x ). Then by Theorem 3.3.2,
dx
f ( x ) = Ceπx
for some constant C. The second piece of given information tells us f (0) = 2. Using this, we can
solve for C:
2 = f (0) = Ce0 = C
Now, we know f ( x ) entirely:
f ( x ) = 2eπt
So, we can evaluate f (2)
f (2) = 2e2π
S-2: To use Corollary 3.3.8, we re-write the differential equation as

dT 9
=7 T´ ´ .
dt 7
9
Now, A = ´ and K = 7, so we see that the solutions to the differential equation have the form
7

9 7t 9
T ( t ) = T (0) + e ´
7 7
for some constant T (0).
We can check that this is reasonable: if

9 7t 9
T ( t ) = T (0) + e ´
7 7
then

dT 9 7t
= 7 T (0) + e
dt 7

9
=7 T+
7
= 7T + 9.
405
S-3: Let Q(t) denote the amount of radioactive material after t days. Then Q(t) = Q(0)ekt . We are
told
Q(8) = 0.8 Q(0)
So,
Q(0)e8k = 0.8 Q(0)

e8k = 0.8
1
ek = 0.8 8
If Q(0) = 100, the time t at which Q(t) = 40 is determined by
t
1 t
kt kt
40 = Q(t) = Q(0)e = 100e = 100 0.8 8 = 100 ¨ 0.8 8
Solving for t:
40 t
= 0.8 8
100
t t
log (0.4) = log 0.8 8 = log(0.8)
8
8 log(0.4)
t= « 32.85 days
log(0.8)
100 grams will decay to 40 grams in about 32.85 days.
S-4: Let t = 0 be the time the boiling water is left in the room, and let T (t) be the temperature of
the water t minutes later, so T (0) = 100. Using Newton’s Law of Cooling, we model the
temperature of the water at time t as
T (t) = [100 ´ A]eKt + A
where A is the room temperature, and K is some constant. We are told that T (15) = 85 and
T (30) = 73, so:
85 = T (15) = [100 ´ A]e15K + A

73 = T (30) = [100 ´ A]e30K + A
Rearranging both equations:
85 ´ A
= e15K
100 ´ A
73 ´ A 2
= e30K = e15K
100 ´ A
406
Using these equations:
2
85 ´ A 2 73 ´ A
= e15K = e30K =
100 ´ A 100 ´ A
(85 ´ A)2
= 73 ´ A
100 ´ A
(85 ´ A)2 = (73 ´ A)(100 ´ A)
852 ´ 170A + A2 = 7300 ´ 173A + A2
173A ´ 170A = 7300 ´ 852
3A = 75
A = 25
The room temperature is 25˝ C.
S-5:
(a) The amount of money at time t obeys

dA
= 0.05A + 2,000 = 0.05[ A ´ (´40,000)]
dt
Using Corollary 3.3.8,
A(t) = [ A(0) + 40,000]e0.05t ´ 40,000
= 90,000 ¨ e0.05t ´ 40,000
where t = 0 corresponds to the year when the graduate is 25.
When the graduate is 65 years old, t = 40, so

A(40) = 90,000 e0.05ˆ40 ´ 40, 000 « $625, 015.05
(b) When the graduate stops depositing money and instead starts withdrawing money at a rate
W, the equation for A becomes
dA
= 0.05A ´ W = 0.05[ A(t) ´ 20W ]
dt
Using Corollary 3.3.8, and assuming that the interest rate remains 5%,
A(t) = [ A(0) ´ 20W ]e0.05t + 20W
= [625, 015.05 ´ 20W ]e0.05t + 20W
Note that, for part (b), we only care about what happens when the graduate starts
withdrawing money. We take t = 0 to correspond to the year when the graduate is 65–so
we’re using a different t from part (a). Then from part (a), A(0) = 625, 025.05.
We want the account to be depleted when the graduate is 85. So, we want
0 = A(20)
0 = 20W + e0.05ˆ20 (625, 015.05 ´ 20W )
0 = 20W + e(625, 015.05 ´ 20W )
20(e ´ 1)W = 625, 015.05e
625, 015.05e
W= « $49, 437.96
20(e ´ 1)
407
S-6: (a) The amount of money at time t obeys
dA
= 0.06A ´ 9,000 = 0.06[ A ´ 150,000]
dt
Using Corollary 3.3.8,
A(t) = [ A(0) ´ 150,000]e0.06t + 150,000

= [120,000 ´ 150,000]e0.06t + 150,000
= ´30,000e0.06t + 150,000
(b) The money runs out when A(t) = 0.
A(t) = 0
150,000 ´ 30,000 e0.06t = 0
30,000 e0.06t = 150,000
e0.06t = 5
0.06t = log 5
log 5
t= « 26.8 yrs
0.06
The money runs out in about 26.8 years.
Remark: without earning any interest, the money would have run out in about 13.3 years.
S-7: Let Q(t) denote the number of bacteria at time t. We are told that Q1 (t) = kQ(t) for some
constant of proportionality k. Consequently, Q(t) = Q(0)ekt (Corollary 3.3.2). We are also told
Q(9) = 3Q(0)
So, Q(0)e9k = 3Q(0)
e9k = 3
1
ek = 3 9
The doubling time t obeys:
Q(t) = 2Q(0)
So, Q(0)ekt = 2Q(0)
ekt = 2
t
39 = 2
t
log 3 = log 2
9
log 2
t=9 « 5.68 hr
log 3
408
S-8: (a) We want our differential equation to have the format of the equation in Corollary 3.3.8:
dv
(t) = ´g ´ kv(t)
dt g
= ´k v(t) +
k g
= ´k v(t) ´ ´
k
g
So, we can use the corollary, with K = ´k, T = v, and A = ´ .
k
g g
v ( t ) = v (0) ´ ´ e´kt ´
k k
g ´kt g
= v0 + e ´
k k
(b)
h g ´kt g i
lim v(t) = lim v0 + e ´
tÑ8 tÑ8
k k
g g
= v0 + lim e´kt ´
k tÑ8 k
Since k is positive:
g g
= v0 + (0) ´
k k
g
=´
k
Remark: This means, as the object falls, instead of accelerating without bound, it approaches
some maximum speed. The velocity is negative because the object is moving in the negative
direction–downwards.
S-1: Since f (0) is closer to g(0) than it is to h(0), you would probably want to estimate
f (0) « g(0) = 1 + 2 sin(1) if you had the means to efficiently figure out what sin(1) is, and if you
were concerned with accuracy. If you had a calculator, you could use this estimation. Also, later
in this chapter we will learn methods of approximating sin(1) that do not require a calculator, but
they do require time.
Without a calculator, or without a lot of time, using f (0) « h(0) = 0.7 probably makes the most
sense. It isn’t as accurate as f (0) « g(0), but you get an estimate very quickly, without worrying
about figuring out what sin(1) is.
Remark: when you’re approximating something in real life, there probably won’t be an obvious
“correct” way to do it. There’s usually a trade-off between accuracy and ease.
S-2: 0.93 is pretty close to 1, and we know log(1) = 0, so we estimate log(0.93) « log(1) = 0.
409
y
0.93 y = f (x)
x
y=0
1
S-3: We don’t know arcsin(0.1), but 0.1 is reasonably close to 0, and arcsin(0) = 0. So, we estimate
arcsin(0.1) « 0.
π ? π
S-4: We don’t know tan(1), but we do know tan = 3. Since « 1.047 is pretty close to 1,
? ? π ? 2 3 3
we estimate 3 tan(1) « 3 tan = 3 = 3.
3
S-5: Since 10.1 is pretty close to 10, we estimate 10.13 « 103 = 1000.
Remark: these kinds of approximations are very useful when you are doing computations. It’s
easy to make a mistake in your work, and having in mind that 10.13 should be about a thousand is
a good way to check that whatever answer you have makes sense.
S-1: The linear approximation is L( x ) = 3x ´ 9. Since we’re approximating at x = 5, f (5) = L(5),

and f 1 (5) = L1 (5). However, there is no guarantee that f ( x ) and L( x ) have the same value when
x ‰ 5. So:
(a) f (5) = L(5) = 6
(b) f 1 (5) = L1 (5) = 3
(c) there is not enough information to find f (0).
S-2: The linear approximation is a line, passing through (2, f (2)), with slope f 1 (2). That is, the
linear approximation to f ( x ) about x = 2 is the tangent line to f ( x ) at x = 2. It is shown below in
red.
y
y = f (x)
x
2
410
S-3: For any constant a, f ( a) = (2a + 5), and f 1 ( a) = 2, so our approximation gives us
f ( x ) « (2a + 5) + 2( x ´ a) = 2x + 5
Since f ( x ) itself is a linear function, the linear approximation is actually just f ( x ) itself. As a
consequence, the linear approximation is perfectly accurate for all values of x.
S-4: We have no idea what f (0.93) is, but 0.93 is pretty close to 1, and we definitely know f (1).
The linear approximation of f ( x ) about x = 1 is given by
f ( x ) « f (1) + f 1 (1)( x ´ 1)
So, we calculate:
f (1) = log(1) = 0
1
f 1 (x) =
x
1 1
f (1) = = 1
1
Therefore,
f ( x ) « 0 + 1( x ´ 1) = x ´ 1
When x = 0.93:
f (0.93) « 0.93 ´ 1 = ´0.07
y
y = x´1
y = f (x)
x
1
?
S-5: We approximate the function f ( x ) = x about x = 4, since 4 is a perfect square and it is close
to 5.
?
f (4) = 4 = 2
1 1 1
f 1 (x) = ? ñ f 1 (4) = ? =
2 x 2 4 4
1
f ( x ) « f (4) + f 1 (4)( x ´ 4) = 2 + ( x ´ 4)
4
1 9
f (5) « 2 + (5 ´ 4) =
4 4
? 9
We estimate 5 « .
4
2
9 81 80
Remark: = , which is pretty close to = 5. Our approximation seems pretty good.
4 16 16
411
?
S-6: We approximate the function f ( x ) = 5 x. We need to centre the approximation about some
value x = a such that we know f ( a) and f 1 ( a), and a is not too far from 30.
? 1
f (x) = 5 x = x 5
1 4 1
f 1 ( x ) = x´ 5 = ? 4
5 5
5 x
?
a needs to be a number whose fifth root we know. Since 5 32 = 2, and 32 is reasonably close to 30,
a = 32 is a great choice.
?5
f (32) = 32 = 2
1 1
f 1 (32) = 4
=
5¨2 80
The linear approximation of f ( x ) about x = 32 is
1
f (x) « 2 + ( x ´ 32)
80
When x = 30:
1 1 79
f (30) « 2 + (30 ´ 32) = 2 ´ =
80 40 40
?
5 79
We estimate 30 « .
40
79 ?
Remark: = 1.975, while 5 30 « 1.97435. This is a decent estimation.
40
S-7: If f ( x ) = x3 , then f (10.1) = 10.13 , which is the value we want to estimate. Let’s take the
linear approximation of f ( x ) about x = 10:
f (10) = 103 = 1000
f 1 ( x ) = 3x2
f 1 (10) = 3(102 ) = 300
f ( a) « f (10) + f 1 (10)( x ´ 10)
= 1000 + 300( x ´ 10)
f (10.1) « 1000 + 300(10.1 ´ 10) = 1030
We estimate 10.13 « 1030. If we calculate 10.13 exactly (which is certainly possible to do by hand),
we get 1030.301.
Remark: in the previous subsection, we used a constant approximation to estimate 10.13 « 1000.
That approximation was easy to do in your head, in a matter of seconds. The linear
approximation is more accurate, but not much faster than simply calculating 10.13 .
S-8: There are many possible answers. One is:

f ( x ) = sin x a=0 b=π
We know that f (π ) = 0 and f (0) = 0. Using a constant approximation of f ( x ) about x = 0, our
estimation is f (π ) « f (0) = 0, which is exactly the correct value. However, is we make a linear
approximation of f ( x ) about x = 0, we get
f (π ) « f (0) + f 1 (0)(π ´ 0) = sin(0) + cos(0)π = π
which is not exactly the correct value.
412
y
linear
const
x
π y = f (x)
Remark: in reality, we wouldn’t estimate sin(π ), because we know its value exactly. The purpose
of this problem is to demonstrate that fancier approximations are not always more accurate. At the
of this section, we’ll talk about how to bound the error of your estimations, to make sure that you
are finding something sufficiently accurate.
S-9: The linear approximation L( x ) of f ( x ) about x = a is chosen so that L( a) = f ( a) and

L1 ( a) = f 1 ( a). So,
1
L1 ( a ) = f 1 ( a ) =
1 + a2
1 1
=
4 1 + a2
?
a=˘ 3
? ?
We’ve narrowed down a to 3 or ´ 3. Recall the linear approximation of f ( x ) about x = a is
a
f ( a) + f 1 ( a)( x ´ a), so its constant term is f ( a) ´ a f 1 ( a) = arctan( a) ´ . We compute this
? ? 1 + a2
for a = 3 and a = ´ 3.
? ? ?
? a ? 3 π 3 4π ´ 27
a = 3 : arctan ( a) ´ = arctan 3 ´ ? 2 = ´ =
1 + a2 1+ 3 3 4 12
? ? ?
? a ? ´ 3 π 3 ´4π + 27
a = ´ 3 : arctan ( a) ´ = arctan ´ 3 ´ ? 2 = ´ + =
1 + a2 1+ ´ 3 3 4 12
?
So, when a = 3,
?
1 4π ´ 27
L( x ) = x +
4 12
? ?
and this does not hold when a = ´ 3. We conclude a = 3.
413
S-1: If Q( x ) is the quadratic approximation of f about 3, then Q(3) = f (3), Q1 (3) = f 1 (3), and
Q2 (3) = f 2 (3). There is no guarantee that f ( x ) and Q( x ) share the same third derivative, though,
so we do not have enough information to know f 3 (3).
f (3) = ´32 + 6(3) = 9

ˇ
1 d 2
(ˇ
f (3) = ´x + 6x ˇˇ = ´2x + 6|x=3 = 0
dx x =3
d2
ˇ ˇ
2
(ˇ d ˇ
f 2 (3) = ´x + 6x ˇ
ˇ = t´2x + 6uˇˇ = ´2
dx2 x =3 dx x =3
S-2: The quadratic approximation of f ( x ) about x = a is
1 2
f ( x ) « f ( a) + f 1 ( a)( x ´ a) + f ( a)( x ´ a)2
2
We subsitute f ( a) = 2a + 5, f 1 ( a) = 2, and f 2 ( a) = 0:
f ( x ) « (2a + 5) + 2( x ´ a) = 2x + 5
So, our approximation is f ( x ) « 2x + 5.
Remark: Our approximation is exact for every value of x. This will always happen with a
quadratic approximation of a function that is quadratic, linear, or constant.
S-3: We approximate the function f ( x ) = log x about the point x = 1. We choose 1 because it is
close to 0.93, and we can evaluate f ( x ) and its first two derivatives at x = 1.
f (1) = 0
1
f 1 (x) = ñ f 1 (1) = 1
x
2 ´1
f (x) = 2 ñ f 2 (1) = ´1
x
So,
1 2
f ( x ) « f (1) + f 1 (1)( x ´ 1) + f (1)( x ´ 1)2
2
1
= 0 + ( x ´ 1) ´ ( x ´ 1)2
2
When x = 0.93:
1 1
f (0.93) « (0.93 ´ 1) ´ (0.93 ´ 1)2 = ´0.07 ´ (0.0049) = ´0.07245
2 2
We estimate log(0.93) « ´0.07245.
Remark: a calculator approximates log(0.93) « ´0.07257. We’re pretty close.
414
S-4: We approximate the function f ( x ) = cos x. We can easily evaluate cos x and sin x (sin x will
1
appear in the first derivative) at x = 0, and 0 is quite close to , so we centre our approximation
15
about x = 0.
f (0) = 1
f 1 ( x ) = ´ sin x
f 1 (0) = ´ sin(0) = 0
f 2 ( x ) = ´ cos x
f 2 (0) = ´ cos(0) = ´1
Using the quadratic approximation f ( x ) « f (0) + f 1 (0)( x ´ 0) + 12 f 2 (0)( x ´ 0)2 :
1
f ( x ) « 1 ´ x2
2
1 1 449
f « 1´ 2
=
15 2 ¨ 15 450

1 449
We approximate cos « .
15 450
449 1

Remark: = 0.9977, while a calculator gives cos 15 « 0.9977786. Our approximation has an
450
error of about 0.000001.
S-5: The quadratic approximation of a function f ( x ) about x = a is
1 2
f ( x ) « f ( a) + f 1 ( a)( x ´ a) + f ( a)( x ´ a)2
2
We compute derivatives.
f (0) = e0 = 1
f 1 ( x ) = 2e2x
f 1 (0) = 2e0 = 2
f 2 ( x ) = 4e2x
f 2 (0) = 4e0 = 4
Substituting:
4
f ( x ) « 1 + 2( x ´ 0) + ( x ´ 0)2
2
2
f ( x ) « 1 + 2x + 2x
S-6: There are a few functions we could choose to approximate. For example:
• f ( x ) = x4/3 . In this case, we would probably choose to approximate f ( x ) about x = 8 (since

8 is a cube, 84/3 = 24 = 16 is something we can evaluate) or x = 1.
415
• f ( x ) = 5x . We can easily figure out f ( x ) when x is a whole number, so we would want to
centre our approximation around some whole number x = a, but then f 1 ( a) = 5a log(5)
gives us a problem: without a calculator, it’s hard to know what log(5) is.
? ?
• Since 54/3 = 5 3 5, we can use f ( x ) = 5 3 x. As in the first bullet, we would centre about
x = 8, or x = 1.
?
There isn’t much difference between the first and third bullets. We’ll go with f ( x ) = 5 3 x, centred
about x = 8.
1
f ( x ) = 5x 3 ñ f (8) = 5 ¨ 2 = 10
5 2 5 ´2 5
f 1 ( x ) = x´ 3 ñ f 1 (8) = 2 =
3 3 12
2 5 2 5 10 5 10 ´5 5
f (x) = ´ x´ 3 = ´ x´ 3 ñ f 2 (8) = ´ 2 =´
3 3 9 9 144
Using the quadratic approximation f ( x ) « f ( a) + f 1 ( a)( x ´ a) + 21 f 2 ( a)( x ´ a)2 :
5 5
f ( x ) « 10 + ( x ´ 8) ´ ( x ´ 8)2
12 288
5 5 275
f (5) « 10 + (´3) ´ (9) =
12 288 32
275
We estimate 54/3 «
32
275
Remark: = 8.59375, and a calculator gives 54/3 « 8.5499. Although 5 and 8 are somewhat far
32
apart, our estimate is only off by about 0.04.
S-7:
(a) For every value of n, the term being added is simply the constant 1. So,
ÿ30
1 = 1 + 1 + ¨ ¨ ¨ + 1. The trick is figuring out how many 1s are added. Our index n takes
n =5
30
ÿ
on all integers from 5 to 30, including 5 and 30, which is 26 numbers. So, = 26.
n =5
If you’re having a hard time seeing why the sum is 26, instead of 25, think of it this way:
there are thirty numbers in the collection t1, 2, 3, 4, 5, 6, . . . , 29, 30u. If we remove the first
four, we get 30 ´ 4 = 26 numbers in the collection t5, 6, . . . , 30u.
(b)
3
ÿ
2(n + 3) ´ n2 = loooooomoooooon
2(1 + 3) ´ 12 + loooooomoooooon 2 ( 3 + 3 ) ´ 32
2(2 + 3) ´ 22 + loooooomoooooon

n =1
n =1 n =2 n =3
= 8 ´ 1 + 10 ´ 4 + 12 ´ 9 = 16
416
(c)
10
ÿ 1 1 1 1 1 1 1 1 1 1 1 1
´ = ´ + ´ + ´ + ´ + ´
n =1
n n+1 1 1 + 1 loooomoooon
loooomoooon 2 2 + 1 loooomoooon
3 3 + 1 loooomoooon
4 4 + 1 loooomoooon
5 5+1
n =1 n =2 n =3 n =4 n =5
1 1 1 1 1 1 1 1 1 1
+ ´ + ´ + ´ + ´ + ´
6 6 + 1 7 7 + 1 8 8 + 1 9 9 + 1 10 10 +1
loooomoooon loooomoooon loooomoooon loooomoooon loooooomoooooon
n =6 n =7 n =8 n =9 n=10
Most of these cancel!

1 1 1 1 1 1 1 1 1 1 1
= ´ + ´ + ´ + ´ + ´ +
1 looomooon
2 2 looomooon
3 3 looomooon
4 4 looomooon
5 5 looomooon
6 6
0 0 0 0 0
1 1 1 1 1 1 1 1 1
´ + ´ + ´ + ´ + ´
7 7 looomooon
looomooon 8 8 looomooon
9 9 loooomoooon
10 10 11
0 0 0 0
1 10
= 1´ =
11 11
(d)
4 4 4 4
ÿ 5 ¨ 2n ÿ 2n 5 ÿ 2n 5 ÿ 1
=5 = =
n =1
4n +1 n =1
4 ¨ 4n 4 n =1 4n 4 n =1 2n
 
5
 1 + 1 + 1 + 1  = 75

=
4 loomo
 2 on loomo
4 on loomo
8 on loo16
moon 64
n =1 n =2 n =3 n =4
S-8: For each of these, there are many solutions. We provide some below.
5
ÿ
(a) 1 + 2 + 3 + 4 + 5 = n
n =1
4
ÿ
(b) 2 + 4 + 6 + 8 = 2n
n =1
5
ÿ
(c) 3 + 5 + 7 + 9 + 11 = (2n + 1)
n =1
7
ÿ
(d) 9 + 16 + 25 + 36 + 49 = n2
n =3
7
ÿ
(e) 9 + 4 + 16 + 5 + 25 + 6 + 36 + 7 + 49 + 8 = ( n2 + n + 1)
n =3
7
ÿ
(f) 8 + 15 + 24 + 35 + 48 = ( n2 ´ 1)
n =3
417
6
ÿ
(g) 3 ´ 6 + 9 ´ 12 + 15 ´ 18 = (´1)n+1 3n
n =1
Remark: if we had written (´1)n instead of (´1)n+1 , with everything else the same, the signs
would have been reversed.
S-9: Let’s start by taking the first two derivative of f ( x ).
f ( x ) = 2 arcsin x ñ f (0) = 2(0) = 0

2 2
f 1 (x) = ? ñ f 1 (0) = = 2
1 ´ x2 1
" *
2 d 1
2 ´2
f (x) = 2(1 ´ x )
dx
3
= 2 ´ 12 (1 ´ x2 )´ 2 (´2x )

(chain rule)
2x
= ? 3 ñ f 2 (0) = 0
1 ´ x2
Now, we can find the quadratic approximation about x = 0.
1 2
f ( x ) « f (0) + f 1 (0) x + f (0) x 2
2
= 2x
f (1) « 2
Our
πquadratic
approximation gives 2 arcsin(1) « 2. However, 2 arcsin(1) is exactly equal to
2 = π. We’ve just made the rather unfortunate approximation π « 2.
2
S-10: From the text, the quadratic approximation of e x about x = 0 is
1
e x « 1 + x + x2
2
So,
1
e = e1 « 1 + 1 + = 2.5
2
We estimate e « 2.5.
Remark: actually, e « 2.718.
S-11:
• First, we’ll show that

(a), (d), (e)
are equivalent:
10
ÿ 10
ÿ
(d) = 2 n = 2(1 + 2 + ¨ ¨ ¨ + 10) = 2(1) + 2(2) + ¨ ¨ ¨ + 2(10) = 2n = (a)
n =1 n =1
418
So (a) and (d) are equivalent.
11
ÿ
(e) = 2 (n ´ 1) = 2(1 + 2 + ¨ ¨ ¨ + 10) = (d)
n =2
So (e) and (d) are equivalent.
• Second, we’ll show that

(b), (g)
are equivalent.
10 10 10 10 n 10
1 ÿ 4n +1 1 ÿ 4 ¨ 4n 4 ÿ 4n
ÿ 4 ÿ
(g) = = = = = 2n = (b)
4 n =1 2n 4 n =1 2n 4 n =1 2n n =1
2 n =1
• Third, we’ll show that

(c), (f)
are equivalent.
14
ÿ 10
ÿ
(f) = (n ´ 4)2 = 12 + 22 + ¨ ¨ ¨ + 102 = n2 = (c)
n =5 n =1
• Now, we have three groups, where each group consists of equivalent expressions. To be
quite thorough, we should show that no two of these groups contain expressions that are
secretly equivalent. They would be hard to evaluate, but we can bound them and show that
no two expressions in two separate groups could possibly be equivalent. Notice that
10
ÿ
2n = 21 + 22 + ¨ ¨ ¨ + 210 ą 210 = 1024
n =1
ÿ10 10
ÿ
2
n ă 102 = 10(100) = 1000
n =1 n =1
ÿ10
n2 = 12 + 22 + ¨ ¨ ¨ 82 + 92 + 102 ą 82 + 92 + 102 = 245
n =1
ÿ10 10
ÿ
2n ă 20 = 200
n =1 n =1
The expressions in the blue group add to less than 200, but the expressions in the green
group add to more than 245, and the expressions in the red group add to more than 1024, so
the blue groups expressions can’t possibly simplify to the same number as the red and
green group expressions.
The expressions in the green group add to less than 1000. Since the expressions in the red
group add to more than 1024, the expressions in the green and red groups can’t possibly
simplify to the same numbers.
419
We group our expressions in to collections of equivalent expressions as follows:
t(a), (d), (e)u, t(b), (g)u, t(c), (f)u
S-1: Since T3 ( x ) is the third-degree Taylor polynomial for f ( x ) about x = 1:
• T3 (1) = f (1)
• T31 (1) = f 1 (1)
• T32 (1) = f 2 (1)
• T33 (1) = f 3 (1)
In particular, f 2 (1) = T32 (1).
T31 ( x ) = 3x2 ´ 10x + 9

T32 ( x ) = 6x ´ 10
T32 (1) = 6 ´ 10 = ´4
So, f 2 (1) = ´4.
S-2: In Question 1, we differentiated the Taylor polynomial to find its derivative. We don’t really
want to differentiate this ten times, though, so let’s look for another way. Unlike Question 1, our
Taylor polynomial is given to us in a form very similar to its definition. The nth degree Taylor
polynomial for f ( x ) about x = 5 is
n
ÿ f ( k ) (5)
Tn ( x ) = ( x ´ 5) k
k!
k =0
So,
n n
ÿ f ( k ) (5) k
ÿ 2k + 1
( x ´ 5) = ( x ´ 5) k
k! 3k ´ 9
k =0 k =0
For any k from 0 to n,
f ( k ) (5) 2k + 1
=
k! 3k ´ 9
In particular, when k = 10,
f (10) (5) 20 + 1
= =1
10! 30 ´ 9
f (10) (5) = 10!
420
S-3: The fourth-degree Maclaurin polynomial for f ( x ) is
1 2 1 1
T4 ( x ) = f (0) + f 1 (0) x + f ( 0 ) x 2 + f 3 ( 0 ) x 3 + f (4) ( 0 ) x 4
2 3! 4!
1 2 1
T3 ( x ) = f (0) + f 1 (0) x + f (0) x 2 + f 3 (0) x 3
2 3!
So, we simply “chop off” the part of T4 ( x ) that includes x4 :
T3 ( x ) = ´x3 + x2 ´ x + 1
S-4: We saw this kind of problem in Question 3. The fourth-degree Taylor polynomial for f ( x )
about x = 1 is
1 2 1 1
T4 ( x ) = f (1) + f 1 (1)( x ´ 1) + f (1)( x ´ 1)2 + f 3 (1)( x ´ 1)3 + f (4) (1)( x ´ 1)4
2 3! 4!
while the third-degree Taylor polynomial for f ( x ) about x = 1 is
1 2 1
T3 ( x ) = f (1) + f 1 (1)( x ´ 1) + f (1)( x ´ 1)2 + f 3 (1)( x ´ 1)3
2 3!
In Question 3 we “chopped off” the term of degree 4 to get T3 ( x ). However, our polynomial is not
in this form. It’s not clear, right away, what the term f (4) ( x ´ 1)4 is in our given T4 ( x ). So, we will
use a different method from Question 3.
One option is to do some fancy algebra to get T4 ( x ) into the standard form of a Taylor polynomial.
Another option (which we will use) is to recover f (1), f 1 (1), f 2 (1), and f 3 (1) from T4 ( x ).
Recall that T4 ( x ) and f ( x ) have the same values at x = 1 (although maybe not anywhere else!),
and they also have the same first, second, third, and fourth derivatives at x = 1 (but again, maybe
not anywhere else, and maybe their fifth derivatives don’t agree). This tells us the following:
T4 ( x ) = x4 + x3 ´ 9 ñ f (1) = T4 (1) = ´7
3 2
T41 ( x ) = 4x + 3x ñ f 1 (1) = T41 (1) = 7
T42 ( x ) = 12x2 + 6x ñ f 2 (1) = T42 (1) = 18
T43 ( x ) = 24x + 6 ñ f 3 (1) = T43 (1) = 30
Now, we can write the third-degree Taylor polynomial for f ( x ) about x = 1:

1 1
T3 ( x ) = ´7 + 7( x ´ 1) + (18)( x ´ 1)2 + (30)( x ´ 1)3
2 3!
= ´7 + 7( x ´ 1) + 9( x ´ 1)2 + 5( x ´ 1)3
Remark: expanding the expression above, we get the equivalent polynomial

T3 ( x ) = 5x3 ´ 6x2 + 4x ´ 10. From this, it is clear that we can’t just “chop off” the term with x4 to
change T4 ( x ) into T3 ( x ) when the Taylor polynomial is not centred about x = 0.
421
S-5: The nth degree Taylor polynomial for f ( x ) about x = 5 is
n
ÿ 1 (k)
Tn ( x ) = f (5)( x ´ 5)k
k!
k =0
We expand this somewhat:
1 (10) 1
Tn ( x ) = f (5) + f 1 ( x ´ 5) + ¨ ¨ ¨ + f (5)( x ´ 5)10 + ¨ ¨ ¨ + f (n) (5)( x ´ 1)n
10! n!
1 (10)
So, the coefficient of ( x ´ 5)10 is f (5). Expanding the given form of the Taylor polynomial:
10!
n/2
ÿ 2k + 1
Tn ( x ) = ( x ´ 5)2k
3k ´ 9
k =0
1 3 11 n+1
= + ( x ´ 5)2 + ¨ ¨ ¨ + ( x ´ 5)10 + ¨ ¨ ¨ + ( x ´ 5) n
´9 ´6
loomoon looooomooooon 6
looooooomooooooon ( 3/2 ) n ´ 9
looooooooooomooooooooooon
k =0 k =1 k =5 k=n/2
Equating the coefficients of ( x ´ 5)10 in the two expressions:
1 (10) 11
f (5) =
10! 6
(10) 11 ¨ 10!
f (5) =
6
S-6: Since T3 ( x ) is the third-degree Taylor polynomial for f ( x ) about x = a, we know the
following things to be true:
• f ( a) = T3 ( a)
• f 1 ( a) = T31 ( a)
• f 2 ( a) = T32 ( a)
• f 3 ( a) = T33 ( a)
But, some of these don’t look super useful. For instance, if we try to use the first bullet, we get
this equation:
2? 3

3 11 ?
a 2 log a ´ =´ e + 3ea ´ 6 ea2 + a3
3 3
Solving this would be terrible. Instead, let’s think about how the equations look when we move
further down the list. Since T3 ( x ) is a cubic equation, T33 ( x ) is a constant (and so T33 ( a) does not
depend on a). That sounds like it’s probably the simplest option. Let’s start differentiating. We’ll
422
need to know both f 3 ( a) and T33 ( a).

3 11
f ( x ) = x 2 log x ´
3

3 2 2 11
1
f (x) = x + 3x 2 log x ´ = 6x2 log x ´ 9x2
x 3
1
f 2 ( x ) = 6x2 ¨ + 12x log x ´ 18x = 12x log x ´ 12x
x
3 1
f ( x ) = 12x ¨ + 12 log x ´ 12 = 12 log x
x
3
f ( a) = 12 log a
Now, let’s move to the Taylor polynomial. Remember that e is a constant.
2? 3 ?
T3 ( x ) = ´ e + 3ex ´ 6 ex2 + x3
3 ?
T3 ( x ) = 3e ´ 12 ex + 3x2
1
?
T32 ( x ) = ´12 e + 6x
T33 ( x ) = 6
T33 ( a) = 6
The final bullet point gives us the equation:
f 3 ( a) = T33 ( a)
12 log a = 6
1
log a =
2
1
a = e2
?
So, a = e.
S-1: If we were to find the 16th degree Maclaurin polynomial for a generic function, we might
expect to have to differentiate 16 times (ugh). But, we know that the derivatives of sines and
cosines repeat themselves. So, it’s enough to figure out the pattern:
f ( x ) = sin x + cos x f (0) = 1
1
f ( x ) = cos x ´ sin x f 1 (0) = 1
f 2 ( x ) = ´ sin x ´ cos x f 2 (0) = ´1
f 3 ( x ) = ´ cos x + sin x f 3 (0) = ´1
f (4) = sin x + cos x f (4) ( 0 ) = 1
Since f (4) ( x ) = f ( x ), our derivatives repeat from here. They follow the pattern:
+1, +1, ´1, ´1.
1 1 1 1 1 1 1 1 1 10 1 11
T16 ( x ) = 1+ x´ x2 ´ x3 + x4 + x5 ´ x6 ´ x7 + x8 + x9 ´ x ´ x
2 3! 4! 5! 6! 7! 8! 9! 10! 11!
1 12 1 13 1 14 1 15 1 16
+ x + x ´ x ´ x + x
12! 13! 14! 15! 16!
423
S-2: A Taylor polynomial gives a polynomial approximation for a function s(t). Since s(t) is itself
a polynomial, any nth-degree Taylor polynomial, with n greater than or equal to the degree of
s(t), will simply give s(t). So, in our case, T100 (t) = s(t) = 4.9t2 ´ t + 10.
If that isn’t satisfying, you can go through the normal method of calculating T100 (t).
s(t) = 4.9t2 ´ t + 10 s(5) = 4.9(25) ´ 5 + 10 = 127.5
1
s (t) = 9.8t ´ 1 s1 (5) = 9.8(5) ´ 1 = 48
s2 (t) = 9.8 s2 (5) = 9.8
The rest of the derivatives of s(t) are identically zero, so they are (in particular) zero when t = 5.
Therefore,
1
T100 (t) = 127.5 + 48(t ´ 5) + 9.8(t ´ 5)2
2
= 127.5 + 48(t ´ 5) + 4.9(t ´ 5)2
We can now check that T100 (t) really is the same as s(t).
T100 (t) = 127.5 + 48(t ´ 5) + 4.9(t ´ 5)2
= 127.5 + 48(t ´ 5) + 4.9(t2 ´ 10t + 25)
= [127.5 + 48(´5) + 4.9(25)] + [48 ´ 4.9(10)]t + 4.9t2
= 10 ´ t + 4.9t2 = s(t)
as expected.
S-3: Let’s start by differentiating f ( x ) and looking for a pattern. Remember that log 2 = loge 2 is a
constant number.
f ( x ) = 2x
f 1 ( x ) = 2x log 2
f 2 ( x ) = 2x (log 2)2
f (3) ( x ) = 2x (log 2)3
f (4) ( x ) = 2x (log 2)4
f (5) ( x ) = 2x (log 2)5
So, in general,
f (k) ( x ) = 2x (log 2)k
We notice that this formula works even when k = 0 and k = 1. When x = 1,
f (k) (1) = 2 (log 2)k
The nth degree Taylor polynomial of f ( x ) about x = 1 is
n
ÿ f ( k ) (1)
Tn ( x ) = ( x ´ 1) k
k!
k =0
n
ÿ 2(log 2)k
= ( x ´ 1) k
k!
k =0
424
S-4: We need to know the first six derivatives of f ( x ) at x = 1. Let’s get started.
f ( x ) = x2 log x + 2x2 + 5 f (1) = 7

1
f 1 ( x ) = ( x2 ) + (2x ) log x + 4x
x
= 2x log x + 5x f 1 (1) = 5
1
f 2 ( x ) = (2x ) + (2) log x + 5
x
= 2 log x + 7 f 2 (1) = 7
f 3 ( x ) = 2x´1 f 3 (1) = 2
f (4) = ´2x´2 f (4) (1) = ´2
f (5) = 4x´3 f (5) ( 1 ) = 4
f (6) = ´12x´4 f (6) (1) = ´12
Now, we can plug in.
1 2 1
T6 ( x ) = f (1) + f 1 (1)( x ´ 1) +f (1)( x ´ 1)2 + f 3 (1)( x ´ 1)3
2 3!
1 (4) 1 1
+ f (1)( x ´ 1)4 + f (5) (1)( x ´ 1)5 + f (6) (1)( x ´ 1)6
4! 5! 6!
1 1
= 7 + 5( x ´ 1) + (7)( x ´ 1) + (2)( x ´ 1)3
2
2 3!
1 1 1
+ (´2)( x ´ 1)4 + (4)( x ´ 1)5 + (´12)( x ´ 1)6
4! 5! 6!
7 1 1 1 1
= 7 + 5( x ´ 1) + ( x ´ 1)2 + ( x ´ 1)3 ´ ( x ´ 1)4 + ( x ´ 1)5 ´ ( x ´ 1)6
2 3 12 30 60
S-5: We’ll start by differentiating and looking for a pattern.
1
f (x) = = (1 ´ x )´1
1´x
Using the chain rule,
f 1 ( x ) = (´1)(1 ´ x )´2 (´1) = (1 ´ x )´2

f 2 ( x ) = (´2)(1 ´ x )´3 (´1) = 2(1 ´ x )´3
f (3) ( x ) = (´3)(2)(1 ´ x )´4 (´1) = 2(3)(1 ´ x )´4
f (4) ( x ) = (´4)(2)(3)(1 ´ x )´5 (´1) = 2(3)(4)(1 ´ x )´5
f (5) ( x ) = (´5)(2)(3)(4)(1 ´ x )´6 (´1) = 2(3)(4)(5)(1 ´ x )´6
Recognizing the pattern,
f (k) ( x ) = k!(1 ´ x )´(k+1)

f (k) (0) = k!(1)´(k+1) = k!
425
The nth degree Maclaurin polynomial for f ( x ) is
n
ÿ f ( k ) (0) k
Tn ( x ) = x
k!
k =0
n
ÿ k! k
= x
k!
k =0
ÿ n
= xk
k =0
S-6: We’ll need to know the first three derivatives of x x at x = 1. This is a good review of
logarithmic differentiation, covered in Section 2.10.
f (x) = xx f (1) = 1
x
log( f ( x )) = log ( x ) = x log x
d d
tlog( f ( x ))u = tx log xu
dx dx
f 1 (x) 1
= x ¨ + log x = 1 + log x
f (x) x
f 1 ( x ) = x x [1 + log x ] f 1 (1) = 1
d d
f 2 (x) = tx x u [1 + log x ] + x x t1 + log xu
dx dx
1
= ( x x [1 + log x ]) [1 + log x ] + x x ¨
x
1
= x x (1 + log x )2 + f 2 (1) = 2
x
" *
3 d x 2 1 x d 2 1
f (x) = tx u (1 + log x ) + +x (1 + log x ) +
dx x dx x

x 2 1 x 2 1
= x [1 + log x ] (1 + log x ) + +x (1 + log x ) ´ 2
x x x

3 1
= x x (1 + log x )3 + (1 + log x ) ´ 2 f 3 (1) = 3
x x
Now, we can plug in:
1 2 1
T3 ( x ) = f (1) + f 1 (1)( x ´ 1) +f (1)( x ´ 1)2 + f 3 (1)( x ´ 1)3
2 3!
1 1
= 1 + 1( x ´ 1) + (2)( x ´ 1) + (3)( x ´ 1)3
2
2 6
1
= 1 + ( x ´ 1) + ( x ´ 1)2 + ( x ´ 1)3
2

1 π
S-7: We note that 6 arctan ? =6 = π. We will find the 5th-degree Maclaurin
3 6
1 1
polynomial T5 ( x ) for f ( x ) = 6 arctan x. Then π = f ? « T5 ? . Let’s begin by finding
3 3
426
the first five derivatives of f ( x ) = 6 arctan x.
f ( x ) = 6 arctan x f (0) = 0

1 1
f (x) = 6 f 1 (0) = 6
1 + x2

2 0 ´ 2x x
f (x) = 6 = ´12 f 2 (0) = 0
(1 + x 2 )2 (1 + x 2 )2
(1 + x2 )2 ´ x ¨ 2(1 + x2 )(2x )

3
f ( x ) = ´12
(1 + x 2 )4
(1 + x ) ´ 4x2
2

= ´12
(1 + x 2 )3
1 ´ 3x2

= ´12 f 3 (0) = ´12
(1 + x 2 )3
(1 + x2 )3 (´6x ) ´ (1 ´ 3x2 ) ¨ 3(1 + x2 )2 (2x )

(4)
f ( x ) = ´12
(1 + x 2 )6
´6x (1 + x2 ) ´ 6x (1 ´ 3x2 )

= ´12
(1 + x 2 )4
x ´ x3

= 144 f (4) ( 0 ) = 0
(1 + x 2 )4
(1 + x2 )4 (1 ´ 3x2 ) ´ ( x ´ x3 ) ¨ 4(1 + x2 )3 (2x )

(5)
f ( x ) = 144
(1 + x 2 )8
(1 + x2 )(1 ´ 3x2 ) ´ 8x ( x ´ x3 )

= 144
(1 + x 2 )5
5x4 ´ 10x2 + 1
= 144 f (5) (0) = 144
(1 + x 2 )5
We now use these values to compute the 5th-degree Maclaurin polynomial for f ( x ).
1 2 1 1 1
f ( 0 ) x 2 + f 3 ( 0 ) x 3 + f (4) ( 0 ) x 4 + f (5) ( 0 ) x 5
T5 ( x ) = f (0) + f 1 (0) x +
2 3! 4! 5!
12 3 144 5
= 6x ´ x + x
6 120
6
= 6x ´ 2x3 + x5
5

1 1
Now, if we want to approximate f ? = 6 arctan ? = π:
3 3

1 1 6 2 6
π= f ? « T5 ? = ? ´? 3+ ? 5
3 3 3 3 5 3
?

1 1
= 2 3 1´ + « 3.156
3¨3 5¨9
Remark: There are numerous methods for computing π to any desired degree of accuracy. Many
of them use the Maclaurin expansion of arctan x. In 1706 John Machin computed π to 100 decimal
digits by using the Maclaurin expansion together with
π = 16 arctan 51 ´ 4 arctan 239
1
.
427
S-8: Let’s start by differentiating, and looking for a pattern.
f ( x ) = x (log x ´ 1) f (1) = ´1

1 1
f (x) = x + log x ´ 1 = log x f 1 (1) = 0
x
1
f 2 ( x ) = = x´1 f 2 (1) = 1
x
f (3) ( x ) = (´1) x´2 f (3) (1) = ´1
f (4) ( x ) = (´2)(´1) x´3 = 2!x´3 f (4) (1) = 2!
f (5) ( x ) = (´3)(´2)(´1) x´4 = ´3!x´4 f (4) (1) = ´3!
f (6) ( x ) = (´4)(´3)(´2)(´1) x´5 = 4!x´5 f (4) (1) = 4!
f (7) ( x ) = (´5)(´4)(´3)(´2)(´1) x´6 = ´5!x´6 f (7) (1) = ´5!
f (8) ( x ) = (´6)(´5)(´4)(´3)(´2)(´1) x´7 = 6!x´7 f (8) (1) = 6!
When k ě 2, making use of the fact that 0! = 1 and (´1)k´2 = (´1)k :
f (k) ( x ) = (´1)k´2 (k ´ 2)!x´(k´1) f (k) (1) = (´1)k (k ´ 2)!
Now we use the standard form of a Taylor polynomial. Since the first two terms don’t fit the
pattern, we add those outside of the sigma.
100 (k )
ÿ f (1)
T100 ( x ) = ( x ´ 1) k
k!
k =0
100 (k)
ÿ f (1)
= f (1) + f 1 (1)( x ´ 1) + ( x ´ 1) k
k!
k =2
100
ÿ (´1)k (k ´ 2)!
= ´1 + 0( x ´ 1) + ( x ´ 1) k
k!
k =2
100
ÿ (´1)k
= ´1 + ( x ´ 1) k
k ( k ´ 1)
k =2
S-9: Recall that

2n
f (k) π4

ÿ π k
T2n ( x ) = x´
k! 4
k =0
Let’s start by taking some derivatives. Of course, since we’re differentiating sine, the derivatives
will repeat every four iterations.
π 1
f ( x ) = sin x f =?
4 2
π 1
f 1 ( x ) = cos x f1 =?
4 2
π 1
f 2 ( x ) = ´ sin x f2 = ´?
4 2
π 1
f 3 ( x ) = ´ cos x f3 = ´?
4 2
428
1 1 1 1 1 1 1 1
So, the pattern of derivatives is ? , ? , ´ ? , ´ ? , ? , ? , ´ ? , ´ ? , etc. This is a little
2 2 2 2 2 2 2 2
tricky to write in sigma notation. We can deal with the “doubles” by separating
the even and odd
powers. The first few terms of T2n that contain even powers of x ´ π4 are
1 1 π 2 1 π 4
? ´ ? x´ + ? x´
loomo 2! 2 4
2on loooooooomoooooooon 4! 2 4
loooooooomoooooooon
k =0 k =2 k =4
Observe that the signs alternate between successive terms. So if we rename k to 2` these terms are
1 1 π 2 1 π 4
? ´ ? x´ + ? x´
loomo 2! 2 4
2on loooooooomoooooooon 4! 2 4
loooooooomoooooooon
`=0 `=1 `=2
(´1)` 2`
and the `th term here is (2`)!?2 x ´ π4 . To verify that this really is the `th term, evaluate this for
` = 0, 1, 2 explicitly. When k = 2n, ` = n so that
ÿ f (k) π
n
4 π k ÿ (´1)` π 2`
x´ = ? x´
k! 4 (2`)! 2 4
0ďkď2n `=0
k even
π

Now for the odd powers. The first few terms of T2n that contain odd powers of x ´ 4 are
1 π 1 π 3 1 π 5
? x´ ´ ? x´ + ? x´
2 4
loooooomoooooon 3! 2 4
loooooooomoooooooon 5! 2 4
loooooooomoooooooon
k =1 k =3 k =5
Observe that the signs again alternate between successive terms. So if we rename k to 2` + 1 these
terms are
1 π 1 π 3 1 π 5
? x´ ´ ? x´ + ? x´
2 4
loooooomoooooon 3! 2 4
loooooooomoooooooon 5! 2 4
loooooooomoooooooon
`=0 `=1 `=2
(´1)` 2`+1
and the `th term here is (2`+1)!?2 x ´ π4 . To verify that this really is the `th term, evaluate this
for ` = 0, 1, 2 explicitly. The largest odd integer that is smaller than 2n is 2n ´ 1 and when
k = 2n ´ 1 = 2` + 1, ` = n ´ 1 so that
ÿ f (k) π
n´1
4 π k ÿ (´1)` π 2`+1
x´ = ? x´
k! 4 (2` + 1) ! 2 4
0ďkď2n `=0
k odd
Putting the even and odd powers together

n n´1
ÿ (´1)` π 2` ÿ (´1)` π 2`+1
T2n ( x ) = ? x´ + ? x´
(2`)! 2 4 (2` + 1) ! 2 4
`=0 `=0
S-10: From Example 3.4.12 in the text, we see that the nth Maclaurin polynomial for f ( x ) = e x is
n
ÿ 1 k x2 x3 x4 xn
Tn ( x ) = x = 1+x+ + + +¨¨¨+
k! 2 3! 4! n!
k =0
429
If n = 157 and x = 1,
157
ÿ 1 1 1 1 1
T157 (1) = = 1+1+ + + +¨¨¨+
k! 2 3! 4! 157!
k =0
Although we wouldn’t expect T157 (1) to be exactly equal to e1 , it’s probably pretty close. So, we
estimate
1 1 1 1
1+ + + +¨¨¨+ « e´1
2 3! 4! 157!
S-11: While you’re working with sums, it’s easy to mistake a constant for a function. The sum
given in this question is some number: π is a constant, and k is an index– if you wrote out all 100
terms of this sum, there would be no letter k. So, the sum given is indeed a number, but we don’t
want to have to add 100 terms to get a good idea of what number it is.
From Example 3.4.14 in the text, we see that the (2n)th-degree Maclaurin polynomial for
f ( x ) = cos x is
n
ÿ (´1)k 2k
T2n ( x ) = ¨x
k =0
(2k)!
5π
If n = 100 and x = , this equation becomes
4
100 2k
(´1)k

5π ÿ 5π
T200 = ¨
4 (2k)!
k =0
4
So, the sum corresponds to the 200th Maclaurin polynomial for f ( x ) = cos x evaluated at x = 5π 4 .
We should be careful to understand that T200 ( x ) is not equal to f ( x ), in general. However, when x
is reasonably close to 0, these two functions are approximations of one another. So, we estimate
100
(´1)k 5π 2k

ÿ 5π 5π 1
= T200 « cos = ´?
2k! 4 4 4 2
k =0
S-1:
y
f ( x + ∆x )
∆y y = f (x)
f (x)
∆x
x
x x + ∆x
430
S-2: Let f ( x ) be the number of problems finished after x minutes of work. The question tells us
that 5∆y « ∆x. So, if ∆x = 15, ∆y « 3. That is, in 15 minutes more, you will finish about 3 more
problems.
Remark: the math behind this problem is intended to be easy! Looking at symbols like ∆y and
f ( x + ∆x ) can be confusing, but the basic idea is pretty simple.
S-3: First, let’s find the first and second derivatives of f when x = 5.
1 1
f 1 (x) = f 1 (5) =
1 + x2 26
´2x ´10
f 2 (x) = 2
f (5) =
(1 + x 2 )2 262
1
The linear approximation for ∆y when ∆x = is
10
1 1 1
∆y « f 1 (5)∆x = ¨ = « 0.003846
26 10 260
1
The quadratic approximation for ∆y when ∆x = is
10
1 2 1 1 1 ´10 1 51
∆y « f 1 (5)∆x + f (5) (∆x )2 = ¨ + ¨ ¨ = « 0.003772
2 26 10 2 262 102 13520
Remark: ∆y = arctan(5.1) ´ arctan(5) « 0.003774.
S-4: (a)
When x is near 4, the linear approximation of ∆y is
∆y « s1 (4)∆x
From the problem, ∆x = 5 ´ 4 = 1. Differentiating,

c
1
? 1 4.9
s (t) = 19.6 ¨ ? = , so
2 x x
c
4.9
s1 (4) =
4
The linear approximation gives us

c
4.9
∆y « (1) « 1.1
4
So moving from 4 metres to 5 metres increases the speed with which you hit the water by about
1.1 metres per second.
431
(b) Again, we’ll use the linear approximation
∆y « s1 ( a)∆x
c
4.9
= ¨ ∆x
a
The difference in height between the first two jumps and between the last two jumps is the same,
∆x. The initial height of the first jump is smaller than the initial height of the second jump. So, the
value corresponding to a is smaller for the first jump than for the second. Therefore, ∆y is larger
between the first two jumps than it is between the last two jumps. So, the increase in speed from
the first jump to the second is larger than the increase in speed from the second jump to the third.
To put that more symbolically, the change in terminal speed between the first two jumps is
c
4.9
∆y « ¨ ∆x
x
while the change in terminal speed between the next two jumps is
c
4.9
∆y « ¨ ∆x
x + ∆x
c c
4.9 4.9
Since ¨ ∆x ą ¨ ∆x (when x and ∆x are positive), the change in terminal speed is
x x + ∆x
greater between the first two jumps than between the next two jumps.
S-1: False. The linear approximation is an approximation. It tells us
∆y « f 1 ( x0 )∆x = f 1 (2)(1) = 25
However, from our definition of ∆y,
∆y = f ( x0 + ∆x ) ´ f ( x0 ) = f (2 + 1) ´ f (2) = 58 ´ 26 = 32
Remark: this is to emphasize that the calculations in this subsection are estimations of error
bounds, rather than actual error bounds. All we can say is that we estimate the error will be no
more than some number–we don’t guarantee it.
In the next subsection, we will introduce an error bound that is guaranteed to be accurate. It is
usually harder to calculate than the estimations in this section.
S-2: When an exact value Q0 is measured as Q0 + ∆Q, Definition 3.4.25 gives us the absolute error
|∆Q|
as |∆Q|, and the percentage error as 100 .
Q0
In our situation, Q0 = 5.83 and Q0 + ∆Q = 6, so ∆Q = 0.17. So, the absolute error is 0.17, and the
percentage error is
0.17
100 « 2.92%
5.83
432
S-3: Since f 1 ( x ) = 6x, when x = 10, f 1 (10) = 60. If ∆y = f (11) ´ f (10), and ∆x = 11 ´ 10, then
the linear approximation tells us
∆y « 60∆x = 60
So, the linear approximates estimates the error in f ( x ) to be about 60.
Since f 2 ( x ) = 6, the quadratic approximation (using f 1 (10) = 60, f 2 (10) = 6, and ∆x = 1) tells us
1 2 1
∆y « f 1 (10)∆x + f (1) (∆x )2 = 60 ¨ 1 + (6)(1)2 = 63
2 2
So, the quadratic approximates estimates the error in f ( x ) to be about 63. (Indeed, the exact value
of f (11) ´ f (10) is 63. It is not a fluke that our estimated error, using a quadratic approximation,
is exactly the same as our actual error. It is a consequence of the fact that f ( x ) is a quadratic
function.)
S-4: Let A be the area of a pen of radius r. Then
A(r ) = πr2
Differentiating with respect to r,
A1 (r ) = 2πr
The exact area desired is A0 . Let the corresponding exact radius desired be r0 .
Using the linear approximation formula, where ∆A is the change in A corresponding to a change
in r of ∆r,
∆A « A1 (r0 )∆r = 2πr0 ∆r

∆A
∆r «
2πr0
What we’re interested in is the percent error r can have. The percent error is:
∆r ∆A
100 « 100
r0 2πr0 ¨ r0
∆A
= 100
2(πr02 )
∆A
= 100
2 ¨ A0
∆A 1

= 100
A0 2
1
ď (2) = 1
2
∆A
(To get the last line, we used the given information that the percent error in the area, 100 ,
A0
must be less than 2%.)
We conclude the error in r cannot be more than 1%.
433
θ
S-5: (a) The area removed represents a proportion of of the entire circle, whose area is
2π
π (32 ) = 9π. So, the area of the sector removed is
θ 9
¨ 9π = θ
2π 2
(b) To find θ from d, we cut our triangle (with angle θ opposite side of length d) into two
equivalent right triangles, as shown below.
3 3 d
2
θ
2
θ d
Using the information that the radius of the circle (also the hypotenuse of the right triangle) is 3,
d
θ d
sin = 2 =
2 3 6
π
Since the question tells us the sector is no more than a quarter of the circle, we know 0 ď θ ď ,
2
θ π θ
so 0 ď ď . This puts well within the range of arcsine.
2 4 2

d
θ = 2 arcsin
6
(c) First, let’s get an expression for the area of the sector in terms of d.

9 9 d
A= θ= 2 arcsin
2 2 6

d
= 9 arcsin
6
Differentiating,
9 1
A1 ( d ) = c 2 ¨ 6
1 ´ d6
9
=?
36 ´ d2
Let ∆A is the error in A corresponding to an error of ∆d in d. Since we measured d to be 0.7
instead of 0.68, in the linear approximation we take ∆d = 0.02 and d0 = 0.68.
∆A « A1 (d0 ) ¨ ∆d
= A1 (0.68) ¨ 0.02
9
=? ¨ 0.02
36 ´ 0.682
« 0.03
434
So, the error in A is about 0.03.
S-6: Suppose we have a function V (h) that gives the volume of water in the tank as a function of
its height.
Let h0 = 0.5 metres, ∆h = ´0.05, and ∆V = V (h0 + ∆h) ´ V (h0 ) = V (0.45) ´ V (0.5). Then, by the
linear approximation,
∆V « V 1 (0.5) ¨ ∆h = ´0.05V 1 (0.5)
In order to solve the problem, we will find a function V (h) giving the volume of water in terms of
the height, then find V 1 (0.5), and finally approximate that the change in the volume of the water
is ∆V « ´0.05V 1 (0.5).
The water in the tank forms a cone. The volume of a cone of height h and radius r is
1 2
V= πr h
3
We need to get rid of the variable r. We can do this using similar triangles. The diagram below
shows the side view of the tank and the water.
1
2
2r
The side view of the tank forms a triangle that is similar to the triangle formed by the side view of
the water, so
1 2r
=
2 h
h
r=
4
Using this, we find our equation for the volume of the water in terms of h.
π h 2 πh3

1
V (h) = πr2 h = h=
3 3 4 48
Differentiating,
πh2
V 1 (h) =
16
0.25π π
V 1 (0.5) = =
16 64
Finally, using the approximation ∆V « ´0.05V 1 (0.5),
´0.05π π
∆V « =´ « ´0.00245 m3
64 1280
We estimate that the volume decreased by about 0.00245 cubic metres, or about 2450 cubic
centimetres.
435
S-7: Let q be the measured amount of the isotope remaining after 3 years. Let h(q) be the half-life
of the isotope that we calculate using q. We measured q = 0.9, but we want to know what the
change in h is if q moves by 0.05. So, let ∆q = ˘0.05, and let ∆h be the corresponding change in h.
The linear approximation tells us
∆h « h1 (0.9)∆q
So,
ˇ ˇ
|∆h| « ˇh1 (0.9)ˇ |∆q| = h1 (0.9) ¨ 0.05
This suggests a plan for solving the problem. We will find the equation h(q) giving the calculated
half-life of the isotope based on the measurement q. Then, we will find h1 (0.9). Finally, the
equation |∆h| = h1 (0.9) ¨ 0.05 will tell us the change in h that corresponds with a change of 0.05 in
our measurement.
Let us find the half-life of the isotope, if after three years q µg is remaining. The amount of the
isotope present after t years is given by
Q(t) = Q(0)e´kt
for some constant k. Let’s take t = 0 to be the time when precisely one µg was present. Then
Q(t) = e´kt
After three years, q is the amount of the isotope remaining, so
q = e´k¨3
1
q 3 = e´k
t t
Q(t) = e´k = q 3
The half-life is the value of t for which Q(t) = 12 Q(0) = 12 .
1 t
= Q(t) = q 3
2
1 t
log = log q 3
2
t
´ log 2 = log q
3
´3 log 2
t=
log q
´3 log 2
So, we calculate the half-life to be . This gives us our first goal: a function h(q) that tells
log q
us the calculated half-life of the element.
´3 log 2
h(q) =
log q
436
Following our plan, we find h1 (0.9).
" *
1 d ´3 log 2
h (q) =
dq log q
d ! )
= ´3 log 2 ¨ (log q)´1
dq
1
= ´3 log 2 ¨ (´1) (log q)´2 ¨
q
3 log 2
=
q log2 q
3 log 2
h1 (0.9) = « 208
0.9 log2 (0.9)
Finally, as outlined in our plan,
|∆h| = h1 (0.9) ¨ 0.05
3 log 2
= « 10.4
18 log2 (0.9)
If our measurement changes by ˘0.05 µg, then we estimate our calculated half-life changes by
about ˘10.4 years. Since our measurement is accurate to within 0.05 µg, that means we estimate
our calculated half-life to be accurate to within about 10.4 years.
´3 log 2
Remark: since h(0.9) = « 19.7, an absolute error of 10.4 years corresponds to a
log 0.9
10.4
percentage error of 100 « 53%. The question did not specify absolute or percentage error.
19.7
Since both make sense, you can use either one.
S-1: From the given information,

|R(10)| = | f (10) ´ F (10)| = | ´ 3 ´ 5| = | ´ 8| = 8
So, (a) is false (since 8 is not less than or equal to 7), while (b), (c), and (d) are true.
Remark: R( x ) is the error in our approximation. As mentioned in the text, we almost never know
R exactly, but we can give a bound. We don’t need the tightest bound–just a reasonable one that is
easy to calculate. If we were dealing with real functions and approximations, we might not know
that |R(10)| = 8, but if we knew it was at most 9, that would be a pretty decent approximation.
Often in this section, we will make simplifying assumptions to get a bound that is easy to
calculate. But, don’t go overboard! It is a true statement to say that our absolute error is at most
100, but this statement would probably not be very helpful as a bound.
S-2: Equation 3.4.33 tells us that, when Tn ( x ) is the nth degree Taylor polynomial for a function
f ( x ) about x = a, then
ˇ ˇ
ˇ f ( n +1) ( c ) ˇ
n +1 ˇ
| f ( x ) ´ Tn ( x )| = ˇ ( x ´ a) ˇ
ˇ
ˇ ( n + 1) ! ˇ
437
for some c strictly between x and a. In our case, n = 3, a = 0, x = 2, and f (4) (c) = ec , so
ˇ ˇ
ˇ f (4) ( c ) ˇ
| f (2) ´ T3 (2)| = ˇ (2 ´ 0)4 ˇ
ˇ ˇ
ˇ 4! ˇ
24 c 2
= e = ec
4! 3
Since c is strictly between 0 and 2, ec ă e2 :
2 2
ď e
3
but this isn’t a number we really know. Indeed: e2 is the very number we’re trying to
approximate. So, we use the estimation e ă 3:
2 2
ă ¨3 = 6
3
We conclude that the error | f (2) ´ T3 (2)| is less than 6.
Now we’ll get a more exact idea of the error using a calculator. (Calculators will also only give
approximations of numbers like e, but they are generally very good approximations.)
ˇ ˇ
ˇ 2 1 2 1 3 ˇˇ
| f (2) ´ T3 (2)| = ˇe ´ 1 + 2 + ¨ 2 + ¨ 2 ˇ
ˇ
2 3!
ˇ ˇ
ˇ 2 4 ˇ
ˇ
= ˇˇe ´ 1 + 2 + 2 +
3 ˇ
ˇ ˇ
ˇ 2 19 ˇ
= ˇe ´ ˇˇ « 1.056
ˇ
3
So, our actual answer was only off by about 1.
Remark: 1 ă 6, so this does not in any way contradict our bound | f (2) ´ T3 (2)| ă 6.
S-3: Whenever you approximate a polynomial with a Taylor polynomial of greater or equal
degree, your Taylor polynomial is exactly the same as the function you are approximating. So, the
error is zero.
S-4: The constant approximation gives
sin(33) « sin(0) = 0
while the linear approximation gives
f ( x ) « f (0) + f 1 (0) x
sin( x ) « sin(0) + cos(0) x
=x
sin(33) « 33
Since ´1 ď sin(33) ď 1, the constant approximation is better. (But both are a little silly.)
438
y
T1 ( x )
T0 ( x )
x
ˇ ˇ
ˇ f ( n +1) ( c ) ˇ
n +1 ˇ
| f ( x ) ´ Tn ( x )| = ˇ ( x ´ a) ˇ
ˇ
ˇ ( n + 1) ! ˇ
6!(2c ´ 5)
for some c strictly between x and a. In our case, n = 5, a = 11, x = 11.5, and f (6) (c) = .
c+3
ˇ ˇ
ˇ 1 6!(2c ´ 5) ˇ
6ˇ
| f (11.5) ´ T5 (11.5)| = ˇˇ (11.5 ´ 11) ˇ
6! c+3
ˇ ˇ
ˇ 2c ´ 5 ˇ 1
= ˇˇ ˇ¨
c + 3 ˇ 26
for some c in (11, 11.5). We don’t know exactly which c this is true for, but since we know that c
lies in (11, 11.5), we can provide bounds.
• 2c ´ 5 ă 2(11.5) ´ 5 = 18
• c + 3 ą 11 + 3 = 14
ˇ ˇ
ˇ 2c ´ 5 ˇ
• Therefore, ˇˇ ˇ = 2c ´ 5 ă 18 = 9 when c P (11, 11.5).
c+3 ˇ c+3 14 7
With this bound, we see

ˇ ˇ
ˇ 2c ´ 5 ˇ 1
| f (11.5) ´ T5 (11.5)| = ˇˇ ˇ¨
c + 3 ˇ 26

9 1
ă « 0.0201
7 26
Our error is less than 0.02.
439
ˇ ˇ
ˇ f ( n +1) ( c ) ˇ
n +1 ˇ
| f ( x ) ´ Tn ( x )| = ˇ ( x ´ a) ˇ
ˇ
ˇ ( n + 1) ! ˇ
for some c strictly between x and a. In our case, n = 2, a = 0, and x = 0.1, so

ˇ ˇ
ˇ f (3) ( c ) ˇ
| f (0.1) ´ T2 (0.1)| = ˇ (0.1 ´ 0)3 ˇ
ˇ ˇ
ˇ 3! ˇ
| f 3 ( c )|
=
6000
for some c in (0, 0.1).
We will find f 3 ( x ), and use it to give an upper bound for
| f 3 ( c )|
| f (0.1) ´ T2 (0.1)| =
6000
when c is in (0, 0.1).
f ( x ) = tan x
f 1 ( x ) = sec2 x
f 2 ( x ) = 2 sec x ¨ sec x tan x
= 2 sec2 x tan x
f 3 ( x ) = 2 sec2 x sec2 x + (4 sec x ¨ sec x tan x ) tan x

= 2 sec4 x + 4 sec2 x tan2 x

1 π
When 0 ă c ă , also 0 ă c ă , so:
10 6
π 1
• tan c ă tan =?
6 3
?
π 3
• cos c ą cos =
6 2
2
• sec c ă ?
3
With these bounds in mind for secant and tangent, we return to the expression we found for our
error.
ˇ ˇ
| f 3 ( c )| ˇ2 sec4 x + 4 sec2 x tan2 xˇ
| f (0.1) ´ T2 (0.1)| = =
6000 6000
4 2 2
2 ?23 + 4 ?23 ?1
3
ă
6000
1
=
1125
1
The error is less than .
1125
440
ˇ ˇ
ˇ f ( n +1) ( c ) ˇ
| f ( x ) ´ Tn ( x )| = ˇ ( x ´ a ) n +1 ˇ
ˇ ˇ
ˇ ( n + 1) ! ˇ
1
for some c strictly between x and a. In our case, n = 5, a = 0, and x = ´ , so
4
ˇ ˇ
ˇ ˇ (6) ˇ
6 ˇ
ˇ f ´ 1 ´ T5 ´ 1 ˇ = ˇˇ f (c) ´ 1 ´ 0 ˇˇ
ˇ ˇ
ˇ 4 4 ˇ ˇ 6! 4 ˇ
ˇ ˇ
ˇ (6) ˇ
ˇ f ( c )ˇ
=
6! ¨ 46
for some c in ´ 14 , 0 . We’ll need to know the sixth derivative of f ( x ).

f ( x ) = log(1 ´ x )
f 1 ( x ) = ´(1 ´ x )´1
f 2 ( x ) = ´(1 ´ x )´2
f 3 ( x ) = ´2(1 ´ x )´3
f (4) ( x ) = ´3!(1 ´ x )´4
f (5) ( x ) = ´4!(1 ´ x )´5
f (6) ( x ) = ´5!(1 ´ x )´6
ˇ ˇ 5!
Plugging in ˇ f (6) (c)ˇ = :
ˇ ˇ
(1 ´ c )6
ˇ ˇ
ˇ 1 1 ˇˇ 5! 1
ˇf ´ ´ T5 ´ ˇ = =
ˇ 4 4 6! ¨ 46 ¨ (1 ´ c)6 6 ¨ 46 ¨ (1 ´ c)6
for some c in ´ 41 , 0 .

We’re interested in an upper bound for the error: we want to know the worst case scenario, so we
can say that the error is no worse than that. We need to know what the biggest possible value of
1 1
6 6
is, given ´ ă c ă 0. That means we want to know the biggest possible value of
6 ¨ 4 ¨ (1 ´ c ) 4
1
. This corresponds to the smallest possible value of (1 ´ c)6 , which in turn corresponds to
(1 ´ c )6
the smallest absolute value of 1 ´ c.
1
• Since ´ ď c ď 0, the smallest absolute value of 1 ´ c occurs when c = 0. In other words,
4
|1 ´ c| ď 1.
• That means the smallest possible value of (1 ´ c)6 is 16 = 1.

1
• Then the largest possible value of is 1.
(1 ´ c )6
1 1 1
• Then the largest possible value of 6
¨ 6
is « 0.0000407.
6 ¨ 4 (1 ´ c ) 6 ¨ 46
441
Finally, we conclude
ˇ ˇ
ˇ f ´ 1 ´ T5 ´ 1 ˇ = 1 1
ˇ ˇ
6 6
ă ă 0.00004
ˇ 4 4 ˇ 6 ¨ 4 ¨ (1 ´ c ) 6 ¨ 46
ˇ ˇ
ˇ f ( n +1) ( c ) ˇ
| f ( x ) ´ Tn ( x )| = ˇ ( x ´ a ) n +1 ˇ
ˇ ˇ
ˇ ( n + 1) ! ˇ
for some c strictly between x and a. In our case, n = 3, a = 30, and x = 32, so
ˇ ˇ
ˇ f (4) ( c ) ˇ
4ˇ
| f (30) ´ T3 (30)| = ˇ (30 ´ 32) ˇ
ˇ
ˇ 4! ˇ
2ˇ ˇ ˇ
= ˇ f (4) ( c ) ˇ
ˇ
3
for some c in (30, 32).
2 ˇˇ (4) ˇˇ
We will now find f (4) ( x ). Then we can give an upper bound on | f (30) ´ T3 (30)| = ˇ f ( c )ˇ
3
when c P (30, 32).
1
f (x) = x 5
1 4
f 1 ( x ) = x´ 5
5
4 9
f ( x ) = ´ 2 x´ 5
2
5
4 ¨ 9 14
f 3 ( x ) = 3 x´ 5
5
4 ¨ 9 ¨ 14 ´ 19
f (4) ( x ) = ´ x 5
54
Using this,
2 ˇˇ (4) ˇˇ
| f (30) ´ T3 (30)| = ˇ f ( c )ˇ
3ˇ ˇ
2 ˇˇ 4 ¨ 9 ¨ 14 ´ 19 ˇˇ
= ˇ´ c 5ˇ
3 54
336
= 19
54 ¨ c 5
Since 30 ă c ă 32,
336 336
ă 19
= 4
54 ¨ 30 5 54 ¨ 303 ¨ 30 5
14
= 4
57 ¨ 9 ¨ 30 5
442
? ?
This isn’t a number we know. We’re trying to find the error in our estimation of 5 30, but 5 30
shows up in our error. From here, we have to be a little creative to get a bound that actually
4
makes sense to us. There are different ways to go about it. You could simply use 30 5 ą 1. We will
be a little more careful, and use the following estimation:
1
14 14 ¨ 30 5
4
= 7
5 ¨ 9 ¨ 30
57 ¨ 9 ¨ 30 5
1
14 ¨ 32 5
ă 7
5 ¨ 9 ¨ 30
14 ¨ 2
ă 7
5 ¨ 9 ¨ 30
14
ă 7
5 ¨ 9 ¨ 15
ă 0.000002
We conclude | f (30) ´ T3 (30)| ă 0.000002.
ˇ ˇ
ˇ f ( n +1) ( c ) ˇ
| f ( x ) ´ Tn ( x )| = ˇ ( x ´ a ) n +1 ˇ
ˇ ˇ
ˇ ( n + 1) ! ˇ
1
for some c strictly between x and a. In our case, n = 1, a = , and x = 0.01, so
π
ˇ ˇ
ˇ f 2 (c) 1 2 ˇˇ

| f (0.01) ´ T1 (0.01)| = ˇ 0.01 ´
ˇ
ˇ
ˇ 2 π ˇ
1 100 ´ π 2 ˇˇ 2 ˇˇ

= ¨ f (c)
2 100π
1 1

for some c in 100 ,π .
Let’s find f 2 ( x ).

1
f ( x ) = sin
x
1

cos

1 ´1 ´ x
f 1 ( x ) = cos ¨ 2 =
x x x2
x2 sin 1x (´x´2 ) + cos 1x (2x )

2
f (x) =
x4
2x cos x ´ sin 1x
1

=
x4
Now we can plug in a better expression for f 2 (c):

1 100 ´ π 2 ˇˇ 2 ˇˇ

| f (0.01) ´ T1 (0.01)| = ¨ f (c)
2 200π
1 100 ´ π 2 ˇ2c cos
ˇ 1
1
ˇ
´ sin
ˇ
c c
= ¨
2 100π c4
443
1 1

for some c in 100 , π .
What we want toˇ do now is find anˇupper bound on this expression containing c,
1 100 ´ π 2 ˇ2c cos 1c ´ sin 1c ˇ

¨ .
2 100π c4
1 1 1
• Since c ě , it follows that c4 ě , so 4 ď 1004 .
100 1004 c
• For any value of x, | cos x| and |ˇsin x| areat most 1.ˇSince |c| ă 1, also
|c cos 1c | ă | cos 1c | ď 1. So, ˇ2c cos 1c ´ sin 1c ˇ ă 3
• Therefore,
2 ˇ ˇ
1 100 ´ π 1 ˇ 1 1 ˇˇ
| f (0.01) ´ Tn (0.01)| = ¨ 4 ¨ ˇ2c cos
ˇ ´ sin
2 100π c c c ˇ
100 ´ π 2

1
ă ¨ 1004 ¨ 3
2 100π
2
3 ¨ 1002 100

= ´1
2 π
3¨1002 100
2
Equation 3.4.33 gives the bound | f (0.01) ´ T1 (0.01)| ď 2 π ´1 .
The bound above works out to approximately fourteen million. One way to understand why the
bound is so high is that sin 1x moves about crazily when x is near zero–it moves up and down
incredibly fast, so a straight line isn’t going to approximate it very well at all.
That being said, because sin 1x is still “sine of something,” we know ´1 ď f (0.01) ď 1. To get a

better bound on the error, let’s find T1 ( x ).

1 1
f ( x ) = sin f = sin(π ) = 0
x π
1

´ cos

1
1
f (x) = x
f 1
= ´π 2 cos(π ) = π 2
x2 π

1 1 1 1
T1 ( x ) = f +f x´
π π π

1
= 0 + π2 x ´
π
= π2 x ´ π
π2
T1 (0.01) = ´π
100
444
Now that we know T1 (0.01), and we know ´1 ď f (0.01) ď 1, we can give the bound
| f (0.01) ´ T1 (0.01)| ď | f (0.01)| + |T1 (0.01)|

ˇ 2 ˇ
ˇπ ˇ
ď 1+ˇ ˇ ´ π ˇˇ
100
ˇ π ˇˇ
= 1 + π ˇ1 ´
ˇ
ˇ
100
ă 1+π
ă 1+4 = 5
A more reasonable bound on the error is that it is less than 5.
Still more reasonably, we would not use T1 ( x ) to evaluate sin(100) approximately. We would
write sin(100) = sin(100 ´ 32π ) and approximate the right hand side, which is roughly
sin(´π/6).
ˇ ˇ
ˇ f ( n +1) ( c ) ˇ
n +1 ˇ
| f ( x ) ´ Tn ( x )| = ˇ ( x ´ a) ˇ
ˇ
ˇ ( n + 1) ! ˇ
1
for some c strictly between x and a. In our case, n = 2, a = 0, and x = , so
2
ˇ 1 1 ˇ ˇ f (c) 1
ˇf ´ T2 ˇ=ˇ ´0 ˇ
ˇ
ˇ 2 2 ˇ ˇ 3! 2 ˇ
ˇ ˇ
ˇ (3) ˇ
ˇ f ( c )ˇ
=
3! ¨ 23
for some c in 0, 12 .

The next task that suggests itself is finding f (3) ( x ).
f ( x ) = arcsin x
1 1
f 1 (x) = ? = (1 ´ x 2 )´ 2
1 ´ x2
1 3
f 2 ( x ) = ´ (1 ´ x2 )´ 2 (´2x )
2
3
= x (1 ´ x 2 )´ 2

3 5 3
3
f (x) = x ´ (1 ´ x2 )´ 2 (´2x ) + (1 ´ x2 )´ 2
2
5 5
= 3x2 (1 ´ x2 )´ 2 + (1 ´ x2 )´ 2 +1
5
= (1 ´ x2 )´ 2 3x2 + (1 ´ x2 )

5
= (1 ´ x2 )´ 2 2x2 + 1

445
ˇ ˇ
ˇ (3) ˇ
ˇ ˇ f ( c ) ˇ
1 1
for some c in 0, 21 ,
ˇ
Since ˇ f ´ T2 ˇ=
2 2 3! ¨ 23
ˇ ˇ
ˇ ˇ
ˇ 1 + 2c2 ˇ
ˇ ˇ
ˇ ? 5 ˇ
1 ´ c2 ˇ
ˇ ˇ
1 + 2c2
ˇ ˇ
ˇ f 1 ´ T2 1 ˇ =
ˇ ˇ ˇ
= ? 5
ˇ 2 2 ˇ 3! ¨ 23
48 1 ´ c2
for some c in 0, 12 .

We want to know what is the worst case scenario-what’s the biggest this expression can be. So,
1 + 2c2 1
now we find an upper bound on ? 5 when 0 ď c ď . Remember that our bound
2
48 1 ´ c2
doesn’t have to be exact, but it should be relatively easy to calculate.
2
1 1 3
• When 0 ď c ď , the biggest 1 + 2c2 can be is 1 + 2 = .
2 2 2
1 + 2c2 3
So, the numerator of ? 5 is at most .
2
48 1 ´ c2
2
1 3
• The smallest 1 ´ c2
can be is 1 ´ = .
2 4
? c !5 ? 5
5 3 3
• So, the smallest 1 ´ c2 can be is = .
4 2
? 5
1 + 2c2 3
• Then smallest possible value for the denominator of ? 5 is 48
2 2
48 1ć
• Then
3
1 + 2c2 2
? 5 ď ? 5
48 1 ´ c2 3
48
2
1 1
=? 5 = ?
3 9 3
1
ă
10
Let’s put together these pieces. We found that
1 + 2c2
ˇ ˇ
ˇ 1 1 ˇˇ
ˇf ´ T2 = ? 5
ˇ 2 2 ˇ
48 1 ´ c2
446
for some c in 0, 21 . We also found that

1 + 2c2 1
? 5 ă
10
48 1 ´ c2
when c is in 0, 12 . We conclude

ˇ ˇ
ˇ
ˇf 1 1 ˇˇ 1
´ T2 ă .
ˇ 2 2 ˇ 10
1 1

For the second part of the question, we need to find f 2 and T2 2 .
Finding f 21 is not difficult.

f ( x ) = arcsin x

1 1 π
f = arcsin =
2 2 6
1

In order to find T2 2 , we need to find T2 ( x ).
1 2
T2 ( x ) = f (0) + f 1 (0) x + f (0) x 2
2
Conveniently, we’ve already found the first few derivatives of f ( x ).
 

1 1 0  2
T2 ( x ) = arcsin(0) + ? x +  ? 3  x
1 ´ 02 2
1 ´ 02
= 0+x+0
=x

1 1
T2 =
2 2
So, the actual error is
ˇ ˇ ˇ ˇ
ˇ f 1 ´ T2 1 ˇ = ˇ π ´ 1 ˇ = π ´ 1
ˇ ˇ ˇ ˇ
ˇ 2 2 ˇ ˇ 6 2ˇ 6 2
A calculator tells us that this is about 0.02.
f ( n +1) ( c )
S-11: Our error will have the form ( x ´ 1)n+1 for some constant c, so let’s find an
( n + 1) !
equation for f (n) ( x ). This has been done before in the text, but we’ll do it again here: we’ll take
several derivatives, then notice the pattern.
f ( x ) = log x
f 1 ( x ) = x´1
f 2 ( x ) = ´x´2
f 3 ( x ) = 2! x´3
f (4) ( x ) = ´3! x´4
f (5) ( x ) = 4! x´5
447
So, when n ě 1,
f (n) ( x ) = (´1)n´1 (n ´ 1)! ¨ xń
Now that we know the derivative of f ( x ), we have a better idea what the error in our
approximation looks like.
ˇ ˇ
ˇ f ( n +1) ( c ) ˇ
| f (1.1) ´ Tn (1.1)| = ˇ (1.1 ´ 1)n+1 ˇ
ˇ ˇ
ˇ ( n + 1) ! ˇ
ˇ ˇ 0.1n+1
= ˇ f ( n +1) ( c )ˇ
ˇ ˇ
( n + 1) !
ˇ ˇ
ˇ n! ˇ 1
= ˇˇ n+1 ˇˇ n+1
c 10 (n + 1)!
1
= n +1
|c| ¨ 10n+1 ¨ (n + 1)
for some c in (1, 1.1)
1
ă
(n + 1)10n+1 ¨ 1n+1
1
=
(n + 1)10n+1
What we’ve shown so far is

1
| f (1.1) ´ Tn (1.1)| ă
(n + 1)10n+1
1
If we can show that ď 10´4 , then we’ll be able to conclude
(n + 1)10n+1
1
| f (1.1) ´ Tn (1.1)| ă ď 10´4
(n + 1)10n+1
That is, our error is less than 10´4 .

1
So, our goal for the problem is to find a value of n that makes ď 10´4 . Certainly,
(n + 1)10n+1
n = 3 is such a number. Therefore, any n greater than or equal to 3 is an acceptable value.
1
S-12: We will approximate f ( x ) = x 7 using a Taylor polynomial. Since 37 = 2187, we will use
x = 2187 as our centre.
We need to figure out which degree Taylor polynomial will result in a small-enough error.
If we use the nth Taylor polynomial, our error will be
ˇ ˇ
ˇ f ( n +1) ( c ) ˇ
| f (2200) ´ Tn (2200)| = ˇ (2200 ´ 2187)n+1 ˇ
ˇ ˇ
ˇ ( n + 1) ! ˇ
ˇ ˇ 13n+1
= ˇ f ( n +1) ( c )ˇ ¨
ˇ ˇ
( n + 1) !
448
for some c in (2187, 2200). In order for this to be less than 0.001, we need
ˇ (n+1) ˇ 13n+1
ˇ ˇ
ˇ f ( c ) ˇ¨ ă 0.001
( n + 1) !
ˇ
ˇ ( n +1) ˇ
ˇ ( n + 1) !
ˇf ( c )ˇ ă
1000 ¨ 13n+1
It’s a tricky thing to figure out which n makes this true. Let’s make a table. We won’t show all the
work of filling it in, but the work is standard.
( n + 1) ! ˇ
ˇ ( n +1) ˇ
ˇ ˇ ˇ ( n + 1) !
n f ( c ) Is ˇ f (n+1) (c)ˇ ă ?
ˇ ˇ
1000 ¨ 13n+1 1000 ¨ 13n+1
ˇ ˇ
1 1 1
0 | f 1 ( c )| = ă
1000 ¨ 13 7c 6/7 7 ¨ 36
2 6 6
1 | f 2 ( c )| = ă Yes!
1000 ¨ 132 13 72 ¨ 313
72 ¨ c 7
That is: if we use the first-degree Taylor polynomial, then for some c between 2187 and 2200,
ˇ ˇ 132
| f (2200) ´ T1 (2200)| = ˇ f 2 (c)ˇ ¨
2!
6 132
= 13
¨
2
72 ¨ c 7
6 132
ă 2 13 ¨
7 ¨3 2
3 ¨ 132
= 2 13 « 0.0000065
7 ¨3
So, actually, the linear Taylor polynomial (or any higher-degree Taylor polynomial) will result in
an approximation that is much more accurate than required. (We don’t know, however, that the
constant approximation will be accurate enough–so we’d better stick with n ě 1.)
Now that we know we can take the first-degree Taylor polynomial, let’s compute T1 ( x ). Recall we
449
1
are taking the Taylor polynomial for f ( x ) = x 7 about x = 2187.
1
f (2187) = 2187 7 = 3
1 6
f 1 ( x ) = x´ 7
7
1 1 1
f (2187) = ? 6
=
7
7 2187 7 ¨ 36
T1 ( x ) = f (2187) + f 1 (2187)( x ´ 2187)
x ´ 2187
= 3+
7 ¨ 36
2200 ´ 2187
T1 (2200) = 3 +
7 ¨ 36
13
= 3+
7 ¨ 36
« 3.00255
?
7
We conclude 2200 « 3.00255.
S-13: If we’re going to use Equation 3.4.33, then we’ll probably be taking a Taylor polynomial.
Using Example 3.4.16, the 6th-degree Maclaurin polynomial for sin x is
x3 x5
T6 ( x ) = T5 ( x ) = x ´ +
3! 5!
so let’s play with this a bit. Equation 3.4.33 tells us that the error will depend on the seventh
derivative of f ( x ), which is ´ cos x:
17
f (1) ´ T6 (1) = f (7) (c)
7!
1 1 ´ cos c
sin(1) ´ 1 ´ + =
3! 5! 7!
101 ´ cos c
sin(1) ´ =
5! 7!
4242 ´ cos c
sin(1) =
7!
for some c between 0 and 1. Since ´1 ď cos c ď 1,
4242 ´ 1 4242 + 1
ď sin(1) ď
7! 7!
4241 4243
ď sin(1) ď
7! 7!
4241 4243
ď sin(1) ď
5040 5040
Remark: there are lots of ways to play with this idea to get better estimates. One way is to take a
π
higher-degree Maclaurin polynomial. Another is to note that, since 0 ă c ă 1 ă , then
3
450
1
ă cos c ă 1, so
2
4242 ´ 1 4242 ´ 21
ă sin(1) ă
7! 7!
4241 8483 4243
ă sin(1) ă ă
5040 10080 5040
If you got tighter bounds than asked for in the problem, congratulations!
S-14: (a) For every whole number n, the nth derivative of e x is e x . So:
4 4
ÿ e0 n ÿ xn
T4 ( x ) = x =
n =0
n! n =0
n!
(b)
4 4
ÿ 1n ÿ 1
T4 (1) = =
n =0
n! n =0
n!
1 1 1 1 1
= + + + +
0! 1! 2! 3! 4!
1 1 1 1 1
= + + + +
1 1 2 6 24
65
=
24
(c) Using Equation 3.4.33,

1 c
e1 ´ T4 (1) = e for some c strictly between 0 and 1. So,
5!
65 ec
e´ =
24 120
65 ec
e= +
24 120
Since e x is a strictly increasing function, and 0 ă c ă 1, we conclude e0 ă ec ă e1 :
65 1 65 e
+ ăeă +
24 120 24 120
Simplifying the left inequality, we see
326
ăe
120
From the right inequality, we see
65 e
eă +
24 120
e 65
e´ ă
120 24
119 65
e¨ ă
120 24
65 120 325
eă ¨ =
24 119 119
451
So, we conclude
326 325
ăeă ,
120 119
as desired.
326 325
Remark: « 2.717, and « 2.731.
120 119
S-1: The third Maclaurin polynomial for f ( x ) is

f 2 (0) 2 f 3 (0) 3 1
f (0) + f 1 (0) x + ¨x + ¨ x = 4 + 3x2 + x3 .
2 6 2
The coefficient of x is f 1 (0) on one side, and 0 on the other, so f 1 (0) = 0.
1
The coefficient of x2 is f 2 (0) on one side, and 3 on the other, so f 2 (0) = 6.
2
S-2: The third Maclaurin polynomial for h( x ) is
h 2 ( 0 ) 2 h (3) ( 0 ) 3 1 2
h (0) + h1 (0) x + ¨x + ¨ x = 1 + 4x ´ x2 + x3
2 3! 3 3
1 (3) 2
The coefficient of x3 is h (0) on one side, and on the other.
3! 3
h (3) ( 0 ) 2 ( 3 ) 2
Thus = , so h (0) = 6 ¨ = 4.
6 3 3
S-3: The third-degree Taylor polynomial for h( x ) about x = 2 is

h2 (2) h3 (2)
h(2) + h1 (2)( x ´ 2) + ¨ ( x ´ 2)2 + ¨ ( x ´ 2)3
2 6
1 1
The coefficient of ( x ´ 2) is h1 (2) in the definition, and in the given function, so h1 (2) = .
2 2
1 2
The coefficient of ( x ´ 2)2 is h (2) in the definition, and 0 in the given function, so h1 (2) = 0.
2
S-4:
(a) For x near 3,

f ( x ) « f (3) + f 1 (3)( x ´ 3) = 2 + 4( x ´ 3)
In particular
f (2.98) « 2 + 4(2.98 ´ 3) = 2 ´ 0.08 = 1.92
(b) For x near 3,

1 2
f ( x ) « f (3) + f 1 (3)( x ´ 3) + f (3)( x ´ 3)2
2
1
= 2 + 4( x ´ 3) ´ 10( x ´ 3)2
2
452
In particular
f (2.98) « 2 + 4(2.98 ´ 3) ´ 5(2.98 ´ 3)2

= 2 ´ 0.08 ´ 0.002 = 1.918
1 2
S-5: Let’s name g( x ) = x1/3 . Then g1 ( x ) = x´2/3 and g2 ( x ) = ´ x´5/3 .
3 9
1 1 2
In particular, g(8) = 2, g (8) = , and g ( x ) ă 0 for all x ą 0.
12
The tangent line approximation to 101/3 is
g(10) « g(8) + g1 (8)(10 ´ 8)

1 13
= 2 + (2) =
12 6
Using the error formula:
g(10) = g(8) + g1 (8)(10 ´ 8) + 12 g2 (c)(10 ´ 8)2
2 13
for some 8 ă c ă 10. Since g2 (c) = ´ 5/3
, g2 (c) is negative, so g(10) is plus some negative
9c 6
quantity. So, the tangent line approximation is too big.
?
S-6: We use the function f ( x ) = x and point a = 1 as the centre of our approximation, since we
can easily calculate
?
f ( a ) = f (1) = 1 = 1.
1
We compute f 1 ( x ) = ? , so
2 x
1 1
f 1 (1) = ? = .
2 1 2
?
So, a linear approximation of 2 = f (2) is
?
2 « T1 (2) = f (1) + f 1 (1) ¨ (2 ´ 1)
1 3
= 1+ = .
2 2
S-7: We use the function f ( x ) = x1/3 and point a = 27 as the centre of our approximation since we
can easily compute
f (27) = 3
1 ´2/3
We compute f 1 ( x ) = x , so
3
1 1
f 1 (27) = ¨ (27)´2/3 =
3 27
453
So, the linear approximation of 261/3 = f (26) is
261/3 « T1 (26) = f (27) + f 1 (27) ¨ (26 ´ 27)

1 80
= 3´ =
27 27
S-8: We use the function f ( x ) = x5 and point a = 10 as the centre of our approximation since we
know that f ( a) = f (10) = 105 .
Since f 1 ( x ) = 5x4 we have f 1 (10) = 50, 000.
So, a linear approximation of 10.15 is
T1 (10.1) = f (10) + f 1 (10) ¨ (10.1 ´ 10)

= 100, 000 + 50, 000 ¨ 0.1 = 105, 000 .
S-9: We use the function f ( x ) = sin( x ) and point a = π as the centre of our approximation since
we know that
sin( a) = f (π ) = sin π = 0.
101π
and π is reasonably close to .
100
1
We compute f ( x ) = cos( x ), so
f 1 (π ) = cos(π ) = ´1.
So, the linear approximation of sin 101π

100 is

101π 101π
f « T1
100 100

1 101π
= f (π ) + f (π ) ¨ ´π
100
π π
= 0 + (´1) ¨ =´ .
100 100
1 1
S-10: Set f ( x ) = arctan( x ). Then f 1 ( x ) = , so f 1 (1) = and
1 + x2 2
π 1
f (1.1) « f (1) + f 1 (1)(1.1 ´ 1) = +
4 20
S-11: Set f ( x ) = (2 + x )3 , so we are approximating f (0.001).

The obvious choice of a is a = 0.
Then f 1 ( x ) = 3(2 + x )2 , so
12 8012
(2.001)3 = f (0.001) « f (0) + f 1 (0)(0.001 ´ 0) = 8 + =
1000 1000
454
Remark: if we had chosen f ( x ) = x3 and a = 2, the result would have been exactly the same.
S-12: We set f ( x ) = (8 + x )2/3 , and choose a = 0 as our centre.

2
Then f 1 ( x ) = (8 + x )´1/3 , so that
3
(8.06)2/3 = f (0.06) « f (0) + f 1 (0) ¨ 0.06

2
= 82/3 + 8´1/3 ¨ 0.06
3
?3 2 2
= 8 + ? ¨ 0.06
338
2
= 22 + ¨ 0.06
3¨2
1
= 4 + ¨ 0.06
3
402
= 4.02 =
100
S-13: We begin by finding the derivatives of f at x = 0.
f ( x ) = (1 ´ 3x )´1/3 f (0) = 1
´1
f 1 ( x ) = (´3) (1 ´ 3x )´4/3 = (1 ´ 3x )´4/3 f 1 (0) = 1
3
´4
f 2 ( x ) = (´3) (1 ´ 3x )´7/3 = 4(1 ´ 3x )´7/3 f 2 (0) = 4
3
´7
f (3) ( x ) = (´3)(4) (1 ´ 3x )´10/3 = 28(1 ´ 3x )´10/3 f (3) (0) = 28
3
Plugging these into the definition of a Taylor Polynomial, we find that the third–order Taylor
polynomial for f around x = 0 is
4 2 28 3
T3 ( x ) = 1 + x + x + x
2! 3!
14
= 1 + x + 2x2 + x3
3
S-14:
• By Equation 3.4.33, the absolute value of the error is

ˇ 3 ˇ ˇ ˇ
ˇ f (c) ˇ ˇ
3ˇ c ˇ
ˇ
ˇ 3! ¨ ( 2 ´ 1 ) = ˇ ˇ
ˇ ˇ 6(22 ´ c2 ) ˇ
for some c P (1, 2).
455
• When 1 ď c ď 2, we know that 18 ď 22 ´ c2 ď 21, and that numerator and denominator are
non-negative, so
ˇ ˇ
ˇ c ˇ c 2 2
ˇ 6(22 ´ c2 ) ˇ = 6(22 ´ c2 ) ď 6(22 ´ c2 ) ď 6 ¨ 18
ˇ ˇ
1 1
= ď
54 50
as required.
• Alternatively, notice that c is an increasing function of c, while 22 ´ c2 is a decreasing

function of c. Hence the fraction is an increasing function of c and takes its largest value at
c = 2. Hence
ˇ ˇ
ˇ c ˇ 2 1 1
ˇ 6(22 ´ c2 ) ˇ ď 6 ˆ 18 = 54 ď 50 .
ˇ ˇ
S-15:
• By Equation 3.4.33, there is c P (0, 0.5) such that the error is
f (4) ( c )
R4 = (0.5 ´ 0)4
4!
1 cos(c2 )
= ¨
24 ¨ 16 3 ´ c
• For any c we have | cos(c2 )| ď 1, and for c ă 0.5 we have 3 ´ c ą 2.5, so that
ˇ cos(c2 ) ˇ
ˇ ˇ
1
ˇ 3 ´ c ˇ ď 2.5 .
ˇ ˇ
• We conclude that
1 1 1 1 1
|R4 | ď = ă = ă
2.5 ¨ 24 ¨ 16 60 ¨ 16 60 ¨ 10 600 500
S-16:
• By Equation 3.4.33, there is c P (0, 1) such that the error is

ˇ 3 ˇ ˇ ˇ
ˇ f (c) 3
ˇ ˇ eć ˇ
ˇ 3! ¨ (1 ´ 0) ˇ = ˇ 6(8 + c2 ) ˇ .
ˇ ˇ ˇ ˇ
• When 0 ă c ă 1, we know that 1 ą eć ą e´1 and 8 ď 8 + c2 ă 9, so

ˇ ˇ
ˇ eć ˇ eć
ˇ ˇ=
ˇ 6(8 + c2 ) ˇ 6(8 + c2 )
1
ă
6|8 + c2 |
1 1 1
ă = ă
6ˆ8 48 40
as required.
456
• Alternatively, notice that eć is a decreasing function of c, while for 0 ă c 8 + c2 is an
increasing function of c. Hence the fraction is a decreasing function of c and takes its largest
value at c = 0. Hence
c
ˇ ˇ
ˇ e ˇ 1 1 1
ˇ 6(8 + c2 ) ˇ ď 6 ˆ 8 = 48 ă 40 .
ˇ ˇ
S-17: (a), (b):

Let f ( x ) = x1/3 and x0 = 27. Then
1 ´2/3 2
f ( x ) = x1/3 f 1 (x) = x f 2 ( x ) = ´ x´5/3
3 9
1 1 1 2 1 2
f (27) = 271/3 = 3 f 1 (27) = ¨ 2 = f 2 (27) = ´ ¨ 5 = ´
3 3 27 9 3 2187
so that, with x = 25,
2
52/3 = f (25) « f (27) + f 1 (27)(25 ´ 27) = 3 ´ « 2.9259 (linear approx)
27
1 2
52/3 = f (25) « f (27) + f 1 (27)(25 ´ 27) + f (27)(25 ´ 27)2
2
2 1 2¨4
= 3´ ´ ¨ « 2.9241 (quadratic app)
27 2 2187
(c) To obtain an error estimate for the linear approximation, we use that
1 2
52/3 = f (25) = f (27) + f 1 (27)(25 ´ 27) + f (z)(25 ´ 27)2
2
for some z between 25 and 27. The error is exactly
ˇ ˇ ˇ ˇ
ˇ1 2
ˇ f (z)(25 ´ 27)2 ˇ = ˇ 1 ´ 2 x´5/3 (´2)2 ˇ = 4 z´5/3 .
ˇ ˇ ˇ
ˇ2 ˇ ˇ2 9 ˇ 9
For z between 25 and 27, z´5/3 is between 25´5/3 and 27´5/3 . The biggest this can be is 25´5/3 , so
4
the maximum possible error is 25´5/3 .
9
To get a better idea of what this number is, we note 2.93 ă 25, so 49 25´5/3 ă 49 2.9´5 = 0.0022.
S-18: The fourth-degree Maclaurin polynomial for f ( x ) is

1 2 1 1
T4 ( x ) = f (0) + f 1 (0) x + f ( 0 ) x 2 + f 3 ( 0 ) x 3 + f (4) ( 0 ) x 4
2 3! 4!
1 2 1
T3 ( x ) = f (0) + f 1 (0) x + f (0) x 2 + f 3 (0) x 3
2 3!
So, we simply “chop off” the part of T4 ( x ) that includes x4 . Since that’s already 0, in this case
T3 ( x ) = T4 ( x ).
T3 ( x ) = 5x2 ´ 9
457
S-19: y is a function of x that obeys
y( x )4 + xy( x ) = x2 ´ 1
By implicit differentiation (and then subbing in x = 2, y(2) = 1)
4y( x )3 y1 ( x ) + y( x ) + xy1 ( x ) = 2x
4y1 (2) + 1 + 2y1 (2) = 4
1
y1 (2) =
2
Differentiating with respect to x a second time and then subbing in
x = 2, y(2) = 1, y1 (2) = 12 :
12y( x )2 y1 ( x )2 + 4y( x )3 y2 ( x ) + y1 ( x ) + y1 ( x ) + xy2 ( x ) = 2
1 1 1
12 ˆ 1 ˆ + 4y2 (2) + + + 2y2 (2) = 2
4 2 2
6y2 (2) = ´2
1
y2 (2) = ´
3
The tangent line approximation to y( x ) at x = 2 is
1
y( x ) « y(2) + y1 (2)( x ´ 2) = 1 + ( x ´ 2)
2
In particular,
1
y(2.1) « y(2) + y1 (2)(2.1 ´ 2) = 1 + (.1) = 1.05
2
The quadratic approximation to y( x ) at x = 2 is
1
y( x ) « y(2) + y1 (2)( x ´ 2) + y2 (2)( x ´ 2)2
2
1 1
= 1 + ( x ´ 2) ´ ( x ´ 2)2
2 6
In particular,
1
y(2.1) « y(2) + y1 (2)(2.1 ´ 2) + y2 (2)(2.1 ´ 2)2
2
1 1
= 1 + (.1) ´ (.1)2 = 1.0483
2 6
At x = 2, y = 1 and y1 = 21 . So the tangent line passes through (2, 1) and has slope 21 . At x = 2,
y2 = ´ 13 , so the graph y = f ( x ) (locally!) looks like a parabola pointing down near x = 2. This
gives the graph fragment below.
Alternatively, we could observe that, near x = 2, y( x ) will be quite close to its quadratic
approximation, 1 + 12 ( x ´ 2) ´ 61 ( x ´ 2)2 .
y
y = 1 + 12 (x − 2)
y = y(x) ≈ 1 + 21 (x − 2) − 61 (x − 2)2
1
2 x
458
S-20: (a) y is a function of x that obeys
1 = x4 + y( x ) + xy( x )4
By implicit differentiation (and then subbing in x = ´1, y(´1) = 1)
0 = 4x3 + y1 ( x ) + y( x )4 + 4xy( x )3 y1 ( x )
0 = ´4 + y1 (´1) + 1 ´ 4y1 (´1)
´1 = y1 (´1)
Differentiating with respect to x a second time and then subbing in x = ´1, y(´1) = 1, and
y1 (´1) = ´1:
0 = 12x2 + y2 ( x ) + 4y( x )3 y1 ( x ) + 4y( x )3 y1 ( x ) + 12xy( x )2 y1 ( x )2 + 4xy( x )3 y2 ( x )

0 = 12 + y2 (´1) ´ 4 ´ 4 ´ 12 ´ 4y2 (´1)
´8 = 3y2 (´1)
8
y2 (´1) = ´
3
The tangent line approximation to y( x ) at x = ´1 is
y( x ) « y(´1) + y1 (´1)( x + 1) = 1 ´ ( x + 1) = ´x
In particular,
y(´0.9) « 0.9
(b) The quadratic approximation to y( x ) at x = ´1 is
1
y( x ) « y(´1) + y1 (´1)( x + 1) + y2 (´1)( x + 1)2
2
4 2
= 1 ´ ( x + 1) ´ ( x + 1)
3
In particular,
4
y(´0.9) « 1 ´ (.1) ´ (.1)2 « 0.8867
3
(c) At x = ´1, the slope of the curve is y1 (´1) = ´1. Its tangent line is falling at 45˝ . At x = ´1,
y2 (´1) = ´ 83 , so the slope of the curve is decreasing as x passes through ´1. Zoomed in very
close, the curve looks like a parabola opening downwards. This gives the figure
y
y = −1 − (x + 1)
y = y(x) ≈ −1 − (x + 1) − 43 (x + 1)2 1
x −1
459
S-21: Let f ( x ) = log x and x0 = 10. Then
1 1
f ( x ) = log x f 1 (x) = f 2 (x) = ´
x x2
1 1 1
f (10) = log 10 « 2.30259 f (10) = f 2 (10) = ´
10 100
so that, with x = 10.3,
0.3
log 10.3 = f (10.3) « f (10) + f 1 (10)(10.3 ´ 10) = 2.30259 + = 2.33259
10
The error in this approximation (excluding the error in the given data log 10 « 2.30259) is
1 2 1
f (z)(10.3 ´ 10)2 for some z between 10 and 10.3. Because f 2 (z) = ´ 2 increases as z increases,
2 z
1 1 1 2 2
it must be between ´ 2 and ´ . This forces f (z)(10.3 ´ 10) to be between
10 10.32 2
1 1 2 1 1 2
´ ¨ 2 (0.3) = ´0.00045 and ´ ¨ (0.3) ă ´0.00042.
2 10 2 10.32
S-22: We begin by finding the values of the derivatives of f at x = 0. We can use the chain rule, or
the formula we found in Question 19, Section 2.14.
x
f ( x ) = ee f (0) = e
1 x ex
f (x) = e e f 1 (0) = e
x
f 2 ( x ) = e x + e2x ee f 2 (0) = 2e
x
f 3 ( x ) = e x + 3e2x + e 3x
ee

(a) L( x ) = f (0) + f 1 (0)( x ´ 0) = e + ex

1 2
(b) Q( x ) = f (0) + f 1 (0)( x ´ 0) + f (0)( x ´ 0)2 = e + ex + ex2
2
(c) Since ex2 ą 0 for all x ą 0, L( x ) ă Q( x ) for all x ą 0.
From the error formula, we know that
1 3
f ( x ) = f (0) + f 1 (0) x + 12 f 2 (0) x2 + f (c) x3
3!
1 c c
e + 3e2c + e3c ee x3
= Q( x ) +
6
1
c
for some c between 0 and x. Since 6 ec + 3e2c + e3c ee is positive for any c, for all x ą 0,
1 c 2c 3c eec x 3 ą 0, so Q ( x ) ă f ( x ).

6 e + 3e + e
1 c 2
(d) Write g( x ) = e x = 1 + x + e x , for some c between 0 and x. For x = 0.1 we have 0 ă c ă 0.1
2!
and 1 ă ec ă e0.1 ă e ă 3. So
e0.1 = f (0.1) = 1 + 0.1 + 12 ec (0.1)2 ą 1 + 0.1 + 21 (1)(0.1)2 = 1.105
0.1 1 c 2 1 2
e = f (0.1) = 1 + 0.1 + 2 e (0.1) ă 1 + 0.1 + 2 (3)(0.1) = 1.115
That is, 1.105 ă e0.1 ă 1.115.
460
S-1:
y
y = f (x)
x
a
When x = 0, the curve y = f ( x ) appears to have a flat tangent line, so the x = 0 is a critical point.
However, it is not a local extremum: it is not true that f (0) ě f ( x ) for all x near 0, and it is not
true that f (0) ď f ( x ) for all x near 0.
To the right of the x-axis, there is a spike where the derivative of f ( x ) does not exist. The x-value
corresponding to this spike (call it a) is a singular point, and f ( x ) has a local maximum at x = a.
S-2:
y
y = f (x)
x
a b
The x-coordinate corresponding to the blue dot (let’s call it a) is a critical point, because the
tangent line to f ( x ) at x = a is horizontal. There is no lower point nearby, and actually no lower
point on the whole interval shown, so f ( x ) has both a local minimum and a global minimum at
x = a.
If a function is not continuous at a point, then it is not differentiable at that point. So, the
461
x-coordinate corresponding to the discontinuity (let’s call it b) is a singular point. Values of f ( x )
immediately to the right of b are lower, and values immediately to the left of b are higher, so f ( x )
has no local (or global) extremum at x = b.
S-3: One possible answer is shown below.
x
2
For every x in the red interval shown below, f (2) ě f ( x ), so f (2) is a local maximum. However,
the point marked with a blue dot shows that f ( x ) ą f (2) for some x, so f (2) is not a global
maximum.
y
x
2
S-4: Critical points are those values of x for which f 1 ( x ) = 0, and singular points are those values
of x for which f ( x ) is not differentiable. So, we ought to find f 1 ( x ). Using the quotient rule,
(1)( x2 + 3) ´ ( x ´ 1)(2x )
f 1 (x) =
( x 2 + 3)2
´x2 + 2x + 3
=
( x 2 + 3)2
( x ´ 3)( x + 1)
=´
( x 2 + 3)2
462
(a) The derivative f 1 ( x ) is zero when x = 3 and when x = ´1, so those are the critical points.
(b) The denominator of f 1 ( x ) is never zero, so the derivative f 1 ( x ) exists for all x and f ( x ) has no
singular points.
(c) Theorem 3.5.4 tells us that local extrema of f ( x ) can only occur at critical points and singular
points. So, the possible points where extrema of f ( x ) may exist are x = 3 and x = ´1.
S-5:
y y y y
x x x x
2 2 2 2
local max neither neither local max
For the first curve, the function’s value at x = 2 (that is, the y-value of the solid dot) is higher than
anything around it. So, it’s a local maximum.
For the second curve, the function’s value at x = 2 (that is, the y-value of the solid dot) is higher
than everything to the left, but lower than values immediately to the right. (On the graph
reproduced below, f ( x ) is higher than everything in the red section, and lower than everything in
the blue section.) So, it is neither a local max nor a local min.
y
x
2
Similarly, for the third curve, f (2) is lower than the values to the right of it, and higher than
values to the left of it, so it is neither a local minimum nor a local maximum.
In the final curve, f (2) (remember–this is the y-value of the solid dot) is higher than everything
immediately to the left or right of it (for instance, over the interval marked in red below), so it is a
local maximum.
y
x
2
S-6: The question specifies that x = 2 must not be an endpoint. By Theorem 3.5.4, if x = 2 not a
463
critical point, then it must be a singular point. That is, f ( x ) is not differentiable at x = 2. Two
possibilities are shown below, but there are infinitely many possible answers.
y y
x x
2 2
S-7: Critical points are those values of x for which f 1 ( x ) = 0, and singular points are those values
of x for which f ( x ) is not differentiable. So, we ought to find f 1 ( x ). Since f ( x ) has an absolute
value sign, let’s re-write it in a version that is friendlier to differentiation. Remember that |X| = X
when X ě 0, and |X| = ´X when X ă 0.
b
f ( x ) = |( x ´ 5)( x + 7)|
" a
( x ´ 5)( x + 7) if ( x ´ 5)( x + 7) ě 0
= a
´( x ´ 5)( x + 7) if ( x ´ 5)( x + 7) ă 0
The product ( x ´ 5)( x + 7) is positive when ( x ´ 5) and ( x + 7) have the same sign, and negative
when they have opposite signs, so
" a
( x ´ 5)( x + 7) if x P (´8, ´7] Y [5, 8)
f (x) = a
´( x ´ 5)( x + 7) if x P (´7, 5)
Now, when x ‰ ´7, 5, we can differentiate, using the chain rule.
$ d
t( x´5)( x +7)u
’
’ dx?
if x P (´8, ´7) Y (5, 8)
& 2 (x´5)(x+7)
’
d
f 1 (x) = dx?t´( x´5)( x +7)u
if x P (´7, 5)
’
’
’ 2 ´( x´5)( x +7)
?? if x = ´7, x = 5
%
$
2x + 2
’ ? if x P (´8, ´7) Y (5, 8)
& 2 (x´5)(x+7)
’
= ? ´2x´2 if x P (´7, 5)
’
’ 2 ´( x´5)( x +7)
% ?? if x = ´7, x = 5
We are tempted to say that the derivative doesn’t exist when x = ´7 and x = 5, but be careful–
we don’t actually know that yet. The formulas we have for the f 1 ( x ) are only good when x is not
´7 or 5.
´2x ´ 2
The middle formula a tells us x = ´1 is a critical point: when x = ´1, f 1 ( x ) is
2 ´( x ´ 5)( x + 7)
given by the middle line, and it is 0. Note that x = ´1 also makes the top formula 0, but f 1 (´1) is
not given by the top formula, so that doesn’t matter.
What we’ve concluded so far is that x = ´1 is a critical point of f ( x ), and f ( x ) has no other
critical points or singular points when x ‰ ´7, 5. It remains to figure out what’s going on at ´7
and 5. One way to do this is to use the definition of the derivative to figure out what f 1 (´7) and
f 1 (5) are, if they exist. This is somewhat laborious. Let’s look for a better way.
464
• First, let’s notice that f ( x ) is defined for all values of x, thanks to that handy absolute value
sign.
• Next, notice f ( x ) ě 0 for all x, since square roots never give a negative value.
• Then if there is some value of x that gives f ( x ) = 0, that x gives a global minimum, and
therefore a local minimum.
• f ( x ) = 0 exactly when ( x ´ 5)( x + 7) = 0, which occurs at x = ´7 and x = 5
• Therefore, f ( x ) has global and local minima at x = ´7 and x = 5
• So, x = ´7 and x = 5 are critical points or singular points by Theorem 3.5.4.
So, all together:

x = ´1 is a critical point, and x = ´7 and x = 5 are critical points or singular points (but we don’t
know which).
Remark: if you would like a review of how to use the definition of the derivative, below we show
that f ( x ) is not differentiable at x = ´7. (In fact, x = ´7 and x = 5 are both singular points.)
f (´7 + h) ´ f (´7)
f 1 (´7) = lim
hÑ0 h
a a
|(´13 + h)(h)| ´ |0|
= lim
hÑ0 h
a
|(´13 + h)(h)|
= lim
hÑ0 h
Let’s first consider the case h ą 0.
a a
|(´13 + h)(h)| (13 ´ h)(h)
lim = lim+
hÑ0 + h hÑ0 h
?
13h ´ h2
= lim+ ?
2
c h
hÑ0
13h ´ h2
= lim+
hÑ0 h2
c
13
= lim+ ´1
hÑ0 h
=8
Since one side of the limit doesn’t exist,
f (´7 + h) ´ (´7)
lim = DNE
hÑ0 h
so f 1 ( x ) is not differentiable at x = ´7. Therefore, x = ´7 is a singular point.
465
S-8: For any real number c, c is in the domain of f ( x ) and f 1 (c) exists and is equal to zero. So,
following Definition 3.5.6, every real number is a critical point of f ( x ), and f ( x ) has no singular
points.
For every number c, let a = c ´ 1 and b = c + 1, so a ă c ă b. Then f ( x ) is defined for every x in

the interval [ a, b], and f ( x ) = f (c) for every a ď x ď b. That means f ( x ) ď f (c) and f ( x ) ě f (c).
So, comparing with Definition 3.5.3, we see that f ( x ) has a global and local maximum AND
minimum at every real number x = c.
S-1: Two examples are given below, but many are possible.
y
x
y = ´x2
a
y = ´ |x|
a
If f ( x ) = ´x2 or f ( x ) = ´ |x|, then f ( x ) has a global maximum at x = 0. Since f ( x ) keeps
getting more and more strongly negative as x gets farther and farther from 0, f ( x ) has no global
minimum.
S-2: Two examples are given below, but many are possible.
y
y = ex
If f ( x ) = e x , then f ( x ) ą 0 for all x. As we move left along the x-axis, f ( x ) gets smaller and
smaller, approaching 0 but never reaching it. Since f ( x ) gets smaller and smaller as we move left,
there is no global minimum. Likewise, f ( x ) increases more and more as we move right, so there
is no maximum.
466
y
y = arctan x + 2

If f ( x ) = arctan( x ) + 2, then f ( x ) ą ´ π2 + 2 ą 0 for all x.

As we move left along the x-axis, f ( x ) gets smaller and smaller, approaching ´ π2 + 2 but never
reaching it. Since f ( x ) gets smaller and smaller as we move left, there is no global minimum.

Likewise, as we move right along the x-axis, f ( x ) gets bigger and bigger, approaching π2 + 2
but never reaching it. Since f ( x ) gets bigger and bigger as we move right, there is no global
maximum.
S-3: Since f (5) is a global minimum, f (5) ď f ( x ) for all x, and so in particular f (5) ď f (´5).
Similarly, f (´5) ď f ( x ) for all x, so in particular f (´5) ď f (5).
Since f (´5) ď f (5) AND f (5) ď f (´5), it must be true that f (´5) = f (5).
A sketch of one such graph is below.
y = f (x)
x
´5 5
S-4: Global extrema will occur at critical or singular points in the interval (´5, 5) or at the
endpoints x = 5, x = ´5.
f 1 ( x ) = 2x + 6. Since this is defined for all real numbers, there are no singular points. The only
time f 1 ( x ) = 0 is when x = ´3. This is inside the interval [´5, 5]. So, our points to check are
x = ´3, x = ´5, and x = 5.
467
c ´3 ´5 5
type critical point endpoint endpoint
f (c) ´19 ´15 45
The global maximum is 45 at x = 5 and the global minimum is ´19 at x = ´3.
S-5: Global extrema will occur at the endpoints of the interval, x = ´4 and x = 0, or at singular or
critical points inside the interval. Since f ( x ) is a polynomial, it is differentiable everywhere, so
there are no singular points. To find the critical points, we set the derivative equal to zero.
f 1 ( x ) = 2x2 ´ 4x ´ 30
0 = 2x2 ´ 4x ´ 30 = (2x ´ 10)( x + 3)
x = 5, ´3
The only critical point inside the interval is x = ´3.

c ´3 ´4 0
type critical point endpoint endpoint
157 1
f (c) 61 3 = 52 + 3 7
The global maximum over the interval is 61 at x = ´3, and the global minimum is 7 at x = 0.
S-1: We compute f 1 ( x ) = 5 x4 ´ 5, which means that f ( x ) has no singular points (i.e., it is

differentiable for all values of x), but it has two critical points:
0 = 5x4 ´ 5
0 = x4 ´ 1 = ( x2 + 1)( x2 ´ 1)
0 = x2 ´ 1
x = ˘1
Note, however, that 1 is not in the interval [´2, 0].

The global maximum and the global minimum for f ( x ) on the interval [´2, 0] will occur at
x = ´2, x = 0, or x = ´1.
c ´2 0 ´1
type endpoint endpoint critical point
f (c) ´20 2 6
So, the global maximum is f (´1) = 6 while the global minimum is f (´2) = ´20.
S-2: We compute f 1 ( x ) = 5x4 ´ 5, which means that f ( x ) has no singular points (i.e., it is
differentiable for all values of x), but it has two critical points:
0 = 5x4 ´ 5
0 = x4 ´ 1 = ( x2 + 1)( x2 ´ 1)
0 = x2 ´ 1
x = ˘1
468
Note, however, that ´1 is not in the interval [0, 2].
The global maximum and the global minimum for f ( x ) on the interval [0, 2] will occur at x = 2,
x = 0, or x = 1.
c 2 0 1
f (c) 12 ´10 ´14
So, the global maximum is f (2) = 12 while the global minimum is f (1) = ´14.
S-3: We compute f 1 ( x ) = 6x2 ´ 12x = 6x ( x ´ 2), which means that f ( x ) has no singular points
(i.e., it is differentiable for all values of x), but it has the two critical points: x = 0 and x = 2. Note,
however, 0 is not in the interval [1, 4].
c 1 4 2
f (c) ´6 30 ´10
So, the global maximum is f (4) = 30 while the global minimum is f (2) = ´10.
S-4: Since h( x ) is a polynomial, it has no singular points. We compute its critical points:
h1 ( x ) = 3x2 ´ 12
0 = 3x2 ´ 12
x = ˘2
Notice as x Ñ 8, h( x ) Ñ 8, and as x Ñ ´8 h( x ) Ñ ´8. So Theorem 3.5.17 doesn’t exactly
apply. Instead, let’s consider the signs of h1 ( x ).
x (´8, ´2) (´2, 2) (2, 8)
h1 ( x ) ą0 ă0 ą0
h( x ) increasing decreasing increasing
So, h( x ) increases until x = ´2, then decreases. That means h( x ) has a local maximum at x = ´2.
The function decreases from ´2 until 2, after which is increases, so h( x ) has a local minimum at
x = 2. We compute f (´2) = 20 and f (2) = ´12.
S-5: Since h( x ) is a polynomial, it has no singular points. We compute its critical points:
h1 ( x ) = 6x2 ´ 24
0 = 6x2 ´ 24
x = ˘2
Notice as x Ñ 8, h( x ) Ñ 8, and as x Ñ ´8 h( x ) Ñ ´8. So Theorem 3.5.17 doesn’t exactly
apply. Instead, let’s consider the signs of h1 ( x ).
x (´8, ´2) (´2, 2) (2, 8)
h1 ( x ) ą0 ă0 ą0
h( x ) increasing decreasing increasing
So, h( x ) increases until x = ´2, then decreases. That means h( x ) has a local maximum at x = ´2.
The function decreases from ´2 until 2, after which is increases, so h( x ) has a local minimum at
x = 2.
469
We compute f (´2) = 33 and f (2) = ´31.
S-6: Suppose that Q is a distance of x from A. Then it is a distance of 18 ´ x from B.

x 18 ´ x
A Q B
12 km
P
?
Using the Pythagorean Theorem, the distance from P to Q is 122 + x2 ?
kilometres, and the buggy
122 + x2
travels 15 kph over this off-road stretch. The travel time from P to Q is hours.
15
The distance from Q to B is 18 ´ x kilometres, and the dune buggy travels 30 kph along this road.
18 ´ x
The travel time from Q to B is hours. So, the total travel time is
30
?
122 + x2 18 ´ x
f (x) = + .
15 30
We wish to minimize this for 0 ď x ď 18. We will test all singular points, critical points, and
endpoints to find which yields the smallest value of f ( x ). Since there are no singular points, we
begin by locating the critical points.
1 1 1
0 = f 1 (x) = ¨ (144 + x2 )´1/2 (2x ) ´
15 2 30
1 x 1
¨? =
15 144 + x2 30
x 1
? =
144 + x 2 2
x 2 1
=
144 + x2 4
4x2 = 144 + x2
12 ?
x= ? =4 3
3
?
So the minimum travel times must be one of f (0), f (18), and f 4 3 .
12 18
f (0) = + = 1.4
15 30
?
122 + 182
f (18) = « 1.44
? 15 ?
? 144 + 144/3 18 ´ 12/ 3
f 4 3 = + « 1.29
15 30
?
So Q should be 4 3 km from A.
470
S-7: Let `, w and h denote the length, width and height of the box respectively. We are told that
4500 4500 1500
`wh = 4500 and that ` = 3w. Hence h = = 2
= . The surface area of the box is
`w 3w w2

1500 1500 6000 2000
A = 2`w + 2`h + 2wh = 2 3w2 + 3w 2 + w 2 = 2 3w2 + = 6 w2 +
w w w w
`w
`h
h wh
`
w
As w tends to zero or to infinity, the surface area approaches infinity. By Theorem 3.5.17 the
2000
minimum surface area must occur at a critical point of w2 + .
w
" *
d 2 2000
0= w +
dw w
2000
= 2w ´ 2
w
2000
2w =
w2
w3 = 1000
w = 10
Therefore,
` = 3w = 30
1500
h= = 15.
w2
The dimensions of the box with minimum surface area are 10 ˆ 30 ˆ 15.
S-8: Let the length of the sides of the square base be b metres and let the height be h metres. The
area of the base is b2 , the area of the top is b2 and the area of each of the remaining four sides is bh
so the total cost is
b2o)n + 1looooomooooon
5o(mo
lo (b2 + 4bh) = 6b2 + 4bh = 72
cost of base cost of 5 sides
Solving for h,
72 ´ 6b2
h=
4b
6 12 ´ b2

=
4 b
3 12 ´ b2

=
2 b
471
The volume is
12 ´ b2

3
V = b2 h = b2 ¨
2 b
3
= 18b ´ b3 .
2
This is the function we want to maximize. Since volume is never negative, the endpoints of the
functions are the values of b that make the volume 0. So, the maximum volume will not occur at
an endpoint, it will occur at a critical point. The only critical point is b = 2:
" *
d 3 3
0= 18b ´ b
db 2
9 2
= 18 ´ b
2
2
b =4

3 12 ´ 4
b = 2, h = =6
2 2
The desired dimensions are 2 ˆ 2 ˆ 6.
S-9: It suffices to consider X and Y such that the line XY is tangent to the circle. Otherwise we
could reduce the area of the triangle by, for example, holding X fixed and reducing Y. So let X
1
and Y be the x– and y–intercepts of the line tangent to the circle at (cos θ, sin θ ). Then = cos θ
X
1 π
and = cos ´ θ = sin θ. The area of the triangle is
Y 2
1 1 1
XY = =
2 2 cos θ sin θ sin(2θ )
y
Y
θ x
X
π 1
This is a minimum when sin(2θ ) is a maximum. That is when 2θ = . Hence X = and
2 cos(π/4)
1 ?
Y= . That is, X = Y = 2.
sin(π/4)
S-10: For ease of notation, we place the semicircle on a Cartesian plane with diameter along the
x-axis and centre at the origin.
472
y
x
x
If x is the point where the rectangle touches the diameter to the right of the y-axis, then 2x is the
width of the rectangle. The origin and the two right corners of the rectangle form a right triangle
with hypotenuse
? R, so by the Pythagorean Theorem, the upper right hand corner of the rectangle
2 2
is at x, R ´ x . The perimeter of the rectangle is given by the function:
a
P( x ) = 4x + 2 R2 ´ x2
So, this is what we optimize. The endpoints of the domain for this function are x = 0 and x = R.
To find the critical points, we differentiate:
2x
P1 ( x ) = 4 ´ ?
R2 ´ x 2
2x
P1 ( x ) = 0 ðñ 4 = ?
R2 ´ x 2
a
x = 2 R2 ´ x 2
x 2 = 4( R2 ´ x 2 )
5x2 = 4R2
2
x=? R
5
Note that since our perimeter formula was defined to work only for x in [0, R], we neglect the
2
negative square root, ´ ? R.
5
Now, we find the size of the perimeter at the critical point and the endpoints:
c 0 R ?2 R
5
type endpoint endpoint critical ? point
P(c) 2R 4R 2 5R
?
So, the largest possible perimeter is 2 5R and the smallest possible perimeter is 2R.
Remark: as a check on the correctness of our formula for P( x ), when x = 0 the rectangle
degenerates to the line segment from (0, 0) to (0, R). The perimeter of this “width zero rectangle”
is 2R, agreeing with P(0). Similarly, when x = R the rectangle degenerates to the line segment
from ( R, 0) to (´R, 0). The perimeter of this “width zero rectangle” is 4R, agreeing with P( R).
S-11: Let the cylinder have radius r and height h. If we imagine popping off the ends, they are
two circular disks, each with surface area πr2 . Then we imagine unrolling the remaining tube. It
473
has height h, and its other dimension is given by the circumference of the disks, which is 2πr.
Then the area of the “unrolled tube” is 2πrh.
h (2πr )h h
πr2
So, the surface area is 2πr2 + 2πrh. Since the area is given as A, we can solve for h:
A = 2πr2 + 2πrh
2πrh = A ´ 2πr2
A ´ 2πr2
h= .
2πr
Then we can write the volume as a function of the variable r and the constant A:
V (r ) = πr2 h
A ´ 2πr2

2
= πr
2πr
1
Ar ´ 2πr3

=
2
This is the function we want to maximize. Let’s find its critical points.
1
V 1 (r ) = A ´ 6πr2

2 c
A
V 1 (r ) = 0 ðñ A = 6πr2 ðñ r =
6π
474
since negative values of r don’t make sense. At this critical point,
c !  c ! c !3 
A 1 A A
V = A ´ 2π 
6π 2 6π 6π
1 A3/2 2πA3/2

= ? ´ ?
2 6π 6π 6π
1 A3/2 A3/2

= ? ´ ?
2 6π 3 6π
A3/2
= ? .
3 6π
We should also check the volume of the cylinder at the endpoints of the function. Since r ě 0, one
endpoint is r = 0. Since h ě 0, and r grows as h shrinks, the other endpoint is whatever value of r
causes h to be 0. We could find this value of r, but it’s not strictly necessary: when r = 0, the
volume of the cylinder is zero, and when h = 0, the volume of the cylinder is still zero. So, the
maximum volume does not occur at the endpoints.
Therefore, the maximum volume is achieved at the critical point, where
A3/2
Vmax = ? .
3 6π
Remark: as a check, A has units m2 and, because of the A3/2 , our answer has units m3 , which are
the correct units for a volume.
S-12: Denote by r the radius of the semicircle, and let h be the height of the recangle.
πr
h h h
2r
Since the perimeter is required to be P, the height, h, of the rectangle must obey
P = πr + 2r + 2h
1
h = ( P ´ πr ´ 2r )
2
475
So the area is
A(r ) = 21 πr2 + 2rh
= 12 πr2 + r ( P ´ πr ´ 2r )
= rP ´ 12 (π + 4)r2
Finding all critical points:
0 = A1 (r ) = P ´ ( π + 4)r
P
r=
π+4
Now we want to know what radius yields the maximum area. We notice that A1 (r ) ą 0 for
P P
ră and A1 (r ) ă 0 for r ą . So, A(r ) is increasing until the critical point, then
π+4 π+4
P
decreasing after it. That means the global maximum occurs at the critical point, r = . The
π+4
maximum area is
1 P2 1 P2
rP ´ (π + 4)r2 = ´ ( π + 4)
2 π+4 2 ( π + 4)2
P2
=
2( π + 4)
Remark: another way to see that the global maximum occurs at the critical point is to compare the
area at the critical point to the areas at the endpoints of the function. The smallest valueof r is 0,
P P
while the biggest is (when the shape is simply a half-circle). Comparing A(0), A ,
π+2 π+2
P
and A is somewhat laborious, but certainly possible.
π+4
S-13:
z
x
y = px
(a) The surface area of the pan is

xy + 2xz + 2yz = px2 + 2xz + 2pxz
= px2 + 2(1 + p) xz
and the volume of the pan is xyz = px2 z. Assuming that all A cm2 is used, we have the constraint
A ´ px2
px2 + 2(1 + p) xz = A or z=
2(1 + p ) x
So
A ´ px2

V ( x ) = xyz = x ( px )
2(1 + p ) x
p
= x ( A ´ px2 )
2(1 + p )
476
Using the product rule,
p
V 1 (x) = x (´2px ) + ( A ´ px2 )

2(1 + p )
p
A ´ 3px2

=
2(1 + p )
d
1 A
The derivative V ( x ) is 0 when x = . The derivative is positive (i.e. V ( x ) is increasing) for
3p
d d
A A
xă and is negative (i.e. V ( x ) is decreasing) for x ą . So the pan of maximum volume
3p 3p
d d c a
A A Ap 2A/3 Ap
has dimensions x = ,y= p = and z = a =? .
3p 3p 3 2(1 + p) A/(3p) 3(1 + p )
(b) The volume of the pan from part (a) is
d ! d ! a 3/2 ?
A A Ap A p
V ( p) = p ? =
3p 3p 3(1 + p ) 3 1+ p
Since
? 1
" ? * 1 ? ? p ´1
d p 2 (1 +p)/ p ´ p p
= =
dp 1+ p (1 + p )2 2(1 + p )2
the volume is increasing with p for p ă 1 and decreasing with p for p ą 1. So the maximum
volume is achieved for p = 1 (a square base).
S-14: (a) We use logarithmic differentiation.
f (x) = xx
log f ( x ) = log ( x x ) = x log x
d d
tlog f ( x )u = tx log xu
dx dx
f 1 (x)

1
=x + log x = 1 + log x
f (x) x
f 1 ( x ) = f ( x ) (1 + log x ) = x x (1 + log x )
(b) Since x ą 0, x x ą 0. Therefore,

1
f 1 ( x ) = 0 ðñ 1 + log x = 0 ðñ log x = ´1 ðñ x =
e
(c) Since x ą 0, x x ą 0. So, the sign of f 1 ( x ) is the same as the sign of 1 + log x.
1 1
For x ă , log x ă ´1 and f 1 ( x ) ă 0. That is, f ( x ) decreases as x increases, when x ă . For
e e
1 1 1
x ą , log x ą ´1 and f ( x ) ą 0. That is, f ( x ) increases as x increases, when x ą . Hence f ( x ) is
e e
1
a local minimum at x = .
e
477
S-15: Call the length of the wire L units and suppose that it is cut ` units from one end. Make the
square from the piece of length `, and make the circle from the remaining piece of length L ´ `.
2
`
The square has perimeter `, so its side length is `/4 and its area is . The circle has
4
L´` 2 ( L ´ `)2

L´`
circumference L ´ `, so its radius is and its area is π = .
2π 2π 4π
The area enclosed by the shapes, when the square is made from a length of size `, is
`2 ( L ´ `)2
A(`) = +
16 4π
We want to find the global max and min for this function, given the constraint 0 ď ` ď L, so we
find its derivative:
` L´` π+4 L
A1 (`) = ´ = `´
8 2π 8π 2π
Now, we find the critical point.
A1 (`) = 0
π+4 L
`=
8π 2π
4L
`=
π+4
.. 4L
` 0 L
. π +4
.. L2 L2
A π4L

A(`) 4π 16 +4
.
It seems obnoxious to evaluate A π4L

+4 , and the problem doesn’t ask for it–but we still have to
figure out whether it is a global max or min.
When ` ă π4L 1 4L 1 4L
+4 , A (`) ă 0, and when ` ą π +4 , A (`) ą 0. So, A (`) is decreasing until ` = π +4 ,
then increasing. That means our critical point ` = π4L +4 is a local minimum.

4
So, the minimum occurs at the only critical point, which is ` = L. This corresponds to
4+π
` 4 4
= : the proportion of the wire that is cut is .
L 4+π 4+π
L2 L2
The maximum has to be either at ` = 0 or at ` = L. As A(0) = ą A( L) = , the maximum
4π 16
has ` = 0 (that is, no square).
x2 ´ 9
S-1: In general, this is false. For example, the function f ( x ) = has no vertical asymptotes,
x2 ´ 9
because it is equal to 1 in every point in its domain (and is undefined when x = ˘3).
478
However, it is certainly possible that f ( x ) has a vertical asymptote at x = ´3. For example,
1
f (x) = 2 has a vertical asymptote at x = ´3. More generally, if g( x ) is continuous and
x ´9
g(´3) ‰ 0, then f ( x ) has a vertical asympotote at x = ´3.
S-2: Since x2 + 1 and x2 + 4 are always positive, f ( x ) and h( x ) are defined over all real numbers.
So, f ( x ) and h( x ) correspond to A( x ) and B( x ). Which is which? A(0) = 1 = f (0) while
B(0) = 2 = h(0), so A( x ) = f ( x ) and B( x ) = h( x ).
That leaves g( x ) and k ( x ) matching to C ( x ) and D ( x ). The domain of g( x ) is all x such that
x2 ´ 1 ě 0. That is, |x| ě 1, like C ( x ). The domain of k ( x ) is all x such that x2 ´ 4 ě 0. That is,
|x| ě 2, like D ( x ). So, C ( x ) = g( x ) and D ( x ) = k ( x ).
S-3: (a) Since f (0) = 2, we solve

b
2= log2 (0 + p)
b
= log2 p
= |log p|
log p = ˘2
p = e˘2
1
p = e2 or p =
e2
1
We know that p is e2 or , but we have to decide between the two. In both cases, f (0) = 2. Let’s
e2
consider the domain of f ( x ). Since log2 ( x + p) is never negative, the square root does not restrict
our domain. However, we can only take the logarithm of positive numbers. Therefore, the
domain is
x such that x + p ą 0
x such that x ą ´p

1 1 1
If p = 2 , then the domain of f ( x ) is ´ 2 , 8 . In particular, since ´ 2 ą ´1, the domain of f ( x )
e e e
does not include x = ´1. However, it is clear from the graph that f (´1) exists. So, p = e2 .
(b) Now, we need to figure out what b is. Notice that b is the end of the domain of f ( x ), which we
already found to be (´p, 8). So, b = ´p = é2 .
(As a quick check, if we take e « 2.7, then é2 = ´7.29, and this looks about right on the graph.)
(c) The x-intercept is the value of x for which f ( x ) = 0:
b
0 = log2 ( x + p)
0 = log( x + p)
1= x+p
x = 1 ´ p = 1 ´ e2
The x-intercept is 1 ´ e2 .
479
(As another quick check, the x-intercept we found is a distance of 1 from the vertical asymptote,
and this looks about right on the graph.)
S-4: Vertical asymptotes occur where the function blows up. In rational functions, this can only
happen when the denominator goes to 0. In our case, the denominator is 0 when x = 3, and in
this case the numerator is 147. That means that as x gets closer and closer to 3, the numerator gets
closer and closer to 147 while the denominator gets closer and closer to 0, so | f ( x )| grows without
bound. That is, there is a vertical asymptote at x = 3.
The horizontal asymptotes are found by taking the limits as x goes to infinity and negative
infinity. In our case, they are the same, so we condense our work.
x (2x + 1)( x ´ 7) 2x3 + ax2 + bx + c

lim = lim
xÑ˘8 3x3 ´ 81 xÑ˘8 3x3 ´ 81
where a, b, ad c are some constants. Remember, for rational functions, you can figure out the end
behaviour by looking only at the terms with the highest degree–the others won’t matter, so we
don’t bother finding them. From here, we divide the numerator and denominator by the highest
power of x in the denominator, x3 .
!
1
2x3 + ax2 + bx + c x3
= lim
xÑ˘8 3x3 ´ 81 1
3 x
a b
2+ x +
x2
+ xc3
= lim
xÑ˘8 3 ´ 81
x3
2+0+0+0 2
= =
3´0 3
2
So there is a horizontal asymptote of y = both as x Ñ 8 and as x Ñ ´8.
3
S-5: Since f ( x ) is continuous over all real numbers, it has no vertical asymptote.
To find the horizontal asymptotes, we evaluate lim f ( x ).
xÑ˘8
lim 103x´7 = lim 10X = 8

xÑ8 XÑ8
looomooon
let X =3x´7
So, there’s no horizontal asymptote as x Ñ 8.
lim 103x´7 = lim 10X

xÑ´8 XÑ´8
loooomoooon
let X =3x´7
1
= lim
1 Ñ8
10´X
X
looooomooooon
let X1 =´X
1
= lim 1
X 1 Ñ8 10X
=0
That is, y = 0 is a horizontal asymptote as x Ñ ´8.
480
S-1: Functions A( x ) and B( x ) share something in common that sets them apart from the others:
they have a horizontal tangent line only once. In particular, A1 (´2) ‰ 0 and B1 (2) ‰ 0. The only
listed functions that do not have two distinct roots are l ( x ) and p( x ). Since l (´2) ‰ 0 and
p(2) ‰ 0, we conclude
A1 ( x ) = l ( x ) B1 ( x ) = p ( x )
Function C ( x ) is never decreasing. Its tangent line is horizontal when x = ˘2, but the curve never
decreases, so C1 ( x ) ě 0 for all x and C1 (2) = C1 (´2) = 0. The only function that matches this is
n( x ) = ( x ´ 2)2 ( x + 2)2 . Since its linear terms have even powers, it is never negative, and its roots
are precisely x = ˘2.
C1 ( x ) = n( x )
For the functions D ( x ) and E( x ) we consider their behaviour near x = 0. D ( x ) is decreasing near
x = 0, so D1 (0) ă 0, which matches with o (0) ă 0. Contrastingly, E( x ) is increasing near zero, so
E1 (0) ą 0, which matches with m(0) ą 0.
D1 ( x ) = o ( x ) E1 ( x ) = m ( x )
S-2: The domain of f ( x ) is all real numbers except ´3 (because when x = ´3 the denominator is
zero). For x ‰ ´3, we differentiate using the quotient rule:
e x ( x + 3) ´ e x (1) ex
f 1 (x) = = ( x + 2)
( x + 3)2 ( x + 3)2
Since e x and ( x + 3)2 are positive for every x in the domain of f ( x ), the sign of f 1 ( x ) is the same as
the sign of x + 2. We conclude that f ( x ) is increasing for every x in its domain with x + 2 ą 0.
That is, over the open interval (´2, 8).
S-3: Since we can’t take the square root of a negative number, f ( x ) is only defined when x ě 1.
Furthermore, since we can’t have zero as a denominator, x = ´2 is not in the domain — but as
long as x ě 1, we also have x ‰ ´2. So, the domain of the function is [1, 8).
In order to find where f ( x ) is increasing, we find where f 1 ( x ) is positive.
?
? +4
2x
´2 x´1
1 2 x´1 ( x + 2) ´ 2( x ´ 1) ´x + 4
f (x) = = ? =?
(2x + 4)2 x ´ 1(2x + 4) 2 x ´ 1(2x + 4)2
The denominator is never negative, so f ( x ) is increasing when the numerator of f 1 ( x ) is positive,

i.e. when 4 ´ x ą 0, or x ă 4. Recalling that the domain of definition for f ( x ) is [1, +8), we
conclude that f ( x ) is increasing on the open interval (1, 4).
S-4: The domain of arctangent is all real numbers. The domain of the logarithm function is all
positive numbers, and 1 + x2 is positive for all x. So, the domain of f ( x ) is all real numbers.
481
In order to find where f ( x ) is increasing, we find where f 1 ( x ) is positive.
2 2x 2 ´ 2x
f 1 (x) = 2
´ 2
=
1+x 1+x 1 + x2
Since the denominator is always positive, f ( x ) is increasing when when 2 ´ 2x ą 0. We conclude

that f ( x ) is increasing on the open interval (´8, 1).
S-1:
concave down
concave up concave down concave up
In the graph above, the concave-up sections are marked in red. These are where the graph has an
increasing derivative; equivalently, where the graph lies above its tangent lines; more
descriptively, where it curves like a smiley face.
Concave-down sections are marked in blue. These are where the graph has a decreasing
derivative; equivalently, where the graph lies below its tangent lines; more descriptively, where it
curves like a frowney face.
S-2: The most basic shape of the graph is given by the last two bullet points:
482
y
x
´5 5
The curve is concave down over the interval (´5, 5), so let’s give it a frowney-face curvature
there.
y
x
´5 5
Finally, when x ą 5 or x ă ´5, our curve should be concave up, so let’s give it smiley-face
curvature there, without changing its basic increasing/decreasing shape.
y
x
´5 5
This finishes our sketch.
S-3: An inflection point is where the concavity of a function changes. It is possible that x = 3 is an
inflection point, but it is also possible that is not. So, the statement is false, in general.
For example, let f ( x ) = ( x ´ 3)4 . Since f ( x ) is a polynomial, all its derivatives exist and are
continuous. f 2 ( x ) = 12( x ´ 3)2 , so f 2 (3) = 0. However, since f 2 ( x ) is something squared, it is
never negative, so f ( x ) is never concave down. Since f ( x ) is never concave down, it never
483
changes concavity, so it has no inflection points.
Remark: finding inflection points is somewhat reminiscent of finding local extrema. To find local
extrema, we first find all critical and singular points, since local extrema can only occur there or at
endpoints. Then, we have to figure out which critical and singular points are actually local
extrema. Similarly, if you want to find inflection points, start by finding where f 2 ( x ) is zero or
non-existant, because inflection points can only occur there (see Question 7). Then, you still have
to check whether those points are actually inflection points.
S-4: Inflection points occur where f 2 ( x ) changes sign. Since f ( x ) is a polynomial, its first and
second derivatives exist everywhere, and are themselves polynomials. In particular,
f ( x ) = 3x5 ´ 5x4 + 13x
f 1 ( x ) = 15x4 ´ 20x3 + 13
f 2 ( x ) = 60x3 ´ 60x2 = 60x2 ( x ´ 1)
The second derivative is negative for x ă 1 and positive for x ą 1. Thus the concavity changes
between concave up and concave down at x = 1, y = 11.
This is the only inflection point. It is true that f 2 (0) = 0, but for values of x both a little larger
than and a little smaller than 0, f 2 ( x ) ă 0, so the concavity does not change at x = 0.
S-5: In order to show that f ( x ) has exactly one inflection point, we will show that is has at least one,
and no more than one.
Let
g( x ) = f 2 ( x ) = x3 + 5x ´ 20.
Then g1 ( x ) = 3x2 + 5, which is always positive. That means g( x ) is strictly increasing for all x. So,
g( x ) can change signs once, from negative to positive, but it can never change back to negative.
An inflection point of f ( x ) occurs when g( x ) changes signs. So, f ( x ) has at most one inflection
point. (At this point, we don’t know that f ( x ) has any inflection points: maybe g( x ) is always
positive.)
Since g( x ) is continuous, we can apply the Intermediate Value Theorem to it. Notice g(3) ą 0
while g(0) ă 0. By the IVT, g( x ) = 0 for at least one x P (0, 3). Since g( x ) is strictly increasing, at
the point where g( x ) = 0, g( x ) changes from negative to positive. So, the concavity of f ( x )
changes. Therefore, f ( x ) has at least one inflection point.
Now that we’ve shown that f ( x ) has at most one inflection point, and at least one inflection point,
we conclude it has exactly one inflection point.
S-6: (a) Let

g( x ) = f 1 ( x )
Then f 2 ( x ) is the derivative of g( x ). Since f 2 ( x ) ą 0 for all x, g( x ) = f 1 ( x ) is strictly increasing
for all x. In other words, if y ą x then g(y) ą g( x ).
Suppose g( x ) = 0. Then for every y that is larger than x, g(y) ą g( x ), so g(y) ‰ 0. Similarly, for
every y that is smaller than x, g(y) ă g( x ), so g(y) ‰ 0. Therefore, g( x ) can only be zero for at
most one value of x. Since g( x ) = f 1 ( x ), that means f ( x ) can have at most one critical point.
Suppose f 1 (c) = 0. Since f 1 ( x ) is a strictly increasing function, f 1 ( x ) ă 0 for all x ă c and
f 1 ( x ) ą 0 for all x ą c.
484
c
x ă c, so x ą c, so
f 1 (x) ă f 1 (c) = 0 f 1 (x)
ą f 1 (c) = 0
Then f ( x ) is decreasing for x ă c and increasing for x ą c. So f ( x ) ą f (c) for all x ă c and
f ( x ) ą f (c) for all x ą c.
y = f (x)
xăc c xąc
f ( x ) decreasing, so f ( x ) increasing, so
f ( x ) ą f (c) f ( x ) ą f (c)
We have concluded that f ( x ) ą f (c) for all x ‰ c, so c is an absolute minimum for f ( x ).

(b) We know that the maximum over an interval occurs at an endpoint, at a critical point, or at a
singular point.
• Since f 1 ( x ) exists everywhere, there are no singular points.

• If the maximum were achieved at a critical point, that critical point would have to provide
both the absolute maximum and the absolute minimum (by part a). So, the function would
have to be a constant and consequently could not have a nonzero second derivative. So the
maximum is not at a critical point.
That leaves only the endpoints of the interval.
S-7: If x = 3 is an inflection point, then the concavity of f ( x ) changes at x = 3. That is, there is
some interval strictly containing 3, with endpoints a and b, such that

Remark: we are leaving unknown whether a ă 3 ă b or b ă 3 ă a. Since we don’t know whether

f ( x ) changes from concave up to concave down, or from concave down to concave up, by
remaining vague we cover both cases.
a 3 b
concave down concave up
485
OR
b 3 a
concave up concave down
Since f 2 ( a) ă 0 and f 2 (b) ą 0, and since f 2 ( x ) is continuous, the Intermediate Value Theorem
tells us that there exists some x strictly between a and b with f 2 ( x ) = 0. So, we know f 2 ( x ) = 0
somewhere between a and b. The question is, where exactly could that be?
• f 2 ( x ) ă 0 (and hence f 2 ( x ) ‰ 0) for all x between a and 3

• f 2 ( x ) ą 0 (and hence f 2 ( x ) ‰ 0) for all x between b and 3
• So, any number between a and b that is not 3 has f 2 ( x ) ‰ 0.
So, x = 3 is the only possible place between a and b where f 2 ( x ) could be zero. Therefore,
f 2 (3) = 0.
Remark: this is why, in general, we set f 2 ( x ) = 0 to find inflection points. (They can also occur
where f 2 ( x ) does not exist.)
S-1: This function is symmetric across the y-axis, so it is even.
S-2: The function is not even, because it is not mirrored across the y-axis.
Assuming it continues as shown, the function is periodic, because the unit shown below is
repeated:
y
y = f (x)
486
Additionally, f ( x ) is odd. In a function with odd symmetry, if we mirror the right-hand portion
of the curve (the portion to the right of the y-axis) across both the y-axis and the x-axis, it lines up
with the left-hand portion of the curve.
y = f (x)
reflected across y-axis
y = f (x)
reflected across both axes
487
Since reflecting the right-hand portion of the graph across the y-axis, then the x-axis, gives us
f ( x ), we conclude f ( x ) is odd.
S-3: Since the function is even, we simply reflect the portion shown across the y-axis to complete
the sketch.
S-4: Since the function is odd, to complete the sketch, we reflect the portion shown across the
y-axis (shown dashed), then the x-axis (shown in red).
488
S-5: A function is even if f (´x ) = f ( x ).
(´x )4 ´ (´x )6
f (´x ) = 2
e(´x)
x4 ´ x6
= 2
ex
= f (x)
So, f ( x ) is even.
S-6: For any real number x, we will show that f ( x ) = f ( x + 4π ).

x + 4π
f ( x + 4π ) = sin( x + 4π ) + cos
2
x
= sin( x + 4π ) + cos + 2π
x 2
= sin( x ) + cos
2
= f (x)
So, f ( x ) is periodic.
S-7: f ( x ) is not periodic. (You don’t really have to justify this, but if you wanted to, you could say
something like this. Notice f (0) = 1. Whenever x ą 10, f ( x ) ą 1. Then the value of f (0) is not
repeated indefinitely, so f ( x ) is not periodic.)
To decide whether f ( x ) is even, odd, or neither, simplify f (´x ):
f (´x ) = (´x )4 + 5(´x )2 + cos (´x )3

= x4 + 5x3 + cos(´x )
= x4 + 5x3 + cos( x )
= f (x)
Since f (´x ) = f ( x ), our function is even.
S-8: It should be clear that f ( x ) is not periodic. (If you wanted to justify this, you could note that
f ( x ) = 0 has exactly two solutions, x = 0, ´5. Since the value of f (0) is repeated only twice, and
not indefinitely, f ( x ) is not periodic.)
To decide whether f ( x ) is odd, even, or neither, we simplify f (´x ).
f (´x ) = (´x )5 + 5(´x )4

= ´x5 + 5x4
We see that f (´x ) is not equal to f ( x ) or to ´ f ( x ). For instance, when x = 1:
• f (´x ) = f (´1) = 4,
• f ( x ) = f (1) = 6, and
489
• ´ f ( x ) = ´ f (1) = ´6.
Since f (´x ) is not equal to f ( x ) or to ´ f ( x ), f ( x ) is neither even nor odd.
S-9: Recall the period of g( X ) = tan X is π.

tan( X + π ) = tan( X ) for any X in the domain of tan X
Replacing X with πx:
tan(πx + π ) = tan(πx ) for any x in the domain of tan(πx )
tan(π ( x + 1)) = tan(πx ) for any x in the domain of tan(πx )
f ( x + 1) = f ( x ) for any x in the domain of tan(πx )
The period of f ( x ) is 1.
S-10: Let’s consider g( x ) = tan(3x ) and h( x ) = sin(4x ) separately. Recall that π is the period of
tangent.
tan X = tan( X + π ) for every X in the domain of tan X
Replacing X with 3x:
tan(3x ) = tan(3x + π ) for every x in the domain of tan 3x
π
tan(3x ) = tan 3 x + for every x in the domain of tan 3x
3
π
g( x ) = g x + for every x in the domain of tan 3x
3
π
So, the period of g( x ) = tan(3x ) is .
3
Similarly, 2π is the period of sine.
sin( X ) = sin( X + 2π ) for every X in the domain of sin( X )
Replacing X with 4x:
sin(4x ) = sin(4x + 2π ) for every x in the domain of sin(4x )
π
sin(4x ) = sin 4 x + for every x in the domain of sin(4x )
2
π
h( x ) = h x + for every x in the domain of sin(4x )
2
π
So, the period of h( x ) = sin(4x ) is .
2
All together, f ( x ) = g( x ) + h( x ) will repeat when both g( x ) and h( x ) repeat. The least common
π π π π
integer multiple of and is π. Since g( x ) repeats every units, and h( x ) repeats every
3 2 3 2
units, they will not both repeat until we move π units. So, the period of f ( x ) is π.
490
S-1: (a) Since we must have 3 ´ x ě 0, this tells us x ď 3. So, the domain is (´8, 3].
(b)
? x 2´x
f 1 (x) = 3´x´ ? =3 ?
2 3´x 2 3´x
For every x in the domain of f 1 ( x ), the denominator is positive, so the sign of f 1 ( x ) depends only
on the numerator.
x (´8, 2) 2 (2, 3) 3
f 1 (x) positive 0 negative DNE
f (x) increasing maximum decreasing endpoint
So, f is increasing for x ă 2, has a local (in fact global) maximum at x = 2, is decreasing for
2 ă x ă 3, and has a local minimum at x = 3.
Remark: this shows us the basic skeleton of the graph. It consists of a single hump.
x
2 3
(c) When x ă 3,
1
f 2 (x) = (3x ´ 12)(3 ´ x )´3/2 ă 0
4
The domain of f 2 ( x ) is (´8, ´3), and over its domain it is always negative (the factor (3x ´ 12) is
negative for all x ă 4 and the factor (3 ´ x )´3/2 is positive for all x ă 3). So, f ( x ) has no inflection
points and is concave down everywhere.
(d) We already found
2´x
f 1 (x) = 3 ? .
2 3´x
This is undefined at x = 3. Indeed,
2´x
lim 3 ? = ´8,
xÑ3 ´ 2 3´x
so f ( x ) has a vertical tangent line at (3, 0).
(e) To sketch the curve y = f ( x ), we already know its intervals of increase and decrease, and its
concavity. We also note its intercepts are (0, 0) and (3, 0).
491
y
(2.2)
x
(3, 0)
increasing decreasing
concave down
S-2:
• Asymptotes:
1 2
lim f ( x ) = lim ´ 4 =0
xÑ˘8 xÑ˘8 x x
So y = 0 is a horizontal asymptote both at x = 8 and x = ´8.
x3 ´ 2
lim f ( x ) = lim = ´8
xÑ0 xÑ0 x4
So there is a vertical asymptote at x = 0, where the function plunges downwards from both
the right and the left.
• Intervals of increase and decrease:
1 8 8 ´ x3
f 1 (x) = ´ + =
x2 x5 x5
The only place where f 1 ( x ) is zero only at x = 2. So f ( x ) has a horizontal tangent at x = 2,
y = 83 . This is a critical point.
The derivative is undefined at x = 0, as is the function.
x (´8, 0) 0 (0, 2) 2 (2, 8)

f 1 (x) negative DNE positive 0 negative
f (x) decreasing vertical asymptote increasing local max decreasing
Since the function changes from increasing to decreasing at x = 2, the only local maximum
is at x = 2.
At this point, we get a rough sketch of f ( x ).
492
horizontal asymptotes y = 0
2, 83

y
x
2
decreasing increasing decreasing
• Concavity:
2 40 2x3 ´ 40
f 2 (x) = ´ =
x3 x6 x6
? ?
) is positive for x ą 3 20 and negative
The second derivative of f ( x? ? for x ă 3 20. So the
curve is concave
? up for x ą 3 20 and concave down for x ă 3 20. There is an inflection
point at x = 3 20 « 2.7, y = 20184/3 « 0.3.
• Intercepts:
Since?f ( x ) is not defined at x = 0, there is no y-intercept. The only x-intercept is

x = 3 2 « 1.3.
• Sketch:
We can add concavity to our skeleton sketched above, and label our intercept and inflection
point (the open dot).
493
y 2, 38

x
? ?
3
2 2 3 20
concave down concave up
S-3:
• Asymptotes:
When x = ´1, the denominator 1 + x3 of f ( x ) is zero while the numerator is 1, so x = ´1 is
a vertical asymptote. More precisely,
lim f ( x ) = ´8 lim f ( x ) = 8
xÑ´1´ xÑ´1+
There are no horizontal asymptotes, because
x4 x4
lim =8 lim = ´8
xÑ8 1 + x 3 xÑ´8 1 + x 3

We note that f 1 ( x ) is defined for all x ‰ ´1 and is not defined for x = ´1. Therefore, the
only singular point for f ( x ) is x = ´1.
To find critical points, we set
f 1 (x) = 0
4x3 + x6 = 0
x 3 (4 + x 3 ) = 0
x3 = 0 or 4 + x3 = 0
?3
x=0 or x = ´ 4 « ´1.6
? ?
3
At these critical points, f (0) = 0 and f (´ 3 4) = 4´34 ă 0. The denominator of f 1 ( x ) is never
negative, so the sign of f 1 ( x ) is the same as the sign of its numerator, x3 (4 + x3 ).
? ? ?
x (´8, ´ 3 4) ´ 3 4 (´ 3 4, ´1) ´1 (´1, 0) 0 (0, 8)
1
f (x) positive 0 negative DNE negative 0 positive
f (x) increasing l. max decreasing VA decreasing l. min increasing
494
Now, we have enough information to make a skeleton of our graph.
x
?
´34 ´1
increasing decr decr increasing
• Concavity:
The second derivative is undefined when x =?´1. It is zero when

12x2 ´ 6x5 = 6x2 (2 ´ x3 ) = 0. That is, at x = 3 2 « 1.3 and x = 0. Notice that the sign of
f 2 ( x ) does not change at x = 0, so x = 0 is not an inflection point.
? ? ?
x (´8, ´1) ´1 (´1, 0) 0 (0, 3 2) 3
2 ( 3 2, 8)
2
f (x) negative DNE positive 0 positive 0 negative
f (x) concave down VA concave up concave up IP concave down
Now we can refine our skeleton by adding concavity.
495
y
x
? ?
´34 ´1 3
2
increasing decr decr increasing
concave down concave up concave down
S-4:
• Asymptotes:
x3 x3
lim =8 lim = ´8
xÑ´8 1 ´ x 2 xÑ8 1 ´ x 2
So, f ( x ) has no horizontal asymptotes.

On the other hand f ( x ) blows up at both x = 1 and x = ´1, so there are vertical asymptotes
at x = 1 and x = ´1. More precisely,
x3 x3
lim =8 lim = ´8
xÑ´1´ 1 ´ x 2 xÑ´1+ 1 ´ x 2
x3 x3
lim =8 lim = ´8
xÑ1´ 1 ´ x 2 xÑ1+ 1 ´ x 2
• Symmetry:
f ( x ) is an odd function, because
(´x )3 ´x3
f (´x ) = 2
= = ´ f (x)
1 ´ (´x ) 1 ´ x2
• Intercepts:
The only intercept of f ( x ) is the origin. In particular, that means that out of the three
intervals where it is continuous, namely (´8, ´1), (´1, 1) and (1, 8), in two of them f ( x )
is always positive or always negative.
496
– When x ă ´1: 1 ´ x2 ă 0 and x3 ă 0, so f ( x ) ą 0.
– When x ą 1: 1 ´ x2 ă 0 and x3 ą 0, so f ( x ) ă 0.
– When ´1 ă x ă 0, 1 ´ x2 ą 0 and x3 ă 0 so f ( x ) ă 0.
– When 0 ă x ă 1, 1 ´ x2 ą 0 and x3 ą 0 so f ( x ) ą 0.
3x2 ´ x4 x 2 (3 ´ x 2 )
f 1 (x) = =
(1 ´ x 2 )2 (1 ´ x 2 )2
The only singular points are x = ˘1, where f ( x ), and hence f 1 ( x ), is not defined. The
critical points are:
f 1 (x) = 0
x2 = 0 or 3 ´ x2 = 0
?
x=0 or x = ˘ 3 « ˘1.7
?
? 3 3
The values of f at its critical points are f (0) = 0, f ( 3) = ´ « ´2.6 and
? 2
? 3 3
f (´ 3) = « 2.6.
2
Notice the sign of f 1 ( x ) is the same as the sign of 3 ´ x2 .
? ? ?
x (´8, ´ 3) ´ 3 (´ 3, ´1) ´1
f 1 (x) negative 0 positive DNE
f (x) decreasing local min increasing VA
? ? ?
x (´1, 0) 0 (0, 3) 3 ( 3, 8)
f 1 (x) positive 0 positive 0 negative
f (x) increasing increasing local max decreasing
Now we have enough information to sketch a skeleton of f ( x ).
497
y
?
3 3
2
x
? ?
´ 3 ´1 1 3
?
´323
decreasing incr increasing incr decreasing
• Concavity:
2x (3 + x2 )
f 2 (x) =
(1 ´ x 2 )3
The second derivative is zero when x = 0, and is undefined when x = ˘1.
x (´8, ´1) (´1, 0) 0 (0, 1) (1, 8)

f 2 (x) positive negative 0 positive negative
f (x) concave up concave down inflection point concave up concave down
Now, we can refine our skeleton.
498
y
?
3 3
2
x
? ?
´ 3 ´1 1 3
?
´323
decreasing incr increasing incr decreasing
concave up cc down concave up concave down
S-5: (a) One branch of the function, the exponential function e x , is continuous everywhere. So
x2 + 3
f ( x ) is continuous for x ă 0. When x ě 0, f ( x ) = , which is continuous whenever
3( x + 1)
x ‰ ´1 (so it’s continuous for all x ą 0). So, f ( x ) is continuous for x ą 0. To see that f ( x ) is
continuous at x = 0, we see:
xÑ0´ xÑ0´
x2 + 3
lim f ( x ) = lim =1
xÑ0+ xÑ0+3( x + 1)
So, lim f ( x ) = 1 = f (0)
xÑ0
Hence f ( x ) is continuous at x = 0, so f ( x ) is continuous everywhere.
499
(b) We differentiate the function twice. Notice
" 2
3( x + 1)(2x ) ´ ( x2 + 3)(3)
*
d x +3
=
dx 3( x + 1) 9( x + 1)2
x2 + 2x ´ 3
=
3( x + 1)2
( x ´ 1)( x + 3)
= where x ‰ ´1
3( x + 1)2
(0 ´ 1)(0 + 3)
Then lim f 1 ( x ) = = ´1 ‰ 1 = e0 = lim f 1 ( x )
xÑ0 + 3(0 + 1)2 xÑ0´
$
x
& e
’ xă0
so 1
f (x) = DNE x=0
( x´1 )( x + 3 )
xą0
’
%
3( x +1)2
Differentiating again,
d2
" 2 * " 2 *
x +3 d x + 2x ´ 3
=
dx2 3( x + 1) dx 3( x + 1)2
3( x + 1)2 (2x + 2) ´ ( x2 + 2x ´ 3)(6)( x + 1) ˜3( x + 1)

=
9( x + 1)4 ˜3( x + 1)
2
( x + 1)(2x + 2) ´ 2( x + 2x ´ 3)
=
3( x + 1)3
8
= where x ‰ ´1
3( x + 1)3
$ x
& e xă0
2 DNE x=0
so f (x) =
8
%
3( x +1)3
xą0
i. The only singular point is x = 0, and the only critical point is x = 1. (When you’re reading off
the expression for f 1 ( x ), remember that the bottom line only applies when x ą 0.)
x (´8, 0) 0 (0, 1) 1 (1, 8)
f 1 (x) positive DNE negative 0 positive
f (x) increasing local max decreasing local min increasing
The coordinates of the local maximum are (0, 1) and the coordinates of the local minimum are
1, 32 .

ii.
When x ‰ 0, f 2 ( x ) is always positive, so f ( x ) is concave up.
iii.
x2 + 3
lim f ( x ) = lim
xÑ8 xÑ8 3x + 3
x + 3x
= lim =8
xÑ8 1 + 3
x
So, there is no horizontal asymptote to the right.

xÑ´8 xÑ´8
500
So, y = 0 is a horizontal asymptote to the left.
Since f ( x ) is continuous everywhere, there are no vertical asymptotes.

(c)
y
(0, 1)
(1, 23 )
increasing decr increasing
concave up concave up
S-6:
• Asymptotes: In the problem statement, we are told:

1 + 2x
lim =0
xÑ˘8 e x2
So, y = 0 is a horizontal asymptote both at x = 8 and at x = ´8.

Since f ( x ) is continuous, it has no vertical asymptotes.
The critical points are the zeroes of 1 ´ x ´ 2x2 = (1 ´ 2x )(1 + x ). That is, x = 12 , ´1.
x (´8, ´1) ´1 (´1, 12 ) 1

2 ( 12 , 8)
f 1 (x) negative 0 positive 0 negative
f ( x ) decreasing local min increasing local max decreasing
At these critical points, f 21 = 2e´1/4 ą 0 and f (´1) = é´1 ă 0.

From here, we can sketch a skeleton of the graph.
x
´1 1
2
decreasing increasing decr increasing
501
• Concavity:
We are told that we don’t have to actually solve for the inflection points. We just need to
know enough to get a basic idea. So, we’ll turn the skeleton of the graph into smooth curve.
x
−1 0.5
Inflection points are points where the convexity changes from up to down or vice versa. It
looks like our graph is convex down for x from ´8 to about ´1.8, convex up from about
x = ´1.8 to about x = ´0.1, convex down from about x = ´0.1 to about x = 1.4 and
convex up from about x = 1.4 to infinity. So there are three inflection points at roughly
x = ´1.8, ´0.1, 1.4.
S-7:
(a) We need to know the first and second derivative of f ( x ). Using the product and chain rules,
2 2
f 1 ( x ) = e´x /2 (1 ´ x2 ). Given to us is f 2 ( x ) = ( x3 ´ 3x )e´x /2 . (These derivatives are also easy to
find using the formula developed in Question 19, Section 2.14.)
2 /2
Since e´x is always positive, the sign of f 1 ( x ) is the same as the sign of 1 ´ x2 . f ( x ) has no
1
singular points and its only critical points are x = ˘1. At these critical points, f (´1) = ´ ? and
e
1
f (1) = ? .
e
x (´8, ´1) ´1 (´1, 1) 1 (1, 8)
f (x) decreasing local min increasing local max decreasing
This, together with the observations that f ( x ) ă 0 for x ă 0, f (0) = 0 and f ( x ) ą 0 for x ą 0 (in
fact f is an odd function), is enough to sketch a skeleton of our graph.
y
?1
e
x
´1 1
´ ?1e
502
2 ? ? ´x2 /2 2
We can factor f 2 ( x ) = ( x3 ´ 3x )e´x /2 = x ( x + ? 3)( x ´ ?
3) e . Since e´x /2 is always positive,
the sign of f 2 ( x ) is the same as the sign of x ( x + 3)( x ´ 3).
? ? ? ? ? ?
x (´8, ´ 3) ´ 3 (´ 3, 0) 0 (0, 3) 3 ( 3, 8)
f 2 (x) negative 0 positive 0 negative 0 positive
f (x) concave down IP concave up IP concave down IP concave up
(b) We’ve already seen that f ( x ) has a local min at x = ´1 and a local max at x = 1.
As x tends to negative infinity, f ( x ) tends to 0, and f ( x ) is decreasing on (´8, ´1). Then f ( x ) is
between 0 and f (´1) = ´1? on (´8, ´1). Then f ( x ) is increasing on (´1, 1) from f (´1) = ´1
e
? to
e
f (1) = ?1 . Finally, for x ą 1, f ( x ) is decreasing from f (1) = ?1 and tending to 0. So when x ą 1,
e e
f ( x ) is between ?1 and 0.
e
So, over its entire domain, f ( x ) is between ´1

? and ?1 , and it only achieves those values at x = ´1
e e
and x = 1, respectively. Therefore, the local and global min of f ( x ) is at (´1, ´1
? ), and the local
e
and global max of f ( x ) is at (1, ?1e ).
(c) In the graph below, open dots are inflection points, and solid dots are extrema.
y
1
?
e
? x
´ 3 ?
´1 1 3
´1
?
e
S-8:
• Symmetry:
f (´x ) = ´x + 2 sin(´x ) = ´x ´ 2 sin x = ´ f ( x )

So, f ( x ) is an odd function. If we can sketch y = f ( x ) for nonnegative x, we can use
symmetry to complete the curve for all x.
• Asymptotes:
Since f ( x ) is continuous, it has no vertical asymptotes. It also has no horizontal asymptotes,
since
lim f ( x ) = ´8 lim f ( x ) = 8
xÑ´8 xÑ8
503
Since f ( x ) is differentiable everywhere, there are no singular points.
f 1 ( x ) = 1 + 2 cos x
So, the critical points of f ( x ) occur when

1
cos x = ´
2
2π
x = 2πn ˘ for any integer n
3
2π 4π 8π 10π
For instance, f ( x ) has critical points at x = ,x= ,x= , and x = .
3 3 3 3
From the unit circle, or the graph of y = 1 + 2 cos x, we see:
´ 2π 2π 2π 2π 4π 4π 4π 8π 8π 8π 10π

x 3 , 3 3 3 , 3 3 3 , 3 3 3 , 3
f 1 (x) positive 0 negative 0 positive 0 negative
f (x) increasing l. max decreasing l. min increasing l. max decreasing
We have enough information to sketch a skeleton of the curve y = f ( x ). We use the pattern
above for the graph to the right of the y-axis, and use odd symmetry for the graph to the
left of the y-axis.
x
´14π ´10π ´8π 2π 4π 8π 10π 14π
3 3 3 3 3 3 3 3
• Concavity:
f 2 ( x ) = ´2 sin x
So, f 2 ( x ) exists everywhere, and is zero for x = π + πn for every integer n.
504
x (0, π ) π (π, 2π ) 2π (2π, 3π ) 3π (3π, 4π )
f 2 (x) negative 0 positive 0 negative 0 positive
f (x) concave down IP concave up IP concave down IP concave up
Using these values, and the odd symmetry of f ( x ), we can refine our skeleton. The closed
dots are local extrema, and the open dots are inflection points occurring at every integer
multiple of π.
x
´14π ´10π ´8π ´4π ´2π 2π 4π 8π 10π 14π
3 3 3 3 3 3 3 3 3 3
S-9: We first compute the derivatives f 1 ( x ) and f 2 ( x ).
f 1 ( x ) = 4 cos x + 4 sin 2x = 4 cos x + 8 sin x cos x = 4 cos x (1 + 2 sin x )

f 2 ( x ) = ´4 sin x + 8 cos 2x = ´4 sin x + 8 ´ 16 sin2 x = ´4(4 sin2 x + sin x ´ 2)
The graph has the following features.
• Symmetry: f ( x ) is periodic of period 2π. We’ll consider only ´π ď x ď π. (Any interval of

length 2π will do.)
• y-intercept: f (0) = ´2
505
π
• Intervals of increase and decrease: f 1 ( x ) = 0 when cos x = 0, i.e. x = ˘ , and when
2
1 π 5π
sin x = ´ , i.e. x = ´ , ´ .
2 6 6
(´π, ´ 5π 5π π
π π
π π
π

x 6 ) ´ 6 , ´ 2 ´ 2 , ´ 6 ´ ,
6 2 2 , π
f 1 (x) negative positive negative positive negative
f ( x ) decreasing increasing decreasing increasing decreasing
π 5π
This tells us local maxima occur at x = ˘ and local minima occur at x = ´ and
2 6
π
x=´ .
6
Here is a table giving the value of f at each of its critical points.
5 π π π
x ´ π ´ ´
6 2 6 2
sin x ´ 12 ´1 ´ 12 1
1 1
cos 2x 2 ´1 2 ´1
f (x) ´3 ´2 ´3 6
From here, we can graph a skeleton of of f ( x ):
x
´π ´ π2 π π
2
´ 5π
6
´ π6
´2
´3
• Concavity: To find the points where f 2 ( x ) = 0, set y = sin x, so f 2 ( x ) = ´4(4y2 + y ´ 2).

Then we really need to solve
4y2 + y ´ 2 = 0 which gives us
?
´1 ˘ 33
y=
8
506
These two y-values map to the following two x-values, which we’ll name a and b for
convenience:
?
´1 + 33
a = arcsin « 0.635
8
?
´1 ´ 33
b = arcsin « ´1.003
8
?
However, these are not the only values of x in [´π, π ] with sin x = ´1˘8 33 . The analysis
above
π πmisses
the others because the arcsine function only returns numbers in the range
´ 2 , 2 . The
?
graph below shows that there should be other values of x with
´1˘ 33
sin x = 8 , and hence f 2 ( x ) = 0.
y = sin x
?
´1+ 33
8
x
b a
?
´1´ 33
8
We can recover the other solutions in [´π, π ] by recalling that

sin( x ) = sin(π ´ x )
?
(see CLP appendix A7). So, if we choose x = arcsin ´1+8 33 « 0.635 to make
?
´1+ 33
sin( x ) = 8 so that f 2 ( x ) = 0, then setting
?
´1 + 33
x = π ´ a = π ´ arcsin « 2.507
8
?
´1+ 33
will also give us sin( x ) = 8 and f 2 ( x ) = 0. Similarly, setting
?
´1 ´ 33
x = π ´ b = π ´ arcsin « 4.145
8
would give us f 2 ( x ) = 0. However, this value is outside [´π, π ]. To find another solution
inside [´π, π ] we use the identity
sin( x ) = sin(´π ´ x )
(which we can obtain from the identity we used above and the fact that
sin(θ ) = sin(θ ˘ 2π ) for any angle θ). Using this, we can show that
?
´1 ´ 33
x = ´π ´ b = ´π ´ arcsin « ´2.139
8
507
also gives f 2 ( x ) = 0.
So, all together, f 2 ( x ) = 0 when x = ´π ´ b, x = b, x = a, and x = π ´ a.
Now, we should compute the sign of f 2 ( x ) while x is between ´π and π. Recall that, if
y = sin x, then f 2 ( x ) = ´4(4y2 + y ´ 2). So, in terms of y, f 2 is a parabola
pointing down,
? ? ?
´1˘ 33 ´1´ 33 ´1+ 33
with intercepts y = 8 . Then f 2 is positive when y is in the interval 8 , 8 ,
?
´1´ 33
and f 2 is negative otherwise. From the graph of sine, we see that y is between 8 and
?
´1+ 33
8 precisely on the intervals (´π, ´π ´ b), (b, a), and (π ´ a, π ).
Therefore, f ( x ) is concave up on the intervals (´π, ´π ´ b), (b, a), and (π ´ a, π ), and f ( x )
is concave down on the intervals (´π ´ b, b) and ( a, π ´ a). So, the inflection points of f
occur at x = ´π ´ b, x = b, x = a, and x = π ´ a.
y
´π ´ b b a πá
x
´π ´ π2 π π
2
´ 5π
6
´ π6
´2
´3
To find the maximum and minimum values of f ( x ) on [0, π ], we compare the values of f ( x ) at its
critical points in this interval (only x = π2 ) with the values of f ( x ) at its endpoints x = 0, x = π.
Since f (0) = f (π ) = ´2, the minimum value of f on [0, π ] is ´2, achieved at x = 0, π and the
π
maximum value of f on [0, π ] is 6, achieved at x = .
2
c
3 x+1
S-10: Let f ( x ) = .
x2
• Asymptotes: Since lim f ( x ) = 8, f ( x ) has a vertical asymptote at x = 0 where the curve

xÑ0
reaches steeply upward from both the left and the right.
lim f ( x ) = 0, so y = 0 is a horizontal asymptote for x Ñ ˘8.
xÑ˘8
• Intercepts: f (´1) = 0.
508
´( x + 2)
f 1 (x) =
3x5/3 ( x
+ 1)2/3
There is a singular point at x = ´1 and a critical point at x = ´2, in addition to a

? 2
discontinuity at x = 0. Note that ( x + 1)2/3 = 3 x + 1 , which is never negative. Note
also that lim f 1 ( x ) = 8, so f ( x ) has a vertical tangent line at x = ´1.
xÑ´1
x (´8, ´2) ´2 (´2, ´1) ´1 (´1, 0) 0 (0, 8)

f 1 (x) negative 0 positive DNE positive DNE negative
f (x) decreasing l. min increasing vertical increasing VA decreasing
This gives us enough information to sketch a skeleton of the curve.
x
´2 ´1
• Concavity:
4x2 + 16x + 10
f 2 (x) =
9x8/3 ( x + 1)5/3
We still have a discontinuity at x = 0, and f 2 ( x ) does not exist at x = ´1. The second
derivative is zero when 4x2 + 16x + 10 = 0. Using the quadratic formula, we find this
? ? 8
occurs when x = ´2 ˘ 1.5 « ´0.8, ´3.2. Note x8/3 = ( 3 x ) is never negative.
? ? ?
x ´8, ´2 ´ 1.5 ´2 ´ 1.5 (´2 ´ 1.5, ´1) ´1
f 2 (x) negative 0 positive DNE
f (x) concave down IP concave up IP
? ? ?
x (´1, ´2 + 1.5) ´2 + 1.5 (´2 + 1.5, 0) (0, 8)
f 2 (x) negative 0 positive positive
f (x) concave down IP concave up concave up
Now, we can refine our skeleton. The closed dot is the local minimum, and the open dots
are inflection points.
509
y
x
´2 ´1
´1
S-11: The parts of the question are just scaffolding to lead you through sketching the curve. Their
answers are given explicitly, in an organized manner, in the “answers” section. In this solution,
they are scattered throughout.
• Asymptotes:
Since the function has a derivative at every real number, the function is continuous for
every real number, so it has no vertical asymptotes. In the problem statement, you are told
lim f ( x ) = 0, so y = 0 is a horizontal asymptote as x goes to infinity. It remains to evaluate
xÑ8
lim f ( x ). Let’s consider the limit of f 1 ( x ) instead. Recall K is a positive constant.
xÑ´8
lim e´x = lim e x = 8

xÑ´8 xÑ8
2
lim K (2x ´ x ) = ´8
xÑ´8
So,
lim K (2x ´ x2 )e´x = ´8

xÑ´8
That is, as x becomes a hugely negative number, f 1 ( x ) also becomes a hugely negative
number. As we move left along the x-axis, f ( x ) is decreasing with a steeper and steeper
slope, as in the sketch below. That means lim f ( x ) = 8.
xÑ´8
510
y
Sketch: various tangent lines to f ( x ),

their slopes getting more strongly negative
as x gets more strongly negative.
We are given f 1 ( x ) (although we don’t know f ( x )):
f 1 ( x ) = Kx (2 ´ x )e´x
The critical points of f ( x ) are x = 0 and x = 2, and there are no singular points. Recall e´x
is always positive, and K is a positive constant.
x (´8, 0) 0 (0, 2) 2 (2, 8)

f (x) decreasing local min increasing local max decreasing
So, f (0) = 0 is a local minimum, and f (2) = 2 is a local maximum.
Looking ahead to part (d), we have a skeleton of the curve.
511
y
x
2
• Concavity:
Since we’re given f 1 ( x ), we can find f 2 ( x ).
f 2 ( x ) = K (2 ´ 2x ´ 2x + x2 )e´x
= K (2 ´ 4x + x2 )e´x
? ?
= K x ´ 2 ´ 2 x ´ 2 + 2 e´x
?
where the last line can be found using the quadratic equation. So, f 2 ( x ) = 0 for x = 2 ˘ 2,
and f 2 ( x ) exists everywhere.
? ? ? ? ? ?
x (´8, 2 ´ 2) 2´ 2 (2 ´ 2, 2 + 2) 2+ 2 (2 + 2, 8)
1
f (x) positive 0 negative 0 positive
f (x) concave up IP concave down IP concave up
Now, we can add concavity to our sketch.
x
? ?
2´ 2 2 2+ 2
decr increasing decreasing
cc up concave down concave up
512
S-12: (a) You should be familiar with the graph of y = e x . You can construct the graph of y = e´x
just by reflecting the graph of y = e x across the y–axis. To see why this is the case, imagine
swapping each value of x with its negative: for example, swapping the point at x = ´1 with the
point at x = 1, etc. Alternatively, you can graph y = f ( x ) = e´x , x ě 0, using the methods of this
section: at x = 0, y = f (0) = 1; as x increases, y = f ( x ) = e´x decreases, with no local extrema;
and as x Ñ +8, y = f ( x ) Ñ 0.
1
y = f (x)
x
1
(b)
Recall that, to graph the inverse of a function, we reflect the original function across the line
y = x. To see why this is true, consider the following. By definition, the inverse function g of f is
obtained by solving y = f ( x ) for x as a function of y. So, for any pair of numbers x and y, we have
f ( x ) = y if and only g(y) = x
That is, g is the function that swaps the input and output of f . Now the point ( x, y) lies on the
graph of f if and only if y = f ( x ). Similarly, the point ( X, Y ) lies on the graph of g if and only if
Y = g( X ). Choosing Y = x and X = y, we see that the point ( X, Y ) = (y, x ) lies on the graph of g
if and only if x = g(y), which in turn is the case if and only if y = f ( x ). So
(y, x ) is on the graph of g if and only if ( x, y) is on the graph of f .
To get from the point ( x, y) to the point (y, x ) we have to exchange x Ø y, which we can do by
reflecting over the line y = x. Thus we can construct the graph of g by reflecting the curve
y = f ( x ) over the line y = x.
513
y
y=x
y = g( x )
1
y = f (x)
x
1
(c) The domain of g is the range of f , which is (0, 1]. The range of g is the domain of f , which is
[0, 8).
(d) Since g and f are inverses,
g( f ( x )) = x
Using the chain rule,
g1 ( f ( x )) ¨ f 1 ( x ) = 1
Since f 1 ( x ) = é´x = ´ f ( x ):
g1 ( f ( x )) ¨ f ( x ) = ´1
We plug in f ( x ) = 21 .

11 1
g ¨ = ´1
2 2

1
g1 = ´2
2
S-13: (a) First, we differentiate.
f ( x ) = x5 ´ x f 1 ( x ) = 5x4 ´ 1 f 2 ( x ) = 20x3
The function and its first derivative tells us the following:
• lim f ( x ) = 8, lim f ( x ) = ´8
xÑ8 xÑ´8
1
• f 1 ( x ) ą 0 (i.e. f is increasing) for |x| ą ?
4
5
514
1
• f 1 ( x ) = 0 (i.e. f has critical points) for x = ˘ ?
4
« ˘0.67
5
1
• f 1 ( x ) ă 0 (i.e. f is decreasing) for |x| ă ?
4
5

1 4
• f ˘?
4
=¯ ? « ¯0.53
5 545
This gives us a first idea of the shape of the graph.
4
?
545
x
´1
? ?1
4 4
5 5
´4
?
545
We refine this skeleton using information from the second derivative.
• f 2 ( x ) ą 0 (i.e. f is concave up) for x ą 0,
• f 2 ( x ) = 0 (i.e. f has an inflection point) for x = 0, and
• f 2 ( x ) ă 0 (i.e. f is concave down) for x ă 0
Thus
• f has no asymptotes
1 1
• f has a local maximum at x = ´ ?
4
and a local minimum at x = ?
4
5 5
• f has an inflection point at x = 0
• f is concave down for x ă 0 and concave up for x ą 0
515
y
4
? 1
545 ?
4
5
x
´1
?
4 ´4
?
5
545
(b) The function x5 ´ x + k has a root at x = x0 if and only if x5 ´ x = ´k at x = x0 . So the number

of distinct real roots of x5 ´ x + k is the number of times the curve y = x5 ´ x crosses the
1 4
horizontal line y = ´k. The local maximum of x5 ´ x (when x = ´ ? 4
) is ? , and the local
5 545
1 4
minimum of x5 ´ x (when x = ? 4
) is ´ ? . So, looking at the graph of x5 ´ x above, we see that
5 545
the number of distinct real roots of x5 ´ x + k is
4
• 1 when |k| ą ?
545
4
• 2 when |k| = ?
545
4
• 3 when |k| ă ?
545
S-14:
(a) You might not be familiar with hyperbolic sine and cosine, but you don’t need to be. We can
graph them using the same methods as the other curves in this section. The derivatives are given
to us:
d e x + e´x d e x ´ e´x
tsinh xu = cosh x = tcosh xu = sinh x =
dx 2 dx 2
2 x
2
d e é ´x d e x + e´x
tsinh xu = sinh x = tcosh xu = cosh x =
dx 2 dx 2
Observe that:
• sinh( x ) has a derivative that is always positive, so sinh( x ) is always increasing. The second
derivative of sinh( x ) is negative to the left of x = 0 and positive to the right of x = 0, so
sinh( x ) is concave down to the left of the y-axis and concave up to its right, with an
inflection point at x = 0.
516
• cosh( x ) has a derivative that is positive when x ą 0 and negative when x ă 0. The second
derivative of cosh( x ) is always positive, so it is always concave up.
• cosh(0) = 1 and sinh(0) = 0.
ex
• lim sinh x = lim cosh x = lim = 8, since lim e´x = 0
xÑ8 xÑ8 xÑ8 2 xÑ8
e x e´x ex
´x
e
• lim sinh x = lim ´ = lim ´ = ´8 and
xÑ´8 xÑ´8
2x 2 xÑ8 2 2
e e´x e´x ex
lim cosh x = lim + = lim + =8
xÑ´8 xÑ´8 2 2 xÑ8 2 2
e´x + e´(´x) e´x + e x

• cosh( x ) is even, since cosh(´x ) = = = cosh( x ), and
2 2
´x
e é ´ ( ´x ) ´x
e é x ´ (e x ´ e´x )
sinh( x ) is odd, since sinh(´x ) = = = = ´ sinh( x )
2 2 2
y y
y = sinh x y = cosh x
1
x x
(b)
• As y runs over (´8, 8) the function sinh(y) takes every real value exactly once. So, for
each x P (´8, 8), define sinh´1 ( x ) to be the unique solution of sinh(y) = x.
• As y runs over [0, 8) the function cosh(y) takes every real value in [1, 8) exactly once. In
particular, the smallest value of cosh(y) is cosh(0) = 1. So, for each x P [1, 8), define
cosh´1 ( x ) to be the unique y P [0, 8) that obeys cosh(y) = x.
To graph the inverse of a (one-to-one) function, we reflect the original function over the line y = x.
Using this method to graph y = sinh´1 ( x ) is straightforward. To graph y = cosh( x ), we need to
be careful of the domains: we are restricting cosh( x ) to values of x in [0, 8). The graphs are
517
y y
y = sinh x
y = sinh´1 x y = sinh´1 x
x x
y = cosh x
y y
y = cosh´1 x y = cosh´1 x
x x
(c) Let y( x ) = cosh´1 ( x ). Then, using the definition of cosh´1 ,
cosh y( x ) = x
We differentiate with respect to x using the chain rule.
d d
tcosh y( x )u = txu
dx dx
y1 ( x ) sinh y( x ) = 1
We solve for y1 ( x ).
1
y1 ( x ) =
sinh y( x )
We want to have our answer in terms of x, not y. We know that cosh y = x, so if we can convert
hyperbolic sine into hyperbolic cosine, we can get rid of y. Our tool for this is the identity, given
in the question statement, cosh2 ( x ) ´ sinh2 ( x ) = 1. This tells us sinh2 (y) = 1 ´ cosh2 (y). Now we
have to decide whether sinh(y) is the positive or negative square root of 1 ´ cosh2 (y) in our
context. Looking at the graph of y( x ) = cosh´1 ( x ), we see y1 ( x ) ą 0. So we use the positive
square root:
1 1
y1 ( x ) = b =?
2
x ´1
cosh2 y( x ) ´ 1
d ´1
Remark: tarccos( x )u = ? , so again the hyperbolic trigonometric function has
dx 1 ´ x2
properties similar to (but not exactly the same as) its trigonometric counterpart.
518
S-1: There are many possible answers. Consider these: f ( x ) = 5x, g( x ) = 2x. Then
f (x) 5x 5 5
lim f ( x ) = lim g( x ) = 8, and lim = lim = lim = = 2.5.
xÑ8 xÑ8 xÑ8 g ( x ) xÑ8 2x xÑ8 2 2
S-2: There are many possible answers. Consider these: f ( x ) = x, g( x ) = x2 . Then

f (x) x 1
lim f ( x ) = lim g( x ) = 8, and lim = lim 2 = lim = 0.
xÑ8 xÑ8 xÑ8 g ( x ) xÑ8 x xÑ8 x
a log 5
S-3: From Example 3.7.20, we know that lim (1 + x ) x = e a , so lim (1 + x ) x = elog 5 = 5.
xÑ0 xÑ0
However, this is the limit as x goes to 0, which is not what we were asked. So, we modify the
functions by replacing x with 1x . If x Ñ 0+ , then 1x Ñ 8.
Taking f ( x ) = 1 + 1xand g(x ) = x log 5, we see:

1
(i) lim f ( x ) = lim 1 + =1
xÑ8 xÑ8 x
(ii) lim g( x ) = lim x log 5 = 8
xÑ8 xÑ8
1
(iii) Let us name = X. Then as x Ñ 8, X Ñ 0+ , so:
x
log 5
1 x log 5

g( x ) 1 1x log 5
lim [ f ( x )] = lim 1 + = lim 1 + = lim+ [1 + X ] X = elog 5 = 5, where in the
xÑ8 xÑ8 x xÑ8 x XÑ0
penultimate step, we used the result of Example 3.7.20.
S-4: If we plug in x = 1 to the numerator and the denominator, we find they are both zero. So, we
have an indeterminate form appropriate for L’Hôpital’s Rule.
x3 ´ e x´1 3x2 ´ e x´1 2

lim = lim =´
sin(πx )
xÑ1 loooomoooon xÑ1 π cos( πx ) π
numÑ0
denÑ0
S-5: Be careful– this is not an indeterminate form!
As x Ñ 0+, the numerator log x Ñ ´8. That is, the numerator is becoming an increasingly huge,
negative number. As x Ñ 0+, the denominator x Ñ 0+, which only serves to make the total
log x
fraction even larger, and still negative. So, lim = ´8.
xÑ0 + x
Remark: if we had tried to use l’Hôpital’s Rule here, we would have come up with the wrong
1
answer. If we differentiate the numerator and the denominator, the fraction becomes x = 1x , and
1
1
lim = 8. The reason we cannot apply l’Hôpital’s Rule is that we do not have an indeterminate
xÑ0+ x
form, like both numerator and denominator going to infinity, or both numerator and
denominator going to zero.
519
S-6: We rearrange the expression to a more natural form:
(log x )2
lim (log x )2 e´x = lim
xÑ8 ex
xÑ8 looomooon
numÑ8
denÑ8
Both the numerator and denominator go to infinity as x goes to infinity. So, we can apply
l’Hôpital’s Rule. In fact, we end up applying it twice.
2 log x
= lim
xe x
xÑ8 loomoon
numÑ8
denÑ8
2/x
= lim
xÑ8 xe x + e x
The numerator gets smaller and smaller while the denominator gets larger and larger, so:
=0
S-7:
x2 2x 2
lim x2 e´x = lim x
= lim x
= lim =0
xÑ8 e on xÑ8 loomo
xÑ8 loomo e x on
e on xÑ8 loomo
numÑ8 numÑ8 numÑ8
denÑ8 denÑ8 denÑ8
S-8:
x ´ x cos x 1 ´ cos x + x sin x sin x + sin x + x cos x
lim = lim = lim
x ´ sin x
xÑ0 looooomooooon 1 ´ cos x
xÑ0 loooooooooomoooooooooon sin x
xÑ0 looooooooooooomooooooooooooon
numÑ0 numÑ0 numÑ0
denÑ0 denÑ0 denÑ0
2 cos x + cos x ´ x sin x

= lim =3
xÑ0 cos x
S-9: If we plug in x = 0 to the numerator and denominator, both are zero, so this is a candidate for
l’Hôpital’s Rule. However, an easier way to evaluate the limit is to factor x2 from the numerator
and denominator, and cancel.
? ? ?
x6 + 4x4 x4 x2 + 4
lim 2 = lim
xÑ0 x cos x xÑ0 x2 cos x
2
?
x x2 + 4
= lim 2
xÑ0 x cos x
?
x2 + 4
= lim
xÑ0 cos x
?
02 + 4
= =2
cos(0)
520
S-10:
(log x )2 2(log x ) 1x log x 1
lim = lim = 2 lim = 2 lim x = 0
x
xÑ8 looomooon xÑ8 1 x on
xÑ8 loomo xÑ8 1
numÑ8 numÑ8
denÑ8 denÑ8
S-11:
1 ´ cos x sin x 1 1
lim 2
= lim = lim =
sin x
xÑ0 loooomoooon xÑ0 2 sin x cos x xÑ0 2 cos x 2
numÑ0
denÑ0
S-12: If we plug in x = 0, the numerator is zero, and the denominator is

1 1 0
sec 0 = = = 1. So the limit is = 0.
cos 0 1 1
Be careful: you cannot use l’Hôpital’s Rule here, because the fraction does not give an
indeterminate form. If you try to differentiate the numerator and the denominator, you get an
expression whose limit does not exist:
1 cos x
lim = lim cos x ¨ = DNE.
xÑ0 sec x tan x xÑ0 sin x
S-13: If we plug x = 0 into the denominator, we get 1. However, the numerator is an

1
indeterminate form: tan 0 = 0, while lim csc x = 8 and lim csc x = ´8. If we use csc x = sin x ,
xÑ0+ xÑ0´
our expression becomes
tan x ¨ ( x2 + 5)
lim
xÑ0 sin x ¨ e x
Since plugging in x = 0 makes both the numerator and the denominator equal to zero, this is a
candidate for l’Hôspital’s Rule. However, a much easier way is to simplify the trig first.
tan x ¨ ( x2 + 5) sin x ¨ ( x2 + 5)
lim = lim
xÑ0 sin x ¨ e x xÑ0 cos x ¨ sin x ¨ e x
x2 + 5
= lim
xÑ0 cos x ¨ e x
02 + 5
= =5
cos(0) ¨ e0
a a
S-14: lim 2x2 + 1 ´ x2 + x has the indeterminate form 8 ´ 8. To get a better idea of what’s
xÑ8
going on, let’s rationalize.
? ? !
a a a a 2x2 + 1 + x2 + x
lim 2x2 + 1 ´ x2 + x = lim 2x2 + 1 ´ x2 + x ? ?
xÑ8 xÑ8 2x2 + 1 + x2 + x
(2x2 + 1) ´ ( x2 + x ) x2 ´ x + 1
= lim ? ? = lim ? ?
xÑ8 2x2 + 1 + x2 + x xÑ8 2x2 + 1 + x2 + x
521
Here, we have the indeterminate form 8 8 , so l’Hôpital’s Rule applies. However, if we try to use it
here, we quickly get a huge mess. Instead, remember how we dealt with these kinds of limits in
the past: factor out the highest power of the denominator, which is x.
x x ´ 1 + 1x

= lim b b
xÑ8
x2 (2 + x12 ) + x2 (1 + 1x )
x x ´ 1 + 1x

= lim b b
xÑ8
x 2 + x2 + 1 + 1x
1
x ´ 1 + 1x
= lim b b
xÑ8
2 + x12 + 1 + 1x
looooooooooomooooooooooon
numÑ8
?
denÑ 2+1
=8
S-15: If we plug in x = 0, both numerator and denominator become zero. So, we have exactly one
of the indeterminate forms that l’Hôpital’s Rule applies to.
sin( x3 + 3x2 ) (3x2 + 6x ) cos( x3 + 3x2 )

lim = lim
sin2 x
xÑ0 looooooomooooooon xÑ0 2 sin x cos x
numÑ0
denÑ0
If we plug in x = 0, still we find that both the numerator and the denominator go to zero. We
could jump in with another iteration of l’Hôpital’s Rule. However, the derivatives would be a
little messy, so we use limit laws and break up the fraction into the product of two fractions. If
both limits exist:
(3x2 + 6x ) cos( x3 + 3x2 ) x2 + 2x 3 cos( x3 + 3x2 )

lim = lim ¨ lim
xÑ0 2 sin x cos x xÑ0 sin x xÑ0 2 cos x
We can evaluate the right-hand limit by simply plugging in x = 0:
3 x2 + 2x
= lim
sin x
2 xÑ0 looomooon
numÑ0
denÑ0
3 2x + 2
= lim
2 xÑ0 cos x

3 2
= =3
2 1
S-16:
log( x3 ) 3 log( x ) 3/x 3
lim 2
= lim 2 = lim =
xÑ1 x ´ 1 x ´1
xÑ1 looomooon xÑ1 2x 2
numÑ0
denÑ0
522
S-17:
• Solution 1.
2 1 ´4 2 ´4
e´1/x x4 x5 x2 x3 2
lim = lim 1/x 2 = lim = lim 1/x2 = lim = lim 2 =0
xÑ0 x4 eomoon xÑ0
xÑ0 lo ´2 1/x2
x3
e eomoon
xÑ0 lo xÑ0 ´2
x3
e1/x
2
xÑ0 e1/x
numÑ8 numÑ8
denÑ8 denÑ8
1 2 2
since, as x Ñ 0, the exponent x2
Ñ 8 so that e1/x Ñ 8 and e´1/x Ñ 0.
• Solution 2.
2
e´1/x e´t t2 2t 2
lim 4
= lim ´2
= lim t
= lim t
= lim t = 0
xÑ0 x t= 12 Ñ8 t e on tÑ8 loomo
tÑ8 loomo e on tÑ8 e
x
numÑ8 numÑ8
denÑ8 denÑ8
S-18:
xe x e x + xe x 1
lim = lim 2
=
tan(3x )
xÑ0 looomooon xÑ0 3 sec (3x ) 3
numÑ0
denÑ0
1
S-19: lim sin2 x = 0, and lim = 8, so we have the form 08 . (Note that sin2 x is positive, so our
xÑ0 xÑ0 x 2
x2
a
root is defined.) This is not an indeterminate form: lim sin2 x = 0.
xÑ0
1 1
S-20: lim cos x = 1 and lim 2
= 8, so lim (cos x ) x2 has the indeterminate form 18 . We want to
xÑ0 xÑ0 x xÑ0
use l’Hôpital, but we need to get our function into a fractional indeterminate form. So, we’ll use a
logarithm.
1
y : = (cos x ) x2
1
1 log cos x
log y = log (cos x ) x2 = 2 log(cos x ) =
x x2
´ sin x
log cos x cos x ´ tan x ´ sec2 x
lim log y = lim = lim = lim = lim
xÑ0 x2
xÑ0 loooomoooon xÑ0 2x 2x
xÑ0 looomooon xÑ0 2
numÑ0 numÑ0
denÑ0 denÑ0
´1 1
= lim =´
xÑ0 2 cos2 x 2
1
Therefore, lim y = lim elog y = e´1/2 = ?
xÑ0 xÑ0 e
S-21:
• Solution 1
y : = e x log x = (e x )log x
lim y = lim (e x )log x
xÑ0+ xÑ0+
523
This has the form 1´8 = 118 , and 18 is an indeterminate form. We want to use l’Hôpital,
but we need to get a different indeterminate form. So, we’ll use logarithms.

lim log y = lim log (e x )log x = lim log x log (e x ) = lim (log x ) ¨ x
xÑ0+ xÑ0+ xÑ0+ xÑ0+
This has the indeterminate form 0 ¨ 8, so we need one last adjustment before we can use
l’Hôpital’s Rule.
1
log x x
= lim+ 1
= lim+ ´1
= lim+ ´x = 0
xÑ0 xÑ0 xÑ0
x on
loomo x2
numÑ´8
denÑ8
Now, we can figure out what happens to our original function, y:

lim y = lim elog y = e0 = 1
xÑ0+ xÑ0+
• Solution 2
x
y : = e x log x = elog x = x x
lim y = lim x x
xÑ0+ xÑ0+
We have the indeterminate form 00 . We want to use l’Hôpital, but we need a different
indeterminate form. So, we’ll use logarithms.
lim log y = lim log( x x ) = lim x log x
xÑ0+ xÑ0+ xÑ0+
Now we have the indeterminate form 0 ¨ 8, so we need one last adjustment before we can
use l’Hôpital’s Rule.
1
log x x
lim y = lim 1
= lim+ ´1
= lim+ ´x = 0
xÑ0+ xÑ0+ xÑ0 xÑ0
x on
loomo x2
numÑ0
denÑ´8
Now, we can figure out what happens to our original function, y:

xÑ0+ xÑ0+
S-22: First, note that the function exists near 0: x 2 is positive, so log( x2 ) exists; near 0, log x2 is
x
negative, so ´ log( x2 ) is positive, so ´ log( x2 ) exists even when x is negative.

Since lim ´ log( x2 ) = 8 and lim x = 0, we have the indeterminate form 80 . We need l’Hôpital,
xÑ0 xÑ0
but we need to manipulate our function into an appropriate form. We do this using logarithms.
x
y : = ´ log( x2 )

 
log ´ log( x2 )

2 x

2

log y = log ´ log( x ) = loomo ´ log( x ) =
x on ¨ log loooomoooon 1
Ñ0 x
Ñ8
loooooooooomoooooooooon
Ñ8
´ 2x
log ´ log( x2 )

´ log( x2 ) ´2x
lim log y = lim = lim ´1 = lim =0
xÑ0 xÑ0 1 xÑ0 xÑ0 log( x 2 )
x
looooooooomooooooooon x2 looomooon
numÑ8 numÑ0
denÑ˘8 denÑ´8
524
Now, we’re ready to figure out our original limit.

xÑ0 xÑ0
S-23: Both the numerator and denominator converge to 0 as x Ñ 0. So, by l’Hôpital,
1 + cx ´ cos x c + sin x
lim 2 = lim 2
ex ´ 1
xÑ0 looooooomooooooon xÑ0 2xe x
numÑ0
denÑ0
The new denominator still converges to 0 as x Ñ 0. For the limit to exist, the same must be true
for the new numerator. This tells us that if c ‰ 0, the limit does not exist. We should check
whether the limit exists when c = 0. Using l’Hôpital:
sin x cos x 1 1
lim 2 = lim 2 = = .
omoxon
2xe
xÑ0 lo xÑ0 e x (4x 2 + 2) 1(0 + 2) 2
numÑ0
denÑ0
So, the limit exists when c = 0.
S-24: The first thing we notice is, regardless of k, when we plug in x = 0 both numerator and
denominator become zero. Let’s use this fact, and apply l’Hôpital’s Rule.
2 2
ek sin(x ) ´ (1 + 2x2 ) 2kx cos( x2 )ek sin(x ) ´ 4x
lim = lim
x4
xÑ0 looooooooooomooooooooooon xÑ0 4x3
numÑ0
denÑ0
2
2k cos( x2 )ek sin(x ) ´ 4
= lim
xÑ0 4x2
When we plug in x = 0, the denominator becomes 0, and the numerator becomes 2k ´ 4. So, we’ll
need some cases, because the behaviour of the limit depends on k.
For k = 2:
2 2
2k cos( x2 )ek sin(x ) ´ 4 4 cos( x2 )e2 sin(x ) ´ 4
lim = lim
xÑ0 4x2 4x2
xÑ0 loooooooooooomoooooooooooon
numÑ0
denÑ0
2 2)
´8x sin( x2 )e2 sin(x ) + 16x cos2 ( x2 )e2 sin(x
= lim
xÑ0 8x
2 2 sin( x2 ) 2
+ 2 cos2 ( x2 )e2 sin(x )

= lim ´ sin( x )e
xÑ0
=2
For k ą 2, the numerator goes to 2k ´ 4, which is a positive constant, while the denominator goes
to 0 from the right, so:
2
lim =8
xÑ0 4x2
525
For k ă 2, the numerator goes to 2k ´ 4, which is a negative constant, while the denominator goes
to 0 from the right, so:
2
lim = ´8
xÑ0 4x2
S-25:
• We want to find the limit as n goes to infinity of the percentage error,

|S(n) ´ A(n)|
lim 100 . Since A(n) is a nicer function than S(n), let’s simplify:
nÑ8 |S(n)| ˇ ˇ
|S(n) ´ A(n)| ˇ A ( n ) ˇ
lim 100 = 100 ˇˇ1 ´ lim ˇ.
nÑ8 |S(n)| nÑ8 S ( n ) ˇ
We figure out this limit the natural way:

ˇ ˇ
ˇ ˇ
ˇ ˇ
4
ˇ ˇ ˇ ˇ
ˇ A(n) ˇ ˇ ˇ 5n ˇ
100 ˇˇ1 ´ lim = 100 ˇ1 ´ lim ˇ
nÑ8 S ( n ) ˇ ˇ
ˇ nÑ8 5n4 ´ 13n3 ´ 4n + log( n ) ˇˇ
ˇ loooooooooooooooomoooooooooooooooonˇ
ˇ numÑ8 ˇ
denÑ8
ˇ ˇ
ˇ 20n3 ˇ
= 100 ˇ1 ´ lim
ˇ ˇ
ˇ 3 2
nÑ8 20n ´ 39n ´ 4 + ˇ 1 ˇ
n
ˇ ˇ
ˇ n 3 20 ˇ
= 100 ˇ1 ´ lim 3 ¨
ˇ ˇ
ˇ nÑ8 n 20 ´ 39 n ´ 4
n3 + n
1 ˇˇ
4
= 100|1 ´ 1| = 0
So, as n gets larger and larger, the relative error in the approximation gets closer and closer
to 0.
• Now, let’s look at the absolute error.
lim |S(n) ´ A(n)| = lim | ´ 13n3 ´ 4n + log n| = 8

nÑ8 nÑ8
So although the error gets small relative to the giant numbers we’re talking about, the absolute
error grows without bound.
S-1: An antiderivative of f 1 ( x ) is a function whose derivative is f 1 ( x ). Our original function f ( x )

has this property, so f ( x ) is an antiderivative of f 1 ( x ), but it’s not the most general. We can add a
constant to f ( x ) without affecting its derivative. The most general antiderivative of f 1 ( x ) is
f ( x ) + C, where C is any constant.
S-2: Notice f ( x ) is nonnegative for an interval covering the left part of the graph, and negative on
the right part of the graph. That means its antiderivative is increasing for the left interval, then
decreasing for the right interval. This applies to A( x ) and C ( x ), but not B( x ).
526
There are only three points where A( x ) has a horizontal tangent line: at its global maximum and
the endpoints of the interval shown. By contrast, C ( x ) has a horizontal tangent line in four places:
at its global maximum, at its inflection point, and at the endpoints of the interval shown. Since
f ( x ) = 0 four times (and these line up with the horizontal portions of C ( x )) we conclude C ( x ) is
the antiderivative of f ( x ).
1
S-3: For any constant n ‰ ´1, an antiderivative of x n is n +1 x
n +1 .
F1 ( x ) = 3x2 + 5x4 + 10x ´ 9

1 3 1 5 1 2 1
F(x) = 3 x +5 x + 10 x ´9 x1 + C
3 5 2 1
= x3 + x5 + 5x2 ´ 9x + C
Remark: we can always check by differentiating:
d
x3 + x5 + 5x2 ´ 9x + C
(
F1 ( x ) =
dx
= 3x2 + 5x4 + 10x ´ 9
= f (x)
so F ( x ) is indeed an antiderivative of f ( x ).
S-4:
3 7
F1 ( x ) =x ´ 18x4 + x
5

3 1 8 1 1
F(x) = x ´ 18 x5 + x2 + C
5 8 5 2
3 8 18 5 1 2
= x ´ x + x +C
40 5 2
1
S-5: For any constant n ‰ 1, an antiderivative of x n is n +1 x
n +1 . The constant n does not have to be
an integer.
? 9
F1 ( x ) = 4 3 x ´
2x2.7
1 9
= 4x 3 ´ x´2.7
2 !
1

1 +1 9 1
F(x) = 4 1 x 3 ´ x (´2.7+1) + C
3 +1
2 ´2.7 + 1

3 4 9 10
=4 x3 ´ x´1.7 + C
4 2 ´17
4 45
= 3x 3 + +C
17x1.7
S-6:
527
• Solution 1: We can re-write f ( x ) to make it a power of x.
1 ´1
F1 ( x ) =x 2
7 !
1

1 1 ´ 2 +1
F(x) = x +C
7 ´ 12 + 1

1 1
= (2) x 2 + C
7
2?
= x+C
7
1 1 ?
• Solution 2: We notice that ? looks a lot like ? , which is the derivative of x. So:
7 x 2 x
d ? ( 1
x = ?
dx 2 x
" *
d 2 ? 2 1
x = ? = f (x)
dx 7 7 2 x
2?
So, an antiderivative of f ( x ) is x. Then the most general antiderivative is
7
2 ?
F(x) = x + C.
7
d x
S-7: We recall e = e x . That is, e x is its own antiderivative. So, a first guess for the
dx
antiderivative of f ( x ) might be itself.
d ! 5x+11 )
e = 5e5x+11
dx
This isn’t exactly right, so we modify it by multiplying by a constant.
" *
d 1 5x+11
e = e5x+11
dx 5
1
This tells us that e5x+11 is an antiderivative of e5x+11 . Therefore, the most general antiderivative
5
+ 1
of e 5x 11 is F ( x ) = e5x+11 + C.
5
S-8: We know the derivatives of sine and cosine. We’ll work from there to build a function whose
derivative is f ( x ). We’ll start by finding an antiderivative of 7 cos(13x ).
d
tsin xu = cos x
dx
d
tsin(13x )u = 13 cos(13x )
"dx *
d 1 13
sin(13x ) = cos(13x ) = cos(13x )
dx 13 13
" *
d 7
sin(13x ) = 7 cos(13x )
dx 13
528
7
So, an antiderivative of 7 cos(13x ) is sin(13x ).
13
d
tcos xu = ´ sin x
dx
d
t´ cos xu = sin x
dx
d
t´ cos(5x )u = 5 sin(5x )
dx
" *
d 3 3
´ cos(5x ) = 5 sin(5x ) = 3 sin(5x )
dx 5 5
3
So, an antiderivative of 3 sin(5x ) is ´ cos(5x ).
5
The most general antiderivative of 3 sin(5x ) + 7 cos(13x ) is
3 7
F ( x ) = ´ cos(5x ) + sin(13x ) + C.
5 13
S-9: We know the derivative of tan x is sec2 x. Modifying this slightly, we see (using the chain rule)
d d
ttan( x + 1)u = sec2 ( x + 1) ¨ tx + 1u
dx dx
= sec2 ( x + 1)
So, tan( x + 1) is an antiderivative of sec2 ( x + 1). Therefore, the most general antiderivative of
sec2 ( x + 1) is F ( x ) = tan( x + 1) + C.
1
S-10: We note that f ( x ) looks similar to .
x
d 1
tlog |x|u =
dx x
d 1
tlog |x + 2|u =
dx x+2
The most general antiderivative of f ( x ) is F ( x ) = log |x + 2| + C.
1
S-11: Our function f ( x ) bears some resemblance to the derivative of arcsine, ? :
1 ´ x2

7 7 1
f (x) = ? =? ?
3 ´ 3x2 3 1 ´ x2
" *
d 7 7 1
? arcsin( x ) = ? ? = f (x)
dx 3 3 1 ´ x2
7
So, the most general antiderivative of f ( x ) is F ( x ) = ? arcsin( x ) + C.
3
1
S-12: We notice that f ( x ) looks similar to the derivative of the arctangent function, .
1 + x2
1 1
f (x) = =
1 + 25x2 1 + (5x )2
529
This gives us a first guess for our antiderivative: perhaps arctan(5x ) will work. We test it by
differentiating, making sure we don’t forget the chain rule.
d 1
tarctan(5x )u = ¨5
dx 1 + (5x )2
We’re close to f ( x ), but we’ve multiplied by 5. That’s easy to take care of: we can divide our
guess by 5.
" * !
d 1 1 1
arctan(5x ) = ¨5
dx 5 5 1 + (5x )2
1
= = f (x)
1 + 25x2
1
So, the most general antiderivative of f ( x ) is F ( x ) = arctan(5x ) + C.
5
S-13: First, let’s find the antiderivative of f 1 ( x ). It’s a polynomial, so we can use the observation
1
from the text that an antiderivative of x n , for any constant n ‰ 1, is n+ 1x
n+1 . Remember that the
most general antiderivative will have an added constant.

3 3 9 2
f (x) = x ´ x + 4x + C
3 2
9
= x3 ´ x2 + 4x + C
2
Use the fact that f (1) = 10 to solve for C.
9
10 = 1 ´ +4+C
2
19
C=
2
All together,
9 19
f ( x ) = x3 ´ x2 + 4x + .
2 2
S-14: First, let’s find the antiderivative of f 1 ( x ). We know that one antiderivative of cos( x ) is sin x.
We might guess that an antiderivative of cos(2x ) is sin(2x ). Check by differentiating:
d
tsin(2x )u = 2 cos(2x )
dx
This is close to f 1 ( x ), but we need to divide by 2.
" *
d 1
sin(2x ) = cos(2x )
dx 2
1
So, f ( x ) = sin(2x ) + C for some constant C. We can find C using the given information
2
f (π ) = π.
1
π = f (π ) = sin(2π ) + C
2
π=C
530
1
Therefore, f ( x ) = sin(2x ) + π.
2
1
S-15: Looking at the table in the notes, we see the antiderivative of is f ( x ) = log |x| + C. The
x
given information f (´1) = 0 lets us find C:
0 = f (´1) = log | ´ 1| + C
0 = log(1) + C
0=C
So, f ( x ) = log |x|.

1 1
Remark: it is true that log x is an antiderivative of , since the derivative of log x is . However,
x x
log x is only defined for positive values of x. Since the given information tells you that f ( x ) is
1
defined when x = ´1, you need to use a more general antiderivative of : log |x| + C.
x
1
S-16: An antiderivative of 1 is x, and an antiderivative of ? is arcsin( x ). So,
1 ´ x2
f ( x ) = arcsin x + x + C. The given information lets us find C.
π
´ = f (1) = arcsin(1) + 1 + C
2
π π
´ = +1+C
2 2
C = ´π ´ 1
So, f ( x ) = arcsin x + x ´ π ´ 1.
S-17: If P(t) is the population at time t, then the information given in the problem is
P1 (t) = 100e2t . Antidifferentiating, we see P(t) = 50e2t + C, where C is some constant. We want
to know for what value of t we get P(t) = P(0) + 300.
P(t) = P(0) + 300

50e + C = 50e0 + C + 300
2t
50e2t = 350
e2t = 7
2t = log(7)
1
t = log(7)
2
1
It takes 2 log 7 hours (about 58 minutes) for the initial colony to increase by 300 individuals.
S-18: If A(t) is the amount of money in your account at time t, then the given information is
t
A1 (t) = 1500e 50
Antidifferentiating,
t
A(t) = 75000e 50 + C
531
for some constant C.
t
That is, at time t, the amount of money in your bank account is 75000e 50 + C dollars, for some
constant C.
S-19: Let P(t) be the amount of power your house has used since time t = 0. If t is measured in
hours, and P(t) in kWh (kilowatt-hours), then P1 (t) is the rate at which your house is consuming
power, in kW. So, the given information is that
π
P1 (t) = 0.5 sin t + 0.25
24
Antidifferentiating,
12 π
P(t) = ´ cos t + 0.25t + C
π 24
Since P(0) is the amount of energy consumed after 0 hours, P(0) = 0, so
12
0 = P (0) = ´ +C
π
12
C=
π
12 h π i
P(t) = 1 ´ cos t + 0.25t
π 24
After 24 hours, your energy consumed is
12
P(24) = [1 ´ cos (π )] + 0.25(24)
π
24
= + 6 « 13.6 kWh
π
d 1
S-20: We differentiate f ( x ) and g( x ) using the chain rule. Recall tarcsin( x )u = ? .
dx 1 ´ x2
2 d ?
f 1 (x) = a ? ¨ x
1 ´ x2 dx
2 1
= ? ¨ ?
1´x 2 x
1
= ?
x ´ x2
1 d
g1 ( x ) = a ¨ t2x ´ 1u
1 ´ (2x ´ 1)2 dx
2
= ?
2
1 ´ 4x + 4x ´ 1
1
= ?
x ´ x2
532
So, the derivative of f ( x ) ´ g( x ) is 0, which implies that f ( x ) and g( x ) differ by a
constant–perhaps a surprising result!
Remark: we ? to show that f ( x ) and g( x ) only differ by a constant. Define
? don’t need calculus
θ = sin´1 x, so that sin θ = x and f ( x ) = 2θ , and then
h πi
sin f ( x ) ´ = ´ cos f ( x ) = ´ cos 2θ = ´[1 ´ 2 sin2 θ ] = 2 sin2 θ ´ 1 = 2x ´ 1
2
Our goal is to take the arcsine of the first and last expressions and conclude
π
f ( x ) ´ = arcsin(2x ´ 1) = g( x ). However, before we can say
2 h π i π π π π
arcsin sin f ( x ) ´ = f ( x ) ´ , we have to check that ´ ď f ( x ) ´ ď . That is, we
2 2 2 2 2
need to show that 0 ď f ( x ) ď π.
?
As 0 ď θ ď π2 (since sin´1 always takes values between ´ π2 and π2 and since x ě 0), we have
that 0 ď f ( x ) = 2θ ď π. So,
π h π i
f ( x ) ´ = arcsin sin f ( x ) ´ = arcsin(2x + 1) = g( x ).
2 2
S-21: The derivative of sin(2x ) (which occurs in the second term) is 2 cos(2x ) (which occurs in the
first term). Similarly, the derivative of cos(3x ) (which occurs in the first term) is ´3 sin(3x )
(which occurs in the second term). So, f ( x ) seems to have come from the product rule.
d
tsin(2x ) cos(3x )u = 2 cos(2x ) cos(3x ) ´ 3 sin(2x ) sin(3x )
dx
Then the antiderivative of f ( x ) is F ( x ) = sin(2x ) cos(3x ) + C.
S-22: The function f ( x ) looks like perhaps it came from the quotient rule. Recall
" *
d u( x ) v ( x ) u1 ( x ) ´ u ( x ) v1 ( x )
=
dx v( x ) (v( x ))2
Then, since the denominator of f ( x ) is ( x2 + 1)2 , we might guess v( x ) = ( x2 + 1). That leaves
u( x ) = e x .
" x *
d e ( x2 + 1)e x ´ e x (2x )
= = f (x)
dx x2 + 1 ( x 2 + 1)2
ex
So, the antiderivative of f ( x ) is F ( x ) = + C.
x2 + 1
S-23: We know that e x is its own antiderivative. The derivative of e g(x) , for some function g( x ), is
g1 ( x )e g(x) . Since 3x2 is the derivative of x3 , f ( x ) fits this pattern. We guess the antiderivative of
3
f ( x ) is F ( x ) = e x + C.
We check by differentiating.
d ! x3 ) 3
e + C = 3x2 e x = f ( x )
dx
3
Indeed, the antiderivative of f ( x ) is F ( x ) = e x + C.
533
S-24: We know the antiderivative of sin x is ´ cos x. Since sin( x2 ) appears in our function, let’s
investigate the derivative of ´ cos( x2 ).
d
´ cos x2 = sin x2 ¨ 2x
(
dx
This differs from f ( x ) only by a constant multiple.

" *
d 5 2 5
= sin x2 ¨ 2x = 5x sin( x2 ) = f ( x )

´ cos x
dx 2 2
5
So, the antiderivative of f ( x ) is F ( x ) = ´ cos x2 + C

2
Remark: as in Question 23, our function f ( x ) involved some function g( x ) as well as g1 ( x ). This
pattern is the basis of an important method of antidifferentiation, called the Substitution Rule.
S-25: For any x in the domain of log( x ), elog x = x. So, f ( x ) = x for every x in its domain. Then its
1
antiderivative is F ( x ) = x2 + C.
2
1
S-26: As in Question 11, we notice that our function is similar to ? , but in this case it
1 ´ x2
doesn’t factor quite as nicely.
7 7
f (x) = ? =b
3´x 2 2
3(1 ´ x3 )
 
7 1
= ? b 
3 1´ x2
3
x2
What we really want under that square root, instead of , is simply x2 . We can get close: we can
3
get something squared.
 
7  1 
f (x) = ?  c 
3 2 
1´ ?x
3
x
Now, the thing that’s squared isn’t x, it’s ? . This gives us a first guess for an antiderivative:
3
7 x
perhaps F ( x ) = ? arcsin ? will work. Let’s try it! Remember to use the chain rule when
3 3
you differentiate.
 
" *
d 7 x 7  1  1
? arcsin ? =?  c 2  ¨ ?3

dx 3 3 3
1´ ?x
3
534
?
We’re very close! We’re only off by a constant, and those?are easy to fix. We’re dividing by 3
when we differentiate, so let’s multiply our function by 3.
 
" *
d x  1  1
7 arcsin ? = 7
c 2  ¨ ?3

dx 3
1´ ?x
3
7
= c
? 2
3 1 ´ ?x3
7
=? = f (x)
3 ´ x2

x
Therefore, the most general antiderivative of f ( x ) is F ( x ) = 7 arcsin ? + C.
3
S-27: Following Example 4.1.7, let V ( H ) be the volume of the solid formed by rotating the
segment of the parabola from x = ´H to x = H. Our plan is to evaluate
V ( H ) ´ V (h)
V 1 ( H ) = lim and then antidifferentiate V 1 ( H ) to find V ( H ). Since we don’t know
hÑH H´h
V ( H ) ´ V (h) (yet) we first find upper and lower bounds on it when h ă H.
For a constant h ă H, V ( H ) ´ V (h) is the volume of the solid inside the larger object (with length
2H) and outside the smaller object (with length 2h, shown below in blue). There are two regions
inside the larger object and outside the smaller.
One of the regions inside the larger object and outside the smaller (blue) object looks like this:
535
y = x2 + 1
x
h H
H´h
Remember we formed this solid by rotating the curve y = x2 + 1 about the x-axis. So, the
cross-section of this solid is a circle, and the radius of the circle when we are x units from the
origin is x2 + 1. So, the largest radius of the little shape shown above, which occurs at the right
end, is H 2 + 1, and the smallest radius (at the left end) is h2 + 1.
The volume of the shape shown above is less than a cylinder of radius H 2 + 1 and height H ´ h,
and it is more than the volume of a cylinder of radius h2 + 1 and height H ´ h. So, the volume of
the shape shown above is between ( H ´ h)π (h2 + 1)2 and ( H ´ h)π ( H 2 + 1)2 cubic units.
Recall that the volume inside the object of length 2H and outside the object of length 2h consists
of two copies of the shape shown above. Therefore:
2( H ´ h ) π ( h2 + 1)2 ă V ( H ) ´ V (h) ă 2( H ´ h ) π ( H 2 + 1)2
Now that we have upper and lower bounds for V ( H ) ´ V (h), we can find V 1 ( H ).
2( H ´ h ) π ( h2 + 1)2 V ( H ) ´ V (h) 2( H ´ h ) π ( H 2 + 1)2

ă ă
H´h H´h H´h
V ( H ) ´ V (h)
2π (h2 + 1)2 ă ă 2π ( H 2 + 1)2
H´h
V ( H ) ´ V (h)
lim 2π (h2 + 1)2 ď lim ď lim 2π ( H 2 + 1)2
hÑH hÑH H´h hÑH
Since the limits on both ends are simply 2π ( H 2 + 1)2 , by the Squeeze Theorem,
V ( H ) ´ V (h)
V 1 ( H ) = lim = 2π ( H 2 + 1)2
hÑH H´h
Now that we know V 1 ( H ), we antidifferentiate to find V ( H ).
V 1 ( H ) = 2π ( H 4 + 2H 2 + 1)

1 5 2 3
V ( H ) = 2π H + H +H +C
5 3
When H = 0, there is no solid, so V (0) = 0. Therefore,

1 5 2 3
V ( H ) = 2π H + H +H .
5 3
536

CLP-1 Differential Calculus Problem Book2

Uploaded by

Copyright:

Available Formats

CLP-1 Differential Calculus Problem Book2

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

CLP-1 Differential Calculus Problem Book2

Uploaded by

Copyright:

Available Formats

CLP-1 D IFFERENTIAL C ALCULUS

Joel F ELDMAN Andrew R ECHNITZER

T HIS DOCUMENT WAS TYPESET ON T UESDAY 23 RD A UGUST, 2022.

• Links to the source files can be found at the text webpage

§§ How to Work Questions

§§ Working with Friends

How to use this book i

3.2 Related Rates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55

4 Towards Integral Calculus 103

II Hints to problems 107

2.8 Derivatives of trigonometric functions . . . . . . . . . . . . . . . . . . . . . 118

III Answers to problems 145

2.6 Using the arithmetic of derivatives - examples . . . . . . . . . . . . . . . . . 161

IV Solutions to problems 215

2.4 Arithmetic of derivatives - a differentiation toolbox . . . . . . . . . . . . . . 271

1.1Ĳ Drawing Tangents and a First Limit

Q[1]: On the graph below, draw:

Q[2]: Suppose a curve y = f ( x ) has tangent line y = 2x + 3 at the point x = 2.

1.2Ĳ Another Limit and Computing Velocity

(b) What is the velocity of the object at time t = 1?

1.3Ĳ The Limit of a Function

Q[2]: Given the function shown below, evaluate lim f ( x ).

Q[3]: Given the function shown below, evaluate:

Q[4]: Draw a curve y = f ( x ) with lim f ( x ) = f (3) = 10.

Q[5]: Draw a curve y = f ( x ) with lim f ( x ) = 10 and f (3) = 0.

Q[6]: Suppose lim f ( x ) = 10. True or false: f (3) = 10.

Q[7]: Suppose f (3) = 10. True or false: lim f ( x ) = 10.

Q[10]: lim sin t

Q[11]: lim log x

1.4Ĳ Calculating Limits with Limit Laws

Q[33]: lim x5 ´ 32x + 15, where x is a constant

Q[44]: Suppose f ( x ) = 2x and g( x ) = 1x . Evaluate the following limits.

Q[48]: Let f ( x ) = 1x and g( x ) = ´1

(b) Evaluate lim f ( x ).

(c) Evaluate lim f ( x ).

(b) Evaluate lim f ( x ).

(c) Evaluate lim f ( x ).

1.5Ĳ Limits at Infinity

Q[2]: Give a polynomial f ( x ) that satisfies lim f ( x ) ‰ lim f ( x ).

Q[4]: Evaluate lim 2x

Q[5]: Evaluate lim 2x

Q[6]: Evaluate lim cos x

Q[7]: Evaluate lim x ´ 3x5 + 100x2 .

Q[4]: Is the following a valid statement?

Q[5]: Is the following a valid statement?

Q[6]: Is the following a valid statement?

Q[7]: Suppose f (t) is continuous at t = 5. True or false: t = 5 is in the domain of f (t).

Q[11]: Find a constant k so that the function

Use the definition of continuity to justify your answer.

2.1Ĳ Revisiting tangent lines

Q[2]: Shown below is the graph y = f ( x ).

(a) (b) (c)

(d) (e) (f)

2.2Ĳ Definition of the derivative

(d) always positive (e) always negative

Q[4](˚): State, in terms of a limit, what it means for f ( x ) = x3 to be differentiable at x = 0.

Q[5]: For which values of x does f 1 ( x ) not exist?

Q[6]: Suppose f ( x ) is a function defined at x = a with

Q[7]: Suppose f ( x ) is a function defined at x = a with