Math 266 Introduction To Calculus II Study Guide

Mathematics 266
Introduction to Calculus II
S TUDY G UIDE
(Revision 8)
Contents
Introduction i
Recommendations for Learning Mathematics . . . . . . . . . . . . . . . . . . . . . . . . i
References in This Study Guide . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . iii
Chapter Review . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . iii
Proofs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . iii
Labs and Quizzes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . iv
Videos . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . iv
Technology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . iv
Unit 1 Inverse Functions: Exponential, Logarithmic and Hyperbolic Functions 1

Objectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
Inverse Functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
Prerequisites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
Exponential Functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
Prerequisites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
Derivative of Exponential Functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
Prerequisites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
Logarithmic Functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
Properties of Exponential and Logarithmic Functions . . . . . . . . . . . . . . . . . . . 25
Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
Derivatives of Logarithmic Functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
Hyperbolic Functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
Hyperbolic Identities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
Finishing This Unit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
Practice Examination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
Unit 2 Inverse Trigonometric and Hyperbolic Functions; L’Hopital’s Rule 43

Objectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
Inverse Trigonometric Functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
Prerequisites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
The Inverse of the Sine Function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
The Inverse of the Cosine Function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
The Inverse of the Tangent Function . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
The Inverse of the Secant Function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61
Inverse of Hyperbolic Functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
The Inverse of the Hyperbolic Sine Function . . . . . . . . . . . . . . . . . . . . . . . . 62
The Inverse of the Hyperbolic Cosine Function . . . . . . . . . . . . . . . . . . . . . . . 64
The Inverse of the Hyperbolic Tangent Function . . . . . . . . . . . . . . . . . . . . . . 67
Representation of Inverse Hyperbolic Functions in Terms of Natural Logarithm . . . . . . 69
Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
Indeterminate Forms and L’Hopital’s Rule . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
Indeterminate Form of Type 0 · ∞ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
Indeterminate Form of Type ∞ − ∞ . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74
Indeterminate Forms of Type 00 , ∞0 and 1∞ . . . . . . . . . . . . . . . . . . . . . . . . 74
Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76
Unit 3 Techniques of Integration 79

Objectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79
Integration by Parts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79
Reduction Formulas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84
Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86
Trigonometric Integrals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87
Prerequisites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87
Other Useful Trigonometric Identities . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94
Trigonometric Substitution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95
Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98
Integration of Rational Functions by Partial Fractions . . . . . . . . . . . . . . . . . . . . . 98
Integration by Rationalizing Substitutions . . . . . . . . . . . . . . . . . . . . . . . . . . 102
Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104
Strategy for Integration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104
Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106
Approximate Integration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107
Prerequisites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107
Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110
Improper Integrals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110
Prerequisites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110
Erroneous Evaluations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115
Comparison Test for Improper Integrals . . . . . . . . . . . . . . . . . . . . . . . . . . . 116
Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118
Unit 4 Applications of Integration 123

Objectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123
Volumes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 124
Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 128
Volumes by Cylindrical Shells . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 128
Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131
Arc Length . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 132
Prerequisites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 132
Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133
Area of a Surface of Revolution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 134
Prerequisites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 134
Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 138
Moments and Centers of Mass . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 138
Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 140
Unit 5 Differential Equations 145

Objectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 145
Modeling with Differential Equations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 145
Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 147
Direction Fields and Euler’s Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 147
Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 148
Separable Equations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 148
Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 151
Integrating Factors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 152
Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 154
Exponential Growth and Decay . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 154
Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 155
Unit 6 Infinite Sequences and Series 157

Objectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 157
Sequences . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 157
Prerequisites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 157
Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 170
Monotonic Sequences . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 170
Prerequisites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 170
Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 178
Important Sequences . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 178
Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 181
Formal Definitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 181
Prerequisites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 181
Sum of Limits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 183
Product of Limits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 184
Limit of the Reciprocal . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 185
Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 191
Series . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 191
Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 202
The Integral Test and Estimates of Sums . . . . . . . . . . . . . . . . . . . . . . . . . . . . 202
Remainder Estimate for the Integral Test . . . . . . . . . . . . . . . . . . . . . . . . . . 205
Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 207
The Comparison Test . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 207
Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 211
Alternating Series . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 211
Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 213
Absolute Convergence and the Ratio and Root Tests . . . . . . . . . . . . . . . . . . . . . 214
Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 220
Representation of Functions as Power Series . . . . . . . . . . . . . . . . . . . . . . . . . 221
Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 227
Taylor and Maclaurin Series . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 228
Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 233
Appendices 237
Appendix A: Mathematical Induction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 237
Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 242
Appendix B: Notation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 244
Appendix C: Answers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 245
Unit 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 245
Unit 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 254
Unit 3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 260
Unit 4 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 268
Unit 5 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 269
Unit 6 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 270
Appendix A . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 275
Appendix D: Practice Examination Solutions . . . . . . . . . . . . . . . . . . . . . . . . . 279
Unit 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 279
Unit 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 282
Unit 3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 287
Unit 4 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 293
Unit 5 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 297
Unit 6 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 300
Introduction
This calculus course is about integration and its applications. In a previous calculus course, you were
introduced to the concept of integration, and learned the u-substitution method. In this course, you will
learn other techniques of integration, study the application of integration in the solution of certain type of
differential equations of first order, and finish with sequences and series. In some sections, we mention,
without references, the prerequisites required to understand the material presented. We recommend that
you have a calculus textbook at hand to consult, and that you review these concepts carefully. (The
Athabasca University Library has several textbooks available.) Note that you must know how to evaluate
limits, differentiate functions and apply the Fundamental Theorem of Calculus.
In this course, the following customized textbook complements this Study Guide:
Lyryx Learning. Calculus Early Transcendentals: An Open Text. Adapted for Athabasca University
Math 266 Calculus II, November 2017 Edition. Based on the original text by D. Guichard. Creative
Commons License (CC BY-NC-SA).
Study this course by following the Study Guide, and do at least the assigned exercises. We encourage you
to try other exercises as well. Answers to most of the exercises from the textbook are given in the section
Selected Exercise Answers of the textbook. Solutions and answers to all the exercises in this Study Guide
are provided in Appendix C. (In Appendix A, we present a brief treatment of mathematical induction that
you will need to do some exercises. Appendix B is devoted to notation. Appendix D contains solutions to
the practice examinations that are given at the end of each unit.)
Recommendations for Learning Mathematics

• Be prepared to practice. Do as many exercises as possible, keep a steady pace and review
background material as you need it.
• Read your Study Guide and textbook carefully. The textbook is not a collection of solved
problems; you are expected to understand the course material and put it to use to solve problems
yourself. Hunting through a section to find a worked-out exercise that is similar to the assigned
problem is totally inappropriate. Solving a mathematical problem requires an understanding of the
concepts discussed.
Read the assigned section of the textbook before you try the corresponding exercises. You should
treat the examples as exercises and try to solve them without the author’s help. Be prepared to check
all the details as you read, and above all, read critically. Take nothing for granted. Never read
mathematics without pencil and paper in hand.
If you do not understand a statement, go back in the section, or to a previous section, to see if you
missed or misunderstood something. If you still fail to understand a concept after puzzling over it for
Mathematics 266: Introduction to Calculus II Study Guide i

a while, mark the place, continue reading, and ask your tutor for help on that point. Remember, too,
that you must learn to make reference notes that are useful to you.
You will be asked to do exercises as concepts arise. Do them at that precise moment, not later; you
must make sure you understand what is presented before going on to the next concept.
Once again, do not limit yourself to the assigned exercises. instead, try to do as many exercises as
you have time for, and do not avoid those that appear to be challenging—the fun in learning is the
satisfaction of being able to respond positively to a challenge. The time you spend reading the Study
Guide and the textbook will save you time in the ling run. You should note any difficulty you have
with the homework, and ask your tutor for help.
• Raise your expectations. We expect a good mastery of the evaluation of techniques for evaluating
limits and for differentiation. However, the discussions of evaluating limits in the Study Guide and
textbook may not present all the details. If you do not fully understand what is presented, review the
material, with the help of your tutor, if necessary. The same applies to differentiation. You are
responsible for ensuring that you get the help you need to master the concepts presented in this
course.
• Learn to evaluate your own work. At the end of each unit, we present a “Practice Examination.”
Pretend that this is a course examination, prepare for it, and write it within the stated time limit. You
are free to write these examinations under different conditions: closed book, open book, consulting
your own notes, etc. If you consult the textbook, the Study Guide, or other materials, take note of the
time you spend doing so; that is, determine whether you can find what you need efficiently (this
strategy will also help you determine whether your notes are useful). Whatever you do, aim to be
efficient with your time. Take note of the concepts you need to consult, and make a point of finding
them quickly – or better still, make sure you know them. Grade yourself based on the solutions
provided in the Appendix D. Put yourself in the place of the marker, and be critical. A marker would
ask the following questions about your examination.
◦ How would you describe the general appearance of the examination?

◦ Are the questions answered in a logical manner?
◦ Are the explanations well presented?
◦ Is the notation correctly used?
◦ Does the student clearly show an understanding of the problem?
Make sure to grade what is actually on the paper—not what you intended to do—because this is what
the marker does. Learn from your mistakes, and question what is being tested. Do not assume that
the course examinations are identical or similar to these practice examinations!
Note: All questions in assignments and examinations are worth several points, because several steps
are required to solve them. Thus you must show all your work and properly justify all your answers.
• Prepare for the examinations, and use good examination strategies when you write them. The
examinations aim to test whether you have met the objectives of the course. The grade assigned
reflects what you know and how well you know it. The onus is on you to show your knowledge and
ii Study Guide Mathematics 266: Introduction to Calculus II

mastery of the subject. You get credit for everything you do know, but a marker will not make
guesses about what you may know.
We strongly recommend that you begin by glancing over all the examination questions, and
answering those you feel most confident about first. Leave to the end those questions that present
special difficulty for you.
If you consider that a question is ambiguous, describe the assumptions you are making to answer it.
You may be familiar with some of problems in the examinations, but be prepared for unfamiliar
questions as well. Do not panic when confronted with an unexpected or unfamiliar problem; read the
question carefully, make sure you understand what is required. Examination questions are not
necessarily difficult; they are designed to test your understanding of concepts, and how well you can
put them to use when solving problems.
In Mathematics it pays to be persistent: people who succeed in mathematics are those who do not give up
after the first attempt at solving a problem, but keep trying and asking themselves, “What else can I try?”
References in This Study Guide

References are found in this Study Guide. References from the textbook, Lyryx Learning. Calculus Early
Transcendentals: An Open Text. Adapted for Athabasca University Math 266 Calculus II, November
2017 Edition. Based on the original text by D. Guichard. Creative Commons License (CC BY-NC-SA),
are so indicated. We have endeavoured to ensure that there is no ambiguity.
Within the Study Guide, examples, theorems, definitions, etc., are numbered; for example, Theorem 3.23.
Note that the first number points to the unit where the item is found. Note as well that within a unit the
numbering of definitions, proportions and theorems is continuous in the Study Guide. So, for example,
Definition 1.2 is followed by Theorem 1.3. The same is true for the examples, figures and exercises; each
follows their own sequencing order.
Definitions marked with an asterisk (*) are not formal definitions, but they do indicate how you should
think about the concept presented. Some formal definitions are provided in the last section of Unit 3.
Chapter Review
Your textbook provides “Additional Exercises” sections. We recommend that you do as many as you can
from these sections.
When you come to the end of a unit, you should feel that you have learned enough to do the extra
practice presented. If you need assistance, consult your tutor.
Proofs
A proof is a logical explanation of why something is true. We present some proofs in this course. AT a
minimum, you should be able to understand the proofs’ arguments, and to prove simple results
independently.
Other formal definitions and proofs can be found in advance calculus textbooks. We recommend:
Mathematics 266: Introduction to Calculus II Study Guide iii

Apostol, Tom A. Mathematical Analysis: A Modern Approach to Advanced Calculus, 2nd ed. New York:
Addison Wesley, 1975.
You may also want to look into the websites below.
Wikipedia: The Free Encyclopedia. Mathematics. Retrieved April 21, 2008, from:
http://en.wikipedia.org/wiki/Mathematics
Wikipedia: The Free Encyclopedia. Mathematical Proof. Retrieved April 21, 2008, from:
http://en.wikipedia.org/wiki/Mathematical_proof
It is fair to say that students are not penalized if they do not master formal definitions and proofs, but are
rewarded if they do.
Labs and Quizzes

The Study Guide will direct you to the Lyryx labs and quizzes at the appropriate times in the course. The
Lyryx exercises open in your Athabasca University course browser. Once you are in the Lyryx system,
locate the lab or quiz for the section of the Study Guide you just completed, then follow the online
instructions.
a
Enter fractions without using a diagonal, like so: .
b
The labs are optional and are intended to provide you with further practice. The quizzes’ grades count
towards your final mark for the course.
When you have completed this course, please click the button that says,“If you have completed all
required course work, use this button to submit your grades to Athabasca University.”
Videos
The videos presented are intended as guided exercises you work out assisted. To benefit from these
exercises, you need to follow them with pencil and paper at hand. Math is learned by doing not by
watching.
Technology
In this course, we do not use technological aids in our work. You are not required to have a particular
calculator, or to use a computer algebra system or software.
In examinations, you are allowed to use a simple calculator, but not a graphing calculator, and not a
programmable calculator.
iv Study Guide Mathematics 266: Introduction to Calculus II

UNIT 1
Inverse Functions: Exponential, Logarithmic and

Hyperbolic Functions
Objectives
When you have completed this unit, you should be able to
1. determine whether a function is invertible.
2. find the inverse of an invertible function.
3. apply the properties of the exponential and logarithmic functions.
4. differentiate exponential functions.
5. differentiate logarithmic functions.
6. differentiate functions using logarithmic differentiation.
7. define hyperbolic functions.
8. differentiate hyperbolic functions.

Z Z 0
f (x)
9. evaluate integrals of the form e f (x) dx and
f (x) 0
dx.
f (x)
Inverse Functions
Prerequisites
To complete this section, you must be able to
• determine the domain and range of a function.

• draw the graphs of basic algebraic functions.
• establish the composition of functions.
• apply the Mean Value Theorem.
• apply the product, multiplication, general power, quotient and chain rules of differentiation.
Mathematics 266: Introduction to Calculus II Study Guide 1

We must study inverse functions in order to be able to integrate functions of the form
Z Z Z Z
dx x 1
x
a dx √ dx and √ dx.
x 4−x 2 4 − x2
Z
dx 1
Observe that to integrate , we must find a function f such that f 0 (x) = .
x x
The function f is not differentiable at x = 0; it is increasing on (0, ∞); and it is decreasing on (−∞, 0).
1
Since f 00 (x) = − , the function is concave down on its domain.
x2
We can see that the graph of f in Figure 1.1, below, satisfies all of these conditions.
1
Figure 1.1: Graph of a function f such that f 0 (x) =
x
As we will find in the next sections, this function is the inverse of an exponential function; therefore, we
must start by finding out what an inverse function is.
To solve the equation
3x − 5 = 1, (1.1)
we could proceed as follows:

6
3x = 1 + 5; 3x = 6; and x= = 2.
3
However, there is another way to solve this equation.
t+5
If we consider the functions f (x) = 3x − 5 and g(t) = (the reason for choosing g is made clear
3
later in this section), then we can see that their composition gives
(3x − 5) + 5
(g ◦ f )(x) = g(f (x)) = = x.
3
Hence, if we apply the function g to Equation (1.1), above, we get
1+5
g(3x − 5) = g(1); g(f (x)) = ; and x = 2.
3
2 Study Guide Mathematics 266: Introduction to Calculus II

Now consider the equation
t+5
= 4. (1.2)
3
We have

t+5
(f ◦ g)(t) = f (g(t)) = 3(g(t)) − 5 = 3 − 5 = t.
3
Then, to solve Equation (1.2), we apply the function f and we get

t+5
f = f (4) t = 3(4) − 5 = 7.
3
We can solve these equations by applying functions because (f ◦ g)(t) = t and (g ◦ f )(x) = x.
In general, if we want to solve an equation of the form f (x) = C, where C is a constant, it is enough to
find a function g(t) such that g(f (x)) = x. The solution, then, is x = g(C).
Before we embark on finding such functions, we must consider the possibility that, for some functions f ,
there is no function g such that g(f (x)) = x.
Let us consider some easy examples.

√
Example 1.1. If x2 = 4, then x = ±2. So, for the function f (x) = x2 , the function g(t) = t seems to
be the correct one, since g(4) = 2.
√
However, g(f (x)) = x2 = |x|, and g(f (x)) 6= x.
√
Example 1.2. If x − 5 = 3, then we add 5 to the equation and square it.
Thus, x = (3 + 5)2 = 82 = 64.

√
For the function f (x) = x − 5, the function g(t) = (t + 5)2 seems to be the correct one. Indeed,
√
g(f (x)) = (f (x) + 5)2 = (( x − 5) + 5)2 = x.
Observe that to obtain g, we take t, add 5, and then square the result.
These examples tell us that only certain functions have the property that allows us to solve equations. We
must define (name) these functions.
Definition 1.1.
a. A function f : D −→ R is left invertible on its domain D if there is a function gl : R −→ D

such that (gl ◦ f )(x) = x for all x in D.
b. A function f : D −→ R is right invertible on its range R if there is a function gr : R −→ D such

that (f ◦ gr )(t) = t for all t in R.
c. A function is invertible if it is both left invertible on its domain and right invertible on its range.

Observe that if f is invertible, (gl ◦ f )(x) = x and (f ◦ gr )(t) = t, then, for t = gr (t), we have
(gl ◦ f )(gr (t)) = gr (t).
Thus,
gl (t) = gl ((f ◦ gr )(t)) = gl (f (gr (t))) = (gl ◦ f )(gr (t)) = gr (t)
and gl = gr .
We then conclude that if f is invertible, there is one function g such that
(f ◦ g)(t) = t and (g ◦ f )(x) = x.
This function g is called the inverse of f , and is denoted by f −1 . Therefore,
(f ◦ f −1 )(t) = t and (f −1 ◦ f )(x) = x.
Remark
Pay attention to this notation: f −1 (x) is the inverse function of the function f ; it is not the reciprocal
1
(f (x))−1 = .
f (x)
If the function f is invertible and f (a) = b, then
a = f −1 (f (a)) = f −1 (b) and f −1 (b) = a.
This statement means that if (a, b) is on the graph of f , then (b, a) is on the graph of f −1 .
With this information, we can find the graph of the function f −1 for any invertible function f .
The points (a, b) and (b, a), in the Figure 1.2, below, are reflections of each other with respect to the line
y = x.
(b,a)
(a,b)
0
y=x
Figure 1.2: Reflection of the point (a, b) with respect to the line y = x

So, to find the graph of f −1 , all we need to do is to reflect the graph of f with respect to the line y = x, as
shown in the Figure 1.3, below,
f -1(x)
f (x)
y=x
Figure 1.3: The function f −1 (x) is the reflection of the function f (x) with respect to the line y = x
The graphs of a function, f , and its inverse, f −1 , are shown in Figure 1.4, below
Figure 1.4: Functions f and g, where g = f −1
Refer back to Example 1.1. If we sketch the graph of the function f (x) = x2 , and then reflect it with
respect to the line y = x, we obtain a parabola x = y 2 . This is not a function.
The points (1, 1) and (−1, 1) are on the graph of f (x) = x2 ; therefore, the points (1, 1) and (1, −1) are
their reflections, and no function can contain these points. Hence, the function f is not invertible.
We conclude that functions that are invertible are functions that do not have points (a, b) and (c, b), where
a 6= c, on their graphs. These functions are called one-to-one or injective.
Definition 1.2.
A function f is injective or one-to-one if, whenever f (a) = f (c), then a = c.
Graphically, a function is one-to-one if no horizontal line intersects its graph at more than one point.
Therefore, a function is one-to-one if
f (a) 6= f (c) whenever a 6= c.

Using this criterion, we can determine that the trigonometric functions are not one-to-one. The function
f (x) = xn for n even is not one-to-one, but it is one-to-one if n is odd. (Why?)
The following theorem indicates that invertible functions are one-to-one.
Theorem 1.3.
Let f : D −→ R be a function with domain D and range R. The inverse of the function f exists if
and only if it is one-to-one, and f −1 : R −→ D.
Proof: If f is invertible, then there is a function f −1 such that
(f ◦ f −1 )(t) = t and (f ◦ f −1 )(x) = x.
If f (a) = f (c) and we apply the function f −1 , we obtain
f −1 (f (a)) = f −1 (f (c)),
and therefore a = c. So f is one-to-one by Definition 1.2 on the preceding page.
Conversely, if f is one-to-one, then for each y in R there is only one x in D such that f (x) = y.
Let us define the function g as g(y) = x; then, g : R −→ D and g(f (x)) = x.
Therefore, g = f −1 , and f is invertible. Q.E.D.1
Intuitively, we can see that if a function is increasing or decreasing on an interval I, then the function is
one-to-one. This intuition is proved in the next theorem.
Theorem 1.4.
If a function f : I −→ R with range R is increasing or decreasing on I, then f is one-to-one, and

hence invertible.
Proof: Let a 6= b in I. We can assume, without loss of generality, that a < b.
By the Mean Value Theorem, there exists an a < c < b such that
f (b) − f (a)
= f 0 (c).
b−a
Since f is increasing or decreasing, f 0 (c) 6= 0, and therefore,
f (b) − f (a) = f 0 (c)(b − a) 6= 0.
1
The end of a proof is indicated by the initials Q.E.D, an abbreviation of the Latin phrase quod erat demonstrandum, “which
was to be proved.”

That is,
f (b) 6= f (a).
Hence, by Definition 1.2 on page 5, the function f is one-to-one. Q.E.D.
Now let us look at the domain and range of invertible functions.
If D and R are the domain and range of the invertible function f

(i.e., f : D −→ R), then f −1 : R −→ D, and
the domain of f = the range of f −1 ;
the range of f = the domain of f −1 .

√
Example 1.3. The function f (x) = x is invertible, and f : [0, ∞) −→ [0, ∞). Therefore, its inverse
function is f −1 : [0, ∞) −→ [0, ∞), and f −1 (t) = t2 because the compositions are
√
(f ◦ f −1 )(t) = t2 = t for t ≥ 0 and
√
(f −1 ◦ f )(x) = ( x)2 = x for x ≥ 0.
The graphs of these functions are shown in Figure 1.5, below.
√
Figure 1.5: Functions f (x) = x and f −1 (t) = t2
Example 1.4. The function f (x) = x2 with domain √ [0, ∞) is invertible (see Figure 1.5). Then
f : [0, ∞) −→ [0, ∞), and its inverse is f −1 (t) = t. [Verify this statement.]
Compare Examples 1.3 and 1.4 with Example 1.1 on page 3, and see how important the domain of an
invertible function is.
Example 1.5. The sine function is not invertible on its domain; so, to make it invertible, we must
restrict
h π πits i domain to some interval on which it is one-to-one. The interval that is universally considered is
− , . The graph of sine on this interval is shown in Figure 1.6, below.
2 2

h π πi
Figure 1.6: Sine function on the interval − ,
2 2
Then
h π πi
sin : − , −→ [−1, 1],
2 2
and its inverse function, written as either sin−1 t or arcsin x is
h π πi
sin−1 : [−1, 1] −→ − , .
2 2
Its graph is shown in Figure 1.7, below.
Figure 1.7: Inverse of the sine function on the interval [−1, 1]
Example 1.6. The domain of the function

3x + 4
f (x) =
2x + 3

3 3
is R − − , and its range is R − , because, if
2 2
3x + 4
=t and we solve for x,
2x + 3
then
4 − 3t 3
x= , and t 6= .
2t − 3 2
Then, the inverse of f is

3 3
f :R−
−1
−→ R − − ,
2 2
and we can verify that

4 − 3t
f −1 (t) = .
2t − 3
We use implicit differentiation and the Chain Rule to find the derivative of the inverse function f −1 of the
function f .
f (f −1 (t)) = t
d d
f (f −1 (t)) = t
dt dt
d
f 0 (f −1 (t)) f −1 (t) = 1
dt
and
d −1 1
f (t) = 0 −1 . (1.3)
dt f (f (t))
d −1
Example 1.7. Let f (x) = x3 + x + 1. We will show you how to obtain the value of f (1) using
dx
the result (1.3). We see that f (0) = 1; hence, f −1 (1) = 0.
d −1 1
f (1) = 0 −1 .
dt f (f (1))
We then have f 0 (x) = 3x2 + 1 and f 0 (f −1 (1)) = f 0 (0) = 1. Thus,
d −1 1 1
f (1) = 0 −1 = 0 = 1.
dt f (f (1)) f (0)
If a function f is increasing (decreasing), then its derivative f 0 is positive (negative). Therefore,
f 0 (f −1 (t)) > 0(< 0) for all t in the domain of f −1 .

Hence
d −1 1
f (t) = 0 −1 > 0(< 0) for all t in the domain of f −1 .
dt f (f (t))
That is, the inverse f −1 is also increasing (decreasing). We state this result in the next proportion.
Proposition 1.5. If f (x) is a continuous and increasing (decreasing) function on an interval I, then
its inverse f −1 (x) is also a continuous and increasing (decreasing) function on I.
The concavity of a function is given by its second derivative. The second derivative of the inverse
function f −1 of the function f is

d2 −1 d 1
f (t) =
dt2 dt f 0 (f −1 (t))

1 00 −1 d −1
= − 0 −1 f (f (t)) f (t)
(f (f (t)))2 dt
00 −1
1 f (f (t))
= − 0 −1
(f (f (t)))2 f 0 (f −1 (t))
f 00 (f −1 (t))
= − 0 −1 .
(f (f (t)))3
Pay attention to this result.
d2 −1 f 00 (f −1 (t))
f (t) = − . (1.4)
dt2 (f 0 (f −1 (t)))3
If a function f is increasing and concave up, then f 0 > 0 and f 00 > 0. By the result (1.4), the second
derivative of its inverse function f −1 is negative. Hence, its inverse function is increasing and concave
down. We state this and similar results below.
If a function is
• increasing and concave up, then its inverse function is increasing and concave down.
• increasing and concave down, then its inverse function is increasing and concave up.
• decreasing and concave up, then its inverse function is decreasing and concave up.
• decreasing and concave down, then its inverse function is decreasing and concave down.
See that in Example 1.5 on page 7 the concavity of sin x and sin−1 x correspond to the results on
concavity, as stated above.
Convince yourself that the reflection of the horizontal line y = a with respect to the line y = x is the
vertical line x = a. Conversely, the reflection of the vertical line x = a with respect to the line y = x is
the horizontal line y = a.

In the next example we learn why this is important.
Example 1.8. Sketch the graph of the function

1
f (x) = for x > 4.
x−4
From the graph of f we obtain the following information
• the domain of f is the interval (4, ∞)

• the range of f is the interval (0, ∞)
• the point (5,1) is on the graph
• the line y = 0 is a horizontal asymptote
• the line x = 4 is a vertical asymptote
• it is decreasing and concave up
Since the graph of the inverse function f −1 (x) is a reflection of f with respect to the line y = x, we have
that
• the domain of f −1 is the interval (0, ∞).

• the range of f −1 is the interval (4, ∞).
• the point (1,5) is on the graph.
• the line x = 0 is a vertical asymptote.
• the line y = 4 is a horizontal asymptote.
• the function f −1 is decreasing and concave up.
With this information we sketch the graph of f −1 (x).
(1,5)
1
Figure 1.8: Graph of the inverse function of f (x) =
x−4

We finish this section with one more example.
Example 1.9. We want to sketch the graph of the inverse function f −1 (x) of the piecewise function
f (x), defined below.

 5x 1

 − if x < −2
 6 3
f (x) =
 3x + 4 if x > −2


 −2 if x = −2
The graph of this function consists of two lines, like so
-8 -2
-2
-7
Figure 1.9: Graph of function of f (x)
To sketch the inverse function we need to find the inverse of both lines. One line possess through the
points (−8, −7) and (−2, −2), so its inverse passes through the points (−7, −8) and (−2, −2). Similarly
the inverse of the other line passes through the points (4, 0) and (−2, −2). The graph of the inverse
function f −1 (x) is shown below.

-2 4
-7
-2
-8
Figure 1.10: Graph of the inverse function f −1 (x)
Summary
1. A function f is invertible if there exists a function f −1 , called the inverse of f , such that the
following cancellation laws hold:
f (f −1 (t)) = t and f −1 (f (x)) = x.
2. The domain of f = the range of f −1 and the range of f = the domain of f −1 .
3. If f is invertible, then its inverse f −1 is also invertible, and its inverse is f ; that is, (f −1 )−1 = f . In
other words, f and f −1 are inverses of one another.
4. If f is continuous on an interval, then its inverse f −1 is also continuous on the same interval.
d −1 1
5. f (t) = 0 −1 .
dt f (f (t))
6. If f (a) = b, then f −1 (b) = a and
d −1 1 1
f (t) = 0 −1 = 0 .
dt t=b f (f (b)) f (a)
d2 −1 f 00 (f −1 (t))
7. f (t) = − .
dt2 (f 0 (f −1 (t)))3

Exercises
1. Read Sections 1.1 Inverse Functions, and 1.2 Derivatives of Inverse Functions of the textbook.
2. a. Draw the graph of the inverse function of the function f (x), whose graph is shown below.
b. Give the domain and range of f (x) and its inverse f −1 (x).
Figure 1.11: Graph of the function f (x)

3. Explain why the inverse of a linear function is also linear.
x2
4. Explain why the function g(x) = x3 + − 2x + 1 is one-to-one on the interval [−1, 2/3] but it is
2
not one-to-one on the interval [0, 2].
d −1
5. Find the value of f (2) if f (x) = x5 − x3 + 2x.
dx
6. Give two closed intervals where the cosine function is one-to-one. Explain.
π π
7. Explain why the tangent function is one-to-one on the interval − , .
2 2
8. Determine whether the functions listed below are one-to-one. Explain.

π 3π
a. f (x) = x + tan(2x) for all x ∈ , .
4 4

 x if x < 1
b. f (x) =
 1 if x ≥ 1
x

 3x
 − + 4 if x < 2
c. f (x) = 4
 1
 if x ≥ 2
x−1
9. For further practice, complete the lab Inverse Functions in the optional Lyryx Labs.

Exponential Functions
Prerequisites
• evaluate limits at infinity.

• apply transformations to sketch functions.
If a is a positive number, then ax is well defined for any number x. Hence, we can define the exponential
function base a as
f (x) = ax .
If a = 1, the exponential function is the constant function 1, and it is considered to be trivial.
The graphs of exponential functions for different bases are as follows.
Figure 1.12: Graph of the exponential function y = ax for 0 < a < 1
For 0 < a < 1, the value of ax is positive and very large as x → −∞ and positive and very small as
x → ∞.
Figure 1.13: Graph of the exponential function y = ax for a > 1
For a > 1, the value of ax is positive and very large as x → ∞ and positive and very small as x → −∞.

1
Figure 1.14: Graph of the exponential function y = ax = 1 for a = 1
For a = 1, the function ax = 1 is the constant function 1.
From these graphs, we can see that the exponential function f (x) = ax is
1. decreasing if a < 1; thus, f 0 (x) < 0 for all x.
2. increasing if a > 1; thus, f 0 (x) > 0 for all x.
3. concave up in any case; thus, f 00 (x) > 0 for all x.
4. in all three cases f (0) = 1.
Furthermore,
5. lim ax = 0 and lim ax = ∞ if a < 1.

x→∞ x→−∞
6. lim ax = 0 and lim ax = ∞ if a > 1.

x→−∞ x→∞
Example 1.10.
x
−x 1 1
lim 2 = lim x = lim =0
x→∞ x→∞ 2 x→∞ 2
and
x
−x 1 1
lim 2 = lim x = lim = ∞.
x→−∞ x→−∞ 2 x→−∞ 2
Example 1.11. To evaluate the limit
3x + 3−x
lim ,
x→∞ 3x − 3−x
we divide the numerator and denominator by 3x .
Thus,
3x + 3−x 1 + 3−2x 1 + 9−x 1+0

lim = lim = lim = = 1.
x→∞ 3x − 3−x x→∞ 1 − 3−2x x→∞ 1 − 9−x 1+0

Example 1.12. For the limit
lim+ (1.1)1/x
x→0
1
if t = , then t → ∞ as x → 0+ . Thus,
x
lim (1.1)t = ∞.
t→∞
x
1
Example 1.13. To sketch the graph of f (x) = 3−x + 1, we sketch g(x) = 3−x = .
3
Figure 1.15: Graph of the exponential function y = 3−x
We sketch the shift of g up by one unit to obtain f .
Figure 1.16: Graph of the exponential function y = 3−x + 1
We sketch its inverse function
Exercises
10. Give the domain of the functions listed below
1
a. √
1 − 4t

2
Figure 1.17: Graph of the inverse function of the exponential function y = 3−x + 1
b. cos(e−t )
11. Evaluate the limits listed below

3ex − e2x
a. lim
x→∞ 3 − e3x
1
b. lim− e 2−x
x→2
c. lim 10−x
x→∞

Derivative of Exponential Functions
Prerequisites
• find f 0 (a) using the limit definition of the derivative function.

• differentiate the composition of functions using the Chain Rule.
In Section 1.5 of the textbook, Derivatives of Exponential and Logarithmic Functions, you can see why
the derivative of the exponential function f (x) = ax is
ah − 1 ah − a0
f 0 (x) = ax lim = ax lim
h→0 h h→0 h
We notice that the limit is the value of the derivative of f (x) = ax at 0, that is,
f 0 (x) = f 0 (0)ax .
Since the number f 0 (0) is the slope of the line tangent to the curve f (x) = ax at (0, 1), we write
ma = f 0 (0), and we see from the graphs of the exponential functions that ma > 0 if a > 1 and ma < 0 if
a < 1. Hence, the derivative of the exponential function is almost equal to itself, except for the number
ma . Thus,
d x
a = ma ax . (1.5)
dx
Ideally, then, we would like to find a number e such that the slope of the tangent line at (0, 1) of the
exponential function f (x) = ex is 1—that is me = 1—so that
d x
e = ex . (1.6)
dx
Since 1 is positive, if the number e exists, it must be greater than 1. By the definition of the derivative,
f 0 (0) is the limit
f (0 + h) − f (0) eh − 1
f 0 (0) = lim = lim = 1.
h→0 h x→0 h
Hence,
eh − 1
≈1 and e ≈ (h + 1)1/h for small h.
h
From the table below, we can see how we can get an approximate value for e.

h→0 (h + 1)1/h
−0.01 2.73199
−0.001 2.71964
−0.0001 2.71841
0.0001 2.71814
0.001 2.71692
0.01 2.70481
We find that e ≈ 2.718, and more importantly, the number e is irrational and
e = lim (h + 1)1/h .
h→0
Since e > 1, the exponential function base e is increasing, and
lim ex = ∞ and lim ex = 0.

x→∞ x→−∞
The composition of the exponential function f (x) = ax with a function g(t) is f (g(t)) = ag(t) , and by the
Chain Rule and the derivative of f , as indicated in Equation (1.5) on page 19, we have
d g(t)
a = g 0 (t)ma ag(t) . (1.7)
dt
In the particular case of the exponential function of base e, we have, by Equation (1.6),
d g(t)
e = g 0 (t)eg(t) . (1.8)
dt
Example 1.14. Applying Equations (1.7) and (1.8) and the rules of differentiation, we have
d −x
a. e = −e−x
dx
d −2x
b. a = −2ma a−2x
dx
√
d √ √ ma a x
c. cos(a x ) = − sin(a x ) √
dx 2 x
√
d √ d 1/x m2 21/x m2 x 2
d.
x
2= 2 =− =−
dx dx x2 x2
2
d ex −1 d x2 −1
e. √ = 2 −1/x
e
dx x 2 dx
2
m2 2−1/x ex −1 2
= 2
+ 2−1/x (2x)ex −1
x
x2 −1
e (m2 + 2x3 )
= √
x2 x 2

From Equations (1.5) and (1.6) on page 19, we have the following antiderivatives:
Z
ax
ax dx = +C (1.9)
ma
and
Z
ex dx = ex + C (1.10)
and from Equations (1.7) and (1.8), above, we have

Z
af (x)
f 0 (x)af (x) dx = +C (1.11)
ma
and
Z
f 0 (x)ef (x) dx = ef (x) + C. (1.12)
Example 1.15. In the examples below, we identify the function f , and then multiply the integrand
function by a constant so that it fits one of the forms of the integrals (1.11) and (1.12) above.
Z Z
1 53x
a. 3x
5 dx = (3)53x dx = + C here f (x) = 3x
3 3m3
Z
b. (sec2 x)etan x dx = etan x + C here f (x) = tan x
Z √ Z
6 x 1 √
c. √ dx = 2 √ (6 x
) dx
x 2 x
√
(2)6 x √
= +C here f (x) = x
m6
Z √ Z
x
e 1
d. 2
dx = − − (e1/x ) dx
x x2
√ 1
= −e1/x + C = − x e + C here f (x) =
x
Summary
1. An exponential function base a > 0 is a function f (x) = ax .
2. An exponential function is continuous for all a, decreasing if a < 1 and increasing if a > 1, and
concave up for all a.
3. lim ax = 0 and lim ax = ∞ if a < 1.

x→∞ x→−∞

4. lim ax = 0 and lim ax = ∞ if a > 1.
x→−∞ x→∞
d x d x
5. a = ma ax where ma = a .
dx dx x=0
6. The number e is equal to e = lim (h + 1)1/h .

h→0
d x
7. me = 1, and e = ex .
dx
8. For any function u, the Chain Rule gives

d u du u d u du u
a = ma (a ) and e = (e ).
dx dx dx dx
Z Z
af (x)
9. 0
f (x)a f (x)
dx = +C and f 0 (x)ef (x) dx = ef (x) + C.
ma
Exercises
12. Read Section 1.3 of the textbook, Exponential Functions.
13. Do exercises 1.2.1 - 1.3.4 of Section 1.3 of the text, Exponential Functions.
14. Differentiate the functions listed below

d 3/x
a. e
dx
d ex
b.
dx 1 − e3x
d
c. cos(3ex − 4x )
dx
d √ 2x
d. xe
dx
d
e. exp(sin(3 − 2x))
dx
15. Find the absolute maximum and minimum value of the function f (x) = x2 ex in the interval [−3, 0].
16. Evaluate the integrals listed below

Z
2
a. xe−2x dx
Z
b. 4−3x dx
Z
c. ex cos(ex ) dx

Z √
1 − e−t
d. dt
et
Z
2x
e. dx
(3 − 2x )2
17. For further practice, complete the lab Exponential Functions in the optional Lyryx Labs.
Logarithmic Functions
ax
From the graphs, we can see that the exponential functions are one-to-one. Thus, the exponential function
of base a is invertible. logathe logarithmic function of base a.
Its inverse is called
1
x
Definition 1.6. 1
The logarithmic function of base a is the inverse function of the exponential function of base a.
This function is denoted by loga x.
The graph of loga x is the reflection with respect to the line y = x of the exponential function ax .
ax
loga x
Figure 1.18: Graph of ax and loga x with a < 1
The graph of loga x with a < 1 is shown in Figure 1.18 above, and the graph of loga x with a > 1 is
shown in Figure 1.19 below.
From these graphs, we see that
1. the domain of loga x is (0, ∞) for any a > 0.
2. loga 1 = 0.
3. loga x is increasing and concave up if a > 1.
4. loga x is decreasing and concave up for a < 1.
5. if a > 1, then lim loga x = ∞ and lim+ loga x = −∞.

x→∞ x→0

ax
loga
1
x
Figure 1.19: Graph of ax and loga x with a > 1
ax
6. if a < 1, then lim loga x = −∞ and lim+ loga x = ∞.
x→∞ x→0
Example 1.16. The domain of the function
f (x) = log10 (x2 − 4)

loga x
consists of all x such that x4 − 4 > 0.
Solving, we find |x| > 2, which gives the set (−∞, 2) ∪ (2, ∞).
Remark
To evaluate limits with logarithmic functions, always refer to their graphs.

2
x
lim log5 ,
x→∞ x−2
x2
we consider t = , and for x → ∞, we have t → ∞. Hence,
x−2
lim log5 (t) = ∞.
t→∞

x
lim log5 ,
x→∞ x2 − 2
x
we consider t = , and for x → ∞, we see that t → 0 and t > 0.
x2 − 2
Hence, t → 0+ , and we have
lim log5 (t) = −∞.

t→0+

Since ax and loga x are inverses of one another, their cancellation laws are as follows.
Cancellation Laws
aloga x = x for x > 0

loga (ax ) = x for any x.
Example 1.19.
a. log4 (45 ) = 5
b. 4log4 6 = 6
c. log3 (27) = log3 (33 ) = 3
Observe that loga x is the number to which we must raise the number a in order to obtain x.
Example 1.20.
a. log2 8 = 3, since 23 = 8.
−3
1
b. log1/2 8 = −3, since = 8.
2
Remark
In this course, we consider logarithmic functions with bases greater than 1 because they are the most
important logarithmic functions; however, you should not forget about logarithmic functions with base a
less than 1!
Properties of Exponential and Logarithmic Functions

Since the exponential function ax and the logarithmic function loga x are inverses of one another, their
properties are “dual” of each other. That is,
1. ax+y = ax ay and loga (xy) = loga x + loga y

ax x
2. a x−y
= y and loga = loga x − loga y
a y
3. (ax )y = axy and loga xr = r loga x
Observe how we apply the cancellation laws to justify these properties.
1. aloga x+loga y = aloga x aloga y = x y; thus, by the cancellation laws,
loga xy = loga (aloga x+loga y ) = loga x + loga y.

aloga x x
2. aloga x−loga y = = ; thus, by the cancellation laws,
aloga y y

x
loga = loga (aloga x−loga y ) = loga x − loga y.
y
3. ar loga x = (aloga x )r = ar ; thus, by the cancellation laws,
loga (ar ) = loga (ar loga x ) = r loga x.
As we discuss later, the properties of the logarithmic functions play an important role in differentiation.
Example 1.21.
lim log4 (3x3 − 2x) − log4 (4 + x3 )

x→−∞
3
3x − 2x
= lim log4 = log4 (3)
x→−∞ 4 + x3
Example 1.22. The solution of the equation
log2 (2x + 3) = log2 (2 − x)
follows from the properties of the logarithmic functions. Thus
log2 (2x + 3) − log2 (2 − x) = 0
and

2x + 3
log2 = 0.
2−x
Note that log2 x = 0 if and only if x = 1. (Why?) Then,

2x + 3
= 1,
2−x
1
and solving, we find x = − .
3
For the particular case of the number e, the logarithmic function of base e is called the natural logarithm
(function), written ln x. That is,
loge x = ln x.
Since e > 1, we have ln x > 0 for x > 1 and ln x < 0 for 0 < x < 1.
For a > 1, the functions ax and loga x are increasing; therefore,
ax < ay for any x<y

and
loga x < loga y for any 0 < x < y.
Example 1.23. To solve 52−x > 16, we apply log5 to this inequality, obtaining
log5 (52−x ) > log5 (16)
and
2 − x > log5 (16).
Therefore,
x < 2 − log5 (16).
Observe that aloga x = x; hence, ln(aloga x ) = ln x, and by the properties of the logarithms
(loga x)(ln a) = ln a, and we conclude that
ln x
loga x = .
ln a
In conclusion, this says that any logarithmic function can be written in terms of the natural logarithmic
function; thus, we need only learn the natural logarithmic function to know any other logarithmic
function.
Exercises
18. Write a summary of logarithmic functions.
19. Explain why ex = ey if and only if x = y.
20. Explain why loga x = loga y if and only if x = y.
21. Explain why the function ln(sin x) is not defined on the interval (−π, 0), but is defined on the
interval (0, π).
22. Read examples 1.10–1.13 of Section 1.4 of the textbook, Logarithms.
23. For further practice, complete the lab Logarithmic Functions in the optional Lyryx Labs.
24. Give the domain of the function ln(tan x).
25. Find the limits listed below
a. lim ln(tan x)
x→0+
b. lim ln(3x2 − 2) − ln(2x + 5x2 )

x→∞
c. lim ln(x2 + 2x − 1)
x→∞

d. lim ln(cos x)
x→0
Derivatives of Logarithmic Functions

Observe that
d d ln x 1 d
loga x = = ln x.
dx dx ln a ln a dx
Thus, it is enough to find the derivative of ln x to obtain the derivative of any logarithmic function.
We have eln x = x; thus,

d ln x d
e = x.
dx dx
By the Chain Rule,
d d
1 = eln x ln x = (x) ln x,
dx dx
and we conclude that
d 1
ln x = for any x > 0.
dx x
1
We almost have the antiderivative we are looking for—namely a function whose derivative is —except
x
that the derivative of ln x exists only for x > 0.
The next example gives the derivative of loga x in terms of the natural logarithm.
Example 1.24. For x > 0,

d 1 d 1 1 1
loga x = ln x = = .
dx ln a dx ln a x x ln a


d 1 sin(ln x)
cos(ln x) = − sin(ln x) =− .
dx x x
d 3x3 − 4x
(3x3 − 4x) log10 x = (9x2 − 4) log10 x + .
dx x ln 10

The composition of ln x with a function g is ln(g(x)), and by the Chain Rule, we have
d 1 g 0 (x)
ln(g(x)) = (g 0 (x)) = . (1.13)
dx g(x) g(x)
In Leibnitz notation, this expression is

d 1 du
ln(u) = .
dx u dx
Example 1.27.
d −1 1
1. ln(−x) = = for x < 0.
dx −x x
d c
2. ln(cx) = for which c and x?
dx x
d d r
3. ln(xr ) = r ln x = for which r and x?
dx dx x
d d 2 − sin x 2
4. ln(x2 cos x) = ln(x2 ) + ln(cos x) = + = − tan x.
dx dx x cos x x
d p g 0 (x)
5. ln( g(x)) = .
dx 2(g(x))
The domain of ln x is (0, ∞). To extend this domain to all nonzero x, we can consider the function ln |x|:
the domain of this function is R − {0}, and

 ln x if x > 0
ln |x| =
 ln(−x) if x < 0
Thus by part (a) of Example 1.27, above, its derivative is


 1

 if x > 0
d x 1
ln |x| = = for all nonzero x.
dx 
 1 x
 if x < 0
x
1
This is the antiderivative we need, because the derivative of ln |x| is for any nonzero x. Hence,
x
Z
1
dx = ln |x| + C.
x
Again, the composition of ln |x| and a function g is ln |g(x)|, and its derivative is
d g 0 (x)
ln |g(x)| = for any x such that g(x) 6= 0.
dx g(x)

The corresponding antiderivative is
Z 0
g (x)
dx = ln |g(x)| + C. (1.14)
g(x)
This antiderivative is very useful, and you should note that the form of the integrand function is the key to
proper differentiation.
Example 1.28. Again, we multiply by a constant so that the integrand function fits the form of the
integral in Equation (1.14), above.
Z Z
1 1 3 ln |3x + 5|
dx = dx = + C.
3x + 5 3 3x + 5 3
Example 1.29.
Z Z
cos x
cot x dx = dx = ln | sin x| + C.
sin x
Since ln | sin x| = − ln |(sin x)−1 )| = − ln | csc x|, we also have

Z
cot x dx = − ln | csc x| + C.
We prefer to keep this antiderivative consistent with the derivatives of cot x and csc x.
Remark
Whenever we integrate a quotient of functions

Z
g(x)
dx,
f (x)
the first thing we do is see if
f 0 (x) = g(x) or f 0 (x) = kg(x) for some constant k.
If this is the case, then

Z Z 0 Z Z
g(x) f (x) g(x) 1 f 0 (x)
dx = dx or dx = dx
f (x) f (x) f (x) k f (x)
and the antiderivative is either

Z 0 Z
f (x) 1 f 0 (x) ln |f (x)|
dx = ln |f (x)| or dx = .
f (x) k f (x) k
Example 1.30. Observe how we manipulate the secant function to obtain its antiderivative.

sec x + tan x sec2 x + sec x tan x
sec x = sec x =
sec x + tan x tan x + sec x

d
tan x + sec x = sec2 x + sec x tan x
dx
Thus,
Z Z
sec2 x + sec x tan x
sec x dx = dx = ln | tan x + sec x| + C.
tan x + sec x
The antiderivative of the cosecant function is obtained in a similar manner.
Example 1.31.
1
Z Z
1 x
dx = dx = ln | ln x| + C.
x ln x ln x
Logarithmic differentiation is a method that uses the properties of the logarithmic function to simplify
calculations. This method is particularly useful when the application of the rules of differentiation yield
cumbersome and difficult operations. It is also the method we must use when the rules of differentiation
do not apply.
Example 1.32. The application of the rules of differentiation to the function
5x3 + 5x2 − x + 1
f (x) = p
cos(x2 ) − 4x4
is difficult and inefficient. However, if we set
5x3 + 5x2 − x + 1
y= p
cos(x2 ) − 4x4
and apply the natural logarithmic function, the properties of the natural logarithmic function change our
original function into a sum of functions. Thus,
!
5x3 + 5x2 − x + 1
ln y = ln p
cos(x2 ) − 4x4
1
= ln(5x3 + 5x2 − x + 1) − ln(cos(x2 ) − 4x4 ).
2
The next step is to differentiate:
y0 15x2 + 10x − 1 2x sin(x2 ) − 16x3

= 3 −
y 5x + 5x2 − x + 1 2(cos(x2 ) − 4x4 )
15x2 + 10x − 1 x sin(x2 ) + 8x3
= 3 − .
5x + 5x2 − x + 1 cos(x2 ) − 4x4

We solve for y 0 and obtain
y 0 = f 0 (x)

5x3 + 5x2 − x + 1 15x2 + 10x − 1 x sin(x2 ) + 8x3
=− p + .
cos(x2 ) − 4x4 5x3 + 5x2 − x + 1 cos(x2 ) − 4x4
Example 1.33. We cannot apply any rule of differentiation to obtain g 0 (x), where g(x) = (2x2 − 1)x ,
so we must use logarithmic differentiation. First we set
y = (2x2 − 1)x .
Applying the natural logarithm, we obtain
ln y = x ln(2x2 − 1).
Differentiating gives us

y0 4x
= ln(2x2 − 1) + x 2
,
y 2x − 1
and we solve for y 0 , as follows:

4x2
0 0 2
y = g (x) = (2x − 1) ln(2x2 − 1) +
x
.
2x2 − 1
Observe that whenever we use logarithmic differentiation, the resulting derivative function is of the form
f 0 (x) = f (x)g(x) for some function g(x).
Any logarithmic function can be written in terms of the natural logarithmic function, just as any
exponential function can be written in terms of the exponential function ex . Indeed,
x
ax = eln(a ) = ex ln a .
Moreover,
d x d x ln a
a = e = (ln a) ex ln a = (ln a)ax .
dx dx
On the other hand, we have, by Equation (1.5) (p. 19 of this Guide) that
d x
a = ma (ax ),
dx
where ma < 1, if a < 1, and ma > 0 if a > 1.
Thus, we conclude that ma = ln a and

d x
a = (ln a) ax .
dx

Example 1.34. We have
d 3x
4 = 3(ln 4)43x
dx
and
Z
x 5x
5 dx = + C.
ln 5
Exercises
26. Write a summary of this section.
27. Read examples 1.14–1.16 of Section 1.5 of the textbook, Derivatives of Exponential and Logarithmic
Functions.
28. Do as many exercises as you can from the section Exercises for 1.5 of the textbook.
29. Read examples 1.17 and 1.18 of Section 1.6 of the textbook, Logarithmic Differentiation.
30. Do exercises 1.6.1–1.6.3 of the section Exercises for Section 1.6 of the textbook.
31. For further practice, complete the lab Derivatives of Logarithmic Functions in the optional Lyryx
Labs.
32. Use logarithmic differentiation to find the derivative of the functions listed below.
√ 2
a. g(x) = x − 1ex (3x + 2)10
b. h(x) = xtan x
x2 − 2x + 1
c. g(x) = √
5 − sin x
d. u(x) = (sin x)x
33. Evaluate the integrals listed below.

Z 6
2 − u2
a. du
1 2u3
Z 1
cos(2x)
b. dx
0 4 − sin(2x)
Z
3ex
c. dx
ex + 1
Z
d. 8x − sec2 (3x) dx
Z
(ln(2x))6
e. dx
x

Z
f. tan x dx
Z
g. cot x dx
Z
h. csc x dx
34. Complete the online Lyryx Assessment quiz titled Unit 1 - I. Your grade on this quiz will count
Hyperbolic Functions
Hyperbolic functions are given in terms of the exponential function ex , and their properties are similar to
those of the trigonometric functions. We use their inverse functions to integrate functions of the form
1 1
√ and .
x2 ± 1 1 − x2
Definition 1.7.
ex − e−x ex + e−x
sinh x = cosh x =
2 2
sinh x ex − e−x cosh x ex + e−x

tanh x = = x coth x = = x
cosh x e + e−x sinh x e − e−x
1 2 1 2
csch x = = x sech x = = x
sinh x e − e−x cosh x e + e−x
The graphs of sinh x and cosh x are shown in Figure 1.4 of Section 1.7 of the textbook, Hyperbolic
Functions. The graph of tanh x is shown below.
From these graphs, we derive the following properties.
Properties of Hyperbolic Sine
1. The domain and range are R.
2. The function is increasing on its domain.
3. sinh 0 = 0.
Properties of Hyperbolic Cosine
1. The domain is R; the range is [1, ∞).

1
-1
Figure 1.20: Graph of tanh x
2. The function is decreasing on (−∞, 0), and increasing on (0, ∞).
3. cosh 0 = 1.
Properties of Hyperbolic Tangent
1. The domain is R; the range is (−1, 1).
2. The function is increasing on its domain.
3. tanh 0 = 0.
4. lim tanh x = 1 and lim tanh x = −1.

x→∞ x→−∞
You can verify that

d d
sinh x = cosh x and cosh x = sinh x.
dx dx
Hence,
d d sinh x cosh2 x − sinh2 x 1

tanh x = = 2 = = sech 2 x.
dx dx cosh x cosh x cosh2 x
In a similar manner, you can obtain the derivatives of the other hyperbolic functions.
Example 1.35. By the General Power Rule

d
sinh3 x = 3 sinh2 x cosh x.
dx
Example 1.36. By the Chain Rule
d sech 2 (ln x)
tanh(ln x) = .
dx x

Hyperbolic Identities
Some hyperbolic identities are listed in the table Useful Hyperbolic Function Properties of Section 1.7 of
the textbook, Hyperbolic Functions. Notice that they are similar to the trigonometric identities. Identities
1, 2 and 5 of this list are proved in Example 1.20 of this same section. In the next example we prove
identities 3 and 4. The rest are left as exercises.
Example 1.37. We have cosh2 x − sinh2 x = 1; thus,
cosh2 x 1 cosh2 x − 1 sinh2 x

coth2 x − csch 2 x = − = = = 1.
sinh2 x sinh2 x sinh2 x sinh2 x
By Definition 1.7 on page 34, we have
(ex + e−x )2 (ex − e−x )2 2e2x + 2e−2x

cosh2 x + sinh2 x = =
4 4
e2x + e−2x
= = cosh(2x)
2

In addition we have other identities listed in the next theorem.
Theorem 1.8. a. coth2 x − 1 = csch2 x
b. 1 − tanh2 x = sech2 x
c. sinh(x + y) = sinh x cosh y + cosh x sinh y
d. cosh(x + y) = sinh x cosh y + cosh x sinh y

tanh x + tanh y
e. tanh(x + y) =
1 + tanh x tanh y
1 + tanh x
f. = e2x
1 − tanh x
Proof: Identities in parts c. and d. are left as exercises. We have the identity cosh2 x − sinh2 x = 1.
1. If we divide it by sinh2 x, we have
cosh2 x − sinh2 x 1
cosh2 x − sinh2 x = 1 ⇒ 2 =
sinh x sinh2 x
This gives
cosh2 x
− 1 = csch2 x ⇒ cot2 x − 1 = csch2 x.
sinh2 x

2. Similarly, if we divide the identity by cosh2 x, we have
cosh2 x − sinh2 x 1
cosh2 x − sinh2 x = 1 ⇒ 2 =
cosh x cosh2 x
This gives
sinh2 x
1− 2 = 1 − tanh2 x = sech2 x ⇒ 1 − tanh2 x = sech2 x.
cosh x
3. We use the identities of parts c. and d.
sinh(x + y) sinh x cosh y + cosh x sinh y

tanh(x + y) = =
cosh(x + y) cosh x cosh y + sinh x sinh y

sinh x cosh y + cosh x sinh y cosh x cosh y
=
cosh x cosh y cosh x cosh y + sinh x sinh y
!
1
= (tanh x + tanh y) cosh x cosh y+sinh x sinh y
cosh x cosh y
tanh x + tanh y
=
1 + tanh x tanh y
4. By definition of hyperbolic tangent, we have
ex − e−x ex + e−x + ex − e−x 2ex

1 + tanh x = 1 + = =
ex + e−x ex + e−x ex + e−x
and
ex − e−x ex + e−x − ex + e−x 2e−x
1 − tanh x = 1 − x = = x
e + e−x ex + e−x e + e−x
Hence,
1 + tanh x 2ex
= −x = e2x .
1 − tanh x 2e
Q.E.D.
In the next theorem we use mathematical induction to prove one more identity.
Theorem 1.9.
For any positive integer n, we have
(cosh x + sinh x)n = cosh(nx) + sinh(nx).

Proof: For n = 1, the identity follows and there is nothing to proof. We assume the identity to be true for
n.
(cosh x + sinh x)n+1 = (cosh x + sinh x)n (cosh x + sinh x)

= (cosh(nx) + sinh(nx))((cosh x + sinh x))
= cosh(nx) cosh x + sinh(nx) sinh x+
sinh(nx) cosh x + cosh(nx) sinh x
= cosh(nx + x) + sinh(nx + x)
= cosh((n + 1)x) + sinh((n + 1)x)
The statement is true for n + 1, and therefore for all n. Q.E.D.
Exercises
35. Sketch the graph of csch x.
1
Hint. Use the graph of sinh x and the fact that csch x = .
sinh x
36. Using the graph of sech x, give the properties of the hyperbolic cosecant function.
37. Sketch the graph of cosh x.

1
Hint. Use the graph of cosh x and the fact that sech x = .
cosh x
38. Using the graph of sech x, give the properties of the hyperbolic secant function.
39. Sketch the graph of coth x.

1
Hint. Use the graph of tanh x and the fact that coth x = .
tanh x
40. Using the graph of coth x, give the properties of the hyperbolic cotangent function.
41. Establish the derivatives listed below

d
a. csch x = −csch x coth x.
dx
d
b. sech x = −sech x tanh x.
dx
d
c. coth x = −csch 2 x.
dx
42. Read Example 1.20 of Section 1.7 of the textbook, Hyperbolic Functions.
43. Study the table Useful Hyperbolic Functions Properties, and read example 1.21 of Section 1.7 of the
textbook, Hyperbolic Functions.
44. Prove the hyperbolic identities listed below. Note that these identities resemble the double angle
formulas of the trigonometric functions.
cosh(2x) + 1
a. cosh2 x =
2
cosh(2x) − 1
b. sinh2 x =
2
c. sinh(x + y) = sinh x cosh y + cosh x sinh y
d. cosh(x + y) = cosh x cosh y + sinh x sinh y
45. Integrate
Z
a. sinh2 x dx
Z
b. cosh2 x dx
46. For further practice, complete the lab Hyperbolic Functions in the optional Lyryx Labs.
Finishing This Unit

1. Review the objectives of this unit and make certain you are able to meet all of them.
2. If there is a concept, definition, example or exercise that it is not yet clear to you, go back and re-read
it. Contact your tutor if you need help.
3. Make a summary of each section of this unit, as if you were going to use your notes in an
examination.
4. Work out the exercises in the two videos Differentiation of Functions, Differentiation 1 and
Differentiation 2.
5. Complete the Practice Examination provided for this unit. Evaluate yourself, checking your answers
against those provided in Appendix D of this Study Guide. Remember: The number of points in a
question may indicate the number of steps in the solution.
6. Complete the online Lyryx Assessment quiz titled Unit 1 – II. Your grade on this quiz will count

Practice Examination
Time: 3 hours
Total points: 80
Passing grade: 55%
1. (4 points) Sketch the graph of a function that is invertible on the interval (−∞, 0) and not invertible
on the interval (0, ∞).
x3
2. (4 points) Find two disjoint intervals where the function f (x) = − 2x − 3x2 + 1 is invertible.
3
6x + 4 3t − 4
3. (4 points) Determine if the function g(x) = is the inverse of the function f (t) = .
3+x 6−t
4. (4 points) Find the intervals where the function h(x) = x2 e4x−2 is increasing, and the intervals
where it is decreasing.
5. (16 points) Find each of the derivatives indicated below.

d 2
a. ln(x2 − 4) + e2−cos(x )
dx
d2 p
b. ln(3 − x2 )
dx2
d
c. sinh(3x − 1) tanh(ex )
dx
6. (4 points) Give the domain of the function ln(sinh x).
7. (12 points) Determine whether the functions listed below are even, odd or neither. Explain.
a. g(x) = ln(cosh x).
b. h(x) = cos(sinh x).
c. f (x) = etanh x .
8. (4 points) Find the equation of the line tangent to the function h(x) = tanh(4 − x2 ) at the point
(2, h(2)).
9. (12 points) Evaluate each of the limits listed below. Justify your answer. If a limit does not exits
explain why.
a. lim etanh x
x→−∞
b. lim ln(cosh x)
x→0
c. lim tanh(ln x)
x→0+
10. (16 points) Integrate.

Z
2 − ex
a. dx
ex − 2x
Z
b. cosh(3 − 5x) dx
Z
c. tanhx dx
Z
dx
d.
cosh x − sinh x

UNIT 2
Inverse Trigonometric and Hyperbolic Functions;

L’Hopital’s Rule
Objectives
1. sketch the graph of inverse trigonometric functions.
2. sketch the graphs of inverse hyperbolic functions.
3. differentiate inverse trigonometric functions.
4. differentiate inverse hyperbolic functions.

Z
dx
5. do integrals of the form √ .
a ± x2
2
Z
dx
6. do integrals of the form .
a2 + x 2
Z
dx
7. do integrals of the form √ .
x ± a2
2
Inverse Trigonometric Functions

Prerequisites
• sketch the graphs of trigonometric functions.

• apply trigonometric identities.
• differentiate all trigonometric functions.
• integrate by u-substitution.

The inverse of the exponential functions give us the antiderivative of the function
1
g(x) = .
x
The inverse of the trigonometric functions will give us the antiderivatives of the functions
1 1
f (x) = √ and f (x) = .
a2 ± x2 a2 + x2
As we indicated earlier, the trigonometric functions are not one-to-one; hence, they are not invertible.
However, as you now know, we can limit them to a portion of their domain where they are one-to-one,
and then proceed to find their inverse function.
The Inverse of the Sine Function

Notation
The inverse of the sine function is denoted by
sin−1 x or arcsin x.
Graph
To make the sine function one-to-one, we restrict it to the closed interval

h π πi
− , .
2 2
The graphs, on this interval, of sine and of its inverse are shown in Figure 2.1, below.
h π πi
Figure 2.1: Graphs of sin x on the interval − , and its inverse sin−1 x (not to scale)
2 2
Properties
From the graph of sin−1 x in Figure 2.1, above, we have the following properties:
1. the domain is [−1, 1].

h π πi
2. the range is − , .
2 2
3. sin−1 (0) = 0.
π
4. sin (−1) = − .
−1
2
π
5. sin−1 (1) = .
2
6. the function is increasing on its domain.
7. the function is concave down on [−1, 0].
8. the function is concave up on [0, 1].
To determine the domain of a composition of functions of the form sin−1 (g(x)), we must consider all x
such that −1 ≤ g(x) ≤ 1. Study the next two examples.
Example 2.1. To find the domain of the function sin−1 (3ex ), we have to find all x so that
1
−1 ≤ 3ex ≤ 1 ⇒ 0 < ex ≤
3
Remember that exponential functions are positive.
When we apply ln x to this inequality, the inequality does not change because ln x is increasing. Thus, we
obtain x ≤ − ln(3). The domain is the interval (−∞, − ln(3)].
Example 2.2. Sketch the graph of the function p(x) = −x2 + 5x − 4 to see that on the interval [1, 4], it
is bounded by 0 and 9/4. That is,
9
0 ≤ −x2 + 5x − 4 ≤ .
4
When we apply sin−1 x to this inequality, the inequality does not change, because sin−1 x is increasing.
However, 9/4 > 1 and and we cannot apply sin−1 x to this value. But, we can find subintervals I of [1, 4]
where p(x) ≤ 1 for all x in I.
√
5 ± 5
−x2 + 5x − 4 = 1 ⇒ x =
2
Again, we refer you to the graph of p(x) to see that
" √ # " √ #
5− 5 5+ 5
0 ≤ −x2 + 5x − 4 ≤ 1 on the intervals 1, , ,4 .
2 2
Thus, for any x in these intervals

π
0 = sin−1 (0) ≤ sin−1 (−x2 + 5x − 4) ≤ sin−1 (1) = .
2

We found the interval 0, π2 where the function sin−1 (−x2 + 5x − 4) is bounded below by 0 and above
by π/2. This interval is the range of the function.
Cancellation Laws
sin(sin−1 x) = x if and only if −1 ≤ x ≤ 1

π π
sin−1 (sin x) = x if and only if − ≤x≤
2 2
In the next example, we show you how to apply the cancellation laws.
Example 2.3. To apply a cancellation law to evaluate

−1 3π
sin sin ,
4
we must exchange
h π πi
3π
sin for sin(x), where x is in − , .
4 2 2
π
3π
Since sin = sin , we have
4 4
π π
−1 3π
sin sin = sin−1 sin = .
4 4 4
Right Triangle
We can associate a right triangle to sin−1 x. Consider that

h π πi
θ = sin−1 x is an angle in the interval − , ,
2 2
then
x
sin(θ) = sin(sin−1 x) = x = ,
1
and the corresponding
√ triangle has opposite side x and hypotenuse 1. By the Pythagorean Theorem, the
adjacent side is 1 − x2 (see Figure2.2, below).
Derivative
By the cancellation law,
sin(sin−1 x) = x.

Figure 2.2: Right triangle with angle θ = sin−1 x
Using implicit differentiation and applying the Chain Rule, we get

d
cos(sin−1 x) sin−1 x = 1.
dx
Solving, we get
d 1
sin−1 x = .
dx cos(sin−1 x)
Since θ = sin−1 x is in [−π/2, π/2], and cos(θ) ≥ 0 for this value of θ, from the straight triangle in
Figure 2.2, above, we have
√
cos(sin−1 x) = cos(θ) = cos(sin−1 x) = 1 − x2 ;
hence,
d 1
sin−1 x = √ .
dx 1 − x2
Example 2.4. By the Quotient Rule
x2 − 3 √
√ − 2x sin−1 x
d sin−1 x 1−x 2 x2 − 3 − 2x 1 − x2 sin−1 x
= = √ .
dx x2 − 3 (x2 − 3)2 (x2 − 3)2 1 − x2
By the Chain Rule,
d g 0 (x)
sin−1 (g(x)) = p .
dx 1 − g(x)2
Example 2.5. By the Chain Rule

d √ 1 1 1
sin−1 ( x) = p √ √ = √ .
dx 1 − ( x)2 2 x 2 x − x2

Integration
Z
dx
√ = sin−1 x + C
1−x 2
and
Z
g 0 (x)
p dx = sin−1 (g(x)) + C. (2.1)
1 − (g(x))2
Example 2.6. For the integral

Z
ex
√ dx,
4 − e2x
d x
we see that e2x = (ex )2 and e = ex .
dx
Thus, for u = ex , du = ex dx, and by Equation (2.1), above,
Z Z x
ex du −1 u −1 e
√ dx = √ = sin = sin + C.
4 − e2x 4 − u2 2 2
Example 2.7. Consider, for a > 0, the integral

Z
dx
√ .
a2 − x 2
We have
s r
√ x2 x 2
a2 − x2 = a2 1− 2 =a 1− .
a a
x dx
For u = , du = , and
a a
Z Z x
dx a du
√ = √ = sin−1 u = sin−1 + C. (2.2)
a2 − x 2 a 1 − u2 a
This is Basic Integral 15 of the table of integrals from the Additional Material section of the textbook.
Example 2.8. We can apply Example 2.7 to the integral

Z
dx
√ .
−1 − x2 + 4x
We first complete the perfect square
−x2 + 4x − 1 = −(x2 − 4x + 4) − 1 + 4 = 3 − (x − 2)2 .

Hence,
Z Z
dx dx
√ = p .
−1 − x2 + 4x 3 − (x − 2)2
√
For u = x − 2, du = dx, and by Equation (2.2), above, with a = 3,
Z Z
dx du
p = √
3 − (x − 2)2 3 − u2

−1 u −1 x−2
= sin √ = sin √ + C.
3 3
The Inverse of the Cosine Function

Notation
The inverse of the cosine function is denoted by
cos−1 x or arccos x.
Graph
To make the cosine function one-to-one, we restrict it to the closed interval [0, π]. The graphs, on this
interval, of cosine and its inverse are shown in Figure 2.3, below.
Figure 2.3: Graphs of the cos x on the interval [0, π] and its inverse cos−1 x
Properties
From the graph of cos−1 x in Figure 2.3, we have the following properties:
1. the domain is [−1, 1].
2. the range is [0, π].

π
3. cos−1 (0) = .
2

4. cos−1 (−1) = π.
5. cos−1 (1) = 0.
6. the function is decreasing on its domain.
7. the function is concave up on [−1, 0].
8. the function is concave down [0, 1].
To determine the domain of functions of the form cos−1 (g(x)), we must find all x so that −1 ≤ g(x) ≤ 1.
See Example 2.9, below.
Example 2.9. To find the domain of the function cos−1 (e2x ) we must find all x so that 0 < e2x ≤ 1
(exponential functions are positive on their domains). The exponential function et is increasing and equal
to 1 for t = 0. Hence, et ≤ 1 for all t ∈ (−∞, 0] and the domain of the function cos−1 (e2x ) is the interval
(−∞, 0] .
Example 2.10. The domain of the function cos−1 (−x2 + 2x) is all x so that −1 √
≤ −x2 + 2x ≤ 1. The
function g(x) = −x + 2x is a concave down parabola and g(x) = 1 for x = 1 ± 2. Moreover,
2
√ √
−1 ≤ g(x) ≤ 1 on the interva [1, − 2, 1 + 2]
When we apply cos−1 x to this inequality, the inequality does change because cos−1 x is decreasing;
hence,
π = cos−1 (−1) ≥ g(x) ≥ cos−1 (1) = 0.
The function cos−1 (−x2 + 2x) is bounded above by π and below by 0. The range of this function is the
interval (0, π].
Cancellation Laws
cos(cos−1 x) = x if and only if −1 ≤ x ≤ 1
cos−1 (cos x) = x if and only if 0≤x≤π

π
cos−1 cos − ,
3
we must exchange
π
cos − for cos(x), where x is in [0, π].
3

Since
π π
cos − = cos ,
3 3
we have
π π π
−1 −1
cos cos − = cos cos = .
3 3 3
Right Triangle
We can associate a right triangle to cos−1 x. Consider that
θ = cos−1 x is an angle in the interval [0, π],
then
x
cos(θ) = cos(cos−1 x) = x = ,
1
√ triangle has adjacent side x and hypotenuse 1. By the Pythagorean Theorem, the
opposite side is 1 − x2 (see Figure 2.4, below).
Figure 2.4: Right triangle with angle θ = cos−1 x
Derivative
By the cancellation law
cos(cos−1 x) = x.

d
− sin(cos−1 x) cos−1 x = 1.
dx
Solving gives us
d 1
cos−1 x = − .
dx sin(cos−1 x)

Since θ = cos−1 x is in [0, π] and sin(θ) ≥ 0 for this value of θ, then, from the right triangle in Figure 2.4,
we have
√
sin(θ) = sin(cos−1 x) = 1 − x2 .
Hence,
d 1
cos−1 x = − √ .
dx 1 − x2
Example 2.12.
d cos−1 x sin−1 x
cos−1 x sin−1 x = √ −√
dx 1 − x2 1 − x2
−1 −1
cos x − sin x
= √ .
1 − x2
By the Chain Rule,
d g 0 (x)
cos−1 (g(x)) = − p .
dx 1 − g(x)2
Example 2.13.
1
d x 1
cos−1 (ln x) = − p =− p .
dx 1 − (ln x)2 x 1 − (ln x)2
Integration
Since
d d
cos−1 x = − sin−1 x
dx dx
the integration of cos−1 x is similar to that of sin−1 x.
The Inverse of the Tangent Function

Notation
The inverse of the tangent function is denoted by
tan−1 x or arctan x.

Notation
The inverse of the tangent function is denoted by
tan−1 x or arctan x.
Graph
Graph
To make the tangent function one-to-one, we restrict it to the open interval
To make
πtheπtangent
function one-to-one, we restrict it to the open interval
π π
−− , , ..
22 22
TheThe
graphs,
graphs,on thisinterval,
on this interval, oftangent
of the the tangent
function function andareitsshown
and its inverse inverse
in are shown in Figure 2.5, below.
Figure 1.11, below.
π
2
− π π
2 2
− π
2
π π
Figure 1.11: Functions tan x and tan−1 x on the interval − , π π
Figure 2.5: Graphs of the unctions tan x on the interval 2 2 − , and its inverse tan−1 x
2 2
Properties
Properties
From the graph of tan−1 x in Figure 1.11, we have the following properties:
From the graph of tan−1 x in Figure 2.5, we have the following properties:
π π
1. the domain is R; the range is − , . π π
1. the domain is R; the range 2is 2 − , .
2 2 Mathematics 266 / Study Guide
2. tan−1 (0) = 0.

π
4. lim tan−1 x = .
x→∞ 2
π
5. lim tan−1 x = − .
x→−∞ 2
6. the function is concave up on the interval (−∞, 0].
7. the function is concave down on the interval [0, ∞).
In the next example we use Property 5.
Example 2.14. Evaluate the limit

−1 −3x2 + 2
lim tan .
x→∞ 2x + 1

We see that
−3x2 + 2 −3 + x22
lim = lim 2 = −∞.
x→∞ 2x + 1 x→∞
x
+ x12
Then,

−1 −3x2 + 2 π
lim tan = lim tan−1 (t) = − .
x→∞ 2x + 1 t→−∞ 2
Cancellation Laws
tan(tan−1 x) = x if and only if x∈R

π π
tan−1 (tan x) = x if and only if − <x<
2 2

−1 7π
tan tan − ,
3
we must exchange
π π
7π
tan − for tan(x), where x is in − , .
3 2 2
π
7π
Since tan − = tan − , we have
3 3
π
−1 7π π
tan tan − = tan−1 tan − =− .
3 3 3
Right Triangle
We can associate a right triangle to tan−1 x. Consider that

π π
θ = tan−1 x is an angle in the interval − , ,
2 2
then
x
tan(θ) = tan(tan−1 x) = x = ,
1
√ triangle has opposite side x and adjacent side 1. By the Pythagorean Theorem, the
hypotenuse is 1 + x (see Figure 2.6, below).
2

Figure 2.6: Right triangle with angle θ = tan−1 x
Derivative
By the cancellation law
tan(tan−1 x) = x

d
sec2 (tan−1 x) tan−1 x = 1.
dx
Solving, we get
d 1
tan−1 x = .
dx sec2 (tan−1 x)
Since θ = tan−1 x is in (−π/2, π/2) and sec(θ) ≥ 0 for this value of θ, from the right triangle in
Figure 2.6, above, we have
sec2 (tan−1 x) = sec2 (θ) = 1 + x2 ;
hence,
d 1
tan−1 x = .
dx 1 + x2
Example 2.16.
x2
− 2x tan−1 x
d tan−1 x 1 + x2 x − 2(1 + x2 ) tan−1 x
= = .
dx x2 x4 x3 (1 + x2 )
By the Chain Rule
d g 0 (x)
tan−1 (g(x)) = .
dx 1 + g(x)2

Example 2.17.
d 1 1
tan−1 (1 − x) = − 2
= 2
.
dx 1 + (1 − x) −x + 2x − 2
Integration
Z
dx
2
= tan−1 x + C
1+x
and
Z
g 0 (x)
dx = tan−1 (g(x)) + C. (2.3)
1 + (g(x))2
Example 2.18. An application of Equation (2.3), above, is

Z Z √ √
dx 1 6 tan−1 ( 6x)
=√ √ dx = √ + C.
1 + 6x2 6 1 + ( 6x)2 6
Example 2.19. By Equation (2.3),

Z
cos x 1 −1 sin x
= tan + C.
9 + sin2 x 3 3
Example 2.20. Consider, for a > 0, the integral

Z
dx
.
a + x2
2
We have
x 2
2 2 x2 2 2
a +x =a 1+ 2 =a 1+ .
a a
x dx
For u = , du = , and
a a
Z Z
dx a du 1 1 −1 x
= = tan−1
u = tan + C. (2.4)
a2 + x 2 a2 (1 + u2 ) a a a
This is Basic Integral 16 from the table of integrals in the Additional Material section of the textbook.
Example 2.21. We can apply Example 2.20 to the integral

Z
dx
2
.
x −x+1

We first complete the perfect square, as in Example 2.8 on page 48, and find that
2
2 1 3
x −x+1= x− + .
2 4
Hence,
Z Z
dx dx
= 2 .
x2 − x + 1 1 3
x− +
4 4
√
1 3
For u = x − , du = dx, and by Equation (2.4) on page 56, with a =
4 2
Z Z
dx 3 du 2 2u
2 + = = √ tan−1 √
1 4 3 3 3
x− u2 +
4 4
 
1
2 x−
2 
−1  4 
= √ tan  √  + C.
3 3 
Therefore,
Z
dx 2 −1 4x − 1
√ = √ tan √ + C.
2
x −x+1 3 2 3
The Inverse of the Secant Function

Notation
The inverse of the secant function is denoted by
sec−1 x or arcsecx.
Graph
To make the secant function one-to-one, we restrict it to the open interval
h π h 3π
0, ∪ π, .
2 2
The graphs, on this interval, of secant and its inverse are shown in Figure 2.7, below.
Choosing these intervals makes it easier to differentiate formulas of the inverse of the secant function.
Properties
From the graph of sec−1 x in Figure 2.7, we have the following properties:

h π h 3π
Figure 2.7: Graphs of the functions sec x on the interval 0, ∪ π, and its inverse sec−1 x
2 2
h π h 3π
1. the domain is (−∞, −1] ∪ [1, ∞); the range is 0, ∪ π, .
2 2
2. sec−1 (−1) = π and sec−1 (1) = 0.
3. the function is decreasing on (−∞, −1] and increasing on [1, ∞).

π
4. lim sec−1 x = .
x→∞ 2
3π
5. lim sec−1 x = .
x→−∞ 2
The domain of a function of the form sec−1 (g(x)) consists of all x such that |g(x)| ≥ 1.
Example√2.22. To find the domain of the function sec−1 (x2 − 2x) we see that g(x) = x2 − 2x = 1 for
x = 1 ± 2, and
√ √
g(x) ≥ 1 for x ≤ 1 − 2 and x ≥ 1 + 2.
√ √
The domain of the function sec−1 (x2 − 2x) are the intervals (−∞, 1 − 2] and [1 + 2, ∞) .
In the next example, we apply Property 4.
tan−1 x
lim
x→∞ sec−1 (x + 2)
We see that lim x + 2 = ∞, then

x→∞

tan−1 x π 2
lim −1
= = 1. .
x→∞ sec (x + 2) 2 π

Cancellation Laws
sec(sec−1 x) = x if and only if |x| ≥ 1

h π h 3π
sec−1 (sec x) = x if and only if x ∈ 0, ∪ π,
2 2
Example 2.24. To apply the cancellation law to evaluate

−1 3π
sec sec ,
4
we must exchange
h π h 3π
3π
sec for sec(x) where x is in 0, ∪ π, .
4 2 2

3π 5π
Since sec = sec , we have
4 4

−1 3π −1 5π 5π
sec sec = sec sec = .
4 4 4
Right Triangle
We can associate a right triangle to sec−1 x. Consider that

h π h 3π
θ = sec−1 x is an angle in the interval 0, ∪ π, ,
2 2
then
x
sec(θ) = sec(sec−1 x) = x =
1
and the corresponding triangle has hypotenuse x and adjacent side 1.
√
By the Pythagorean Theorem, the opposite side is x2 − 1 (see Figure 2.8, below).
Figure 2.8: Right triangle with angle θ = sec−1 x

Derivative
By the cancellation law,
sec(sec−1 x) = x.

d
sec(sec−1 x) tan(sec−1 x) sec−1 x = 1.
dx
Solving, we get
d 1
sec−1 x = .
dx x tan(sec−1 x)
h π h 3π
Note that θ = sec −1
x is in 0, ∪ π, , and tan(θ) ≥ 0 for this value of θ.
2 2
Therefore, from the right triangle in Figure 2.8 on the previous page, we have
√
tan(θ) = tan(sec−1 x) = x2 − 1.
Hence,
d 1
sec−1 x = √ .
dx x x2 − 1
Example 2.25.
d sec x
sec x sec−1 x = sec x tan x sec−1 x + √ .
dx x x2 − 1
By the Chain Rule
d g 0 (x)
sec−1 (g(x)) = p .
dx g(x) g(x)2 − 1
Example 2.26.
d sec2 x
sec−1 (tan x) = √ .
dx tan x tan2 x − 1
Integration
Z
dx
√ = sec−1 x + C
2
x x −1

and
Z
g 0 (x)
p dx = sec−1 (g(x)) + C.
2
g(x) (g(x)) − 1
Example 2.27.

1
Z Z dx
dx 1 a
√ = x r x 2
x x 2 − a2 a
−1
a a
1
−1 x
= sec + C.
a a
This is Basic Integral 17 from the table of integrals in the Additional Material section of the textbook.
Exercises
1. Consider the cotangent function restricted to the interval (0, π).
Give the graph, state the corresponding properties, establish the cancellation laws, give the
corresponding right triangle, and the derivative and integration of its inverse function.
π i 3π i
2. Consider the cosecant function restricted to the interval 0, ∪ π, .
2 2
Give the graph, state the corresponding properties, establish the cancellation laws, give the
corresponding right triangle, and the derivative and integration of its inverse function.
3. Read the examples of Section 2.1 of the textbook, Inverse of Trigonometric Functions
4. Do the exercises of the section Exercises for 2.1 of the textbook.
5. Do the exercises in the two videos, Limits of Inverse Tangent Function. Videos Limit 1 and Limit 2.
6. Do the exercises in the video Area below the Curve. Video Area 1.
7. For further practice, complete the lab Inverse Trigonometric Functions in the optional Lyryx Labs.

Inverse of Hyperbolic Functions
The Inverse of the Hyperbolic Sine Function
Notation
The inverse of the hyperbolic sine function is denoted by
sinh−1 x.
Graph
The hyperbolic sine function is one-to-one in its domain. You can see this in the graphs of sinh x and its
inverse sinh−1 x shown on Figure 2.5 of Section 2.2 of the textbook, Inverse Hyperbolic Functions.
Properties
From the graph of sinh−1 x, we have the following properties:
1. the domain and range are R.
2. sinh−1 (0) = 0.
3. the function is increasing in its domain.
Cancellation Laws
sinh(sinh−1 x) = x for all x ∈ R
sinh−1 (sinh x) = x for all x ∈ R
Hyperbolic Sine in Terms of the Natural Logarithm
If y = sinh−1 x, then
ey − e−y
x = sinh(sinh−1 x) = sinh y = .
2
Solving for y, we get
0 = ey − e−y − 2x.
Multiplying by ey gives us
0 = e2y − 2xey − 1.
Setting u = ey , we get
u2 − 2xu − 1 = 0.

By the quadratic equation,
√
2x ± 4x2 + 4 √
u= = x ± x2 + 1;
2
therefore,
√
ey = x ± x2 + 1.
√
Since ey > 0 and x − x2 + 1 < 0 for any x, we have
√
ey = x + x2 + 1,
and applying the natural logarithm gives us

√
sinh−1 x = ln(x + x2 + 1).
Derivative
For the derivative, we have

d d √
sinh−1 x = ln(x + x2 + 1)
dx dx
1 + x(x2 + 1)−1/2
= √
x + x2 + 1
√
x2 + 1 + x
=√ √
x2 + 1(x + x2 + 1)
√ √ !
x2 + 1 + x 2
x +1−x
=√ √ √
x2 + 1(x + x2 + 1) x2 + 1 − x
1
=√
x2 + 1
Example 2.28.
√ 1 1
sinh−1 ( x) = √ √ = √ .
2 x x+1 2 x2 + x
Integration
Z
dx
√ = sinh−1 x + C
2
x +1
Z
g 0 (x)
p dx = sinh−1 (g(x)) + C (2.5)
2
g(x) + 1

Example 2.29. By Equation (2.5),
Z Z
5x 1 ln 5(5x )
√ dx = p dx
52x + 1 ln 5 (5x )2 + 1
sinh−1 (5x )
= + C.
ln 5
Example 2.30. For a > 0,
Z Z
dx 1
√ = r (2.6)
2
x +a 2 x 2
a +1
a
1
Z
a −1 x
= r dx = sinh + C.
x 2 a
+1
a
Compare this result with Integral 72 from the table of integrals in the Additional Material section of the
textbook.
The Inverse of the Hyperbolic Cosine Function

Notation
The inverse of the hyperbolic cosine function is denoted by
cosh−1 x.
Graph
The hyperbolic cosine function is one-to-one on the interval [0, ∞). Its graph is shown on Figure 2.5 of
Section 2.2 of the textbook, Inverse Hyperbolic Functions.
Properties
From the graph of cosh−1 x, we have the following properties:
a. the domain is [1, ∞); the range is [0, ∞).
b. cosh−1 (1) = 0.
c. the function is increasing on its domain.
d. lim cosh−1 x = ∞.
x→∞

Cancellation Laws
cosh(cosh−1 x) = x for all x > 1
cosh−1 (cosh x) = x for all x > 0
Hyperbolic Cosine in Terms of the Natural Logarithm
If y = cosh−1 x, then x > 1 and
ey + e−y
x = cosh(cosh−1 x) = cosh y = .
2
0 = ey + e−y − 2x.
0 = e2y − 2xey + 1.
Setting u = ey , we get
u2 − 2xu + 1 = 0.
By the quadratic equation,

√
2x ± 4x2 − 4 √
u= = x ± x2 − 1;
2
therefore,
√
ey = x ± x2 − 1.
√
Since ey > 0 and x − x2 − 1 < 0 for any x > 1, we have
√
ey = x + x2 − 1,

√
cosh−1 x = ln(x + x2 − 1).

Derivative

d d √
cosh−1 x = ln(x + x2 − 1)
dx dx
1 + x(x2 − 1)−1/2
= √
x + x2 − 1
√
x2 − 1 + x
=√ √
x2 − 1(x + x2 − 1)
√ √ 2
x2 − 1 + x x −1−x
=√ √ √
x2 − 1(x + x2 − 1) x2 − 1 − x
1
=√
2
x −1
Example 2.31.
2e2x 2e2x
cosh−1 (e2x ) = p =√ .
(e2x )2 − 1 e4x − 1
Integration
Z
dx
√ = cosh−1 x + C
2
x −1
Z
g 0 (x)
p dx = cosh−1 (g(x)) + C
2
g(x) − 1
Example 2.32.
Z Z p
dx 1
p = (ln x)2 − 1 dx = cosh−1 (ln x) + C.
2
x (ln) − 1 x

Z Z
dx 1
√ = r (2.7)
2
x −a 2 x 2
a −1
a
1
Z x
a
= r = cosh−1 + C.
x 2 a
−1
a

Compare this result with Integral 81 from the table of integrals in the Additional Materials section of the
textbook.
The Inverse of the Hyperbolic Tangent Function

Notation
The inverse of the hyperbolic sine function is denoted by
tanh−1 x.
Graph
The hyperbolic tangent function is one-to-one on its domain. Its graph is shown in Figure 2.5 of Section
2.2 of the textbook, Inverse Hyperbolic Functions.
Properties
From the graph of tanh−1 x, we have the following properties:
1. the domain is (−1, 1); the range is R.
2. tanh−1 (0) = 0.
4. lim tanh−1 x = −∞ and lim− tanh−1 x = ∞.

x→−1+ x→1
Cancellation Laws
tanh(tanh−1 x) = x for all |x| < 1
tanh−1 (tanh x) = x for all x ∈ R
Hyperbolic Tangent in Terms of the Natural Logarithm
If y = tanh−1 x, then |x| < 1 and
ey − e−y
x = tanh(tanh−1 x) = tanh y = .
ey + e−y
0 = (x − 1)ey + (x + 1)e−y .

x+1
e2y = ,
1−x

−1 1 1+x
tanh x = ln for |x| < 1.
2 1−x
Derivative

d −1 d 1 1+x
tanh x = ln
dx dx 2 1−x
1 d
= (ln(1 + x) − ln(1 − x))
2 dx

1 1 1
= +
2 1+x 1−x
1
=
1 − x2
Example 2.34.
d cosh x
tanh−1 (sinh x) = .
dx 1 − sinh2 x
Integration
Z
dx
= tanh−1 x + C
1 − x2
Z
g 0 (x)
dx = tanh−1 (g(x)) + C.
1 − (g(x))2

1
Z Z
dx a −1 x
= x 2 dx = tanh + C. (2.8)
a2 − x2 a
1−
a

Representation of Inverse Hyperbolic Functions in Terms of Natural Loga-
rithm
Your textbook gives a list of inverse hyperbolic functions expressed in terms of natural logarithm in the
table titled, Logarithmic Definitions of Inverse Hyperbolic Functions, in Section 2.2, Inverse Hyperbolic
Functions.
In the next two examples, we show you how two of these representations are obtained. We leave the
others as exercises.
Example 2.36. We want to show that

√
sinh−1 x = ln(x + x2 + 1).
Let y = sinh−1 x. Then,
ey − e−y
x = sinh(sinh−1 x) = sinh y = ⇒ ey − 2x − e−y = 0.
2
We want to solve for y from this equation; to achieve this we multiply it by ey
e2y − 2xy − 1 = 0.
If we set w = ey and consider x as a constant we obtain a quadratic with coefficients

a = 1, b = −2x, c = −1
√
2 2x ± 4x2 + 4 √
w − 2xw − 1 = 0 ⇒ w = = x ± x2 + 1.
2
√
Since ey > 0 the value of w cannot be x − x2 + 1 < 0 (why?). Thus,
√ √
ey = x + x2 + 1 ⇒ y = sinh−1 x = ln(x + x2 + 1).
Example 2.37. See that the domain of tanh−1 x is the interval (−1, 1). Hence, we want to show that for
−1 < x < 1

−1 1 1+x
tan x = ln .
2 1−x
Let y = tanh−1 x. Then,
ey − e−y
x = tanh(tanh−1 x) = tanh y = y −y
⇒ x(ey + e−y ) = ey − e−y .
e +e
As in the previous example we want to solve for ey from this equation; to achieve this we again multiply
it by ey
xe2y + x = e2y − 1 ⇒ (x − 1)e2y = −x − 1.

We solve

2y −x − 1 x+1 x+1
e = = ⇒ 2y = ln .
x−1 1−x 1−x
Therefore

−1 1 1+x
y = tanh x = ln .
2 1−x
Exercises
6. Read Section 2.2 of the textbook, Inverse of Hyperbolic Functions.
8. Explain why the composition tanh−1 (cosh x) is undefined for all x.
9. Give the domain of ln(tanh−1 x).
10. Prove the formulas

√
a. cosh−1 x = ln(x + x2 − 1) for x ≥ 1.

−1 1 x+1
b. coth x = ln for |x| > 1.
2 x−1
√
1 + 1 − x2
c. sech x = ln
−1
for 0 < x ≤ 1.
x
√ !
1 1 + x 2
d. csch −1 x = ln + for x 6= 0.
x |x|
11. Evaluate the integrals listed below

Z
a. csch 5 x coth x dx
Z
b. tanh x dx
Z
sinh x
c. dx
sinh2 x + 1
Z
d. sech 2 x tanh2 x dx
Z
csc2 x
e. dx
2 − 3 coth x

Indeterminate Forms and L’Hopital’s Rule
L’Hopital’s Rule is a technique we use to evaluate limits of certain types. The determination of these
types requires the evaluation of algebraic and trigonometric limits. You will find it helpful to have the
graphs of the trigonometric, exponential, logarithmic and hyperbolic functions handy as you work
through this section.
The difficulty in evaluating the limit

sin x
lim
x→0 x
is that the function

sin x
f (x) =
x
is not continuous at zero, it cannot be simplified, and
lim sin x = 0 and lim x = 0.

x→0 x→0
That is, the limit is of type 00 .
We have a similar situation with the limit

ex
lim ,
x→∞ x
which is of type ∞
∞
. Why?
It is precisely to the evaluation of these types of limits that we apply l’Hopital’s Rule.
Definition 2.1. a. If lim f (x) = 0 and lim g(x) = 0, then the limit
x→a x→a
f (x)
lim
x→a g(x)
is an indeterminate form of type 00 .
b. If lim f (x) = ±∞ and lim g(x) = ±∞, then the limit

x→a x→
f (x)
lim
x→a g(x)
is an indeterminate form of type ∞

∞
.
Definition 2.1 extends naturally to side limits and infinite limits; that is, where x → a is replaced by the
side limits x → a+ , x → a− , or by the infinite limits x → ∞, x → −∞.

Example 2.38. The limit
ln x
lim+ −1
x→0 tanh (x − 1)
is of type ∞
∞
because
lim ln x = −∞ (see the graph of ln x),

x→0+
and for t = x − 1, we have t → −1+ , since
lim x − 1 = −1 and x − 1 > −1 for x → 0+ .

x→0+
Hence,
lim tanh−1 (x − 1) = lim+ tanh−1 t = −∞

x→0+ t→−1
−1
(see the graph of tanh t).

1
sin
x
lim x
x→∞ 3
is neither of type 00 nor ∞∞
. Observe that

1
lim sin = lim+ sin t = 0
x→∞ x t→0
and
lim 3x = ∞.
x→∞
L’Hopital’s Rule: If the limit
f (x)
lim
x→a g(x)
is an indeterminate form of type 0

0
or ∞
∞
, then
f (x) f 0 (x)
lim = lim 0 .
x→a g(x) x→a g (x)
L’Hopital’s Rule is also valid if x → a is replaced by
x → a+ , x → a− , x → ∞ or x → −∞.

ex
lim
x→∞ x
is of type ∞
∞
; hence, by l’Hopital’s Rule,
ex ex
lim = lim = ∞.
x→∞ x x→∞ 1
Example 2.41. As you can see from Example 2.38 on page 72, we can apply l’Hopital’s Rule to the
limit
ln x
lim+ −1 .
x→0 tanh (x − 1)
Hence,
1
ln x x
lim+ −1 = lim+
x→0 tanh (x − 1) x→0 1
(2x − x2 )
2x − x2
= lim+ = lim+ 2 − x = −∞.
x→0 x x→0
Other indeterminate forms can be changed into indeterminate forms of type 0

0
or type ∞
∞
Indeterminate Form of Type 0 · ∞

Definition 2.2. If
lim f (x) = 0 and lim g(x) = ±∞

x→a x→a
then the limit
lim f (x)g(x)
x→a
is an indeterminate form of type 0 · ∞.
If the limit
lim f (x)g(x)
x→a
is of type 0 · ∞, then the limits
f (x) 1/g(x)
lim or lim
x→a 1/g(x) x→a f (x)
are of type 0
0
or ∞
∞
. Why?

We can then proceed to apply l’Hopital’s Rule to evaluate either limit
f (x) f 0 (x)(g(x)2 )
lim f (x)g(x) = lim = lim −
x→a x→a 1/g(x) x→a g 0 (x)
or
g(x) g 0 (x)(f (x)2 )
lim f (x)g(x) = lim = lim − .
x→a x→a 1/f (x) x→a f 0 (x)
d 1 h0 (x)
Remember that =− .
dx h(x) h(x)2
Note: This strategy is also valid if x → a is replaced by
x → a+ , x → a− , x → ∞ or x → −∞.
For this type of indeterminate form, we have a choice of limit, naturally we tend to use the easiest.
Indeterminate Form of Type ∞ − ∞

If lim f (x) = ∞ and lim g(x) = ∞, then
x→a x→a
lim f (x) + lim g(x) = ∞

x→a x→a
and
lim −f (x) − lim g(x) = −∞.

x→a x→a
However, a problem arises with the evaluation of the limit
lim f (x) − lim g(x).

x→a x→a
In this case, we try—by an algebraic method or by rationalizing—to convert the difference of these
functions into a quotient such that the resulting quotient function yields an indeterminate form of type 0
0
or ∞
∞
, and we can apply l’Hopital’s Rule.
This strategy is also valid if x → a is replaced by
x → a+ , x → a− , x → ∞ or x → −∞.
Indeterminate Forms of Type 00 , ∞0 and 1∞

The properties of the natural logarithm are used in the evaluation of limits of the form
lim f (x)g(x) .
x→a
Observe that

1. if y = f (x)g(x) , then ln y = g(x) ln f (x) for all x such that f (x) > 0.
2. since ex is a continuous function,
lim eu = elimu→a u .
u→a
Using these observations, we can see that if lim ln y = lim g(x) ln f (x), then
x→a x→a
lim f (x)g(x) = lim y = lim eln y = elimx→a ln y . (2.9)

x→a x→a x→a
We can apply l’Hopital’s Rule if the limit
lim ln y = lim g(x) ln f (x)

x→a x→a
is of type 0 · ∞. Note that this is the case if the limit lim f (x)g(x) is and indeterminate form of type 00 ,
x→a
type ∞0 or type 1∞ . Why?
Note: If lim f (x) = −∞, then the limit lim ln f (x) is undefined.
x→a x→a
Definition 2.3. The limit
lim f (x)g(x)
x→a
is an indeterminate form of type
a. 00 if lim f (x) = 0 and lim g(x) = 0

x→a x→a
b. ∞ 0
if lim f (x) = ∞ and lim g(x) = 0
x→a x→a
c. 1 ∞
if lim f (x) = 1 and lim g(x) = ±∞
x→a x→a
From Definition 2.3, above, we have



 eL if lim ln y = L


 x→a
lim f (x)g(x)
= ∞ if lim ln y = ∞
x→a 
 x→a


 0 if lim ln y = −∞
x→a
Where
lim ln y = lim g(x) ln f (x).

x→a x→a

Exercises
13. Review the tutorials Limit 1 and Limit 2.
14. Read the examples of Section 2.4 of the textbook, L’Hopital’s Rule.
15. Do as many exercises as you can from Exercises for 2.4 of the textbook.
16. For further practice, complete the lab titled L’Hopital’s Rule in the optional Lyryx Labs.
Finishing This Unit

examination.
against those provided in Appendix D of this Study Guide.
5. Complete the first assignments and submit your work to your tutor for grading. Remember: The
number of points in a question may indicate the number of steps in the solution.
6. Complete the online Lyryx Assessment quiz titled Unit 2. Your grade on this quiz will count towards
your final mark for the course.

Time: 3 hours
Total points: 94
Passing grade: 55%
1. (12 points) Find the domain and range of the functions listed below.
a. cosh−1 (x2 − x).

1
b. sin−1
x−1
c. cos−1 (ln x)
2. (6 points) Prove the identity
x2 − 1
tanh(ln x) = .
x2 + 1
3
3. (6 points) If cosh x = find the values of the other hyperbolic functions.
2
4. (12 points) Find each of the derivatives indicated below.
d 4x3 + sin−1 x
a.
dx ln x
d
b. cosh−1 (3x + 1)
dx
d2
c. 2
tan−1 (ex )
dx
5. (16 points) Integrate
Z
dx
a. √
2
2x − 6x + 3
Z
dx
b. √
2 − 10x − 2x2
6. (12 points) Integrate the definite integrals listed below.
Z ln(2)
dx
a. √
0 x2 − 4
Z e−1
b. tanh3 x sech2 x dx
0
Z 1
c. sinh2 x dx
−1

7. (25 points) Evaluate each of the limits listed below. Justify your answer. If a limit does not exits
explain why.
1 1
a. lim+ − x
x→0 x e −1
b. lim [cos(2/x)]x
x→∞
cos−1 x − π
c. lim + √
x→−1 x+1
d. lim tanh−1 (cos x)
x→0
ln(sin x)
e. lim+
x→0 ln(tan x)
8. (5 points)
a. Identify the region below the curve y = tanh(x − 1) on the interval [0, 2].
b. Find the area of the region identified in part a.

UNIT 3
Techniques of Integration
Integration is not straightforward; there is no rule or method that guarantees that we can obtain an
indefinite integral. We must rely on the basic antiderivatives (as listed in the table of integrals in the
Additional Material section of the textbook), in practice and in intuition—that is, on knowledge and
experience.
Basically, there are two techniques of integration: integration by parts and integration by substitution.
The key in both is to identify a differential in the integrand function. In this unit, we start with parts; then,
we discuss techniques that are particular cases of the substitution method, which you should have learned
in a previous calculus course.
Objectives
1. apply different techniques of integration.
2. determine proper and improper integrals.
3. apply the Midpoint Rule to approximate definite integrals.
4. apply the Trapezoidal Rule to approximate definite integrals.
5. apply Simpson’s Rule to approximate definite integrals.
6. evaluate improper integrals.
Integration by Parts
We can integrate by parts because of the Product Rule.
We have
d
f (x)g(x) = f (x)g 0 (x) + f 0 (x)g(x)
dx

then
Z Z Z
0 0 0
f (x)g (x) + f (x)g(x) dx = f (x)g (x) dx + f 0 (x)g(x) dx
= f (x)g(x)
and
Z Z
0
f (x)g (x) dx = f (x)g(x) − f 0 (x)g(x) dx. (3.1)
Observe that the form of the integrand function f (x)g 0 (x) dx is the product of a function f (x) and a
differential g 0 (x) dx. Hence, to integrate by parts—that is, to apply Equation (3.1), we need two things:
1. the integrand function must have the form f (x)g 0 (x) dx, and
2. we must know the antiderivative g(x) of g 0 (x).
Example 3.1. The integrand function in

Z
x cos x dx
has two possible forms:
a. f (x) = x and g 0 (x) = cos x, thus g(x) = sin x, or

x2
b. f (x) = cos x and g (x) = x, thus g(x) = .
0

2
Z
ln x dx
can be thought of as the product (1) ln x, and we have two possibilities:
a. f (x) = ln x and g 0 (x) = 1, thus g(x) = x, or
b. f (x) = 1 and g 0 (x) = ln x, but in this case we do not know the antiderivative of g(x), which is our
original problem.

Z
x ln x dx
can be considered the product of f (x) = x and g 0 (x) = ln x; however, we do not know the antiderivative
of g.
x2
Hence, we consider instead f (x) = ln x and g 0 (x) = x, so g(x) = .
2

When we apply Equation (3.1), above, we get another integral, one that we expect will, at the least, be
easier to integrate than the original one.
Example 3.4. From Example 3.1 on the facing page, we have two possibilities for Equation (3.1):
Z Z
x cos x dx = x sin x − sin x dx
or
Z Z
x2 cos x x2
x cos x dx = − − sin x dx.
2 2
We know the antiderivative of sin x, and the integral

Z 2
x
− sin x dx
2
R
is more complicated than the original integral, x cos x dx. Hence, the former is the best choice, and we
have
Z Z
x cos x dx = x sin x − sin x dx = x sin x + cos x + C.
Example 3.5. From Example 3.3 on the preceding page, we have

Z Z
x2 ln x 1 x2 x2 ln x x2
x ln x dx = − dx = − + C.
2 x 2 2 4
Example 3.2 on the facing page illustrates a possible way to find the antiderivative of a function whose
antiderivative is not a basic function.
Example 3.6. Exercise 3.1.6 in Exercises for 3.1 in the textbook. For
Z
ln x dx
we have
f (x) = ln x and g 0 (x) = 1.
Hence,
1
f 0 (x) = and g(x) = x.
x
By parts,
Z Z
ln x dx = x ln x − dx = x ln x − x.
You may want to check this result by taking the derivative of x ln x − x.

Example 3.7. For
Z
sin−1 x dx
we have
f (x) = sin−1 x and g 0 (x) = 1.
Hence,
1
g(x) = x and f 0 (x) = √ .
1 − x2
By parts,
Z Z √
−1 −1 x
sin x dx = x sin x− √ dx = x sin−1 x + 1 − x2 + C.
1 − x2
Z
x
Note: The integral √ dx is solved by substitution, with u = x2 .
1−x 2
Example 3.8. In the integral

Z
sec3 x dx,
we take
sin3 x = sec x sec2 x and f (x) = sec x, g 0 (x) = sec2 x.
Thus,
f 0 (x) = sec x tan x and g(x) = tan x.
By parts,
Z Z
3
sec x dx = sec x tan x − sec x tan2 x dx.
The integral
Z Z Z
sec x tan x dx = sec x(sec x − 1) dx = sec3 x − sec x dx.
2 2
We then have
Z Z Z
3 3
sec x dx = sec x tan x − sec x dx − sec x dx.

which gives us
Z Z
3
2 sec x dx = sec x tan x − sec x dx
= sec x tan x − ln | sec x + tan x|,

Z
sec x tan x ln | sec x + tan x|
sec3 x dx = − + C.
2 2
Integration by parts is applied to the product of two functions, but keep in mind that
Z
(f (x))2
f (x)f 0 (x) dx = + C.
2
and for r 6= −1,
Z
(f (x))r+1
(f (x))r f 0 (x) dx = + C.
r+1
Thus, when the integrand functions have these forms, integration by parts may be correct but not
efficient, or even entirely inappropriate.

Z
sin x cos x dx,
we have f (x) = sin x and g 0 (x) = cos x. But f 0 (x) = g 0 (x); hence, the integrand function is of the form
f (x)f 0 (x), and we do not need to apply parts. We simply have
Z
sin2 x
sin x cos x dx = + C.
2
[Try solving this integral by parts to see what we mean by “not efficient.”]

Z √
3
x 3 − x2 dx,
integration by parts gives

√
f (x) = 3 − x2 , g 0 (x) = x
3
and
2x x2
f 0 (x) = − , g(x) = .
3(3 − x2 )2/3 2

You can see the difficulty of this method since
Z √ √
2 3 Z
3 2
x 3 − x2 2x3
x 3 − x dx = − − dx.
2 6(3 − x2 )2/3
Instead, we should see that the integrand can be changed into the form
f (x)1/3 f 0 (x) dx
with the help of constants:

Z
1
− (3 − x2 )1/3 (−2x) dx,
2
and therefore the answer is
3(3 − x2 )4/3
− + C.
8
Reduction Formulas
In the next example, we show you how to apply integration by parts to obtain a reduction formula for
sinn x for any positive integer n ≥ 2.
Example 3.11. For any integer n ≥ 2, we have the reduction formula

Z Z
n cos x sinn−1 x n − 1
sin x dx = − + sinn−2 x dx.
n n
Let
f (x) = sinn−1 x and g 0 (x)dx = sin x dx,
then
f 0 (x) = (n − 1) sinn−2 x cos x dx and g(x) = − cos x.
The integration by parts gives
Z Z
n
sin x dx = − sin n−1
x cos x + (n − 1) sinn−2 x cos2 x dx.
To evaluate the integral on the right, we replace cos2 x with 1 − sin2 x, since cos2 + sin2 x = 1. Thus,
sinn−2 x(1 − sin2 x) = sinn−2 x − sinn x.

Hence,
Z
(n − 1) sinn−2 x cos x dx
Z Z Z
n−2 n n−2
= (n − 1) sin x − sin x dx = (n − 1) sin x dx − (n − 1) sinn x dx.
As in Example 3.8 on page 82, we can add the last term on the right to the left. Thus, we have
Z Z
n n−1
n sin x dx = − sin x cos x + (n − 1) sinn−2 x dx.
Dividing by n, we obtain the reduction formula as indicated. This is integral 28 from the table of
integrals.
In the next two examples, we apply the reduction formula for n equal to 2 and 4.
Example 3.12. For n = 2 we have

Z Z
2 cos x sin2−1 x 2 − 1
sin x dx = − + sin2−2 x dx
2 2
Z
cos x sin x 1
=− + dx
2 2
cos x sin x x
=− + + C.
2 2
You may also know that this same integral can be solved by using the double angle formula. In this case,
we obtain
Z Z
2 1 1 sin(2x)
sin x dx = 1 − cos(2x) dx = x− + C.
2 2 2
To reconcile these two results, we use a trigonometric identity, as follows:

1 sin(2x) 1 2 sin x cos x cos x sin x x
x− = x− =− +
2 2 2 2 2 2

Example 3.13. For n = 4, we apply the reduction formula and use Example 3.12:
Z Z
4 cos x sin4−1 x 4 − 1
sin x dx = − + sin4−2 x dx
4 4
Z
cos x sin3 x 3
=− + sin2 x dx
4 4

cos x sin3 x 3 cos x sin x x
=− + − +
4 4 2 2
3
cos x sin x 3 cos x sin x 3x
=− − + + C.
4 8 8
Other reduction formulas are left as exercises. You may also want to look at the reduction formulas 30,
31, 32, 33, 47 in the table of integrals.
Exercises
1. Read examples 3.37–3.40 of Section 3.1 of the textbook, Integration by Parts.
3. For further practice, complete the lab titled Integration by Parts in the optional Lyryx Labs.
4. Use integration by parts to obtain the reduction formulas listed below for any integer n ≥ 2.
Z Z
cosn−1 x sin x dx n − 1
a. n
cos x dx = + cosn−1 x dx.
n n
Z Z
tan x secn−2 x n − 2
b. n
sec x dx = + secn−2 x dx.
n−1 n−1
Z Z
c. (ln x) dx = x(ln x) − n (ln x)n−1 dx
n n
Z Z
d. n x n x
x e dx = x e − n xn−1 ex dx
Z Z
x(x2 + a2 )n 2na2
e. 2 2 n
(x + a ) dx = + (x2 + a2 )n−1 dx
2n + 1 2n + 1
5. Use integration by parts to obtain the reduction formula for any integer n ≥ 2 of
Z
cscn x dx.
6. a. Apply transformations to sketch the graph of the curve y = 2 ln(x + e)
b. Give the equation of the line which passes through the points (−e, 0) and (0, 2)
c. Find the area between the curve in part a. and the line of part b.

Trigonometric Integrals
Prerequisites
To complete this section, you must know the trigonometric identities and the antiderivative of a power
function for r 6= −1:
Z
[f (x)]r+1
[f (x)]r f 0 (x) dx = + C.
r+1
The trigonometric integrals considered in this section are of the form
Z
A. sinm x cosn x dx
Z
B. tanm x secn x dx
Z
C. cotm x cscn x dx
for any positive integers m and n.
For some values of m and n, these integrals are of the form

Z
[f (x)]k+1
[f (x)]k f 0 (x) dx = + C. (3.2)
k+1
Example 3.14. For n = 1 in A., above, we have the integral
Z
sinm x cos x dx,
which fits the form of Equation (3.2), with f (x) = sin x, f 0 (x) = cos x and k = m. Thus
Z
sinm+1 x
sinm x cos x dx = + C.
m+1
Similarly for m = 1 in A., we have an integral that fits the form of Equation (3.2), with f (x) = cos x,
f 0 (x) = − sin x and k = n. Therefore,
Z Z
cosn+1 x
sin x cos x dx = − (− sin x) cosn x dx = −
n
+ C.
n+1
Example 3.15. For n = 2 in B., the integral fits the form of Equation (3.2), with f (x) = tan x,
f 0 (x) = sec2 x and k = m. Thus,
Z
tanm+1 x
tanm x sec2 x dx = + C.
m+1

Example 3.16. For m = 1 and n > 1 in B., the integral fits the form of Equation (3.2), with
f (x) = sec x, f 0 (x) = tan x sec x, and k = n − 1. Thus,
Z Z
secn x
tan x sec x dx = (tan x sec x) secn−1 x dx =
n
+ C.
n
Example 3.17. For n = 2 in C. on page 87, the integral fits the form of Equation (3.2), with
f (x) = cot x, f 0 (x) = − csc2 x and k = m. Thus,
Z Z
cotm+1 x
cot x csc x dx = − cotm x(− csc2 x) dx = −
m 2
+ C.
m+1
Example 3.18. For m = 1 and n > 1 in C., the integral fits the form of Equation (3.2), with
f (x) = csc x, f 0 (x) = − cot x csc x, and k = n − 1. Thus,
Z Z
cscn x
cot x csc x dx = − (− cot x csc x) cscn−1 x dx = −
n
+ C.
n
For some other cases of m and n, we use trigonometric identities to change the integrand function and
make it fit any one of the integrals in Examples 3.14 to 3.18 on pages 87–88.
Example 3.19. If n = 2s + 1 is odd in A., then
sinm x cos2s+1 x = sinm x cos2s x cos x = sinm x(cos2 x)s cos x.
Using the trigonometric identity cos2 = 1 − sin2 x, we get
sinm x(1 − sin2 x)s cos x.
Multiplying this expression out yields a sum of products of powers of sin x and cos x; that is, each one of
the terms of this sum is of the form of Example 3.14 on page 87. Hence,
Z Z
2s+1
m
sin x cos x dx = sinm x(1 − sin2 x)s cos x dx.
Similarly, if m = 2s + 1 is odd, then
sin2s+1 x cosn x = sin x(sin2 x)s cosn x.
We use the trigonometric identity sin2 x = 1 − cos2 x and we have
sin x(1 − cos2 x)s cosn x.
Multiplying this expression out yields a sum of products of powers of cos x and sin x; that is, each one of
the terms of this sum is of the form of Example 3.14. Hence,
Z Z
2s+1
sin x cos x dx = sin x(1 − cos2 x)s cosn x dx.
n


Example 3.19 shows that when either m or n is odd, we factor sin x or cos x, and apply the identity
sin2 x + cos2 x = 1.
Z
Example 3.20. For the integral sin3 x cos3 x dx, we can factor either sin x or cos x.
Factoring sin x, we have
sin3 x cos3 x = sin2 x cos3 x sin x

= (1 − cos2 ) cos3 x sin x
= cos3 x sin x − cos5 x sin x
and
Z Z
3 3
sin x cos x dx = cos3 x sin x − cos5 x sin x dx
cos4 x cos6 x
=− + + C.
4 6
Factoring cos x, we have
sin3 x cos3 x = sin3 x cos2 x cos x

= sin3 x(1 − sin2 ) cos x
= sin3 x cos x − sin5 x cos x
and
Z Z
3 3
sin x cos x dx = sin3 x cos x − sin5 x cos x dx
sin4 x sin6 x
= − + C.
4 6
Example 3.21. If n = 2s is even in B. on page 87, then we factor sec2 x and use the identity
tan2 x + 1 = sec2 x:
tanm x sec2s x = tanm x(sec2s−2 x) sec2 x

= tanm x(tan2 +1)s−1 sec2 x.
This product yields a sum of products of powers of tan x and sec2 x; that is, the sum fits the form of
Example 3.15 on page 87. Hence,
Z Z
tan x sec x dx = tanm x(tan2 +1)s−1 sec2 x dx.
m 2s

Example 3.22. If m = 2s + 1 is odd in B., then we can factor tan x sec x and use the identity
tan2 x = sec2 x − 1:
tan2s+1 x secn x = (tan2 x)s secn−1 x(tan x sec x)

= (sec2 x − 1)s secn−1 x(tan x sec x).

This product yields a sum of products of powers of sec x and tan x sec x; that is, the sum fits the form of
Example 3.16 on page 87. Hence,
Z Z
2s+1
tan x sec x dx = (sec2 x − 1)s secn−1 x(tan x sec x) dx.
n


Z
tan3 x sec4 x dx,
we can either factor sec2 x or tan x sec x.
If we factor sec2 x, then
tan3 x sec4 x = tan3 x sec2 x sec2 x

= tan3 x(tan2 x + 1) sec2 x
= tan5 x sec2 x + sec2 x.
Hence,
Z Z Z
3 4 5 2
tan x sec x dx = tan x sec x dx + sec2 x dx
tan6 x
= + tan x + C.
6
If we factor tan x sec x, then
tan3 x sec4 x = tan2 x sec3 x(tan x sec x)

= (sec2 −1) sec3 x(tan x sec x)
= sec5 x(tan x sec x) − sec3 x(tan x sec x)
hence
Z Z Z
3 4 5
tan x sec x dx = sec x(tan x sec x) dx − sec3 x(tan x sec x) dx
sec6 x sec4 x
= − + C.
6 4
For the remaining cases, we use trigonometric identities to enable us to use the reduction formulas of
powers of sin x and sec x. See the reduction formulas in the previous section, in particular Example 3.11
on page 84 and Exercise 4 of the exercises on page 86.
Example 3.24. If m and n = 2s are even in A. on page 87, then we use the identity cos2 x = 1 − sin2 x:
sinm x cos2s x = sinm x(cos2 x)s = sinm x(1 − sin2 x)s .
This product yields a sum of powers of sin x. Each one of the terms of this sum can be integrated using
the reduction formula of sin x.

Example 3.25. If m = 2s is even in B. on page 87, then we use the identity tan2 x = sec2 x − 1 to
obtain a product yielding a sum of powers of sec x:
tan2s x secn x = (tan2 x)s secn x = (sec2 x − 1)s secn x.
Again, each one of the terms of this sum can be integrated using the reduction formula of sec x.

Z
tan5 x sec6 x dx,
we can apply either Example 3.21 or Example 3.22 on page 89. We then have
Z Z
a. 5 6
tan x sec x dx = tan5 x sec4 x sec2 x dx
Z
= tan5 x(tan2 +1)2 sec2 x dx
Z Z
= tan x sec x dx + 2 tan7 x sec2 x dx
9 2
Z
+ tan5 x sec2 x dx
tan10 x 2 tan8 x tan6 x
= + + +C
10 8 6
Z Z
b. tan x sec x dx = (sec2 −1)2 sec5 x(tan x sec x) dx
5 6
Z Z
= sec x(tan x sec x) dx − 2 sec7 x(tan x sec x) dx
9
Z
+ sec5 x(tan x sec x) dx
sec10 x 2 sec8 x sec6 x
= − + +C
10 8 6
Either answer is correct.
For an integral of the form

Z
tanm x dx,
we must use a combination of trigonometric identities and parts.

If m = 2s is even, then we have
tan2s x = (tan2 x)s = (sec2 x − 1)s ,
which yields a sum of terms, each of which is a power of sec x or a constant.
If m = 2s + 1 is odd, then we factor tan x and we have
tan2s+1 x = (tan2s x)(tan x)

= (tan2 x)s tan x = (sec2 x − 1)s tan x.
Example 3.27. By the reduction formula of sec x (Exercise 4 of the exercises on page 86)
Z Z
4
tan x dx = (sec2 x − 1)2 dx
Z
= sec4 x − 2 sec2 x + 1 dx
Z Z
tan x sec2 x 2
= + sec x dx − 2 sec2 x dx + x
2
3 3
tan x sec2 x 4 tan x
= − +x+C
3 3
Example 3.28. In each of the integrals below, we indicate the strategy we would use.
Z
a. sin5 x cos6 x dx as in Example 3.19 on page 88
Z
b. tan3 x sec x dx as in Example 3.22 on page 89
Z
c. tan x sec4 x dx as in Example 3.21 on page 89
Z
d. tan2 x sec4 dx as in Example 3.21 on page 89
or Example 3.25 on the previous page
In the next example, we illustrate the application of the integration of trigonometric functions to find the
area between curves.
Example 3.29. We want to find the area between the curves y = sin x and y = 2 sin2 x on the interval
[0, π/2].
We have to find the limits of integration; that is, solve the equality
π
sin x = 2 sin2 x for 0 ≤ x ≤
2

Thus,
sin x(1 − 2 sin x) = 0 ⇒ sin x = 0 or sin x = 1/2.

π
One solution is x = 0, and another is . We establish that on the interval [0, π/6] the function y = sin x
6
is above the function y = 2 sin x, and on the interval [π/6, π/2] the function y = 2 sin x is above the
2
function y = sin x. Thus, the area is given by the integral

Z π/6 Z π/2
2
sin x − 2 sin x dx + 2 sin2 x − sin x dx.
0 π/6
We apply the double angle identity and sin x − 2 sin2 x = sin x − 1 + cos(2x)
Z π/6
sin(2x) π/6
sin x − 1 + cos(2x) dx = − cos x − x +
0 2 0
√ √ √
3 π 3 3 π
=− − + +1=− − +1
2 6 4 4 6
We leave for you to check that
Z π/2
√
2 π π 3
2 sin x − sin x dx = − − .
π/6 2 6 4
√
π 3
So the area is 1 + − .
6 2
Other Useful Trigonometric Identities

To evaluate integrals of the form
Z
a. sin(mx) cos(nx) dx
Z
b. sin(mx) sin(nx) dx
Z
c. cos(mx) cos(nx) dx
We apply the corresponding trigonometric identities

1
sin A cos B = [sin(A − B) + sin(A + B)] (3.3)
2
1
sin A sin B = [cos(A − B) − cos(A + B)] (3.4)
2

1
cos A cos B = [cos(A − B) + cos(A + B)] (3.5)
2
These identities follow from the identities of the sum and difference of sine and cosine. We will deduce
the first equation and leave the others to you.
sin(A − B) + sin(A + B) = sin A cos B − cos A sin B + sin A cos B + cos A sin B
= 2 sin A cos B
and the Equation (3.3) follows.
Example 3.30. To evaluate the integral

Z
cos(3x) sin(7x) dx
we apply the equation 3.3 with A = 7x, B = 3x. Hence,

Z Z
1
sin(7x) cos(3x) dx = sin(4x) + sin(10x) dx
2

1 cos(4x) cos(10x)
=− + + C.
2 4 10
Exercises
6. Read examples 3.41–3.52 of Section 3.2 of the textbook, Power of Trigonometric Functions.
7. Do as many exercises as you can from section Exercises for 3.2 of the textbook.
8. Explain the strategy you would use to integrate

Z
cotm x cscn x dx
a. if n is even.
b. if m is even.
9. Explain the strategy you would use to integrate

Z
cotm x dx
a. if m is even.

b. if m is odd.
10. Evaluate the indefinite integrals listed below

Z
a. cot2 x csc2 x dx
Z
b. cot3 x dx
Z
c. [sin(3x) − sin(4x)] cos(3x) dx
11. For further practice, complete the lab titled Trigonometric Integrals in the optional Lyryx Labs.
Trigonometric Substitution
The technique of trigonometric substitution is used when the integrand function involves terms of the
form
√ √ √
a2 − x 2 , a2 + x2 or x 2 − a2 .
This method is a variation of the substitution method. Once the substitution is made, the integrand
function no longer has a square root and it becomes a trigonometric integral.
From the identity 1 − sin2 x = cos2 x, we have, for a > 0,
a2 − a2 sin2 x = a2 cos2 x,
√
and for the term a2 − x2 , the substitution suggested is x = a sin θ, so that
√ p
a2 − x2 = a2 − a2 sin2 x = a| cos θ|.
If we restrict the range of θ to the interval where cos θ is positive, that is

h π πi
− , ,
2 2
we have
√
a2 − x2 = a cos θ.
√
For the term x2 + a2 , we have the identity
a2 tan2 θ + a2 = a2 sec2 θ.
The substitution is x = a tan θ, and

√ p
x2 + a2 = a2 tan2 x + a2 = a| sec θ|.

For θ restricted to the interval where sec θ is positive, namely
π π
− , ,
2 2
we have
√
x2 + a2 = a sec θ.
√
For the term x2 − a2 , we have the identity
a2 sec2 θ − a2 = a2 tan2 θ.
The substitution is x = a sec θ, and

√ √
x2 − a2 = a2 sec2 x + a2 = a| tan θ|.
h π h 3π
For θ restricted to the interval where tan θ is positive, namely 0, and π, , we have
2 2
√
x2 − a2 = a tan θ.
Before applying this method, it is worthwhile to check if the integrand function is of the form shown in
Equation (3.2) on page 87. This strategy will save you work.

Z √
x x2 − 4
we have, for f (x) = x2 − 4, f 0 (x) = 2x.
Hence, this integral is of the form given in Equation (3.2), and

Z Z
1 √ 2 1
x − 4(2x) dx = f (x)1/2 f 0 (x) dx
2 2
f (x)3/2 (x2 − 4)3/2
= = + C.
3 3
In the next example, we see that trigonometric substitution is not necessary either.

Z
dx
√
5 − 4x − x2
we complete the perfect square first. Thus,
−(x2 + 4x − 5) = −(x2 + 4x + 4 + 4 − 5) = −(x + 2)2 + 1.

We make a substitution with u = x + 2, then du = dx. We set up the integral in terms of u
Z Z
dx du
√ = √ = sin−1 u.
5 − 4x − x 2 1 − u2
And
Z
dx
√ = sin−1 (x + 2).
5 − 4x − x 2
The next example shows that sometimes a preliminary substitution is necessary.

Z √
x 1 + x4 dx
We make a preliminary substitution u = x2 , then du = 2x dx. The integral in terms of u is

Z
1 √
1 + u2 du.
2
√
We apply the trigonometric substitution u = tan θ. Thus, du = sec2 θ and 1 + u2 = sec θ. The integral
in terms of θ is
Z
1
sec3 θ dθ.
2
We can apply a reduction integral as in the exercise on page 86
Z Z
1 3 tan θ sec θ 1
sec θ dθ = + sec θ dθ
2 4 4
tan θ sec θ ln | sec θ + tan θ|
= +
4 4
√ √
With this substitution sec θ = 1 + u2 = 1 + x4 and tan θ = x2 . Thus,
Z √ √ √
4
x2 1 + x4 ln | 1 + x4 + x2 |
x 1 + x dx = + + C.
4 4

Study the table in Section 3.3 of the section, Trigonometric Substitution, which summarizes the
trigonometric substitutions described in this section.

Exercises
12. Read examples 3.52–3.57 of Section 3.3 of the textbook, Trigonometric Substitution.
14. Integrate, for a > 0,

Z √ 2
x − a2
dx.
x
Hint: See integral 79 in the table of integrals of the textbook.
15. Use trigonometric substitution to integrate, for a > 0,

Z √ 2
x − a2
dx.
x2
Hint. See integral 80 in the table of integrals of the textbook.
16. For further practice, complete the lab titled Trigonometric Substitution in the optional Lyryx Labs.
Integration of Rational Functions by Partial Fractions

Recall that a rational function is of the form
P (x)
f (x) = ,
Q(x)
where P (x) and Q(x) are polynomial functions.
We already know how to integrate some rational functions, called partial fractions; in particular, we will
use the cases considered below.
Case 3.1.
Z
A A ln(ax + b)
dx = + C.
ax + b a
Case 3.2. For a positive integer k 6= 1, to integrate
Z
A
dx,
(ax + b)k
we use the substitution u = ax + b, and we obtain

Z Z
A A
k
dx = u−k du
(ax + b) a
Au1−k A(ax + b)1−k
= = + C.
a(1 − k) a(1 − k)

Case 3.3. For a positive integer k 6= 1, to integrate
Z
Ax
2
dx
(ax + b)k
we use the substitution u = ax2 + b. Then du = 2ax dx,

Z Z
Ax A
dx = u−k du
(ax2 + b)k 2a
Au1−k A(ax2 + b)(1−k)
= = + C,
2a(1 − k) 2a(1 − k)
and
Z Z
Ax A 1 A ln |u| A ln |ax2 + b|
2
dx = du = = + C.
ax + b 2a u 2a 2a
Case 3.4. For the integral
Z
A
2
dx,
ax + b
we have
Z √
A 1 A ax
2
dx = √ tan−1 √ + C.
a x + (b/a) ab b
Z
A
2
dx
ax + bx + c
with b2 − 4ac < 0 (i.e., the quadratic ax2 + bx + c cannot be factored—it is “irreducible”), we complete
the perfect square and use substitution.

Z
3
2
dx,
x +x+1
we complete the perfect square
2
2 1 3
(x + x + 1) = x+ + ,
2 4
and we use the substitution

1
u=x+ .
2

Hence, du = dx, and
Z Z
3 1
dx = 3 du
x2 + x + 1 3
u2 +
4
6 −1 2u
=√ tan √
3 3
 
1
6 2 x + 2 
−1 
=√ tan  √ 
3 3 

6 2x + 1
= √ tan−1 √ + C.
3 3
Z
Ax
2
dx
ax + bx + c
with b2 − 4ac < 0 (i.e., the quadratic ax2 + bx + c cannot be factored—it is “irreducible”), we complete
the perfect square and use substitution.

Z
3x
2
dx,
x +x+1
we complete the perfect square
2
2 1 3
(x + x + 1) = x+ + ,
2 4
and we use the substitution

1
u=x+ .
2
Hence
1
du = dx and x = u − .
2
Thus,
Z Z
3x u − 21
dx = 3 du
2
x +x+1 u2 + 34
Z Z
u 3 1
=3 2 3 − 2 3 du.
u +4 2 u + 4

The first integral is as in Case 3.3 on page 99 and the second as in Case 3.4 on page 99; therefore,
Z
3x
2
dx
x +x+1
1
!
3 6 2 x +
= ln |x2 + x + 1| + √ tan−1 √ 2 + C.
2 2 3 3
Rational functions are classified into “proper” and “improper” types.
P (x)
Definition 3.1. A rational function f (x) = is
Q(x)
a. proper if the degree of the polynomial P is less than the degree of the polynomial Q.
b. improper if the degree of the polynomial P is greater than or equal to the degree of the
polynomial Q.
If a rational function f (x) = P (x)/Q(x) is improper, then long division yields a decomposition into the
sum of a polynomial S(x) and a proper rational function R(x)/Q(x); that is,
P (x) R(x)
f (x) = = S(x) + .
Q(x) Q(x)
This fact indicates that to integrate rational functions, it is enough to integrate proper rational functions.
An important theorem in algebra says that any proper rational function can be expressed as a sum of
partial fractions of the form
A Ax + B
or
(ax + b)k (ax2 + bx + c)k
for an integer k ≥ 0 and b2 − 4ac < 0.
Therefore, an integral of a proper rational function can be reduced to a sum of integrals of the same form
as Cases 3.1 to 3.6 on pages 98–100.
Example 3.36. The rational function
4x3 + 6x2 + 3x + 1
(2x + 1)2 (x2 + x + 1)
is proper. Why?
We have two linear terms, (2x + 1) and (2x + 1)2 , to which we associate constants A and B, and one
quadratic term, x2 + x + 1, to which we associate a linear term Cx + D. Hence,
4x3 + 6x2 + 3x + 1 A B Cx + D
2 2
= + 2
+ 2 .
(2x + 1) (x + x + 1) 2x + 1 (2x + 1) x +x+1
Doing the operations yields

4x3 + 6x2 + 3x + 1
(2x + 1)2 (x2 + x + 1)
A(2x + 1)(x2 + x + 1) + B(x2 + x + 1) + (Cx + D)(2x + 1)2
= .
(2x + 1)2 (x2 + x + 1)
Doing the operations in the numerator gives us
4x3 + 6x2 + 3x + 1 = (2A + 4C)x3 + (3A + B + 4C + 4D)x2

+ (3A + B + 4D + C)x + (A + B + D).
Thus,
2A + 4C = 4 3A + B + 4C + 4D = 6
3A + B + 4D + C = 3 A + B + D = 1
You should verify that

2 1
A = 0, B= , C=1 and D= .
3 3
The integral of this function takes the form of Cases 3.2, 3.6 and 3.5; thus,
Z
4x3 + 6x2 + 3x + 1
(2x + 1)2 (x2 + x + 1)
Z Z Z
2 x dx
= 2
dx + 2
dx + 2
.
3(2x + 1) x +x+1 3(x + x + 1)
From these examples, the answer is

1 √
− + ln x2 + x + 1
3(2x + 1)

1 −1 2(x + 1/2) 2 −1 2x + 1
+ √ tan √ + √ tan √ + C.
3 3 3 3 3
Integration by Rationalizing Substitutions

p
An interesting case of the substitution method of integration is when we set u = g(x). Observe that in
this case
g 0 (x) g 0 (x)
du = p dx = dx ⇒ 2u du = g 0 (x) dx.
2 g(x) 2u

This p
property of the derivative of the square root allowp
us to integrate functions with the substitution
u = g(x) when the integrand function has the form g(x). The end result of the integrand function in
n
terms of u may be reduced to a rational function.
Example 3.37. Consider the integral

Z
dx
√ .
x− x+2
√
We make the substitution u = x + 2. Hence,
dx
x = u2 − 2, du = √ and 2u du = dx.
2 x+2
The integral in terms of u is
Z Z
2u u
du = 2 du.
(u2 − 2) − u u2 − u − 2
This is a proper rational function, and u2 − u − 2 = (u + 1)(u − 2). Its decomposition in partial fractions
is
u A B
= + .
(u + 1)(u − 2) u+1 u−2
Multiplying by (u + 1)(u − 2),
u = (A + B)u − 2A + B.
1 2
Solving for A, B, we get A = and B = . The integral is
3 3
Z Z Z
u 2 du 4 du
2 du = +
(u + 1)(u − 2) 3 u+1 3 u−2
2 ln |(u + 1)| 4 ln |(u − 2)|
= + .
3 3
The answer in terms of x is
√ √
2 ln |( x + 2 + 1)| 4 ln |( x + 2 − 2)|
+ + C.
3 3
Example 3.38. Consider the integral
Z
x
√ √ dx.
x− 3x
We have radicals √
of degrees 1/2 and 1/3. Their least common multiple of 2 and 3 is 6. So we make the
substitution u = 6 x. For this value of u we have
√ √ 1
x = u6 x = u3 3
x = u2 du = dx ⇒ 6u5 = dx.
6x5/6

Z
u11
6 du.
u3 − u2
This is an improper rational function. Making a long division we obtain
u11 8 7 6 5 4 3 2 u2
= u + u + u + u + u + u + u + u + 1 + .
u3 − u2 u3 − u2
u2 1
The proper fraction is3 2
is equal to . Thus,
u −u u−1
Z Z
u11 1
6 3 2
du = 6 u8 + u7 + u6 + u5 + u4 + u3 + u2 + u + 1 + du
u −u u−1
9
u u8 u7 u6 u5 u4 u3 u2
=6 + + + + + + + + u + ln |u − 1|
9 8 7 6 5 4 3 2

√ √ √ √
x x x 3 x x 6 x x x5/6 x2/3 x x1/3 √ √
6 + + + + + + + + x + ln | x − 1|
6 6
9 8 7 6 5 4 3 2
A further factorization yields

√ √ √ √ √ √
x 3
x 6
x 1 √
3
x 3
x 6
x 1
6x + + + +6 x + + +
9 8 7 6 5 4 3 2
√ √
+ 6 x + 6 ln | 6 x − 1| + C.
Exercises
17. Read examples 3.58–3.63 of Section 3.4 of the textbook, Rational Functions.
19. For further practice, complete the lab titled Partial Fractions in the optional Lyryx Labs.
Strategy for Integration

It takes practice and patience to integrate, because there is no clear technique or method. You need to
start by keeping in mind the basic integral formulas listed in the table of integrals of the textbook. Add to

these the following general forms, where the constant of integration has been omitted.
Z
(f (x))2
f (x)f 0 (x) dx = (3.6)
2
Z
(f (x))r+1
f (x)r f 0 (x) dx = for r 6= −1 (3.7)
r+1
Z 0
f (x)
dx = ln |f (x)| (3.8)
f (x)
Remember that we can always include a constant to force an integrand function to fit one of these forms.
Example 3.39. The integral

Z
e2x
dx
1 − e2x
is almost of the form shown in Formula (3.8), above. The derivative of the denominator function is
1 − e2x = −2e2x .
Hence, we have
Z
1 2e2x ln |1 − e2x |
− − dx = − + C.
2 1 − e2x 2
In the next examples, we show you how we manipulate the integrand function up to the point where is
easy to conclude without doing the integral fully.

Z
sin3 x
dx
cos7 x
We see that the derivative of cosine is negative sine; thus,
sin3 x sin x(sin2 x)

=
cos7 x cos7 x
sin x(1 − cos2 x)
=
cos7 x
sin x sin x
= +
cos x cos5 x
7
= cos−7 x sin x + cos−5 x sin x
We can apply the integral (3.7), above.

Z r
4−x
dx
x+4

We need to manipulate the integrand function so get rid of the square root in the numerator or
denominator. Thus,
4−x (4 − x)(4 − x) (4 − x)2

= = .
x+4 (x + 4)(4 − x) 16 − x2
The denominator is promising in this form since its square root suggest a trigonometric substitution.
r
(4 − x)2 4−x 4 x
2
=√ =√ −√ .
16 − x 16 − x 2 16 − x 2 16 − x2
These two functions can be integrated applying Example 2.7 on page 48 and the integral (3.7) above.

Z √
3
xe x dx
√ √
we can integrate eu and 3
x is a radical, so this suggest the substitution u = 3
x. Thus,
1
x = u3 du = dx ⇒ 3u2 du = dx.
3x2/3
Z
3 u5 eu du.
We can apply the reduction formula in Exercises 3 on page 86.
Exercises
20. Do the exercises in the six videos, Strategy for Integration. Videos Integration 1, Integral 2,
Integral 3, Integral 4, Integral 5, and Integral 6.
21. Evaluate the integrals listed below.

Z √
1 + ln x
a. dx.
x ln x
Z
ln(tan x)
b. dx.
sin x cos x
Hint. Manipulate the integrand function.
Z √
c. cos 6x dx.
Z
t3 + 1
d. dt.
2t3 − t2

Z
dx
e. dx.
ex − e−x
Z
ln(x + 1)
f. dx.
x2
Z
dx
g. .
1 + 2ex − e−x
Approximate Integration
Prerequisites
To complete this section, you must be able to state the definition of the definite integral.
There are other resources that enable us to integrate or approximate the integral of elementary functions:
1. the table of integrals in the Additional Material section of the textbook
2. computer algebra systems
3. the Midpoint Rule in Riemann Sums section of the textbook
4. the Trapezoidal Rule in the Numerical Integration section of the textbook
5. Simpson’s Rule in Numerical Integration section of the textbook.
You may wonder why we need to learn techniques of integration if there are integration tables that we can
and should use. The answer lies in the fact that not all integrals fit one of the integrals in the tables. An
integral may have a resemblance or similarity to one of the known integrals, but we need to be able to
make justifiable changes to make the formula fit exactly. It may also be the case that the integral is not
similar to any of the integrals of the table. In either case, the techniques of integration help us to
manipulate integrals and to make good use of the tables, or when this is not possible, to use our
understanding of antiderivatives. Be aware also that not all integration tables are correct: they may
contain mistakes we should be able to correct.
We cover only the last three rules in this course. In more advanced courses, you may want or need to use
the first two options.
The approximations presented in this unit are consequences of the definition of the definite integral:
Z b X
n
f (x) dx = lim f (x∗i )∆xi
a n→∞
i=1
where
• the function f is continuous on [a, b].

• the interval [a, b] is divided in subintervals a = x0 < x1 < · · · < xn = b.
• the number x∗i is in the subinterval [xi−1 , xi ]; that is, xi−1 < xi ∗ < xi for i = 1, . . . n.

• ∆xi = xi − xi−1 is the length of the subinterval [xi−1 , xi ] for i = 1, . . . n.
If ∆xi is constant for every i, that is, ∆xi = ∆x for all i, then
Z b X
n
f (x) dx ≈ f (x∗i )∆x (3.9)
a i=1
= ∆x[f (x1 ∗) + f (x2 ∗) + · · · + f (xn ∗)].
The larger the n, the better the approximation.
For the Midpoint Rule, we take

b−a 1
∆x = and xi ∗ = x̄i = (xi−1 + xi ),
n 2
the midpoint of the subinterval [xi−1 , xi ].
Hence
Z b
f (x) dx ≈ Mn = ∆x[f (x̄1 ) + f (x̄2 ) + · · · + f (x̄n )].
a
If, in Equation (3.9), above, we take xi ∗ = xi , the left endpoint of the interval [xi−1 , xi ], then,
Z b X
n
f (x) dx ≈ Ln = f (xi−1 )∆x.
a i=1
If we take xi ∗ = xi , the right endpoint of the interval [xi−1 , xi ], then,

Z b X
n
f (x) dx ≈ Rn = f (xi )∆x.
a i=1
The Trapezoidal Rule results from averaging the two approximations Ln and Rn and by taking
(b − a)
∆x = and xi ∗ = a + i∆x.
n
Hence,
Z b
∆x
f (x) dx ≈ Tn = [f (x0 ) + 2f (x1 ) + · · · + 2f (xn−1 ) + f (xn )].
a 2
You will see the details in the exercises of this unit.
If EM is the error in the Midpoint Rule, then the distance between the actual value of the definite integral
and the approximation obtained is EM . The error in the Trapezoidal Rule, ET , is similar. The
approximation error for each of these rules uses the second derivative of the function.

If |f 00 (x)| ≤ K for a ≤ x ≤ b, then
K(b − a)3 K(b − a)3

|ET | ≤ and |EM | = .
12n 24n2
Simpson’s Rule tends to be more exact than the Midpoint and Trapezoidal Rules, and the idea is to
approximate a curve y = f (x) on the interval [a, b] with parabolas on subintervals [xi−1 , xi ] of the
interval [a, b]. In this case, we must divide the interval [a, b] into an even number of subintervals; that is, n
is even, and we take
b−a
∆x = .
n
Hence,
Z b
∆x
f (x) dx ≈ Sn = [f (x0 ) + 4f (x1 ) + 2f (x3 ) + 4f (x4 )+
a 3
· · · + 2f (xn−2 ) + 4f (xn−1 ) + f (xn )].
The error in the approximation ES uses the fourth derivative of the function.
If
|f (4) (x)| ≤ K for a ≤ x ≤ b,
then
K(b − a)5
ES ≤ .
180n4
Summarizing the Approximations

b−a xi−1 − xi
Mid- For ∆x = and x̄i = the midpoint of the interval [xi−1 , xi ]
n 2
point
Rule. Z b
f (x) dx ≈ Mn = ∆x[f (x̄1 ) + f (x̄2 ) + · · · + f (x̄n )].
a
b−a
Trape- For ∆x = and xi = a + i∆x
n
zoidal
Rule. Z b
∆x
f (x) dx ≈ Tn = [f (x0 ) + 2f (x1 ) + 2f (x2 ) + · · · + 2f (xn−1 ) + f (xn )].
a 2

b−a
Simp- For n even and ∆x =
n
son’s
Rule. Z b
f (x) dx ≈ Sn
a
∆x
= [f (x0 ) + 4f (x1 ) + 2f (x2 ) + 4f (x3 ) + · · · + 2f (xn−2 ) + 4f (xn−1 ) + f (xn )].
3
Error Bounds
Suppose |f 00 (x)| ≤ K for a ≤ x ≤ b. If ET and EM are the errors of the Trapezoidal and Midpoint Rules,
then
K(b − a)3 K(b − a)3

|ET | ≤ and |EM | ≤ .
12n2 24n2
Suppose that |f ( 4)(x)| ≤ K for a ≤ x ≤ b. If ES is the error of the Simpson’s Rule, then
K(b − a)5
|ES | ≤ .
180n4
Exercises
22. Read Section 3.5 of the textbook, Riemann Sums.
24. Read Section 3.6 of the textbook, Numerical Integral of the textbook.
Improper Integrals
Prerequisites
To complete this section, you must be able to state and use the Fundamental Theorem of Calculus.
The Fundamental Theorem of Calculus states that for a continuous function f on a closed interval [a, b]
the definite integral is a number,
Z b
f (x) dx = F (b) − F (a),
a
where F 0 (x) = f (x). Such integrals are called proper integrals.

Definition 3.2. An integral on an interval J of a function f is improper if either the interval J is
infinite or the function f is discontinuous at finitely many values of J.
Improper integrals are limits, and we need to understand the type of limits they are.
Definition 3.3. Improper integrals of Type 1

Z t
a. If f (x) dx exists for every number t ≥ a, then
a
Z ∞ Z t
f (x) dx = lim f (x) dx
a t→∞ a
if this limit exists as a finite number.

Z b
b. If f (x) dx exists for every number t ≤ b, then
t
Z b Z b
f (x) dx = lim f (x) dx
−∞ t→−∞ t
The improper integrals

Z ∞ Z c
f (x) dx and f (x) dx
c −∞
are convergent if the corresponding limits are finite, and divergent if they either do not exist or
are (positive or negative) infinity.
c. If both
Z ∞ Z c
c −∞
are convergent, then we define

Z ∞ Z c Z ∞
f (x) dx = f (x) dx + f (x) dx
−∞ −∞ c
for any real number c.

Z ∞
cos x dx
0

is an improper integral of Type 1, and
Z ∞ Z t t

cos x dx = lim cos x dx = lim sin x = lim sin t.
0 t→∞ 0 t→∞ 0 t→∞
This limit does not exist; therefore, the given integral is divergent.

Z ∞
ex dx
1
is an improper integral of Type 1, and

Z ∞ Z t t
x
e dx = lim ex dx = lim ex = lim et − e = ∞.
1 t→∞ 1 t→∞ 1 t→∞
Thus, the integral is divergent.

Z ∞
dx
−
0 1 + x2
is an improper integral of Type 1,
Z ∞ Z t
dx dx
− = lim −
0 1 + x2 t→∞ 0 1 + x2
t π π

= lim cot−1 x = lim cot−1 t − = −
t→∞ 0 t→∞ 2 2
and the integral is convergent. Similarly,
Z 0 Z 0
dx dx
− 2
= lim −
−∞ 1+x t→−∞ t 1 + x2
0 π π
−1
= lim cot x = lim − cot−1 t = − .
t→−∞ t t→−∞ 2 2
Both integrals are convergent; hence,
Z ∞ Z 0 Z ∞
dx dx dx π π
− 2
= − 2
+ − 2
= − − = −π.
−∞ 1+x −∞ 1+x 0 1+x 2 2
Definition 3.4. Improper integrals of Type 2.
a. If f is continuous on [a, b) and discontinuous at b, then

Z b Z t
f (x) dx = lim− f (x) dx,
a t→b a

b. If f is continuous on (a, b] and discontinuous at a, then
Z b Z b
f (x) dx = lim+ f (x) dx,
a t→a t
The improper integral

Z b
f (x) dx
a
is convergent if the corresponding limit is finite, and divergent if the limit either does not exist or
is (positive or negative) infinity.
c. If f is discontinuous at c, a < c < b, and both

Z c Z b
a c
are convergent, then

Z b Z c Z b
f (x) dx = f (x) dx + f (x) dx.
a a c

Z 1
ln x dx
0
is an improper integral of Type 2, since ln x is discontinuous at 0. Thus,

Z 1 Z 1
ln x dx = lim+ ln x dx
0 t→0 t
1

= lim+ x ln x − x = lim+ −1 − t ln t + t = −1
t→0 t t→0
Note that we use l’Hopital’s Rule to evaluate the limit

ln t
lim+ t ln t = lim+ = lim −t = 0.
t→0 t→0 1/t t→0+
The integral is convergent.

Z 6
dx
dx
0 x−3
1
is an improper integral of Type 2, since is discontinuous at x = 3.
x−3

Thus, we apply part c. of Definition 3.4 on page 112
Z 6 Z 3 Z 6
dx dx dx
dx = dx + dx
0 x−3 0 x−3 3 x−3
and the limits are

Z 3 Z t
dx dx
dx = lim− dx
0 x−3 t→3 0 x−3
t

= lim− ln |x − 3| = lim− ln |t − 3| − ln | − 3|
t→3 0 t→3
t − 3

= lim− ln = −∞
t→3 3
Z 6
dx
You may want to check that the limit dx is infinity. Hence, the integral is divergent.
3 x−3

Z 3
dx
2
−3 x + x − 6
is an improper integral of Type 2 because it is discontinuous at −3 and 2. Thus, we need to consider three
improper integrals:
Z 0 Z 2 Z 3
dx dx dx
2
+ 2
+ 2
.
−3 x + x − 6 0 x +x−6 2 x +x−6
Each of these improper integrals is a limit.
Since we have
Z Z
dx 1 1 ln |x + 3|
2
= − dx = ln |x − 2| − ,
x +x−6 x − 2 5(x + 3) 5
we apply the properties of limits and obtain

Z 0 Z 0
dx 1 1
2
= lim − dx.
−3 x + x − 6 t→−3+ t x − 2 5(x + 3)
But
Z 0 0
1
lim+ dx = lim+ ln |x − 2|
t→−3 t x−2 t→−3 −3
= lim+ ln(2) − ln |t − 2| = ln(2) − ln(5)

t→−3

and
Z 0 0
1 1
lim dx = lim+ ln |x + 3|
t→−3+ t 5(x + 3) t→−3 5 t
1
= lim+ (ln(3) − ln(t + 3)) = ∞.
t→−3 5
Hence,
Z 0
dx
= −∞,
−3 x2 + x − 6
and the improper integral is divergent.

Z ∞
dx
0 x
is an improper integral of both Type 1 and Type 2, because it is over an infinite interval and it is not
continuous at 0. Hence, we consider two improper integrals:
Z ∞ Z 1 Z ∞
dx dx dx
= + .
0 x 0 x 1 x
From Example 3.81 of Section 3.7, Improper Integrals, of the textbook we conclude that the improper
integral is divergent.
Erroneous Evaluations
Warning: It is important that you understand that improper integrals are limits and that you must state
this fact clearly. Do not skip steps in the evaluation of improper integrals, and above all, do not be
“creative” about notation. All of the following erroneous “evaluations,” or evaluations that are similar to
them, will receive no credit, even if the answer is correct.
Z ∞ ∞ 1
dx 1 1 π
1. 2
= tan (x/2) = tan−1 (0) + tan−1 (∞) = .
−1
0 x +4 2 0 2 2 4
There are three errors in this evaluation.
• The improper integral is a limit; thus, it should be

Z ∞ Z t
dx dx
2
= lim 2
.
0 x +4 t→∞ 0 x +4
• The symbol ∞ is not a number, but an upper limit of integration must be a number. The
expression
1 ∞

tan (x/2) does not make any sense.
−1
2 0

• For the same reason, we do not evaluate a function at a symbol, only at numbers, and
1
tan−1 (∞) is mathematically incorrect.
2
Z ∞ Z 0 Z ∞
dx dx dx
2. 2
= 2
+ 2
−∞ x +4 −∞ x + 4 0 x +4
1 0 ∞
−1 x 1 −1 x
= tan + tan
2 2 −∞ 2 2 0
1 1
= − tan−1 (−∞) + tan−1 (∞) = π.
2 2
This is a combination of the previous errors.
3. The following example is even worse.

Z ∞ ∞
dx 1 −1
2
= tan (x/2)
−∞ x + 4 2 −∞
1 1
= tan−1 (∞) − tan−1 (−∞) = π.
2 2
4. It is not true that

Z ∞ Z t
f (x) dx = lim f (x) dx,
−∞ t→∞ −t
so never try to do this.

Z 1 1
dx
5. = ln |x| = ln(1) − ln(0) = ∞.
0 x 0
Here we have two errors, the limit of the improper integral is not stated, and it is not true that
ln(0) = −∞. Because 0 is not in the domain of ln x.
6. If we fail to notice the discontinuities of the integrand functions, we may mistakenly treat an
improper integral as if it were a definite integral. For example, it is not true that
Z 6 6
dx
= ln |x − 3| = ln |3| − ln | − 3| = 0.
0 x−3 0
We know from Example 3.47 on page 113 that this integral is divergent.
Comparison Test for Improper Integrals

The Comparison Test for improper integrals of type 1 is stated in Section 3.7, Improper Integrals, of your
textbook together with a plausible explanation of its veracity. We will re-state here for reference.

Theorem 3.5. If f (x) and g(x) are continuous functions on an interval [a, ∞), and
f (x) ≥ g(x) ≥ 0 on this interval,
then
Z ∞ Z ∞
a. If f (x) dx is convergent, then g(x) dx is convergent.
a a
Z ∞ Z ∞
b. If g(x) dx is divergent, then f (x) dx is divergent.
a a
A similar theorem is true for improper integrals of type 1.
Theorem 3.6. If f (x) and g(x) are continuous functions on an interval (a, b], and
f (x) ≥ g(x) ≥ 0 on this interval,
then
Z b Z b
a. If f (x) dx is convergent, then g(x) dx is convergent.
a a
Z b Z b
b. If g(x) dx is divergent, then f (x) dx is divergent.
a a
Pay attention to the inequalities when applying this theorem. The reverse of this theorem it is not
necessarily
Z true. Z Z
∞ ∞ ∞
If g(x) dx is convergent, the limit f (x) dx may or may not be convergent, and if f (x) dx is
a Z ∞ a a
divergent, g(x) dx may or may not be divergent.

a
In the next examples we show you how we apply these theorems.
Example 3.50. The exponential function ex is increasing and for x ≥ 1 we have that x ≤ x2 ; hence,
2
−x ≥ −x2 and therefore e−x ≥ e−x . The integral
Z ∞
2
e−x dx
1
is improper of type 1. Moreover,

Z ∞ Z t t
−x −x −x
e dx = lim e dx = lim −e
0 t→∞ 1 t→∞ 1
−t −1
= lim −e + e−1 = e
t→∞
Z ∞ Z ∞
2
The integral e −x
dx is convergent and by the comparison test the integral e−x dx is also
1 1
convergent.

2
The function e−x is continuous on the interval [0,1]. Therefore, the integral
Z ∞
2
e−x dx
0
is convergent.

Z π/2
dx
0 x sin x
if improper of type 2. And for 0 < x ≤ π/2 we have 0 < sin x ≤ 1. Thus,
1 1
0 < x sin x ≤ x ⇒ ≤ .
x x sin x
To apply the comparison test we need to show that the integral

Z π/2
dx
0 x
is divergent.
Z π/2 Z π/2
dx dx
= lim+ = lim+ ln(π/2) − ln t = ∞.
0 x t→0 t x t→0
So the integral
Z π/2
dx
0 x sin x
is also divergent.
Exercises
24. Read examples 3.78–3.89 of Section 3.7 of the textbook, Improper Integrals.
26. Do as many exercises as you can from the section 3.8 Additional Exercises of the textbook.
27. For further practice, complete the lab Approximation of Integrals in the optional Lyryx Labs.
28. For further practice, complete the lab Improper Integrals in the optional Lyryx Labs.
29. Evaluate the integral

Z ∞
dx
√ .
0 x(1 + x)

Z ∞
30. a. Show that the improper integral x dx is divergent.
−∞
b. Show that
Z t
lim x dx = 0.
x→∞ −t
Note. This shows you that we cannot define

Z ∞ Z t
x dx = lim x dx.
−∞ x→∞ −t
Finishing This Unit

examination.
5. Complete the second assignments and submit your work to your tutor for grading.
Remember: The number of points in a question may indicate the number of steps in the solution.

Time: 3 hours
Total points: 64
Passing grade: 55%
1. (8 points)
a. Use integration by parts to integrate for a, b > 0

Z √
x a + bx dx.
b. Use a method other than integration by parts to integrate the integral in part (a), above.
2. (10 points) Complete the perfect square to integrate the integrals below. If you use an integral from
the table of integrals, identify it by number, and indicate how you are applying it.
Z √
a. 3x2 + 2x dx
Z √
b. x2 x2 + 2x + 5 dx
3. (10 points) Integrate the trigonometric integrals below. If you use an integral from the table of
integrals, identify it by number, and indicate how you are applying it.
Z
a. sin3 x cos(2x) dx
Z
b. tan5 x sec3 x dx
4. (10 points) Use any method to integrate the integrals below. If you use an integral from the table of
integrals, identify it by number, and indicate how you are applying it.
Z √
x
a. √ dx
1+ 3x
Z
sec x cos(2x)
b. dx
sin x + sec x
5. (6 points) Use Simpson’s Rule with n = 6 subintervals to approximate the integral
Z π/2 p
1 + sin2 x dx.
−π/2
Express your answer to at least four decimal places.
6. (6 points)

a. Explain why the integral
Z 2
dx
3
−1 x
is improper.
b. Determine whether the integral in part (a) is convergent or divergent. Evaluate it if it is

convergent, if it is not, explain why.
7. (5 points) Use the Comparison Theorem to determine whether the integral

Z ∞
dx
√ dx
1 1 + x4
is convergent or divergent.
8. (4 points) Explain why Simpson’s Rule is more accurate than the Trapezoidal Rule.
9. (5 points) Evaluate the integral

Z ∞
ln x
dx.
1 x3

UNIT 4
Applications of Integration
The applications presented in this unit are consequences of the definition of the definite integral
Z b X
n
f (x) dx = lim f (x∗i )∆xi
a n→∞
i=1
where
• the function f is continuous on [a, b].

• the interval [a, b] is divided in subintervals a = x0 < x1 < · · · < xn = b.
• the numbers x∗i is in the subinterval [xi−1 , xi ]; that is, xi−1 < xi ∗ < xi for i = 1, . . . , n.
• ∆xi = xi − xi−1 is the length of the subinterval [xi−1 , xi ] for i = 1, . . . , n.
Since the definite integral is the limit of a sum and the differential ∆x is very small, we can, in a loose
way, interpret the definite integral as an infinite “sum” of f (x) ∆x for very small ∆x, with x running
through the interval [a, b]. In all the applications presented, we aim to represent what we are looking for
in this form.
Objectives
1. find the volume of a solid generated by a region rotated about a vertical line using the disk method or
cylindrical shells.
2. find the volume of a solid generated by a region rotated about a horizontal line using the disk method
or cylindrical shells.
3. find the volume of a solid using its cross section.
4. find the arc length of a curve.
5. find the area of a surface of revolution.
6. find the center of mass of a flat plate.

Volumes
To find the volume of a solid, we can consider the sum of areas of the slices that form the volume, as
shown in Figure 4.19: Volume of a pyramid approximated by rectangular prisms.
Definition 4.1. If the solid S lies between x = a and x = b and has a cross-sectional area A(x) on
the plane P that passes through x and is perpendicular to the x-axis, then its volume is given by the
definite integral
Z b X
n
A(x) dx = lim A(xi ∗)∆x,
a n→∞
i=1
whenever A(x) is continuous on the interval [a, b].
Similarly, if we consider a solid S that lies between y = c and y = d, and has a cross-sectional area
A(y) on the plane P that passes through y and is perpendicular to the y-axis, then its volume is given
by the definite integral
Z d X
n
A(y) dy = lim A(yi ∗)∆y,
c n→∞
i=1
whenever A(y) is continuous on the interval [c, d].
Example 4.1. Let us consider cross sections on planes perpendicular to the y-axis. Imagine a cone
obtained by rotating a line around the y-axis (see Figure 4.1, below).
Figure 4.1: Line hr to be rotated around the y-axis
The base of the cone is at y = 0, and the tip is at y = h. The cone can be thought of as a series of circles,
each of which has a radius of
r(y − h)
R=− ,
h
as indicated in Figure 4.1.
Thus, the area of each circle is
πr2 (y − h)2
A(y) = πR2 =
h2
and the volume of the cone is
Z
πr2 h 2 πr2 y 3 2 2 h πr2 h
(y − h) dy = − y h + h y = .
h2 0 h2 3 0 3
If the solid is generated by rotating a region R around the x axis, then the cross-sectional areas on a plane
perpendicular to the x-axis are disks or washers—see Figure 4.21: A solid of rotation; Figure 4.22: A
region that generates a cone, approximating the volume by circular disks; and Figure 4.23: Solid with a
hole, showing the outer cone and the shape to be removed to form a hole.
To find the volume of this solid, we must find the areas of the cross sections. To do so, we make a sketch
of the solid and one cross section, and find the radius of the disk, or the outer and inner radii of the
washer.
Example 4.2. Consider the region R in Figure 4.2, below.
Figure 4.2: Region R to be rotated around the x-axis (Example 4.2)
If we rotate this region around the x-axis, we have a washer. The outer radius is given by the second
coordinate of (a, b) and the inner radius is given by the second coordinate of (c, d). The area of the
washer is
A = πb2 − πd2 = π(b2 − d2 ).
The same is true if a region is rotated around the y-axis. The cross-sectional areas on a plane
perpendicular to the y-axis are disks or washers, as shown in Figure 4.3 below
Example 4.3. If we rotate the region R of Figure 4.4, below, around the y-axis we also have a washer.
The outer radius is the first coordinate of (c, d) and the inner radius is the first coordinate of (a, b). The
area of the washer is A = πc2 − πa2 .
Example 4.4. See the region R3 , below. If we rotate this region around the x axis, a washer is produced
by a cross section perpendicular to the x-axis. See Figure 4.5
√
The outer radius is x and the inner radius is x3 ; hence, the area of the washer is
√
A(x) = π(( x)2 − (x3 )2 ) = π(x − x6 ),

∆𝑦
(𝑥, 𝑦)
Figure 4.3: Region R rotating around the y-axis, creating a disk of radius x and ∆y thickness
Figure 4.4: Region R to be rotated around the y-axis (Example 4.3)
and the volume is given by the definite integral

Z 1
π x − x6 dx.
0
If we rotate this same region around the y axis to create a washer, we take a cross section perpendicular to
√
the y-axis. The outer radius is 3 y and the inner radius is y 2 (see Figure 4.5, above). The area of this
washer is
A(y) = π(y 2/3 − y 4 ),

Figure 4.5: Region R3 with perpendicular lines to the x and y axis.
so the volume is given by the definite integral

Z 1
π y 2/3 − y 4 dy.
0
The same method applies to regions rotated around any vertical or horizontal lines.
Example 4.5. If we rotate the region R3 of Example 4.4 around the horizontal
√ line y = −1, then the
cross section perpendicular to this line gives a washer with outer radius x + 1 and inner radius x3 + 1.
Why?
The volume of the solid is

Z 1
√
π ( x + 1)2 − (x3 + 1)2 dx.
0
If we rotate the region R3 around the vertical line x = 2, then the cross section perpendicular to this line
√
gives a washer with outer radius 2 − y 2 and inner radius 2 − 3 y and the volume is
Z 1
√
π (2 − y 2 )2 − (2 − 3
y)2 dy.
0
Example 4.6. You can verify that the the volume of the region R3 rotated around the line x = −1 is
given by the integral
Z 1
√
π ( 3 y + 1)2 − (y 2 + 1)2 dy.
0
Remarks
1. Observe that the disk or washer method requires an integral in terms of x if the region is rotated
around a horizontal line, and in terms of y if the region is rotated around a vertical line.

2. It is good practice to sketch the region and give the coordinates of the endpoint or endpoints of the
segment that generates the disk or the washer.
Exercises
1. Set up the integral (do not integrate) that gives the volume of the region bounded by y = cos x,
x = 0, x = π/2 around the y-axis.
2. Find the volume of the region bounded by y = sin x, x = 0, x = π around the line y = −1
3. For further practice, complete the lab Volumes—Disks in the optional Lyryx Labs.
Volumes by Cylindrical Shells

In Exercise 3 at the end of the previous section, we found that, in order to apply the disk method to find
the volume of the solid generating by rotating the region bounded by y = cos x, x = 0, x = π/2 around
the y-axis, we must consider the function x = cos−1 x. However, we cannot always solve for x from a
function y = f (x), or for y from a function x = g(y). In these cases, we may be able to use the
cylindrical shell method.
If we consider a solid generated by a region R rotated around the y-axis, a cross section on a plane
parallel to the y-axis gives a vertical line on the xy-plane. See Figure 4.24 of the textbook: Computing
volumes with “shells.”
This line, rotated around the y-axis, gives a cylindrical shell. The area A of this shell is
A = π(2r)h
where r is the radius and h is the height of the cylinder. To see how this works, examine Figure 4.6, below
Figure 4.6: Cylinder of height h, radius r and surface area A = h × 2πr (not to scale)
If we “cut” the cylindrical shell and flatten it, the area corresponds to the rectangle with dimensions
π(2r) (circumference) and h the height of the cylinder.
If this shell is very thin, with a thickness of ∆x, then the volume v of the cylindrical shell is
v = 2πrh∆x.

The radius and height for a function y = f (x) corresponds to r = x and h = f (x), respectively, as shown
in Figure 4.7, below.
Figure 4.7: Cylinder of height f (x), radius x, and area A = f (x) × 2πx
The volume of the solid V is approximately the sum of the volume of all of the very thin cylindrical
shells that make up the solid; therefore,
X
n
V ≈ 2πxf (x)∆x.
i=1
If the number of cylindrical shells is very large—that is, if we have infinitely many of them—then we
have a definite integral, and the volume is
Z b X
n
V = 2πxf (x) dx = lim 2πxf (x)∆x.
a n→∞
i=1
Similarly, for a solid generated by a region R rotated around the x-axis, a cross section on a plane
parallel to the x-axis gives a horizontal line on the xy-plane. See Figure 4.8, below.
The volume of the cylindrical shell in such a case is
(circumference)(length)(thickness) = (2πy)(g(y)(∆y)
(see Figure 4.8, below), and the volume of the solid is

Z d X
n
V = 2πyg(y) dy = lim 2πyg(y)∆y.
c n→∞
i=1
Example 4.7. The volume of the region bounded by y = cos x, x = 0, x = π/2 around the y-axis is
Z π/2
2π x cos x dx.
0

Figure 4.8: Cylinder of length g(y), radius y, and area A = g(y) × 2πy
Integrating by parts, we have

Z π/2 π/2 Z π/2

2π x cos x dx = 2πx sin x − 2π sin x dx = π 2 − 2π.
0 0 0
You can see that with the shell method the resulting integral is easier than with the disk method.
Example 4.8. Consider the solid generated by rotating the region bounded by x = 2y − y 2 and x = 0
around the x-axis (see Figure 4.9, below).
Figure 4.9: Solid generated by rotation of the region bounded by x = 2y − y 2 and x = 0 about the x-axis
With respect to the point (2y − y 2 , y) on the curve, the circumference of the cylinder is 2πy (the second
coordinate), and the height (length) is 2y − y 2 (the first coordinate); hence, the volume is
Z 2
V = 2π (2y − y 2 )y dy.
0
8π
You can verify that this integral is equal to .
3
Example 4.9. Consider the region bounded by the line x + y = 3 and the parabola x = 4 − (y − 1)2
(make a sketch of this region). We want to find the volume of the solid generated by rotating this region
around the x-axis, using the shell method.

The points of intersection of the line and the parabola, (0, 3) and (3, 0), are found by solving the equation
3 − y = 4 − (y − 1)2 . Hence, the range of y is the interval [0, 3].
Now we take a horizontal line in the region, and its endpoints are (3 − y, y) and (4 − (y − 1)2 , y). This
line, rotated around the x-axis, generates a cylinder with radius 2πy and height
4 − (y − 1)2 − (3 − y) = 3y − y 2 . The volume is given by the integral
Z 3
2π (3y − y 2 )y dy.
0
27π
You can verify that the volume is .
2
If this same region is rotated around the line y = −2, then the same horizontal line rotated around this
line gives a cylinder with radius y + 2 and height 3y − y 2 . Hence, the volume is
Z 3
63π
2π (3y − y 2 )(y + 2) dy = .
0 2
If we were to find the volume of the solid of the same region by rotating it around the y-axis, then the
shell method is not efficient.Why?
But the washer method is. The outer radius of the washer is 4 − (y − 1)2 , and the inner radius is 3 − y.
Hence, the volume is
Z 3 Z 3
2 2 2 108π
π (4 − (y − 1) ) − (3 − y) dy = π 18y − 3y 2 − 4y 3 + y 4 dy = .
0 0 5
Remarks
1. Observe that the shell method requires an integral in terms of x, if the region is rotated around a
vertical line, and in terms of y if the region is rotated around a horizontal line.
2. It is good practice to sketch the region and give the coordinates of the endpoint or endpoints of the
segment that generates the cylinder.
3. If we are able to apply either method, disk or shell, we choose the one that gives us the integral that
is easiest to evaluate.
Exercises
3. Read the examples in Section 4.1 of the textbook, Volume.
5. Do the exercises in the four videos, Volumes of Revolution. Videos Volume 1, Volume 2, Volume 3,
and Volume 4.
6. For further practice, complete the lab Volumes—Shells in the optional Lyryx Labs.

Arc Length
Prerequisites
To complete this section, you must be able to apply the Mean Value Theorem.
The next application of the integral is the evaluation of the length of a smooth curve y = f (x) on an
interval [a, b]. The idea is to consider the lengths of straight segments joining two different points on the
curve, as shown in Figure 4.27: Approximating arc length with line segments of the textbook. The sum of
all of these lengths is approximately equal to the length of the curve.
Let any two points on the curve be
Pi = (xi , f (xi )) and Pi+1 = (xi+1 , f (xi+1 ));
then, the length of the segment Pi Pi+1 is

p
Li = (xi+1 − xi )2 + (f (xi+1 ) − f (xi ))2 (4.1)
Since we can take (xi+1 − xi ) to be very small, we have the differential
∆xi = (xi+1 − xi ),
and we use the Mean Value Theorem on the interval
[xi , xi+1 ].
Hence, there is an xi < xi ∗ < xx+1 such that
f (xi+1 ) − f (xi ) = f (xi ∗)(xi+1 − xi ) = f (xi ∗)∆xi .
So, from Equation (4.1),

q p
Li = ∆x2i + f (xi ∗)2 ∆x2i = 1 + f (xi ∗)2 ∆xi
Hence, the length L of the curve is approximately equal to the sum of all these segments:
X
n
p
L≈ 1 + f (xi ∗)2 ∆xi .
i=1
For infinitely many of these segments, we have

X
n
p Z bp
L = lim 1 + f (xi ∗)2 ∆x i = 1 + f (x)2 dx.
n→∞ a
i=1

Similarly, for a smooth curve x = g(y) on an interval [c, d], the length of the curve is
Z d p
L= 1 + g(y)2 dy.
c
Example 4.10. The length of the arc of y = sin x on [0, π] is given by the definite integral
s 2
Z π Z π√
d
1+ sin x dx = 1 + cos2 x dx
0 dx 0
We will not try to evaluate this integral.
If a curve y = f (x) is smooth on the interval [a, b], then it is smooth on the interval [a, x] for any
a < x < b. The length of the curve on this interval [a, x] gives the arc length function s(x), and
Z xp
s(x) = 1 + f 0 (t)2 dt.
a
Example 4.11. The function y = ln(cos x) is continuous on [0, π/2), and the arc length function is
defined for 0 ≤ x < π/2. It is given by
Z xp
s(x) = 1 + tan2 x dx = ln | sec x + tan x|.
0
The length of this curve on the interval [0, π/4] is

π π π √

s = ln sec + tan = ln | 2 + 1|
4 4 4
and on the interval [0, π/3] is
π π π √

s = ln sec + tan = ln |2 + 3|.
3 3 3
Note: The interval for the arc length function may be infinite; that is, if a curve y = f (x) is smooth on
the interval [a, ∞), then it is smooth in the interval [a, x] for any a < x. The length of the curve on this
interval [a, x] gives the arc length function s(x) defined for any a < x:
Z xp
s(x) = 1 + f 0 (t)2 dt.
a
Exercises
7. Read Section 4.2 of the textbook, Arc Length.
9. For further practice, complete the lab Arc Length in the optional Lyryx Labs.

Area of a Surface of Revolution
Prerequisites
To complete this section, you must be able to apply the Mean Value Theorem.
Our next application of the integral is to find the surface area of a solid generated by rotating a region in
the xy-plane around the x-axis or y-axis. In this case, we add the areas of thin bands around the solid, as
shown in the Figure 4.10, below.
𝑃𝑖+1
𝑃𝑖
Figure 4.10: Solid showing a band around the x-axis
For a solid generated by rotating the region below the curve y = f (x) on the interval [a, b], the area A of
one of the thin bands is equal to the area of a frustum of a cone with slant height l and lower and upper
radii r1 and r2 , respectively (see Figure 4.30: The area of a frustum of the textbook. This area is
A = 2πrl,
where
1
r = (r1 + r2 )
2
is the average radius of the band (you will see the details later). Figure 4.11 shows the relationship
between the curve y = f (x) and the radii and slant height of one of these bands.
We have Pi = (xi , f (xi )) and Pi+1 = (xi+1 , f (xi+1 )); the upper radius is f (xi ) and the lower radius is
f (xi+1 ). As in the previous section, if we apply the Mean Value Theorem to the function y = f (x) on the
interval [xi , xi+1 ], we have xi < xi ∗ < xi+1 such that
f (xi+1 ) − f (xi ) = f 0 (xi ∗)(xi+1 − xi ) = f 0 (xi ∗)∆xi .

Figure 4.11: Relationship between the curve y = f (x) and the radii and slant height of a thin band of
the solid generated by the curve
The slant height is the arc length of the curve y = f (x) on the interval [xi , xi+1 ], and as we saw in the
previous section, this value is
p
l = 1 + f 0 (xi ∗)2 ∆xi ,
and the average radius of the band is

1
r = (f (xi+1 ) + f (xi )).
2
Therefore, the surface area of the band is

1 p
A = 2π (f (xi+1 + f (xi ) 1 + f 0 (xi ∗)2 ∆xi .
2
For a very thin interval [xi , xi+1 ], we have f (xi ) ≈ f (xi ∗) and f (xi+1 ) ≈ f (xi ∗); hence,
1 1
r = (f (xi+1 + f (xi )) ≈ (2f (xi ∗)) = f (xi ∗),
2 2
p
A ≈ 2πf (xi ∗) 1 + f 0 (xi ∗)2 ∆xi .
The area of the surface of revolution is

Z b p X
n
p
0 2
2πf (x) 1 + f (x) dx = lim 2πf (xi ∗) 1 + f 0 (xi ∗)2 ∆xi .
a n→∞
i=1
Similarly, if the solid is generated by a region bounded by x = g(y) y = c, y = d around the y-axis, then
the surface area of this solid is
Z d p
2πf (y) 1 + f 0 (y)2 dy.
c

Example 4.12. This example relates to Example 4.10 on page 133. The surface area of the solid
generated by rotating the curve of y = sin x on [0, π] is given by the integral
Z π √
2π sin x 1 + cos2 x dx.
0
We evaluate this integral by substitution and set u = cos x. Then, du = − sin x; for x = 0, u = 1; and for
x = π, u = −1. Hence, by trigonometric substitution,
Z 1√ √ √ !
u 1 + u2 ln(u + 1 + u 2) 1

2π 1 + u2 du = 2π +
−1 2 2 −1
s√
√ 2+1
= 2π 2 + 2π ln √ .
2−1
To apply the correct formula for the area of a surface generated by rotating a curve, we must identify the
radius of the “band” created.
Let y = f (x) be a positive and continuous function on an interval [a, b].
If the curve y = f (x) rotates around the x-axis, then the radius of the “band” is f (x). Thus, the area of
the surface is given by
s 2
Z b
dy
2πf (x) 1 + dx (4.2)
a dx
If the curve y = f (x) rotates around the y-axis, then the radius of the “band” is x. Thus, the area of the
surface is given by
s 2
Z b
dy
2πx 1 + dx (4.3)
a dx
Let x = g(y) is a positive and continuous function on an interval [c, d].
If the curve x = g(x) rotates around the x-axis, then the radius of the “band” is y. Thus, the area of the
surface is given by
s 2
Z d
dx
2πy 1 + dy (4.4)
c dy

If the curve x = g(x) rotates around the y-axis, then the radius of the “band” is g(y). Thus, the area of the
surface is given by
s 2
Z d
dx
2πg(y) 1 + dy (4.5)
c dy
Example 4.13. We want to find the surface area generated by rotating the curve y = ln x for 1 ≤ x ≤ e,
about the y-axis.
We apply formula 4.3. Since x is positive, we have

r √
p 1 1 + x2
1 + f 0 (x)2 = 1 + 2 =
x x
and we have to evaluate the integral
Z e √ ! Z e√
1 + x2
2π x dx = 2π 1 + x2 dx.
1 x 1
We can apply trigonometric substitution or the formula 68 of the table of integrals.

Z e√
x √ 1 √ e
2π 1 + x2 dx = 2π 1 + x2 + ln(x + 1 + x2 )
1 2 2 1
√ √ e

= π[x 1 + x2 + ln(x + 1 + x2 )]
1
√ √ √ √
2 2
= π[e 1 + e + ln(e + 1 + e ) − 2 − ln(1 + 2)]
" √ !#
√ √ e + 1 + e 2
= π e 1 + e2 − 2 + ln √
1+ 2
In the next example, we only give the integral that gives the area of the surface generated by rotating
around one of the axes.
Example 4.14. a. The curve y = cos−1 x, 0 ≤ x ≤ 1, around the y-axis. We apply formula 4.3 with
s 2 r r
dy 1 2 − x2
1+ = 1+ = .
dx 1 − x2 1 − x2
Hence,
Z 1
r
2 − x2
2π x dx.
0 1 − x2

π
b. If we consider x = cos y, 0 ≤ y ≤ , we apply formula 4.5
2
s 2 q
dx
1+ = 1 + sin2 y.
dy
Hence,
Z π/2 q
2π cos y 1 + sin2 y dy.
0
Example 4.15. a. The curve y = cosh x, 0 ≤ x ≤ 1, around the x-axis. We apply formula 4.2 with
s 2 p
dy
1+ = 1 + sinh2 x = cosh x.
dx
Hence,
Z 1
2π cosh2 x dx.
0
1 + e2
b. If we consider x = cosh−1 y, 1 ≤ y ≤ = b, we apply formula 4.4
2e
s 2 r
dx 1 y
1+ = 1+ =p .
dy y2 −1 y2 − 1
Hence,
Z b
y2
2π p dy.
1 y2 − 1
Exercises
10. Read examples 4.96–4.97 of Section 4.3 of the textbook, Surface Area.
12. For further practice, complete the lab Area of a Surface of Revolution in the optional Lyryx Labs.
Moments and Centers of Mass

Read Section 4.4 of the textbook, Center of Mass.
Let y = f (x) be a continuous function on the interval [a, b]. If we consider the area R under this curve
y = f (x) as a plate, then this plate has a centroid (x̄, ȳ) given by

Rb Rb [f (x)]2
a
xf (x) dx a 2
dx
x̄ = R b ȳ = R b
a
f (x) dx a
f (x) dx
Z b
See that A = f (x) dx is the area of R. Hence,
a
Z b Z b
1 1 [f (x)]2
x̄ = xf (x) dx ȳ = dx
A a A a 2
If the region R lies between two curves f (x) and g(x), where f (x) ≥ g(x) on the interval [a, b], then the
centroid of R is
Z b Z b
1 1 [f (x)]2 − [g(x)]2
x̄ = x[f (x) − g(x)] dx ȳ = dx
A a A a 2
Z b
where A = (f (x) − g(x)) dx.
a
Example 4.16. The area of the region R bounded by sin x on the interval [0, π/2] is
Z π/2 π/2

sin x dx = − cos x = 1.
0 0
The centroid (x̄, ȳ) of the region R is

Z π/2 π/2

x̄ = x sin x dx = −x cos x + sin x = 1.
0 0
and
Z
π/2
2 1 sin(2x) π/2 π
ȳ = sin x dx = x− = .
0 2 2 0 4
Example 4.17. We want to find the centroid of the region R bounded by the curve y = sin x and the
2x
line y = .
π
We start by finding the points of intersection of the curve and the line
2x π
sin x = ⇒ x = 0, x = .
π 2

We need to establish the area A of R
Z π/2
2x x2 π/2 π 4−π
A= sin x − dx = − cos x − = 1 − = .
0 π π 0 4 4
The centroid (x̄, ȳ) of the region R is

Z π/2
4 2x
x̄ = x sin x − dx
4−π 0 π

4 2x3 π/2
= −x cos x + sin x −
4−π 3π 0

4 π2 12 − π 2
= 1− = .
4−π 12 3(4 − π)
And
Z π/2
4 2 4x2
ȳ = sin x − 2 dx
2(4 − π) 0 π
Z π/2
2 1 − cos(2x) 4x2
= − 2 dx
4−π 0 2 π

2 1 sin(2x) 4x3 π/2
= x− − 2
4−π 2 2 3π 0
2 π π
= =
4−π 6 12 − 3π
Exercises
13. Find the centroid of the region R bounded by the curves sin x and cos in the interval [0, π/4].
15. For further practice, complete the lab Center of Mass in the optional Lyryx Labs.
Finishing This Unit

examination.
4. Complete the “Practice Examination” provided for this unit. Evaluate yourself, checking your
answers against those provided in “Appendix D” of this Study Guide. Remember: The number of
points in a question may indicate the number of steps in the solution.


Time: 2 hours
Total points: 60
Passing grade: 55%
1. (16 points)
a. Sketch the region R bounded by the curves
π 2 cos x π 2
y= and y = x− .
4 2
b. Indicate the method you use to set up the integral (do not evaluate it) that gives the volume of the
solid generated by rotating the region R in part (a) when it is rotated around
i. the x-axis.
ii. the y-axis.
iii. the line y = −1.
iv. the line x = 3π/4.
2. (8 points)
a. Identify the region bounded by the ellipse 4x2 + 9y 2 = 36.
b. Find the volume of the solid given that the cross sections perpendicular to the x-axis are
equilateral triangles.
3. (5 points) Find the volume of the solid generated by revolving the region between the x-axis and the
curve y = e−x for x ≥ 0 when it is rotated around the x-axis.
4. (6 points)
a. Show that the arc length s(t) function of the curve y = cosh−1 x is given by an improper
integral. Explain.
b. Evaluate the arc length function in part (a) at t = 4.
5. (4 points) Show that the curves y = sin−1 x and y = cos−1 have the same length on their domains.
6. (6 points) Give the arc length function of the curve 7xy = y 4 + 48 from the point P (7, 1).
7. (6 points) Find the surface area of a sphere with radius r.
Hint: A sphere is generated by a circle of radius r rotated around either the y or the x axis.
8. (4 points) Give the surface area of a cone with radius r and height h.
Hint: Which curve generates a cone when it is rotated around the y-axis?

9. (5 points) Find the centroid of the region bounded by the curves y = x2 and y = −x2 + 2.

UNIT 5
Differential Equations
In this unit, we continue with another application of calculus to solve problems that have a differential
equation as a mathematical model (just as some problems are represented by an algebraic equation). A
differential equation involves one or several functions and its or their derivatives. That is, the
mathematical model corresponds to situations that change, and we want to be able to predict as much as
we can about this change.
Objectives
1. apply direction fields and Euler’s method to estimate the solution of differential equations.
2. solve separable differential equations.
3. solve linear differentiable equations using the method of integrating factors.
4. solve problems of exponential growth and decay.
Modeling with Differential Equations

A differential equation is an equality that involves derivatives and functions. The following expressions
are differential equations:
dP
sin(at) = y 0 = kP + ctP y 0 = x2 y 00 .
dt
where a, k, c are constants. The expressions below are not differential equations.
x3 + y 3 = 6y (x2 − sin(ax)y) = 4
Note: When working with differential equations, it is customary to use a, b, . . . , m, n as constants and
o, . . . , t, x, y, z as variables. However, you should always identify the constants and variables correctly.
Since a differential equation involves one or more functions, we must identify the independent and
dependent variables.

Example 5.1. In the equation sin(at) = y 0 , we have the function sin(at); hence, the independent
variable is t, and the equation can be written as
dy
sin(at) = .
dt
The independent variable in the equation
dP
= kP + ctP,
dt
is t, and the equation is
dP
= kP (t) + ctP (t).
dt
In the differential equation y 0 = x2 y 00 , we have the independent variable x and the function x2 ; hence, the
equation is
dy d2 y
= x2 2 .
dx dx
The order of a differential equation is the highest derivative that occurs in the equation. For example, the
equation sin(at) = y 0 is of the first order, and the equation y 0 = x2 y 00 is of the second order. A function f
is a solution of a differential equation if the equation is satisfied when we replace y = f (x) and the
corresponding derivatives of y.
When we want to solve a differential equation, we want to find each possible function that is a solution of
the equation.
Example 5.2. The function P (t) = kect , with k a constant, is a solution of the differential equation
dP
= cP (t),
dt
where c is a constant. You can see that
dP
= kcect ,
dt
and, replacing in the differential equation,
dP
= kcect = c(kect ) = cP (t).
dt
This example shows that there are infinitely many solutions to the differential equation: as many as the
constant k. For example, P (t) = 3ect and P (t) = πect are solutions. We say that P (t) = kect is the
general solution to the equation.
When we are not interested in a general solution, it is because we are looking for a solution that satisfies
other required conditions. One such condition is the initial condition of the function; that is, the value of

the solution function at a given value. The finding of the solution of a differential equation with a
required initial condition is called an initial-value problem.
Example 5.3. The initial-value problem
y 0 + (tan x)y = cos2 x y(0) = −1
indicates that the required solution function f (x) satisfies the differential equation, and must be such that
f (0) = −1.
The function f (x) = sin x cos x − cos x is a solution to this initial-value problem. Indeed we have that
f (0) = −1 and f 0 (x) = cos2 x − sin2 x + sin x; thus,
f 0 (x) + (tan x)f (x)

= cos2 x − sin2 x + sin x + (tan x)[sin x cos x − cos x]
= cos2 x − sin2 x + sin x + sin2 x − sin x = cos2 x.
Exercises
1. For further practice, complete the lab Modeling with Differential Equations in the optional Lyryx
Labs.
Direction Fields and Euler’s Method

There is no single, definitive method for solving differential equations. However, there is a method based
on the geometry of the solutions by means of direction fields, and there is a numerical approach called
“Euler’s method.”
Direction fields are applied to initial-value problems of the form
y 0 = F (x, y),
where the function F (x, y) is a function of two variables, x and y; for instance F (x, y) = cos x + sin y.
We must remember that the derivative of a function is the slope of a line tangent to the function at a given
point; hence, for a fixed (x0 , y0 ), we have
y 0 (x0 ) = F (x0 , y0 ) = k,
and at the point (x0 , y0 ), the slope of the solution curve (which we do not know) is k. If we draw several
small lines with the slope k at the points (x0 , y0 ), we obtain a directed field—these line segments give the
direction of the solution curve, and allow us at least to visualize the shape of the solution curve.
To apply this method, we must be careful and systematic. We make a table with the values for each point
(x0 , y0 ), and draw a short line with the corresponding slope.
Euler’s method also applies to initial-value problems of the form
y 0 = F (x, y) with y(x0 ) = k

The solution is also based on finding lines that are tangent to the solution curve. Remember that a tangent
line at the point (x0 , y0 ) is
L(x) = y 0 (x0 )(x − x0 ) + y0 .
Euler’s method does not give the solution of the differential equation, but does give an approximation.
Again, we must understand the basic idea and apply it correctly so that we can find a good approximation.
Exercises
1. Read Section 5.4 of the textbook, Approximation.
3. For further practice, complete the lab Euler’s Method in the optional Lyryx Labs.
Separable Equations
In this section, we consider a certain type of differential equation that can be solved explicitly.
A first-order differential equation is separable if it is of the form

dy
= g(x)f (y).
dx
This type of equation can be “separated” into
1
h(y)dy = g(x)dx, where h(y) = .
f (y)
For indefinite integrals, if G(x) is the antiderivative of g(x), then

Z
G (x) = g(x) and
0
g(x) dx = G(x);
thus,
Z
d d
g(x) dx = G(x) = g(x).
dx dx
Z
If y is implicit in terms of x (i.e., y = u(x)), and h(y) dy = H(y), then, by the Chain Rule,
Z
d d dH dy
h(y) dy = H(y) =
dx dx dy dx
Z
d dy dy
= h(y) dy = h(y) .
dy dx dx

We conclude that if we take the indefinite integral in the equation, we have
Z Z
dy
h(y) dy = g(x) dx, and h(y) = g(x);
dx
therefore,
dy
= g(x)h(y).
dx
Method
Step 1
Separate the variables by rewriting the equation in the differential form
h(y)dy = g(x)dx.
Step 2
Integrate both sides of the equations in Step 1 (the left side with respect to y and the right side with
respect to x)
Z Z
h(y)dy = g(x)dx.
Step 3
If H 0 (y) = h(y) and G0 (x) = g(x), then the equation
H(y) = G(x) + C
will generally define a family of solutions implicitly. In some cases, it may be possible to solve this
equation explicitly for y.
Example 5.4. Let us solve the equation

√
du 1+ r
= √
dr 1+ u
√ √
This equation is of the form (1 + u) du = (1 + r) dr, and the integration gives
Z Z
√ √
(1 + u) du = (1 + r) dr.
Thus,
2u3/2 2r3/2
u+ =r+ + C.
3 3
The general solution u is given implicitly.

Example 5.5. Chemical Reactions. Consider a large number (a + b) of two types of molecules, a and b,
per unit volume. These molecules react to form a new molecule. If x(t) denotes the number of new
molecules per unit volume at time t, then the chemical law of mass action states that
dx
= (a − x)(b − x) > 0.
dt
find the expression for x(t).
In the beginning, at t = 0, x(0) = 0 and x0 (0) = ab. Near the end of the reaction, when x reaches a or b,
new molecules are no longer formed. Now,
dx
= dt.
(a − x)(b − x)
Integrating both sides

Z Z
dx
= dt.
(a − x)(b − x)
the integral on the right is t + C. We apply partial fractions to the integral on the left
Z Z b − x
dx 1 1 1 1
= − dx = ln + C.
(a − x)(b − x) b−a a−x b−x b−a a−x
Since, x < a and x < b

1 b−x
ln + C = t.
b−a a−x

1 b
When t = 0, x = 0, so C = ln .
b−a a
Therefore,

1 b−x 1 b 1 a(b − x)
t= ln − ln = ln .
b−a a−x b−a a b−a b(a − x)
Solving for x we have

(b−a)t a(b − x) e(b−a)t − 1
e = ⇒ x(t) = ab .
b(a − x) be(b−a)t − a

Typically, a mixing problem involves a tank or container of fixed capacity filled with a thoroughly mixed
solution of some substance, such as salt. If a solution of a given concentration enters the container at a
fixed rate a0 (t) and the mixture, thoroughly stirred, leaves at a fixed rate b0 (t), it may be that a(t) 6= b(t).
If y(t) is the amount of the substance in the container at time t, then y 0 (t) is the rate at which the
substance is being added minus the rate at which it is being removed. That is, y 0 (t) = a0 (t) − b0 (t). The
solution of y(t) leads to a first-order separable differential equation.
Example 5.6. Mixing Problem. A tank holds 500 gal of brine containing 200 lb of dissolved salt. If
brine containing 0.1 lb of salt per gallon runs into the tank at the rate of 10 gal/min and the thoroughly
mixed brine flows out of the tank at the same rate, how much salt remains in the tank at the end of 1 hour?
The rate of input is (0.1)(10) = 1 lb/min. Let x(t) be the amount of salt (in lbs) in the tank after t
x
minutes. Then the concentration of the salt (in pounds of salt per gallon) is (the tank always contains
500
500 gal). Thus,
Z
dx dt
=
50 − x 50
we integrate both sides
t
− ln |50 − x| = + C.
50
When t = 0, x = 200. Hence, C = − ln(150).
t |50 − x|
ln |50 − x| − ln(150) = − ⇒ = e−t/50 .
50 150
At the end of one hour t = 60, and x > 50. Thus,
|50 − x| = 150e−1.2 ⇒ x = 50 + 150e−1.2 = 50 + 150(0.3012) = 95.
After one hour 95 lbs of salt remain in the tank.
Exercises
4. Read Section 5.1 of the textbook, First Order Differential Equations.
6. A tank contains 1000 L of pure water. Brine that contains 0.05 kg of salt per litre of water enters the
tank at a rate of 5 L/min. Brine that contains 0.04 kg of salt per litre of water enters the tank at a rate
of 10 L/min. The solution is kept thoroughly mixed and drains from the tank at a rate of 15 L/min.
How much salt is in the tank after half an hour?
7. For further practice, complete the lab Separable Equations in the optional Lyryx Labs.

Integrating Factors
A first-order differential equation is called linear if can be expressed in the form
dy
+ p(x)y = q(x)
dx
It is assumed that the functions p(x) and q(x) are continuous on a common interval J, and the general
solutions is valid on J. One method for solving such a differential equation is based on the observation
that, if we define
Z
H(x) = p(x) dx,
then
µ = eH(X)
is an integrating factor, and

Z
d H(X) d
e = eH(X) p(x) dx = eH(X) p(x) = µp(x).
dx dx
Thus,
d dy dµ dy
(µy) = µ + y = µ + µp(x).
dx dx dx dx
dy
If we multiply + p(x)y = q(x) by µ, we get
dx
dy
µ + µp(x)y = µq(x).
dx
Combining these last two equations, we have
d
(µy) = µq(x).
dx
We solve this equation for y by integrating both sides with respect to x and
Z Z
1
µy = µq(x) dx thus y= µq(x) dx.
µ
This y is a general solution of the equation.
Definition 5.1. The function

R
p(x) dx
µ=e

is called the integrating factor of the equation
dy
+ p(x)y = q(x).
dx
The method of finding the general solution is called the method of integrating factors. A summary of this
method is given below. We recommend you follow each step when solving these types of differential
equations.
Method
Step 1
Find the integrating factor

Z R
p(x) dx then µ=e p(x) dx
.
We take the constant of integration to be 0, since any µ will do.
Step 2
Multiply both sides of the equation by µ and express the result as

d
(µy) = µq(x).
dx
Step 3
Integrate both sides of the equation in Step 2, and solve for y. Include a constant of integration.
Example 5.7. Applying the method of integrating factors, we solve the equation
dy
+ 4y = e−3x .
dx
Here p(x) = 4 and q(x) = e−3x . These functions are continuous on R, and the general solution is for all
real numbers.
Step 1
Z
dx = 4x and µ = e4x
Step 2
dy d d 4x dy
µ + 4µy = µe−3x and (µy) = e y = e4x + 4e4x y.
dx dx dx dx
Hence,
dy
e4x + 4e4x y = e4x e−3x = ex .
dx

We write this equation as
d 4x
(e y) = ex .
dx
Step 3
e4x y = ex + C and y = e−3x + C e−4x
If in Example 5.7, above, we have the initial value y(0) = 2, then we have y(0) = 1 + C = 2, and C = 1.
The solution, therefore, is y = e−3x + e−4x .
Exercises
8. Use the method of integrating factors to solve the linear equations below.
a. y 0 + y = cos(ex )
1
b. y0 + y + =0
1 − ex
c. xy 0 − 2y = x3 ex , y(1) = 0
9. Read Section 5.2 of the textbook, First Order Homogeneous Linear Equations.
11. For further practice, complete the lab Integral Factors in the optional Lyryx Labs.
Exponential Growth and Decay

In this section, we discuss problems whose mathematical model is given by the differential equation
y 0 = ky.
These problems are of growth (population or compound interest) and decay (radioactivity or Newton’s
law of cooling).
The solution of the initial-value problem
y 0 = ky y(0) = y0
is
y = y0 ekt
If k > 0, then the solution function y is increasing and the equation corresponds to a problem of growth;
if k < 0, then the solution function is decreasing and the equation corresponds to a decay problem.

The initial value y(0) is, for example, the initial population or principal invested for a problem of growth;
and or the initial amount of radioactive material or temperature of a cooling object for a decay problem.
For problems of decay, the term half-life is the time that an initial amount of a substance takes to be
reduced to one half. Hence, if the half-life of a substance is m0 , and y(t) = mekt is the solution to the
decay problem, then m is the initial amount of the substance and
1 m
y(m0 ) = m; that is mekm0 = .
2 2
Solving for k, we get
ln 2
k=− .
m0
Once we have the value of the constant k, we can find the solution to this decay problem:
− ln 2
t m m
y(t) = me m 0 =
t
= t .
ln 2 2 m0
e m0
Exercises
12. Read Example 5.110 of the textbook, Population Growth and Radioactive Decay.
13. Do exercises 5.1.16–5.1.19 from the section Exercises for 5.1 of the textbook.
14. For further practice, complete the lab Exponential Growth in the optional Lyryx Labs.
Finishing This Unit

examination.
5. Complete the third assignments and submit your work to your tutor for grading. Remember: The
number of points in a question may indicate the number of steps in the solution.

Time: 1.5 hours
Total points: 30
Passing grade: 55%
1. (4 points) Find all nonzero values of a and b such that the function y = a sin(bx) satisfies the
differential equation y 00 + 4y = 0.
2. (5 points) Determine whether the functions

7
y1 = e−x + 2e3x and y2 = xe3x
4
satisfy the equation
y 00 − 2y 0 − 3y = 7e3x .
3. (5 points) Use Euler’s method to estimate the values of the solution of the initial value problem
y 0 = y − x and y(0) = 2, setting a step size of 0.1 on the interval [0, 0.5].
4. (6 points) Show that the equation
0 x2 y − y
y =
y+1
is separable, and solve it with the initial value y(3) = 1.
5. (5 points) A thermometer is taken from a room where the temperature is 72◦ F to the outside, where
the temperature is 32◦ F. After half a minute, the thermometer reads 50◦ F. What will the thermometer
read at t = 2 minutes? How long will it take for the thermometer to read 35◦ F?
6. (5 points) A scientist determines that in exactly 5 days a 10 milligram sample of a certain

radioactive substance decays to 4.5 milligrams. Based on these data, what is the half-life of this
substance?

UNIT 6
Infinite Sequences and Series
The representation of a function as an infinite series gives us one more method to use when we are
approximating or evaluating integrals.
Objectives
1. determine whether sequences converge or diverge.
2. apply appropriate tests to determine the convergence of infinite series.
3. find Taylor series of a given function.
4. determine the convergence of Taylor series.
5. differentiate and integrate using infinite series.
Sequences
Prerequisites
1. define the factorial of 0 as 1 (i.e., 0! = 1) and the factorial of a positive number n as

n! = 1 · 2 · 3 . . . (n − 1) · n.
2. evaluate limits of algebraic and trigonometric functions and apply the Squeeze Theorem.
An infinite sequence can be thought of as an ordered infinite list of numbers, called terms.
The first term of the sequence is denote as a1 , the second as a2 , the third as a3 , and in general, the n-th
term as an . Hence, the sequence is written
a1 , a2 , a3 , . . . , an , . . .
or
{a1 , a2 , a3 , . . . , an , . . . }.

Example 6.1. Consider the following sequences
a. s1 = {2, 4, 8, 16, 32, 64, . . . }
b. s2 = {1, 2, 6, 24, 120, 720, 5040, . . . }

1 2 3 4
c. s3 = , , , ,...
2 3 4 5
d. s4 = {−1, 1, −1, 1, −1, 1, −1, . . . }

4 1 4 1
e. s5 = , , , ,...
10 102 103 104
f. p = {2, 3, 5, 7, 11, 13, 17, . . . }

1 1 1 1
g. C= , , , ,...
2 2 2 2
When listing the terms of a sequence, we must provide enough terms to enable a reader to identify the
next and subsequent terms. For example, the next term in the sequence s4 is 1; in the sequence s5 , the
4
next term is 5 .
10
The letter n is called the index of the sequence, and the index can be any letter, such as k or i. The n-th
term of a sequence is identified by establishing a relationship between each number n and the
corresponding term an .
In the sequence s1 , the first term a1 is 2, the second term a2 is 4 = 2(2), the third term a3 is 8 = 23 , the
fourth term a4 is 16 = 24 ; from here, we see that the n-th term an is 2n . Thus, an = 2n .
In the sequence s2 , the term a1 is 1, a2 is 2, a3 is 6 = 3(2), a4 is 24 = 4(3)(2), a5 is 120 = 5(4)(3)(2);

from here, we see that an is n(n − 1)(n − 2) . . . (3)(2) = n!. Thus, an = n!.
In the sequence s4 , the term a1 is −1, a2 is 1, a3 is −1, a4 is 1; from here, we see that an is 1 if n is odd
and −1 if n is even. Hence, in general, the an term is (−1)n . Thus, an = (−1)n .
The sequence p lists in increasing order the prime numbers; in this case, there is no known relationship
between a number n and the an term of this sequence.
1 1 1
The sequence C is the constant sequence , and in this case, an = for all n ≥ 1. Thus, an = .
2 2 2
If the n-th term of a sequence an is well identified, then the sequence can be written as
{an } or {an }∞
n=1 .
This expression is known as the brace notation with index n.

Example 6.2. In Example 6.1, the brace notation of the first four sequences are as follows:

n n
s1 = {2 }, s2 = {n!}, s3 = , s4 = {(−1)n }.
n+1
1 4
The even terms of the sequence s5 are n
and the odd terms are n , which can be expressed as
10 10
1 4
a2n = and a2n+1 = .
102n 102n+1

1
The constant sequence C can be written as .
2
Definition 6.1.
Two sequences {an } and {bn } are equal if
an = b n for all n.
Example 6.3. The sequences
t1 = {2, 4, 6, 8, 10, . . . } and t2 = {1, 2, 4, 6, 8, 10, . . . }
are not equal, because their nth terms are not equal. For instance, their first terms are 2 and 1,
respectively.
Example 6.4. The sequences

cos(nπ) (−1)n
and
n n
are equal, because

cos(nπ) 1 1 1 (−1)n
= −1, , − , , . . . =
n 2 3 4 n
The brace notation allow us to describe a sequence as a function f with domain the natural numbers N
(positive integers) to the real numbers R, represented as
f : N −→ R.
Thus, f (n) = an for all n ≥ 1.
Example 6.5. From Example 6.2 on the preceding page, the function associated with the sequence s1 is
f (n) = 2n , that associated with the sequence s2 is f (n) = n!, that associated with the sequence s4 is

f (n) = (−1)n , and that associated with the sequence s5 is

 1
 n if n is even
f (n) = 10

 4 if n is odd.
10n
The function associated with a sequence has a graph represented by dots in the Cartesian plane.
Figure 6.1, below, shows the graphs of the functions associated with sequences s1 and s4 of Example 6.1
on page 158.
Figure 6.1: Graphs of sequences s1 and s4 of Example 6.1
We can also make a visual representation of a sequence by plotting its terms on the real line. Figure 6.2,
below, shows the representation of the sequences s2 and s5 of Example 6.1 on page 158.
Figure 6.2: Graphs of sequences s2 and s5 of Example 6.1
These visual representations help us to understand the behaviour of the terms of the sequence as n
increases. From Figure 6.1, we “see” that the terms of the sequence s1 increase without bound as n
increases, while sequence s4 oscillates between 1 and −1. Similarly, Figure 6.2 shows us that the
sequence s5 decreases and is bounded by 0. When the terms of a sequence {an } approach a number L as
n increases, we say that the limit of the sequence is L, and we write
lim an = L.
n→∞
Definition* 6.2.

The limit of a sequence {an } is L if we can make the terms an as close as L as we want by taking n
sufficiently large. We write
lim an = L or an → L as n → ∞.
n→∞
In this case we say that the sequence converges or is convergent.
The function
f : N −→ R
associated with a sequence {an } can be extended to a real valued function if we can define a function
F : R −→ R
such that
F (n) = f (n) = an
for any positive integer n.
Example 6.6. The function f (n) = 2n is associated with the sequence {2n }. It is extended to the
function F (x) = 2x for any real number x, as F (n) = 2n = f (n) for any positive integer n.
Example 6.7. The function f (n) = n! associated with the sequence {n!} cannot be extended to a real
valued function, because n! is only defined for nonnegative integers.
Proposition 6.3.
Let F : R −→ R be the extension of the sequence f : N −→ R.
If lim F (x) = L then lim f (n) = L.

x→∞ n→∞
x n
Example 6.8. The function F (x) = (x ∈ R), is the extension of the function f (n) =
x+1 n+1
(n ∈ N).

n
Hence, the sequence converges to 1, because
n+1
x
lim = 1,
x→∞ x+1
and therefore,
n
lim = 1.
n→∞ n + 1
Definition* 6.4.

The limit of a sequence {an } is infinite if we can make the terms an as large as we want by taking n
sufficiently large. We write
lim an = ∞.
n→∞
The limit of a sequence {an } is negative infinite if we can make the terms an negatively small, as
small as we want, by taking n sufficiently large. We write
lim an = −∞.
n→∞
In either case we say that the sequence diverges or is divergent.
The proposition below is similar to Proposition 6.3.
Proposition 6.5.
Let F : R −→ R be the extension of the function f : N −→ R.
If lim F (x) = ±∞ then lim f (n) = ±∞.

x→∞ n→∞
Example 6.9. From the visual representation of the sequence {n!}, we conclude that
lim n! = ∞.
n→∞
Example 6.10. From Example 6.6 on the preceding page, we have
lim 2n = ∞,
n→∞
because, for the extension function F , we have lim 2x = ∞.

x→∞
Definition* 6.6.
The limit of a sequence {an } does not exists if the limit of the sequence
lim an
n→∞
is not a number, or is not infinity or negative infinity. In this case, we also say that the sequence is
divergent.
Example 6.11. The sequence {(−1)n } oscillates between 1 and −1; hence, its limit does not exist, and
the sequence is divergent.
Example 6.12. There is no function associated with the sequence
{0, 1, 0, 0, 1, 0, 0, 0, 1, 0, 0, 0, 0, 1, . . . }.

Observe that it does not matter how large n is, the term an may be 0 or it may be 1. Hence, we cannot
make all the terms an close to either 0 or 1 for large values of n. That is, if an = 0 for large n, then there
will be k > n such that ak = 1. The sequence is divergent and the limit does not exist.
In calculus, it is very important to determine whether a sequence converges or diverges. There are several
theorems and propositions that we can apply to answer this question. Propositions 6.3 on page 161
and 6.5, above, are useful when we have a function associated with the sequence and this function can be
extended to all real numbers. We also have the following results, their formal proofs will be provided in
the next section.
Operations for sequences are performed term-by-term; that is, if {an } and {bn } are two sequences and c
is a constant, then their sum {an + bn } is the sequence
{a1 + b1 , a2 + b2 , a3 + b3 , . . . }.
Similarly, their product and quotient are as follows:
{an bn } = {a1 b1 , a2 b2 , a3 b3 , . . . }
and

an a1 a2 a3
= , , ... .
bn b1 b2 b3
Finally, the sequence {can } is {ca1 , ca2 , ca3 , . . . }.
When we work with sequences, we pay attention to the eventual behaviour of the terms of the sequence.
That is, the behaviour of the sequence is not dictated by all of the terms of the sequence, but only by the
“tail” of the terms of the sequence.
If {an }∞
n=1 is a sequence, then for a positive integer k > 1,
{an }∞ ∞
n=k = {ak , ak+1 , ak+2 . . . } = {an+k }k=1
is a sub-sequence of {an }; that is, a “tail” of the sequence.
Definition 6.7.
A sequence {an }∞
n=1 eventually or ultimately has the property P if the sub-sequence {an }n=k has the
∞
property P .
Example 6.13. The sequence {−4, −3, −2, −1, 0, 1, 2, 3, . . . } is eventually positive, because an > 0
for n > 5.
Example 6.14. Two sequences {an } and {bn } are eventually equal if an = bn for all n > k for some k.
Remark

Since the limit of a sequence is determined by the terms an for sufficiently large n, we see that if a
sequence {an } is convergent, then the limit of the sequence {an }∞
n=1 and the sub-sequence {ak }k=n are
∞
the same.
n=1 is divergent, then the sub-sequence {bn }n=k is also divergent.

Similarly, if a sequence {bn }∞ ∞
Theorem 6.8. Law of Limits
If {an } and {bn } are convergent and c is a constant, then
a. lim an ± bn = lim an ± lim bn

n→∞ n→∞ n→∞
b. lim can = c lim an

n→∞ n→∞

c. lim an bn = lim an lim bn
n→∞ n→∞ n→∞
an lim an
d. lim = n→∞ if lim bn 6= 0
n→∞ bn lim bn n→∞
n→∞
r
e. lim arn = lim an if r > 0 and {an } is eventually positive.
n→∞ n→∞
Example 6.15. By Example 6.8 on page 161,

n
lim = 1.
n→∞ n + 1
By Proposition 6.3 on page 161,

n
lim= 0.
n→∞ en
Hence, Theorem 6.8 gives the conclusions below.

n n n n
1. lim ± n = lim ± lim n = 1 ± 0 = 1
n→∞ n + 1 e n→∞ n + 1 n→∞ e
n 1 n 1 n 1 1
2. lim = lim = lim = (1) =
n→∞ 2n + 2 n→∞ 2 n + 1 2 n→∞ n + 1 2 2

n n n n
3. lim = lim lim = 1(0) = 0
n→∞ n+1 en n→∞ n + 1 n→∞ en
n n
n+1 n
lim n 0
4. lim n = lim en = n→∞ e = =0
n→∞ e n→∞ n 1
lim
n+1 n→∞ n + 1
6 6
n n
5. lim = lim = 16 = 1
n→∞ n+1 n→∞ n + 1

Example 6.16. The law of limits does not apply to the evaluation of the limit
en
lim
n→∞ n + 1
because
lim en = ∞ and lim n + 1 = ∞.

n→∞ n→∞
In this case we apply l’Hopital’s Rule and obtain

en ex
lim = lim = lim ex = ∞
n→∞ n + 1 x→∞ x + 1 x→∞
Remark Observe that Theorem 6.8 on the facing page applies only to convergent sequences; therefore,
we cannot apply this theorem to the limit
n!
lim ,
n→∞ en
expressing it as
lim n!
n→∞
lim en
n→∞
because, by Example 6.9 on page 162,
lim n! = ∞.
n→∞
We will see later how to evaluate this limit.
The following result refers to the odd and even terms of a sequence. If sub-sequences of odd and even
terms behave differently, then the sequence is divergent.
Theorem 6.9.
Let {an } be a sequence, {a2n } be its sub-sequence of even terms, and {a2n+1 } be its sub-sequence of
odd terms. Then,
lim an = L
n→∞
if and only if
lim a2n = L = lim a2n+1

n→∞ n→∞

Example 6.17. Consider the sequence s5 of Example 6.1 on page 158. Its sub-sequence of even terms is

1
,
10n
and its sub-sequence of odd terms is

4
.
10n
By Proposition 6.3 on page 161, we have

1 1
lim = lim = 0,
n→∞ 10n x→∞ 10x
and
4 4
lim n
= lim x = 0.
n→∞ 10 x→∞ 10
Thus, by Theorem 6.9, we conclude that the sequence s5 is convergent and its limit is 0.
Note that Theorem 6.9 can be used to determine whether or not a sequence diverges: if the limits of the
sub-sequences of even terms and odd terms are not equal, then the sequence is divergent.
Example 6.18. The sequence of even terms of the sequence s4 in Example 6.1 on page 158 is {1} and
the sequence of odd terms is {−1}. Therefore, by Theorem 6.9 on the previous page the sequence s4 is
divergent.
The next theorem is useful when we have positive and negative terms.
Theorem 6.10.
If
lim |an | = 0,
n→∞
then
lim an = 0.
n→∞
Example 6.19. Consider the sequence

1 1 1 1 1 1 1 1
−1, − , − , , , , − ,− ,− ,... .
4 8 16 32 64 128 256 512
We can see that the absolute values of the terms of this sequence give the sequence

1
.
2n

1
lim = 0,
n→∞ 2n
and by Theorem 6.10, above, the given sequence is convergent and its limit is 0.
Example 6.20. Let us evaluate the limit of the sequence

(−1)n n3
.
2n3 + 4n − 1
By applying Theorem 6.10, we have the following sequence of absolute values:

(−1)n n3 n3
3 = .
2n + 4n − 1 2n3 + 4n − 1
n3 x3 1
lim = lim = 6= 0;
n→∞ 2n3 + 4n − 1 x→∞ 2x3 + 4x − 1 2
hence, the theorem does not apply.
Instead, we apply Theorem 6.9 on page 165. The sub-sequences of odd and even terms are

n3 n3
− 3 and .
2n + 4n − 1 2n3 + 4n − 1
Again, by Proposition 6.3 on page 161,
n3 1
lim 3
=
n→∞ 2n + 4n − 1 2
and
n3 1
lim − 3
=− .
n→∞ 2n + 4n − 1 2
Therefore, the sequence is divergent.
The next theorem is known as the Squeeze Theorem for Sequences. You may wish to review the Squeeze
Theorem for Functions.
Theorem 6.11.
Let {an }, {bn }, {cn } be three sequences such that, eventually,
an ≤ c n ≤ b n .

If
lim an = lim bn = L.
n→∞ n→∞
then
lim cn = L.
n→∞

3n
Example 6.21. Let us find the limit of the sequence .
n!
We see that Propositions 6.3 (p. 161) and 6.5 (p. 162) do not apply. The sequence has positive terms, so
Theorem 6.10 (p. 166) does not apply. To consider the even and odd terms does not offers a good
possibility either. So we will try to apply the Squeeze Theorem. We need two sequences {an } and {bn }
such that eventually
3n
an ≤ ≤ bn .
n!
We have
3n 3 · 3 · 3...3
= ;
n! 1 · 2 · 3 . . . (n − 1) · n
so, for n > 4 we have

3 3 · 3 · 3...3 3·3·3 3 · 3 · 3...3 3 9 3
< 1 and = ≤ .
n 1 · 2 · 3 . . . (n − 1) · n 1·2·3 4 · 5 . . . (n − 1) n 2 n
If we take

27
{an } = {0} and {bn } = ,
2n
we conclude that for n > 4

3n 27
0≤ ≤ ,
n! 2n
and by Proposition 6.3 on page 161,
27 27
lim = lim = 0.
n→∞ 2n x→∞ 2x
Hence, by the Squeeze Theorem,

3n
lim = 0.
n→∞ n!
Some sequences are defined recursively; that is, the an terms are given in terms of the previous terms ak
for k < n. The best-known recursive sequence is the Fibonacci sequence, but there are many others.

Example 6.22. Let a1 and a2 = 2. The Fibonacci sequence is defined as
an+1 = an−1 + an .
The first few terms of the sequence are
1, 2, 3, 5, 8, 13, 21, 34, ....
Example 6.23. Let

1
a1 = 1 and an+1 = 1 + .
1 + an
This expression gives us the terms of the sequence
1 1 1 3
a2 = 1 + =1+ =1+ = ,
1 + a1 1+1 2 2
1 1 2 7
a3 = 1 + =1+ =1+ = ,
1 + a2 1 + 3/2 5 5
1 1 5 17
a4 = 1 + =1+ =1+ = ,
1 + a3 1 + 7/5 12 12
and so on.
Example 6.24. Let a1 = 1, a2 = 2, and

1
an = an−1 + .
an−2
The next few terms of the sequence are
1
a3 = a2 + =2+1=3
a1
1 1 7
a4 = a3 + =3+ =
a2 2 2
1 7 1 23
a5 = a4 + = + =
a3 2 3 6
and so on.
In this course, we consider only one way of evaluating the limits of recursive sequences.

By the remark on page 163,
lim an = lim an+1 .

n→∞ n→∞
If we assume (and this is the key here—our assumption must be true) that
lim an = L,
n→∞
then, by the law of limits

1 1 L+2
L = lim an = lim 1 + =1+ = .
n→∞ n→∞ 1 + an 1+L 1+L
√
Solving for L, we have L =√ 2. If the recursive sequence of Example 6.23 on the previous page is
convergent, then its limit is 2.
This method does not apply to the recursive sequence of Example 6.24 on the preceding page.
All the results presented in this section hold for sequences that eventually fulfil the required conditions.
Exercises
1. Give an example of a sequence that is eventually constant.
2. Find the limit of each of the sequences listed below.

(−1)n (n + 1)
a.
n
n
2
b. {2 + − }
π

2n
c. tan−1
2n + 1

sin n
d. √
n
Monotonic Sequences
Prerequisites
To complete this section, you must be able to do proofs using mathematical induction.
In the previous section, we considered results that allow us to find the limit of a sequence. The results in
this section give conditions that allow us to determine whether a sequence is convergent or divergent, but
in the case of convergent sequences, they do not enable us to establish the value of their limits.

Any two consecutive terms of a sequence are an and an+1 .
Definition 6.12.
A sequence {an } is
a. strictly increasing if an < an+1 for n ≥ 1.
b. increasing if an ≤ an+1 for n ≥ 1.
c. strictly decreasing if an > an+1 for n ≥ 1.
d. decreasing if an ≥ an+1 for n ≥ 1,
A sequence that is strictly increasing is increasing; similarly, a strictly decreasing sequence is

decreasing. A sequence that is either increasing or decreasing is called monotonic. If a sequence is
strictly increasing or strictly decreasing, it is described as strictly monotonic. Thus, a sequence that is
strictly monotonic is monotonic.
Example 6.25.
a. The sequence

4 1 4 1
s5 = , , , ,...
10 102 103 104
is strictly decreasing; hence, strictly monotonic, and therefore, monotonic.
b. The sequence {1, 2, 2, 3, 3, 3, 4, 4, 4, 4, 5, 5, 5, 5, 5, . . . } is increasing; hence, monotonic. However, it

is not strictly increasing.
c. The sequence {n!} = {1, 2, 6, 24, 120, . . . } is strictly increasing; hence, strictly monotonic, and
therefore, monotonic.
d. The sequence {(−1)n 2n } = {−2, 4, −8, 16, −32, 64, . . . } is neither increasing nor decreasing;
hence, it is not monotonic.
We can also consider sequences that are eventually increasing; that is, sequences such that
an ≤ an+1 for n ≥ k for some k.
Similarly, there are sequences that are eventually strictly increasing, eventually decreasing and eventually
strictly decreasing.
Example 6.26. The sequence {−1, 0, 1, −2, 2, −3, −6, −12, −24, −48, . . . } is eventually strictly
decreasing, since an > an+1 for n > 5.
In Example 6.25, we identified the monotonicity of the sequences by listing the first few terms of each
one. However, this method is not always conclusive, nor is it desirable, since it is prone to yield the
wrong conclusion. We have other, more reliable tests.

Test of monotonicity: Let {an } be a sequence. The properties in the table below hold for n > k for some
k.
Difference of Ratio of Conclusion

consecutive terms consecutive terms
an
an − an+1 < 0 <1 strictly increasing
an+1
an
an − an+1 ≤ 0 ≤1 increasing
an+1
an
an − an+1 > 0 >1 strictly decreasing
an+1
an
an − an+1 ≥ 0 ≥1 decreasing
an+1
If the function associated with the sequence is increasing (decreasing), then the sequence is increasing
(decreasing).
Example 6.27. Two consecutive terms of the sequence

3
n n3 (n + 1)3
are an = and an+1 = .
n! n! (n + 1)!
If we take their ratio, we have
n3 3
n! n (n + 1)! n3 (n + 1)!
= = .
(n + 1)3 n! (n + 1)3 (n + 1)3 n!
(n + 1)!
If we simplify, we see that
an n3
= .
an+1 (n + 1)2
Since
n3
lim =∞
n→∞ (n + 1)2
by Proposition 6.5 on page 162, there is a k such that
n3
>1 for n > k.
(n + 1)2
You may want to check that k = 5.

Thus
an
> 1 for n>k for some k.
an+1
This conclusion indicates that the sequence is eventually strictly decreasing.

2n − 3
Example 6.28. Consider the sequence .
3n + 4
The function associated with this sequence is
2x − 3
f (x) = .
3x + 4
To determine if it is increasing or decreasing, we take its derivative
17
f 0 (x) = > 0;
(3x + 4)2
thus, f is increasing, and the sequence is strictly increasing.
Example 6.29. Two consecutive terms of the sequence

1
n+
n
are
1 1
an = n + and an+1 = n + 1 + .
n n+1
We consider the difference of these two consecutive terms

1 1 1
an − an+1 = n + − n + 1 + = −1 + <0
n n+1 n(n + 1)
for n > 0, then the sequence is strictly increasing.
Note that we could also consider the function associated with this sequence.
Definition 6.13.
A sequence {an } is bounded above if there is a number M such that
an ≤ M for n ≥ 1.
If
an ≤ M for n ≥ k for some k
then the sequence is eventually bounded above.

A sequence {an } is bounded below if there is a number m such that
an ≥ m for n ≥ 1.
If
an ≥ M for n ≥ k for some k
then the sequence is eventually bounded below.
A sequence is bounded it is bounded both above and below. A sequence is eventually bounded if it is
both eventually bounded above and below.
Example 6.30. A sequence of positive terms is bounded below, and the bound is 0. Similarly a
sequence of negative terms is bounded above by 0.
Example 6.31. The sequence

(−1)n
n
is bounded above by 1, and below by −1, because

(−1)n 1

= ≤ 1.
n n
Hence,
(−1)n
−1 ≤ ≤ 1 for all n ≥ 1.
n

1
n+
n
is bounded below by 0, and it is not bounded above. Why?

3
n
n!
is bounded below by 0, and it is eventually strictly decreasing, by Example 6.27 (p. 172). Hence, it is
bounded above by 4. Why?
The following theorem, known as the Monotonic Sequence Theorem, relates what intuitively we suspect,
if a sequence is increasing and bounded above, then it is convergent. Similarly if a sequence is decreasing
and bounded below, then it is convergent.

Theorem 6.14.
If a sequence {an } is eventually increasing, then there are two possibilities:
a. the sequence is bounded above by M , and in this case the sequence is convergent and its limit L
is less than or equal to M .
b. the sequence is not bounded above, and in this case the sequence is divergent and lim an = ∞.
n→∞
If a sequence {an } is eventually decreasing, then there are two possibilities:
a. the sequence is bounded below by m, and in this case the sequence is convergent and its limit L
is greater than or equal to m.
b. the sequence is not bounded below, and in this case the sequence is divergent and
lim an = −∞.
n→∞
This theorem says that any monotonic bounded sequence is convergent.

5
n
Example 6.34. Let us apply Theorem 6.14 to the sequence .
en
First, we must find out if the sequence is monotonic.
The function associated with this sequence is
x5
f (x) = x ,
e
and its derivative is
x4 ex (5 − x)
0
f (x) = < 0, for x > 5.
e2x
Hence, the sequence is eventually strictly decreasing. Since the terms are all positive, it is bounded below
by 0, and by Theorem 6.14, the sequence is convergent.
Let us try to find the value of this sequence.
Let L be the limit of this sequence; then, by Theorem 6.14, L ≥ 0.
We take the ratio of two consecutive terms:

5 n+1
an n e n5
= = e.
an+1 en (n + 1)5 (n + 1)5
Hence,
n5
an = (ean+1 ).
(n + 1)5

If we take the limit, we have
n5
L = lim an = lim (ean+1 ) = (1)eL.
n→∞ n→∞ (n + 1)5
If L > 0, we have a contradiction, namely 1 = e; hence, L must be 0.
Example 6.35. Let

√ √
a1 = 2 and an+1 = 2 + an .
We will use mathematical induction to show that this recursive sequence is strictly increasing and
bounded above by 3.
√
For n = 2, we have a1 = 1 < 2 + 1 = a2 , and a1 < a2 < 3.
We assume that a1 < a2 < · · · < an ≤ 3. We need to show that an < an+1 ≤ 3.
√ √
Since an−1 < an , we have 2 + an−1 < 2 + an ; thus, an < an+1 .
The statement above is true for n + 1. Therefore, the sequence is strictly increasing and bounded by 3. By
Theorem 6.14 on the preceding page, the sequence is convergent, with limit L such that L ≤ 3. Taking
the limit, we have
√ √
L = lim an+1 = lim 2 + an = 2 + L.
n→∞ n→∞
Solving, we obtain L2 − L − 2 = 0 and L = 2, −1.
Since L is positive, the limit is 2.
The next theorem is useful when we can compare two sequences.
Theorem 6.15.
Let {an } and {bn } be two sequences. If
an ≤ b n for n > k for some k,
and if
lim an = ∞,
n→∞
then
lim bn = ∞.
n→∞

Example 6.36. On page 165, we promised to evaluate the limit of the sequence

n!
.
en
The terms of the sequence are positive and

n! 1 · 2 · 3 . . . (n − 1)n 2 3 · 4 . . . (n − 1)n 2n
= = 2 ≥ 2 for n > 2.
e n e · e...e e e · e...e · e e e
By Proposition 6.5 on page 162,

2n
lim = ∞,
n→∞ e3
and by Theorem 6.15, above, we conclude that the sequence is divergent and
n!
lim = ∞.
n→∞ en
Example 6.37. As in Example 6.36, we have the sequence

n
n
en
of positive terms, and

nn n · n · n...n nnn n n
= = ... ≥ for n > 2.
en e · e · e...e eee e e
n
lim = ∞,
n→∞ e
and by Theorem 6.15, we conclude that the sequence is divergent and
nn
lim = ∞.
n→∞ en
Example 6.38. By Example 6.25 on page 171, the sequence

4 1 4 1
s5 = , , , ,...
10 102 103 104
is strictly decreasing and bounded below by 0; hence, it is convergent with limit L. Since
4
an ≤ ,
10n
by Proposition 6.3 on page 161,
4
lim = 0.
n→∞ 10n

By the Squeeze Theorem (see Theorem 6.11 on page 167), we conclude that the limit L is 0.
Exercises
2. For further practice, complete the lab Sequences in the optional Lyryx Labs.
Important Sequences
In this section, we consider some important sequences that we will need for our study of series.
Example 6.39. We want to determine for which values of r the sequence {rn }, called geometric,
converges. The easy cases are when r = 1, or r = 0, since the sequence becomes constant, at 1 or 0, and
therefore convergent. If r = −1, then the sequence is {(−1)n }, and as we saw in Example 6.11 on
page 162, this sequence is divergent.
If r > 0, we can consider the function associated with the sequence f (x) = rx , which is an exponential
function. By Propositions 6.3 on page 161 and 6.5 on page 162, we have

 ∞ if r > 1
lim rn =
n→∞  0 if 0 < r < 1
If −1 < r < 0, then |rn |. By Proposition 6.3
lim |rn | = 0,
n→∞
and by Theorem 6.10 on page 166, we conclude that the limit of the sequence in this case is 0.
If r < −1, then the sub-sequence of even terms {r2k } = {(r2 )k } are all positive, and r2 > 1; hence, by
the previous case, the sub-sequence is divergent. The sub-sequence of odd terms is
{r2k+1 } = {r(r2n } = {r(r2 )n }.
Again, we have
lim (r2 )n = ∞,
n→∞
and since r < 0 we have
lim r(r2 )n = −∞.

n→∞
Hence, the limit of the sequence does not exists—the sequence is divergent.

We conclude, then, that


 ∞ if r>1




 0 if |r| < 1
n
lim r =
n→∞ 
 1 if r=1




 dne if r ≤ −1
Therefore, the sequence converges for |r| < 1 or r = 1 and diverges for |r| > 1 or r = −11 .
Example 6.40. For any a > 0,
lim a1/n = 1.
n→∞
If y = a1/x is the function associated with the sequence, then

ln a ln a
ln y = and lim = 0.
x x→∞ x
By Proposition 6.3, we have
lim a1/n = lim a1/x = e0 = 1.

n→∞ x→∞
Example 6.41. We will investigate for which values of r is the sequence

1
nr
is convergent. The trivial case is r = 0.
If r = 1, then the sequence is the harmonic sequence

1
.
n
By considering the function associated with this sequence, we conclude that the sequence is convergent
and the limit is 0.
If r < 0, then

1
= {n−r }.
nr
Since −r > 0, then by Proposition 6.5 on page 162, the sequence is divergent with an infinite limit.
1
The abbreviation “dne” is the short form of “does not exist.”

If r > 0, then by Proposition 6.5, the sequence is convergent with limit 0. Hence,


 ∞ if r < 0


1
lim = 0 if r > 0
n→∞ nr 


 1 if r = 0
The sequence is convergent for r ≥ 0.
Example 6.42. The sequence in Example 6.21 on page 168 is a particular case of the sequence
n
r
.
n!
If r = 0 the sequence is constant 0.
If r > 0, then we have
rn+1 r(rn ) r
an+1 = = = an .
(n + 1)! (n + 1)n! n+1
From here, we conclude that

an+1 r
= <1
an n+1
for n > r. Hence, the sequence is eventually decreasing, and it is bounded below by 0. By Theorem 6.14
on page 175 the sequence is convergent. If the limit is L, then L ≥ 0, and taking the limit, we have
r
L = lim an+1 = lim an = 0(L) = 0.
n→∞ n→∞ n + 1
If r < 0, then
n n
r |r|
= .
n! n!
By the previous case, the limit of this sequence of absolute values is 0, and by Theorem 6.10 on
page 166, the limit of the sequence without the absolute values is also 0. Hence, for any r, the sequence is
convergent and
rn
lim = 0.
n→∞ n!
Example 6.43. For any real number a

a n
lim 1 + = ea ,
n→∞ n
by Proposition 6.3 on page 161. Let
a x
y = 1+
x

be the function associated with the sequence. Thus,
a
ln y = x ln 1 + = x[ln(x + a) − ln x].
x
By l’Hopital’s Rule, we have
ln(x + a) − ln x
lim x[ln(x + a) − ln x] = lim
x→∞ x→∞ 1
x
1 1
−
= lim x + a x
x→∞ 1
− 2
x
ax2
= lim = a.
x→∞ x2 − ax
Therefore,
a x
lim 1 + = ea .
x→∞ x
Exercises
3. Read Section 6.1 of the textbook, Sequences.
Formal Definitions
Prerequisites
For this section you need to know the triangle inequality.
In this (optional) section, we provide formal definitions of “limit,” and work through some proofs of the
theorems we used in the earlier sections of this unit. It is our intention to show you how the intuitive
definitions (Definitions 6.2 on page 160 and 6.4 on page 161) are formalized, and how the theorems we
use are consequences of these definitions. We recommend that you review the formal definition of the
limit of a function at infinity.
The number L is the limit of a sequence {an } if we can make the terms of the sequence an as close to L
as we want by taking a sufficiently large n. The visual representation of a limit is shown in Figure 6.3,
below.
Figure 6.3: Visual representation of a limit

Observe that this statement says that for a sufficiently large N > 0, the terms of the sequence aN , aN +1 ,
aN +2 . . . are close to L; that is, an is close to L for all n ≥ N .
We turn now to formalize the idea that the terms of the sequence an can be as close to L as we want.
Two numbers are very close if their distance is very small. Mathematically the distance between two
numbers a and b is |a − b|.
Figure 6.4: Distance between two numbers, a and b
You can see, in Figure 6.4, above, that if a < b (i.e., b − a > 0), then b − a is the distance between a and
b. Conversely, if a > b (i.e., a − b > 0), then a − b is the distance between a and b. Both of these values
coincide with |a − b|, since

 b − a if b − a > 0
|a − b| =
 a − b if a − b > 0
Thus, the distance between a and b is very small if, for a small number ε 2 we have |a − b| < ε.
Now consider the set of all numbers whose distance to a number b is less than ε. This set is expressed as
all numbers x such that
|x − b| < ε.
Figure 6.5 shows this set.
Figure 6.5: Set of numbers x such that |x − b| < ε
The statement “the terms of the sequence an are as close to L as we want” can be expressed
mathematically by saying that if we identify a number ε that is as small as we want it to be, then
|an − L| < ε.
2
In mathematics is customary to use Greek letters to denote small numbers, ε is the Greek letter epsilon.

Finally to say that “we can make an close to L” can be expressed by saying that even if ε is very small,
we can always find terms an whose distance to L is less than ε. Hence. the formal definition of a limit is
as given below.
Definition 6.16.
The number L is the limit of the sequence {an } if, given ε > 0 there is a number N > 0 such that
|an − L| < ε for all n > N.
Figure 6.6, below, shows all the terms an , for n > N , whose distance to L is less than ε.
Figure 6.6: Terms an , for n > N , whose distance to L is less than ε
We recommend that you compare this definition with the formal definition of the limit
lim f (x) = L
x→∞
for a function f .
We will prove some of the results from this section using Definition 6.16.
Sum of Limits
Let
lim an = L and lim bn = M

n→∞ n→∞
be two convergent sequences. Thus,
lim an + bn = L + M.
n→∞
Proof. Given ε > 0, we want to show that there is an N > 0 such that for all n > N
|an + bn − (L + M )| < ε.
We assume that
lim an = L and lim bn = M ;

n→∞ n→∞
hence, by Definition 6.16,

1. for any ε1 > 0, there is N1 > 0 such that |an − L| < ε1 for all n > N1 .
2. for any ε2 > 0, there is N2 > 0 such that |bn − M | < ε2 for all n > N2 .
By the triangle inequality, we have
|an + bn − (L + M )| ≤ |an − L| + |bn − M |,
which suggests that since the conditions above hold for any given ε1 and ε2 , we should take
ε
ε1 = = ε2 .
2
If we take
N = max{N1 , N2 },
then
N > N1 and N > N2 ,
and for n > N , the two conditions above hold; therefore
|an + bn − (L + M )| ≤ |an − L| + |bn − M |

ε ε
< ε1 + ε2 = + = ε. Q.E.D.
2 2
Product of Limits
Let
lim an = L and lim bn = M

n→∞ n→∞
be two convergent sequences.
Thus,
lim an bn = L M.
n→∞
Proof. Given ε > 0, we want to show that there is an N > 0 such that for all n > N
|an bn − L M | < ε.
We assume that
1. for any ε1 > 0 there is N1 > 0 such that |an − L| < ε1 for all n > N1 .
2. for any ε2 > 0 there is N2 > 0 such that |bn − M | < ε2 for all n > N2 .

See how we use the triangle inequality to obtain what we want.
|an bn − L M | = |an bn − bn L + bn L − L M |
≤ |bn (an − L) + L(bn − M )|
≤ |bn ||an − L| + |L||bn − M |
We want this last inequality to be less than any ε > 0, and we can achieve this result by taking particular
values for ε1 > 0 and ε2 > 0 in Statements 1 and 2, above.
ε
In Statement 1, we take ε1 = > 0, and we find that there is an N1 > 0 such that
2(1 + |M |)
ε
3. |an − L| < for any n > N1 .
2(1 + |M |)
In Statement 2, we take ε2 = 1, and we find that there is a N2 > 0 such that for any n > N2 we have
|bn − M | < 1.
Since |bn | = |bn − M + M | ≤ |bn − M | + |M |, we conclude that
4. |bn | ≤ 1 + |M | for any n > N2 .

ε
Again, in Statement 2 we take ε3 = > 0, and we find that there is an N3 > 0 such that
2|L|
ε
5. |bn − M | < for any n > N3 .
2|L|
Hence, if N = min(N1 , N2 , N3 ), then for any n > N , Statements 3, 4 and 5 are valid, and therefore,
|an bn − LM | ≤ |bn ||an − L| + |L||bn − M |

ε ε
< (1 + |M |) + |L| = ε.
2(1 + |M |) 2|L| Q.E.D.
Limit of the Reciprocal

If
lim an = M and M 6= 0,
x→∞
then
1 1
lim = .
x→∞ an M
Proof. We want to show that for any given ε > 0, we can find an N > 0 such that for any n > N , we
have
1 1

− <ε
an M

We know that lim an = M ; hence,
x→∞
1. For any ε1 > 0, there is an N1 > 0 such that for any n > N1 we have |an − M | < ε1 .
To relate what we know to what we want to show, we do the operations:

1 1 M − an 1

− = = |an − M |
an M an M |an M |
From Statement 1, if |an − M | < ε1 then, by the properties of the absolute value,
ε1 > |an − M | = |M − an | ≥ |M | − |an |.
Thus,
|an | > |M | − ε1 .
Therefore,
1 1
< ,
|M an | |M |(|M | − ε1 )
and
1 ε1
|an − M | < .
|an M | |M |(|M | − ε1 )
If we want
ε1
ε= ,
|M |(|M | − ε1 )
we solve for ε1 , and we have
ε|M |2
ε1 = > 0.
1 + |M |ε
For this value of ε1 , in Statement 1, there is an N such that if n > N , then
ε|M |2
|an − M | < .
1 + |M |ε
We conclude that
1 1 1 ε1

− = |an − M | < = ε.
an M |an M | |M |(|M | − ε1 ) Q.E.D.
Finally, we can use the fact that

an 1
= an
bn bn
and conclude the limit of the quotient from the product of limits and the limit of the reciprocal.

If
lim an = L and lim bn = M, where M 6= 0,

x→∞ x→∞
then
an 1 1 1 L
lim = lim an = lim an lim =L = .
x→∞ bn x→∞ bn x→∞ x→∞ bn M M
Next, we prove Theorem 6.9 on page 165.
Theorem 6.17. Let {an } be a sequence, {a2k } its sub-sequence of even terms and {a2k+1 } its
sub-sequence of odd terms.
lim an = L
n→∞
if and only if
lim a2k = L = lim a2k+1 .

k→∞ k→∞
Proof. We assume that lim an = L, and we want to show that

n→∞

k→∞ k→∞
By Definition 6.16 on page 183, for any ε > 0 there is an N > 0, such that
|an − L| < ε for any n>N (6.1)
There is a positive integer K0 such that 2k > N for any k > K0 , and therefore, by Equation (6.1),
|a2k − L| < ε for k > K0 .
Similarly there is a positive integer K1 such that 2k + 1 > N for any k > K1 , then by Equation (6.1),
|a2k+1 − L| < ε for k > K1 .
By Definition 6.16 on page 183,

k→∞ k→∞
Conversely, we assume that
lim a2k = L = lim a2k+1 ,

k→∞ k→∞
and we want to show that lim an = L. Hence, by Definition 6.16 on page 183,
n→∞
1. for any ε > 0, there is a K0 > 0 such that |a2k − L| < ε for k > K0 , and

2. for any ε > 0, there is a K1 > 0 such that |a2k+1 − L| < ε for k > K1 .
If N = max{K0 , K1 }, and we take n > N , then n > K0 and n > K1 . Thus Statements 1 and 2 hold, and
we have two cases to consider.
Case 1
If n = 2k is even, then, since Statement 1 holds, |an − L| < ε.
Case 2
If n = 2k + 1 is odd, then Statement 2 holds, and |an − L| < ε.
In either case, we have |an − L| < ε for any n > N , and we conclude that lim an = L. Q.E.D.
n→∞
Theorems 6.10, 6.11 and 6.18 Theorem 6.10 (p. 166) and Theorem 6.11 (the Squeeze Theorem, p. 167)
are direct consequences of the following result.
Theorem 6.18. If {an } and {bn } are two convergent sequences, and
an ≤ b n for all n > k for some k,
then
lim an ≤ lim bn .
n→∞ n→∞
Proof. Let M = lim bn and L = lim an .

n→∞ n→∞
First, we will assume that L > M , an assumption that will take us to a contradiction. From this
contradiction, we conclude that our assumption is incorrect, and that what is true is that L ≤ M .
By the sum of limits we have
lim an − bn = L − M.
x→∞
If L > M , then L − M > 0. We have two hypothesis, set out in Statements 1 and 2, below.
1. For any ε1 > 0, there is an N1 > 0 such that
|bn − M | < ε1 for any n > N1 .
2. For any ε2 > 0, there is an N2 > 0 such that
|an − bn − (L − M )| < ε2 for any n > N2 .
In particular, we take ε2 = L − M > 0 in Statement 2, and we find that there is an N2 > 0 such that
3. |an − bn − (L − M )| < L − M for any n > N2 .

If we take N = min(N1 , N2 , k) then Statements 1 and 3 are true, and for any n > N we have
−(L − M ) < an − bn − (L − M ) < L − M.
Adding (L − M ), this conclusion implies that
0 < an − bn < 2(L − M ).
Therefore bn < an for n > N > k, contrary to our assumption.
Thus, L ≤ M . Q.E.D.
Theorem 6.19. If lim |an | = 0, then lim an = 0.

n→∞ n→∞
Proof. Since for any number x it is true that
−|x| ≤ x ≤ |x|,
we have
−|an | ≤ an ≤ |an | for all n.
By Theorem 6.18, above,
0 = lim −|an | ≤ lim an ≤ lim |an | = 0.

n→∞ n→∞ n→∞
Therefore, lim an = 0. Q.E.D.

n→∞
Theorem 6.20. Let {an }, {bn }, {cn } be three sequences such that, eventually,
an ≤ c n ≤ b n .
If
lim an = lim bn = L
n→∞ n→∞
then
lim cn = L.
n→∞
Proof. By Theorem 6.18,
L = lim an ≤ lim cn ≤ lim bn = L.

n→∞ n→∞ n→∞
Hence, lim cn = L. Q.E.D.

n→∞

The limit of a sequence {an } is infinite if we can make the terms of the sequence an as large as we want
by taking a sufficiently large n. The visual representation of an infinite limit is shown in Figure 6.7
Figure 6.7: Infinite limit
The statement “we can make the terms of the sequence an as large as we want” can be expressed by
saying that the terms of the sequence an are larger than M for an M that is as large as we want. That is,
given even a very large M > 0, we have an > M .
The statement “by taking a sufficiently large n” can be expressed mathematically by saying that there is
an N > 0 such that for any n > N , we have an > M .
Definition 6.21.
The limit of the sequence {an } is infinity if, for a given M > 0, there is an N > 0 such that
an > M for any n > N.
Figure 6.8 shows Definition 6.21.
Figure 6.8: Limit of a sequence an > M for any n > N
Next, we prove Theorem 6.15 on page 176.
Theorem 6.22. Let {an } and {bn } be two sequences. If
an ≤ b n for n > k for some k,
and if
lim an = ∞,
n→∞
then
lim bn = ∞.
n→∞
Proof. If lim an = ∞, then for any M > 0 there is an N0 > 0 such that an > M for all n > N0 .
n→∞
If we take N = max(N0 , k), then bn ≥ an > M for all n > N .
Hence, lim bn = ∞. Q.E.D.

n→∞

Exercises
5. Use Definition 6.16 on page 183 to prove the statements below.
a. If lim an = L and c is a nonnegative constant, then lim can = cL.

n→∞ n→∞
b. If {an } is an increasing sequence, lim bn = M and an ≤ bn for all n ≥ 1, then the sequence
n→∞
{an } is convergent and lim an = L ≤ M .
n→∞
6. a. Give the formal definition of the limit lim an = −∞.

n→∞
b. Use the definition in part (a), above, to prove the statement below.
Let {an } and {bn } be two sequences, such that
an ≥ b n for n > k for some k.
If lim an = −∞, then lim bn = −∞.

n→∞ n→∞
7. Use Definition 6.21 on the preceding page to prove the statements below.
a. If lim an = ∞ and lim bn = ∞, then lim an + bn = ∞.

n→∞ n→∞ n→∞
b. If lim an = ∞ and c > 0 is a constant, then lim can = ∞.

n→∞ n→∞
Series
An infinite series is the limit of a sequence, and you must not forget this fact.
Definition 6.23.
If {ak } is a sequence, then the sum of the nth partial sum sn of this sequence is defined as
X
n
s n = a1 + a2 + · · · + an = ak .
k=1
The partial sum sn is determined by the sum from the first to the nth terms of the sequence {ak }.
Example 6.44. If {k} is a sequence of positive integers, then
X
n
sn = 1 + 2 + 3 + · · · + n = k.
k=1
Hence, for example,
s6 = 1 + 2 + 3 + 4 + 5 + 6.

nno
Example 6.45. The partial sum s7 of the sequence is
en
1 2 7
s7 = + 2 + ··· + 7.
e e e
Example 6.46. If we consider the recursive sequence
1
a1 = 1 and an+1 = 1 + ,
a1
then the partial sum s5 is

1 1 1 1 1
s5 = 1 + 1 + + 1+ + 1+ + 1+ + 1+
a1 a2 a3 a4 a5

1 2 3
= 1 + (1 + 1) + 1 + + 1+ + 1+
2 3 5
1 2 3 37
=6+ + + =
2 3 5 5
We guess that
1 2 3 5 7
s7 = 8 + + + + + .
2 3 5 7 9
For any n, we guess that
1 2 3 2n − 5
sn = (n + 1) + + + + ··· + .
2 3 5 2n − 3
Definition 6.24.
For a given sequence {an }, its infinite series or series, denoted by
X ∞
X
an = an ,
n n=1
is the limit of the sequence {sn } of partial sums; that is,
X ∞
X X
n
an = an = lim sn = lim ak .
n→∞ n→∞
n n=1 k=1
∞
X
Note: You must understand this notation. an is not the infinite sum of the terms of the sequence
n=1
{an }—it simply does not make any sense to interpret it in this way. It is true that you occasionally see
∞
X
a1 + a2 + a3 + . . . an + · · · = an ,
n=1

but this notation is the result of laziness. We use it because we do not want constantly to write
X
n
lim ak .
n→∞
k=1

n!
Example 6.47. The series of the sequence is defined as
en
X∞ Xn
n! k!
n
= lim k
.
n=1
e n→∞
k=1
e
Example 6.48. The series of the recursive sequence

1
a1 = 1 and an+1 = 1 + ,
a1
from Example 6.46, is
∞
X 1 2 2n − 5
an = lim sn = lim (n + 1) + + + ··· + .
n=1
n→∞ n→∞ 2 3 2n − 3
Observe that if k < r, then the difference sr − sk is the sum of the r − k terms of the sequence; that is,
sr − sk = ak+1 + ak+2 + · · · + ar .
Hence, sn − sn−1 = an .
We may also consider a sequence with 0 as an index in the first term; that is,
{a0 , a1 , a2 , . . . }.
In this case, the partial sums are also indexed from 0,
s0 = a0 , s1 = a0 + a1 and sn = a0 + a1 + · · · + an .
The existence of the limit of the sequence of partial sums determines whether the series is convergent or
divergent.
Definition 6.25.
If the sequence of partial sums {sn } of the sequence {an } is convergent, then the series is said to be
convergent, if the limit of the sequence is either infinite or non-existent, then the series is said to be
divergent.

X
If the series an is convergent and s is the limit of the partial sum sequence, then we write
n
X
an = s which indicates that lim sn = s.
n→∞
n
The number s is called the sum of the series. If the sum of the series is not known, but it is known that the
series is convergent, then we write
X
an < ∞.
n
If the limit of the sequence of partial sums is infinity or does not exist, then we write

X  lim sn = ±∞ or
an = ∞ which indicates either n→∞
 lim s does not exist
n n
n→∞
∞
X
The convergence of a series is determined by the convergence of a partial sum an for some k.
n=k
That is, if for some k,

∞
X
an = lim tn
n→∞
n=k
where tn = ak + ak+1 + · · · + an , then

∞
X X
k−1 ∞
X
an = an + an .
n=1 n=1 n=k
Indeed, if
∞
X
an = lim sn
n→∞
n=k
where sn = a1 + a2 + · · · + an , then tn = sn − sk−1 and

∞
X
an = lim sn = lim tn + sk−1
n→∞ n→∞
n=1
X
k−1 ∞
X
= sk−1 + lim tn = an + an .
n→∞
n=1 n=k

Hence,
∞
X
an is convergent or divergent
n=1
if and only if
∞
X
an is convergent or divergent.
n=k
The evaluation of the sums of infinite series is for the most part elusive. Here, we present some known
sums and a method to find the sum of telescoping series.
Definition 6.26.
The geometric series is a series of the form

∞
X
arn−1 ,
n=1
where a and r are any nonzero real numbers.
The geometric series converges if

∞
X a
|r| < 1 and arn = .
n=0
1−r
It is divergent for |r| ≥ 1. It is important that you understand the argument used to find this limit in
Theorem 6.137: Geometric Series Convergence.
Observe that to find the sum of a geometric series, we must identify the numbers a and r. We do so by
looking at the first few terms of the series,
∞
X
arn−1 = a + ar + ar2 + ar3 + . . . .
n=1
The first and second terms of the series give these numbers.
X∞
1
Example 6.49. To determine whether the series is geometric, we write the first few terms
n=0
3n
X∞
1 1 1 1 1
n
= 1 + + 2 + 3 + 4 + ....
n=0
3 3 3 3 3

The series is geometric, with
1
a=1 and r= .
3
Therefore, the sum of the given series is equal to
1 3
= .
1 2
1−
3
Example 6.50. The series
∞
X (−6)n
n=1
5n
looks like a geometric series. We must identify the numbers a and r.

∞
X 2 3
(−6)n 6 6 6
=− + − + − + ......
n=1
5n 5 5 5
2
6 6 6 6 6
=− + − − + − − + ....
5 5 5 5 5
We can see that

6
a = − = r.
5
This geometric series is
X∞ n−1
6 6
− − .
n=1
5 5
Since |r| > 1, the series is divergent.
Numbers with periodic decimal representation are rational numbers (fractions). The geometric series is
used to find the representation as a fraction of such a number.
Example 6.51. The number 3.42125125125 · · · = 3.42125 is periodic, with period 125. Hence, it can
be represented as a fraction.
42 125 125
3.42125 = 3 + 2
+ 5 + 8 + ....
10 10 10
We concentrate on the terms of the period of the number
2
125 125 125 125 125 1 125 1
5
+ 8 + 11 · · · = 5 + 5 3 + 5 ....
10 10 10 10 10 10 10 103

This is a geometric series with
125 1
a= and r= ,
105 103
and
125
X∞ n−1
125 1 105 125
= = .
105 103 1 99900
n=1 1−
103
The fraction for this number is

42 125 341783
3+ 2
+ = .
10 99900 99900
A series is called “telescoping” if the terms of the sequence of partial sums reduces itself to one or two
terms as a result of the cancelation of terms. It is always possible to find the sum of a telescoping series.

X∞
1 1
−
n=1
n n+2
is a telescoping series because some of the terms in the partial sum cancel:

1 1 1 1 1 1
sn = − + − + −
1 1+2 2 2+2 3 3+2

1 1 1 1
+ ··· + − + −
n−1 n+1 n n+2

1 1 1 1 1 1 1
= 1− + − + − + −
3 2 4 3 5 4 6

1 1 1 1
+ ··· + − + −
n−1 n+1 n n+2
1 1 1 2 1 1
=1+ − − = − −
2 n+1 n+2 3 n+1 n+2
Hence,
3 1
sn = − ,
2 (n + 2)(n + 1)
and
X∞
1 1 3 1 3
− = lim − = .
n=1
n n + 2 n→∞ 2 (n + 2)(n + 1) 2

We also have well-known divergent series. One of them is the harmonic series:
X∞
1
= ∞.
n=1
n
You can see why in Example 6.142: Harmonic Series.

∞
X 1
n=1
n+2
is a variation of the harmonic series. We can see that

∞
X X1 ∞
1
= .
n=1
n + 2 n=3 n
Hence, the series is divergent.
Theorem 6.27. Test for Divergence

X
If an < ∞, then lim an = 0.
n→∞
n
This theorem says that if
lim an 6= 0,
n→∞
then the series

X
an
n
must be divergent.
This result, known as the “test for divergence,” is the first test we try to apply when we want to determine
if a series is convergent. The proof is given in Section 6.2 Series as Theorem 6.139: Divergence Test.
Theorem 6.28.
X
If lim an 6= 0, or it does not exist, then an < ∞.
n→∞
n
Example 6.54. You can apply the divergence test to see that each of the series below is divergent.
X
a. ln(n)
n

∞
X k2
b.
k=1
(2 − k)k
∞
X
c. cos(πk)
k=1
The laws of limits imply the following theorem.
Theorem 6.29.
If
X X
an < ∞, bn < ∞ and c is a nonzero constant,
n n
then each of the following series is convergent:

X X
a. can = c an .
n n
X X X
b. (an ± bn ) = an ± bn .
n n n
We must be careful with the application of this theorem. The series

X X
an and bn
n n
must be convergent to allow us to conclude that

X X X
(an ± bn ) = an ± bn .
n n n
We found, in Example 6.52 on page 197, that the series

X∞
1 1
−
n=1
n n+2
is convergent; however the series

X∞ ∞
X
1 1
and
n=1
n n=1
n+2
are divergent, and it is not true that

X∞ ∞
X ∞
1 1 1 X 1
− = − .
n=1
n n + 2 n=1 n n=1 n + 2

This discussion also shows that if two series
X X
an and bn
n n
are divergent, then it is not true that

X
(an ± bn )
n
is divergent.
Example 6.55. The geometric series

X∞ ∞
X
1 5
and −
n=1
2n n=1
6n
are convergent; hence

X∞ X∞ X∞ X∞
1 5 1 1
n
+ − n = n
+ (−5)
n=1
2 n=1
6 n=1
2 n=1
6n
X∞
1 5 16
= n
− n
= .
n=1
2 6 11
∞ k
X ∞
X ∞
X
2 3 2k 3
k
+ 6= + ,
k=1
5 k k=1
5k k=1
k
because
∞
X 2k
is convergent [why?]
k=1
5k
and
∞
X 3
is divergent [why?]
k=1
k
We also have theorems that help us to decide if a series is divergent.
Theorem 6.30.
∞
X X1
If an < ∞, then = ∞.
n n
an

Proof. By Theorem 6.27 on page 198,
lim an = 0,
n→∞
and therefore,
1
lim = ±∞.
n→∞ an
X1
Hence, by Theorem 6.28 on page 198, the series is divergent. Q.E.D.
n
a n
Theorem 6.31.
If
∞
X
an = ∞ and c is a nonzero constant,
n
then
∞
X
can = ∞.
n
Proof. The limit of partial sums lim sn is ±∞ or does not exist; therefore, the limit lim csn is also
n→∞ n→∞
equal to ±∞ or does not exist. Q.E.D.
Theorem 6.32.
X X X
If an < ∞ and bn = ∞, then an + bn = ∞.
n n n
Proof. If we assume that

X
an + bn < ∞,
n
then by Theorem 6.29 on page 199 the series

X
(−an ) < ∞
n
and
X X X X
∞> an + b n + (−an ) = an + bn + (−an ) = bn .
n n n n

This contradiction indicates that our assumption is incorrect and
X
an + bn = ∞.
n Q.E.D.
Example 6.57. In Example 6.56 on page 200, we have the series

∞ k
X
2 3
+ ,
k=1
5k k
and by Theorem 6.32, this series is divergent.
Note: Often, when the the determination of the convergence of a series is incorrect, it is because either
the theorems are applied incorrectly (under the incorrect assumptions) or because we try to “guess” by
looking at the terms of the series.
For the next sections, you will need to have a list of convergent and divergent series. Make a list of the
most important series we presented in this section. Always apply theorems properly to decide the
convergence of a series. Do not rely in your “feelings” or “intuition.”
Exercises
8. Read Section 6.2 of the textbook, Series.
9. Do the exercises in the section Exercises for 6.2 of the textbook.

∞
X 2
10. Show that the series is telescopic, and give its sum.
n=1
n2 + 4n + 3
11. For further practice, complete the lab Series in the optional Lyryx Labs. Take note that in this lab the
“nth term test” is what we call the Divergence Test (Theorem 6.27 on page 198).
The Integral Test and Estimates of Sums

The integral test results from the link between improper integrals and series when the terms of the series
are nonnegative and the sequence of partial sums is decreasing. The test is stated here, and you should
read Example 6.143 of Section 6.3 of the textbook, The Integral Test.
Theorem 6.33. The Integral Test
Let f be the function associated with the sequence {an }. If f is continuous and decreasing on the
interval [k, ∞) for some positive integer k, then
Z ∞ ∞
X
a. if f (x) dx is convergent, then an is convergent.
k i=1

Z ∞ ∞
X
b. if f (x) dx is divergent, then an is divergent.
k i=1
Proof. Example 6.143 of the textbook gives you an idea behind of the proof of this theorem. We have a
decreasing continuous function f (x) associated to the sequence. That is, f (n) = an for all natural
numbers n.
𝑦 = 𝑓(𝑥)
𝑦 = 𝑓(𝑥)
𝑎2 𝑎3 𝑎4 𝑏1 𝑏2 𝑏3
𝑏𝑛
𝑎𝑛
1 2 3 4 5 6 7 ⋯ ⋯ 1 2 3 4 5 6 7 ⋯ ⋯
Figure 6.9: Areas of an (below the curve) and bn (above the curve)
In the above figure 6.9, the area of a2 is equal to f (2). So a2 = f (2), a3 = f (3), ...an = f (n) and we can
see that the sum of these areas is less than the area below the curve. Similarly,
b1 = f (1), b2 = f (2), ...bn = f (n), and the sum of the areas is bigger than the area below the curve.
Hence,
X
n Z n X
n−1
ak ≤ f (x) dx ≤ bk .
k=2 1 k=1
Z ∞
i. If f (x) dx is convergent; for any n
1
X
n Z n Z ∞
an ≤ f (x) dx ≤ f (x) dx = M.
k=2 1 1
and for any n

X
n Z n
s n = a1 + ≤ f (x) dx ≤ M.
k=2 1
So the sequence of partial sums sn is bounded above and
sn+1 = sn + an+1 ≥ sn

because an+1 = f (n + 1) ≥ 0. This shows that the sequence sn is increasing and bounded above and
therefore convergent by the Monotonic Sequence Theorem (Theorem 6.14 on page 175); thus, the
series is convergent.
Z ∞ Z ∞
ii. If f (x) dx is divergent, then f (x) dx → ∞ because f (x) ≥ 0. And
1 1
Z n X
n−1
f (x) dx ≤ bk = sn−1 .
1 k=1
so, sn−1 → ∞ and the series is divergent.
Q.E.D.
Observe that the conclusion of this test is about convergence, and not about the sum of the series.
The primary example of the application of this test is the p-series. You should put it on your list of
convergent and divergent series.
Example 6.58. A p-series (p > 0) is of the form

X∞
1
p
,
n=1
n
and it is convergent if p > 1 and divergent if p ≤ 1.

X∞
3 3
a. √ is a p-series with p = ; hence, it is convergent.
n=1
n n 2
X∞
1 1
b. √ is a p-series with p = ; hence, it is divergent.
n=1
n 2
Example 6.59. We need to find the values of p > 0 for which the series
∞
X ln n
.
n=1
np
converges and diverges.
The function associated with the sequence of the series is

ln x
f (x) = ,
xp
and
1 − p ln x
f 0 (x) = < 0 for x ≥ e1/p .
xp+1
Hence, the function is decreasing.

For p = 1 we have the improper integral
Z ∞ Z t
ln x ln x
dx = lim dx
1 x t→∞ 1 x
(ln x)2 t (ln t)2
= = lim = ∞.
2 1 t→∞ 2
Hence, the series is divergent.
For p 6= 1, we use parts in the following evaluation

Z ∞ Z t
ln x ln x
p
dx = lim dx
e1/p x t→∞ e1/p xp

1 1−p t1−p e1−p/p e1−p/p
= lim t ln t − − +
t→∞ 1 − p 1−p p(1 − p) (1 − p)2

1 1 e1−p/p e1−p/p
= lim t1−p ln t − − + .
t→∞ 1 − p 1−p p(1 − p) (1 − p)2


 0 if
1 1−p<0
lim t1−p ln t − =
t→∞ 1−p  ∞ if 1 − p > 0.
Hence, the integral is convergent if 1 − p < 0, and divergent for 1 − p > 0. In conclusion, the series is
convergent for p > 1 and divergent for p ≥ 1.
X
If an is a convergent series with sum s, then
n
s = lim sn and sn ≈ s.
n→∞
The distance from the actual value of the sum s and an approximation sn is called the remainder, and we
denote it as Rn ; that is,
Rn = |s − sn |.
When the sequence of partial sums is increasing, then sn < s, and Rn = s − sn . This is the case when the
terms of the sum are positive.
If we can apply the integral test, then it is possible to control the value of the remainder Rn , and so
estimate how far the approximation of sn is from the actual value of the function s.
Remainder Estimate for the Integral Test

Let f be the function associated with the sequence an .

X
If f is positive, continuous and decreasing on [k, ∞) for some integer k, and the series an is
n
convergent, then
Z ∞ Z ∞
f (x) dx ≤ Rn ≤ f (x) dx
n+1 n
for any n > k. This relationship controls the size of the remainder.
Therefore, for any n > k

Z ∞ Z ∞
sn + f (x) dx ≤ s ≤ sn + f (x) dx.
n+1 n
This relationship gives a lower and upper bound for the sum s; that is, an interval on which the sum is
located.
Example 6.60. In Example 6.58 on page 204, we saw that the series
∞
X 3
√
n=1
n n
is convergent.
If we wish to find an n such that the remainder Rn is less than 1, we evaluate

Z ∞ t
3 −1/2 6
3/2
dx = lim −6x = lim −6t−1/2 + 6n−1/2 = √ .
n x t→∞ n t→∞ n
Hence, we want n such that

6
Rn ≤ √ < 1.
n
Solving for n, we have n > 36. We then have s − s37 < 1, which says that
X37
3 3 3 3 3
s37 = 3/2
= + 3/2 + 3/2 + · · · + 3/2 ≈ s,
n=1
n 1 2 3 37
and the distance from s37 to s is less than 1.
If we use s5 to approximate the value of s, we have

Z ∞ Z ∞
3 3
s5 + 3/2
dx ≤ s ≤ s5 + 3/2
dx.
5+1 x 5 x
Thus,
6 6
s5 + √ ≤ s ≤ s5 + √ ,
6 5

and
3 3 3 3 3
s5 = + 3/2 + 3/2 + 3/2 + 3/2 ≈ 8.102454.
1 2 3 4 5
This equation gives the interval 10.551944 ≤ s ≤ 10.785736, and
R5 = s − s5 ≈ 2.5663858 < 2.6.
The middle point of the interval [10.551944, 10.785736] gives another, better approximation, and we find
that
s ≈ 10.6688.
Exercises
12. Read Section 6.3 of the textbook, Integral Test.
14. For further practice, complete the lab Integral Test in the optional Lyryx Labs.
The Comparison Test

To apply the comparison test, you will need a list of convergent and divergent series.
Theorem 6.34. The Comparison Test
Let {an } and {bn } be two sequences of eventually positive terms.

X
a. If bn < ∞ and an ≤ bn for all n > k for some positive integer k, then
n
X
an < ∞.
n
X
b. If bn = ∞ and an ≥ bn for all n > k for some positive integer k, then
n
X
an = ∞.
n
Proof. Let
X
n X
n
sn = ai tn = bi .
i=1 i=1
Since the terms of the sequences are positive, the sequences sn and tn are increasing:

a. if an ≤ bn , then sn ≤ tn , and by Theorem 6.15 on page 176, the sequence sn ,
X X
and hence the series an , is convergent if bn is convergent.
n n
b. if an ≥ bn , then sn ≥ tn , and by Theorem 6.15, the sequence sn , and hence the

X X
series an , is divergent if bn is divergent.
n n
Q.E.D.
To
Xapply this test, we need a sequence bn with which to compare an , and we must know if the series
bn is convergent or divergent. Here is where we use our list of convergent and divergent series. To
choose bn , we try several things, one of which is to identify the expression bn in the sequence an that
“dominates.” Another is to consider bn as the sequence that most resembles the sequence an .
One of the difficulties with this test is to obtain the correct inequality; the test is inconclusive if
X
an ≥ bn and bn is convergent,
n
or if
X
an ≤ b n and bn is divergent.
n
X 1
Example 6.61. Consider the series .
n
2n + 3
1 1
The part of the term that is dominant is .
2n + 3 2n
1 1
However, the inequality we have is ≤ ,
2n + 3 2n
X 1
and the series is divergent.
n
2n
Hence, the comparison test does not apply.

∞
X 1 + sin n
Example 6.62. The series may be considered as
n=0
10n
X∞
1 sin n
n
+ n
.
n=0
10 10
X∞
1 1
The series n
is geometric, with r = < 1, and therefore convergent.
n=0
10 10

sin n
Furthermore, is not eventually positive.
10n
Hence, we cannot consider the sum as
X∞
1 sin n
n
+ ,
n=0
10 10n
∞
X 1 + sin n
and we must apply the comparison test to .
n=0
10n

1 + sin n
The terms of the sequence are positive for all n ≥ 0.
10n
1
The term that dominates is .
10n
We compare an arbitrary term of the sum with the dominant term:
1 + sin n 2
n
≤ n.
10 10
X∞
2
Since the series is convergent, the given series is also convergent.
n=0
10n
X 1
Example 6.63. For the series , we look into the sequence
n
nn

1 1 1 1 1
= ... ≤ .
nn n n n n2
X 1
The series 2
is convergent (p-series with p > 1).
n
n
Therefore, by the comparison test, the given series is convergent.
In Example 6.60 on page 206, we saw the difficulty of applying the comparison est because of inequality
between the sequences. In such cases, we try to apply the limit comparison test.
Theorem 6.35. The Limit Comparison Test
Let {an } and {bn } be two sequences of eventually positive terms.
If
an
lim =c
n→∞ bn
then

a. both series converge or both series diverges if c is a positive number.
X X
b. an < ∞ if c = 0 and bn < ∞.
n n
X X
c. an = ∞ if c = ∞ and bn = ∞.
n n

1
(2n + 3) 2n
lim = lim = 1 > 0.
n→∞ 1 n→∞ 2n + 3
2n
X 1 X 1
Since = ∞, the series is also divergent.
n
2n n
2n + 3
∞
X ln n ln n
Example 6.65. For the series we know that lim = 0.
n=1
n3 n→∞ n

1
Hence, we consider the sequence :
n2
n
ln
lim n3 = lim ln n = 0.
n→∞ 1 n→∞ n
n 2
X1
The p-series is convergent.
n
n2
X ln n
Therefore, by the limit comparison test, the series is convergent.
n
n3
Proposition 6.36.
If the sequence {an } is eventually positive and lim nan 6= 0, then

n→∞
X
an = ∞.
n
Proof. Since
an
0 6= lim nan = lim ,
n→∞ n→∞ 1
n

X1 X
and the harmonic series is divergent, by the limit comparison test, the series an is
n
n n
divergent. Q.E.D.
Exercises
15. Read Section 6.5 of the textbook, Comparison Tests.
17. For further practice, complete the lab Comparison Test in the optional Lyryx Labs.
Alternating Series
In this section, we consider sequences where the terms alternate between positive and negative numbers.
More precisely, if {bn } is a sequence of positive terms, then the sequences
{(−1)n bn } = {−b1 , b2 , −b3 , . . . }
and
{(−1)n−1 bn } = {b1 , −b2 , b3 , −b4 , . . . }
are alternating sequences. The corresponding series

X X
(−1)n bn and (−1)n−1 bn
n n
are called alternating series.
Example 6.66. Observe that

n πn o
{cos(nπ)} = {−1, 1, −1, 1, −1, . . . } = sin .
2
Hence, if {bn } is a sequence of positive terms, the series

X X n
cos(nπ)bn and sin π bn
n n
2
are alternating
Theorem 6.37.
X
The alternating series (−1)n bn is convergent if the following two conditions hold:
n

a. the sequence {bn } is eventually decreasing; that is,
bn+1 ≤ bn for all n>k
for some positive integer k; and
b. lim bn = 0.
n→∞
The proof of this theorem is given in Theorem 6.150: Alternating Series Test of the textbook.
One important example of a convergent alternating series is the alternating harmonic series:

(−1)n
.
n
See the first example in Section 6.4 of the textbook, Alternating Series.
Example 6.67. We apply the alternating test to the series

X (−1)n
.
n
nn
1 1
From Example 6.63 on page 209, we have ≤ .
nn n2
1 1
Since lim 2
= 0 , by the Squeeze Theorem, lim n = 0.
n→∞ n n→∞ n
We have
1
nn (n + 1)n+1 1
= = 1+ (n + 1) ≥ 1,
1 nn n
(n + 1)n+1

1
and therefore the sequence is decreasing, and by the alternating test, the given series is
nn
convergent.
An alternating series does not satisfy the conditions of the alternating test if either the sequence {bn } is
not decreasing or the limit lim bn 6= 0.
n→∞
If lim bn 6= 0, then the limit lim (−1)n bn is nonzero or does not exist. In either case, the divergence test
n→∞ n→∞X
implies that the alternating series (−1)n bn is divergent.
n
If the sequence {bn } is not decreasing, then we cannot conclude anything about the alternating series
X
(−1)n bn ; that is, the alternating test is inconclusive.
n

X
In the example below, we consider a convergent alternating series (−1)n bn such that the limit
n
lim bn = 0 and the sequence {bn } is not decreasing.
n→∞
Example 6.68. Let {bn } be a sequence where


 1

 n2 n is odd
bn =

 1

n is even
n3
It is clear that the sequence {bn } is not eventually decreasing.

1
We have bn ≤ for any n, odd or even.
n2
X
Hence, by the Squeeze Theorem lim bn = 0, and by the comparison test, the series bn is convergent.
n→∞
n
Observe that since bn ≥ 0,

2
0 ≤ (−1)n bn + bn ≤ 2bn ≤ .
n2
X2 X
The p-series 2
is convergent, and by the comparison test, the series (−1)n bn + bn is convergent.
n
n n
Therefore,
X X X X
(−1)n bn + bn − bn = (−1)n bn + bn − bn = (−1)n bn
n n n n
is convergent.
From this observation, we recommend that, when you apply the alternating test, you evaluate the limit
lim bn first. If this limit does not exist, the alternating series is divergent. If the limit is 0, then check
n→∞
whether the sequence {bn } is decreasing; if it is, the alternating series is convergent; if it is not, no
conclusion is possible.
If an alternating series is convergent because it satisfies the two conditions of the alternating test, then it
is possible to estimate its remainder. In this case |Rn | ≤ bn+1 .
If an alternating series is convergent for a reason other than the alternating test, then it is not true that
|Rn | ≤ bn+1 .
Exercises
18. Read Section 6.4 of the textbook, Alternating Test.

20. For further practice, complete the lab Alternating Test in the optional Lyryx Labs.
Absolute Convergence and the Ratio and Root Tests

X
For the convergence of a series an with no positive terms, we consider the corresponding series
X n
|an | of absolute values, because one of the most interesting things about series is that, if the series of
n X X
absolute values |an | is convergent, then so is the series an , without the absolute values.
n n
This fact opens several possibilities; for example, we can apply the tests we already have forXpositive
terms, or we can apply two new tests, known as the ratio test and the root test, to the series |an |.
n
Definition 6.38.
X
A series an is
n
X
a. absolutely convergent if |an | < ∞.
n
X X
b. conditionally convergent if an < ∞ and |an | = ∞.
a n
Observe that a series is either absolutely convergent or conditionally convergent but not both. Certainly a
series can be neither.
Example 6.69. The alternating harmonic series

X (−1)n
n
n
is conditionally convergent. See Example 1 on page 291.
Example 6.70. The alternating series

X (−1)n
n
nn
is absolutely convergent because

X (−1)n X 1
n = < ∞,
n
n n
nn
by Example 6.67 on page 212.

Example 6.71. You can verify that the alternating series
X (−1)n
√
n
n
is convergent by the alternating test.

X 1
The series of absolute values is divergent because it is a p-series ,
n
n1/2
1
with p = < 1.
2
X (−1)n
Hence, the series √ is conditionally convergent.
n
n
Example 6.72. Since

n n n
lim ≥ lim = ∞,
n→∞ 5 n→∞ 5
X n n
by the Divergence Test, both the alternating series − and the series of absolute values
5
X n n n
are divergent.
n
5
X n n
Thus, the series − is neither absolutely nor conditionally convergent.
n
5
Theorem 6.39.
X X
If |an | < ∞, then an < ∞.
n n
The proof of this theorem is in Theorem 6.157: Absolute Convergence, Section 6.6 of the textbook,
Absolute Convergence.
When the terms of a sequence are negative and the series is not alternating, we consider the series of
absolute values.

X cos(3n)
n
2n
is not alternating, and it has positive and negative terms.

X cos(3n)
Hence, we consider the series .
n
2n

cos(3n) 1

We have n
≤ n.
2 2
X 1
The series n
is convergent. Why?
n
2
X cos(3n)
Therefore, by the comparison test, the series is convergent.
n
2n
X cos(3n)
Hence, the series is absolutely convergent, and by Theorem 6.39 on the preceding page, the
n
2n
series is convergent.
The ratio and root tests apply to series of absolute values terms.
Theorem 6.40. The Ratio Test

a X
n+1
a. If lim = L < 1, then the series an is absolutely convergent (and therefore
n→∞ an n
convergent).
a a X
n+1 n+1
b. If lim = L > 1 or lim = ∞, then the series an = ∞.
n→∞ an n→∞ an n
a
n+1
c. If lim = 1, the ratio test is inconclusive; that is, no conclusion can be drawn about the
n→∞ an X
convergence or divergence of an .
n
The proof of this theorem is in Theorem 6.160: The Ratio Test, Section 6.6 of the textbook.
Theorem 6.41. The Root Test

p X
a. If lim |an | = L < 1, then the series an is absolutely convergent (and therefore
n→∞
n
convergent).
p p X
b. If lim |an | = L > 1 or lim |an | = ∞, then the series an = ∞.
n→∞ n→∞
n
p
c. If lim |an | = 1, the ratio test is inconclusive; that is, no conclusion can be drawn about the
n→∞ X
convergence or divergence of an .
n
The proof of this theorem is left as an exercise.
The ratio test is particularly useful when the terms of the series involve factorials.

Example 6.74. For the sequence
X n!
,
n
2n
we apply the ratio test and find
(n + 1)!
n+1 (n + 1)!2n n+1
lim 2 = lim = lim = ∞.
n→∞ n! n→∞ n!2n+1 n→∞ 2
2n
By the ratio test the series is divergent.
Observe that we did not use the absolute value in this example, because the terms of the series are
positive.
Example 6.75. If we apply the ratio test to the series

X (−1)n √n
2
,
n
1 + n
we have
√ √
an+1 (−1)n+1 n + 1 1 + n2 (1 + n2 ) n + 1
= √ = √ ,
an 1 + (n + 1)2 (−1)n n (1 + (n + 1)2 ) n
Hence,
r
1 + n2 n
lim = 1.
n→∞ n2 + 2n + 2 n+1
In this instance, the ratio test is inconclusive. We try next to see if the series is absolutely convergent.
We have
X (−1)n √n X √n
2
= 2
.
n
1 + n n
1 + n
X 1
We apply the Limit comparison test with the convergent series 3/2
.
n
n
Thus,
√
n
2 n2
lim 1 + n = lim = 1 > 0.
n→∞ 1 n→∞ 1 + n2
n3/2
We conclude that the series is absolutely convergent, and therefore convergent.

It is possible to estimate the remainder of a series if the series is convergent because of the ratio test.
P an+1
Proposition 6.42. Let an be a series with positive terms and let rn = . Suppose that
X an
lim rn = L < 1, so the series an is convergent by the ratio test. Let Rn be the remainder of the
n→∞
series
Rn = an+1 + an+2 + an+3 . . .
a. If {rn } is a decreasing sequence and rn+1 < 1, then

an+1
Rn ≤ .
1 − rn+1
b. If {rn } is an increasing sequence, then

an+1
Rn ≤ .
1−L
Proof. a. We factor an+1 from Rn , and use the fact that rk+1 ≤ rk for all k

an+2 an+3 an+4
Rn = an+1 1 + + + ...
an+1 an+1 an+1

an+2 an+3 an+2 an+4 an+3 an+2
= an+1 1 + + + ...
an+1 an+2 an+1 an+3 an+2 an+1
= an+1 (1 + rn+1 + rn+2 rn+1 + rn+3 rn+2 rn+1 + . . . )
2 3
≤ an+1 (1 + rn+1 + rn+1 + rn+1 + ...)
X∞
k
= an+1 rn+1
k=0
∞
X
The series k
rn+1 is geometric (Definition 6.26 on page 195) with a = 1 and r = rn+1 . Since
k=0
1
rn+1 < 1, the series is equal to . Hence,
1 − rn+1
an+1
Rn ≤ .
1 − rn+1
b. Since rn → L as n → ∞, we have, rn < L for all n. As before
Rn = an+1 (1 + rn+1 + rn+2 rn+1 + rn+3 rn+2 rn+1 + . . . )

≤ an+1 (1 + L + L2 + L3 . . . )
X∞
= an+1 Lk
k=0

∞
X
The series Lk is again geometric with a = 1 and r = L. Since L < 1, the series is equal to
k=0
1
. Hence,
1−L
an+1
Rn ≤ .
1−L
Q.E.D.
We apply this proposition to estimate the error in the value of a series by using a partial sum of the series.
Example 6.76. Consider the series

X∞
1
.
n=1
n2n
We know that this series is convergent. We want to estimate the error in using the partial sum s5 as an
approximation to the sum of the series.
1 1 1 1 1 661
s5 = + + + + = ≈ 0.68854.
2 8 24 64 160 960
n2n n
The ratios rn = n+1
= form an increasing sequence because
(n + 1)2 2(n + 1)
n+1 n (n + 1)2 − n(n + 2) 1

rn+1 − rn = − = = > 0.
2(n + 2) 2(n + 1) 2(n + 1)(n + 2) 2(n + 1)(n + 2)
And
n 1
lim rn = lim = .
n→∞ n→∞ 2(n + 1) 2
We can apply part b. of the previous proposition

a6 2 1
R5 ≤ = 6
= ≈ 0.00521.
1 − 1/2 6(2 ) 192
P∞ 661 1
This says that the sum of the series 1
n=1 n2n is approximately equal to with an error of .
960 192
Moreover, the error in using sn as an approximation to the sum if the series S is
an+1 2
S − sn = Rn = = .
1 − 1/2 (n + 1)2n+1

If we want to estimate the value of S with an error less than 0.00005, we must find n so that
Rn < 0.00005. Thus,
1
Rn = < 0.00005 ⇒ (n + 1)2n > 20, 000.
(n + 1)2n
We use trial and error or the graph of g(x) = (x + 1)2x to find that for n = 11, we have
(11 + 1)211 = 24, 576. That is,
X11
1
s11 = n
≈ 0.693109 ⇒ S ≈ 0.693109 with an error of less than 0.00005.
n=1
n2
Pay attention to the observations below in the application of tests for series.
1. The integral, comparison, and limit comparison tests are for positive terms.
2. You can apply the integral, comparison, limit comparison, ratio and root tests to the absolute value of
the terms of the series.
3. The alternating test is conclusive only when the two conditions (discussed on page 211) are fulfilled.
a a
n+1 n
4. In the ratio test, the limit must be not .
an an+1
Exercises
21. Read Section 6.6 of the textbook, Absolute Convergence.
23. Read Section 6.7 of the textbook, The Ratio and Root Test.
24. Do the exercises of section Exercises for 6.7 of the textbook.
25. Proof the Root Test 6.41 on page 216.
26. Use Proposition 6.42 on page 218 of this Study Guide to estimate the error in the approximation of
X∞
n
the sum of the first 8 terms of the sum .
n=1
2n
27. Do the exercises in the two videos, Strategy for Testing Series. Videos Series 1 and Series 2.
28. For further practice, complete the lab Ratio and Root Tests in the optional Lyryx Labs.
29. Complete the online Lyryx Assessment quiz titled Unit 6 - I. Your grade on this quiz will count

Representation of Functions as Power Series
2
The function ex does not have an elementary antiderivative, but it can be expressed as a power series.
That is, for all x,
∞
X
x2
e = cn x n .
n=0
An important achievement in the study of series is a theorem that says that

Z ∞ Z
X
x2
e dx = cn xn dx.
n=0
2
Note that we may not have an elementary antiderivative for ex , but we can integrate
Z
cn xn dx,
and we can try to find out the convergence of the series of integrals.
Moreover, we have
Z k Z
X
x2
e dx ≈ cn xn dx.
n=0
The larger the k, the better the approximation.
This approximation is true for any function that can be expressed as a power series and is the main reason
we study power series.
If |x| < 1, then

∞
X
xn
n=0
is a geometric series, and

∞
X 1
xn = .
n=0
1−x
This equation gives a representation of the function

1
f (x) =
1−x
as a power series, but is true only for |x| < 1.
We can use this discussion to find the power representation of other functions that resemble f .

Example 6.77. The function
3
g(x) =
1 − 9x2
1
is similar to f (x) = . You can see that, if y = 9x2 , then
1−x
X∞
3 1 1
g(x) = 2
=3 2
=3 =3 yn
1 − 9x 1 − 9x 1−y n=0
∞
X ∞
X ∞
X
= 3y n = 3(9x2 )n = 27x2 n.
n=0 n=0 n=0
1
Since this is true if |y| < 1, we have 9x2 < 1 and |x| < .
3

1 1 1
Hence, the function g is expressed as a power series only if |x| < , which gives the interval − , .
3 3 3
Definition 6.43.
A power series around the number a is a series of the form

∞
X
cn (x − a)n .
n=0
The terms of the sequence {cn } are the coefficients of the power series.
Note that if
∞
X
cn (x − a)n
n=0
is a power series, then for x = a, the series is convergent to c0 . Hence, any power series is convergent for
at least one number.
Theorem 6.44.
If
∞
X
cn (x − a)n
n=0
is a power series, then one and only one of the statements given below holds.
a. The series converges for x = a only. In this case, the radius of convergence of the power series
is 0.

b. The series converges on an interval (a − R, a + R) for some number R. In this case, the radius
of convergence of the power series is R.
c. The series converges for all x. In this case, the radius of convergence is infinity.
Note that the interval (a − R, a + R) is also denoted |a − x| < R.
1
Example 6.78. The power series of Example 6.77 is around 0, and its radius of convergence is R = .
3
Example 6.79. The power series

∞
X xn
n=0
n!
1
has coefficients cn = , and by the ratio test, the series converges for all x because
n!
xn+1 n! 1

lim = lim |x| = 0 < 1.
n→∞ (n + 1)! xn n→∞ n+1
Hence, the radius of convergence of the power series is infinity.
The following theorem is the main one of this unit.
Theorem 6.45.
If the power series

∞
X
cn (x − a)n
n=0
has radius of convergence R > 0, then the function

∞
X
f (x) = cn (x − a)n
n=0
= c0 + c1 (x − a) + c2 (x − a)2 + · · · + cn (x − a)n + . . .
is differentiable on the interval (a − R, a + R) or on R and

∞
X
a. f (x) = 0
ncn (x − a)n−1
n=1
= c1 + c2 (x − a)
+ c3 (x − a)2 + · · · + ncn (x − a)n−1 + . . . .

Z ∞
X (x − a)n+1
b. f (x) dx = C + cn
n=0
n+1
= C + c0 (x − a)
(x − a)2 (x − a)n+1
+ c1 + · · · + cn + ....
2 n+1
The radius of convergence of each power series is R.
1
Example 6.80. In Example 6.78, we found R = to be the radius of convergence of the power series
3
X ∞
1
= 27x2 n,
1 − 9x2 n=0
and by Theorem 6.45, above,

18x d 1
2 2
=
(1 − 9x ) dx 1 − 9x2
d
= (27 + 27x2 + 27x6 + 27x8 + . . .
dx
= 27(2)x + 27(6)x5 + 27(8)x7 + . . .
X∞ ∞
X
2n−1
= 27(2n)x = 54n(x2n−1 ),
n=1 n=1
and
Z
3 1 + 3x dx
ln =
2 1 − 3x 1 − 9x2
Z Z Z Z
= C + 27 dx + 27x dx + 27x dx + 27x8 dx + . . .
2 6
27 27 27
= C + 27x + x3 + x7 + x9 + . . .
3 7 9
X 27
=C+ (x2n+1 ).
n=0
2n + 1
1
The radius of convergence of each series is R = .
3
The next example is an interesting application of Theorem 6.45 on the preceding page.

Example 6.81. To obtain the power series of ln(1 − x), we see that this function is the antiderivative of
1
. Hence, for |x| < 1
1−x
Z Z
dx
− ln(1 − x) = = (1 + x + x2 + x3 + . . . ) dx
1−x
x2 x3
=x+ + + ··· + C
2 3
X∞ X∞
xn+1 xn
=C+ = +C
n=0
n + 1 n=1
n
To find the value of C we set x = 0 on this equation and we obtain − ln(1 − 0) = C. thus, C = 0 and
∞
X xn
ln(1 − x) = − for |x| < 1
n=1
n
The radius of convergence is R = 1.
To obtain the power series of ln(x − 1) we see that ln(x − 1) = ln(1 − (2 − x)). Hence,
∞
X (2 − x)n
ln(x − 1) = ln(1 − (2 − x)) = −
n=1
n
(2 − x)2 (2 − x)3
= −(2 − x) − − ...
2 3
(x − 2)2 (x − 2)3
= (x − 2) − + + ...
2 3
X∞
(−1)n
= n
n=1
(x − 2)n
This is for |x − 2| < 1; that is, 1 < x < 3.
This example shows how Theorem 6.45 can be used to find power series of functions.
If power series are added, subtracted, multiplied or divided, they behave like polynomials. In the next
examples, we show you how power series are used to find the first few terms (and some of the most
important ones) of their product and division.

∞
X ∞
X
xn (−1)n x2n
ex = cos x = .
n=0
n! n=0
(2n)!
The power series of ex cos x is the product of their series; that is,

x x x2 x3 x2 x4
e cos x = 1 + + + ... 1− + ...
1! 2! 3! 2! 4!

We make a vertical multiplication
x2 x3 x4 x5
1 +x + 2
+ 6
+ 24
+ 120
...
x2 x4
1 − 2
+ 24
...
x2 x3 x4 x5
1 +x + 2
+ 6
+ 24
+ 120
...
x2 x3 x4 x5
− 2
− 2
− 4
− 12
...
x4 x5
+ 24
+ 24
...
.....................................
x3 x4 x5
1 +x − 3
− 6
− 30
...
x3 x4 x5
Thus, ex cos x = 1 + x − − − ...
3 6 30
sin x
Example 6.83. For tan x = , we make a long division of the first few terms of
cos x
∞
X ∞
X
(−1)n x2n+1 (−1)n x2n
sin x = and cos x = .
n=0
(2n + 1)! n=0
(2n)!
You must verify that
x3 2x5
tan x = x + + + ....
3 15
We still have to investigate the radius of convergence of the

P product and division
P of power series. There is
a theorem that states that if both power series of f (x) = an and g(x) = bn are convergent for
|x| < R, then the product of f (x)g(x) is convergent for |x| < R. For the division, we require b0 6= 0; the
resulting series is convergent for sufficiently small |x|.
In some instances we may apply power series to evaluate limits.

1 − cos x
lim
x→0 1 + x − ex
using power series, we find the power series of the numerator and denominator

x2 x4 x6 x2 x4 x6
1 − cos x = 1 − 1 − + − + ... = − + − ...
2! 4! 6! 2! 4! 6!

x x 2 x3 x2 x3
1+x−e =1+x− 1+x+ + + ... = − − + ...
2! 3! 2! 3!

Hence,
x 2 4 6
1 − cos x 2!
− x4! + x6! − . . .
lim = lim 2 3
x→0 1 + x − ex x→0 − x − x + . . .
2! 3!
1 2 3
2!
− x4! + x6! − . . .
= lim = −1
x→0 − 2!1 − 3!x + . . .
If we make a long division we only need the first constant terms of the series because the limit is at zero.
Thus,
1 − cos x x
= −1 + . . .
1+x−e x 3
and again
1 − cos x
lim = −1.
x→0 1 + x − ex
Exercises
30. Read Section 6.8 of the textbook, Power Series.
32. Read Section 6.9 of the textbook, Calculus with Power Series.
34. Find the ratio of convergence of the series listed below.

∞
X
a. n!xn .
n=0
X∞
(−3)n xn
b. √
n=0
n+1
∞
X (−1)n x2n
c. J0 (x) = (The Bessel function)
n=0
22n (n!)2
1 − cos(2x)
35. a. Use the Maclaurin series of cos x and the identity sin2 x = to obtain the Maclaurin
2
series of sin2 x.
b. Use part b and the identity sin2 x + cos2 x = 1 to obtain the Maclaurin series of cos2 x.

36. Use Maclaurin series to approximate the value of the integral
Z 1
2
e−x dx
0
correct to within an error of 0.001.
37. For further practice, complete the lab Power Series in the optional Lyryx Labs.
Taylor and Maclaurin Series

In this section, you will learn how to find the power series of certain class of functions.
First, let us assume that a function f can be expressed as a power series

∞
X
cn (x − a)n ,
n=0
with radius of convergence of R. We want to find the coefficients cn .
Applying Theorem 6.45 on page 223 several times, we have

∞
X
f (x) = cn (x − a)n = c0 + c1 (x − a) + . . . .
n=0
f 0 (x) = c1 + 2c2 (x − a) + 3c3 (x − a)2 + . . . ,
and f 0 (a) = c1 .
f 00 (x) = 2c2 + 3(2)c3 (x − a) + 4(3)c4 (x − a)2 + . . . ,
and f 00 (a) = 2c2 .
f 00 (a)
Thus, c2 = .
2!
f 000 (x) = 3(2)c3 + 4(3)(2)c4 (x − a) + . . . ,
and f 000 (a) = 3(2)c3 .
f 000 (a)
Thus c3 = .
3!
f (4) (x) = 4(3)(2)c4 + . . . and f (4) (a) = 4(3)(2)c4 .

f (4) (a)
Thus c4 = .
4!
In general, we see that
f (n) (a)
cn = ,
n!
and the power series is
∞
X f (n) (a)
f (x) = (x − a)n .
n=0
n!
Definition 6.46.
If f has a power series representation (expansion) at a, then

∞
X f (n) (a)
f (x) = (x − a)n
n=0
n!
f 0 (a) f 00 (a) f 000 (a)
= f (a) + (x − a) + (x − a)2 + (x − a)3 + . . .
1! 2! 3!
f (n) (a)
+ (x − a)n + . . .
n!
This power representation is called the Taylor series of the function f at a (or about a, or centred at
a).
The Taylor series of a function at 0 is called the Maclaurin series of that function.
The significance of the centre number a in a Taylor series is in the approximation given by the power
series. We have
f 0 (a) f 00 (a) f (n) (a)
f (x) ≈ f (a) + (x − a) + (x − a)2 + · · · + (x − a)n .
1! 2! n!
The larger the n, the better the approximation. Hence, the function f is approximated by polynomials
f 0 (a) f 00 (a) f (n) (a)

Tn (x) = f (a) + (x − a) + (x − a)2 + · · · + (x − a)n ,
1! 2! n!
which are called Taylor polynomials of degree n.
Observe that
f (x) = lim Tn (x).

n→∞
For n = 1 we have
f (x) ≈ f (a) + f 0 (a)(x − a),

which is the linear approximation we learned earlier. This line is the closest (tangent) line to the curve
y = f (x) at x = a.
For n = 2 we have
f 00 (a)
f (x) ≈ f (a) + f 0 (a)(x − a) + (x − a)2 ,
2!
which is the closest quadratic (parabola) to the curve at x = a.
The larger the n the better the approximation of the polynomial to the curve at a.
1
Example 6.85. In Example 6.79 on page 223, we have, for all |x| < ,
3
X ∞
1
= 27x2 n.
1 − 9x2 n=0
1
This is the Maclaurin series of the function .
1 − 9x2
1
Thus, for any |x| <
3
1
≈ 27 + 27x2 + 27x4 + · · · + 27x2n .
1 − 9x2
Since x = 0.1 is close to 0, we have
1
≈ 27 + 27(0.1)2 + 27(0.1)4 + 27(0.1)6 = T6 (0.1).
1 − 9(0.1)2
If we were to approximate the value of the function

1
f (x) =
1 − 9x2
at some number close to 2 by a polynomial of degree 4, we would have to find the Taylor series of this
function around 2.
It is tempting to think that, since we know how to differentiate, we can always find the Taylor series of
any given function, but we can do so only under certain conditions.
Definition 6.47.
If Tn (x) is the Taylor polynomial of a function f at a, and
f (x) = lim Tn (x),

n→∞

then we define the remainder Rn (x) of the Taylor series as
Rn (x) = f (x) − Tn (x).
If lim Rn (x) = 0, then

n→∞
lim Tn (x) = lim f (x) − Rn (x) = f (x) − lim Rn (x) = f (x).

n→∞ n→∞ n→∞
This is the condition under which a function f is equal to its Taylor series.
Theorem 6.48.
If f (x) = Tn (x) + Rn (x), where Tn (x) is the Taylor polynomial of f at a of degree n, and
lim Rn (x) = 0 for|x − a| < R,

n→∞
then f is equal to the sum of its Taylor series on the interval |x − a| < R.
In general, it is not trivial to show that lim Rn (x) = 0. The result shown below helps.
n→∞
Theorem 6.49. Taylor’s Inequality.
If
|f (n+1) (x)| ≤ M for |x − a| ≤ d,
then the remainder Rn (x) of the Taylor series satisfies the inequality
M
|Rn (x)| ≤ |x − a|n+1 for |x − a| ≤ d.
(n + 1)!
M
If lim |x − a|n+1 = 0, then, by Theorem 6.18 on page 188, we have
n→∞ (n + 1)!
lim |Rn | = 0.
n→∞
and by Theorem 6.10 on page 166, we conclude that
lim Rn = 0.
n→∞
Note: Do not make the mistake of assuming that the remainder of the Taylor series is
∞
X
Ti (x).
i=n+1

This is not true, the remainder is as defined in Definition 6.47 on page 230.
In the exercises, you are not required to show that the remainder tends to 0 to find the Taylor series of a
given function. However, keep in mind that, for other functions, you may have the opportunity to work
with, you may have to check this first.
π
Example 6.86. The Taylor series of the function cos x centered at is found as follows.
2
d

cos x = − sin x = −1 and c1 = −1
dx π/2 π/2
d2

cos x = − cos x = 0 and c2 = 0
dx2 π/2 π/2
d3 1

cos x = sin x = 1 and c 3 =
dx3 π/2 π/2 3!
d4

4
cos x = cos x = 0 and c4 = 0
dx π/2 π/2
d5 1

5
cos x = − sin x = −1 and c5 = −
dx π/2 π/2 5!
We then have
1 1
c2n = 0, c4n+1 = − and c3n+1 = .
(4n + 1)! (4n + 3)!
Hence,
π c2 π 2 c7 π 7
T7 (x) = c0 + c1 x − + x− + ··· + x−
2 2! 2 7! 2
and

π 3 π 5 π 7
π
x− x− x−
T7 (x) = − x − + 2 − 2 + 2 .
2 3! 5! 7!
Power series can also be used to evaluate limits, and we can multiply and divide Taylor polynomials.
That is, if Tn (x) and Gn (x) are the Taylor polynomials of functions f and g, then
f (x) ≈ Tn (x) for |x − a| < R1 ,
g(x) ≈ Gn (x) for |x − a| < R2

and the Taylor polynomials of the product and quotient are
f (x)g(x) ≈ Tn (x)Gn (x) for |x − a| < R
and
f (x) Tn (x)
≈ for |x − a| < R
g(x) G( x)
where R = min{R1 , R2 }.
In the table below, we give important Maclaurin series for ease of reference. We include the series in
Examples 6.170 and 6.171: Maclaurin Series of the textbook, and leave others as exercises.
function Maclaurin Series interval of

convergence
∞
X n
x
ln(1 − x) − (-1,1)
n=1
n
∞
X
1
xn = 1 + x + x2 + x 3 . . . (-1,1)
1−x n=0
X∞
xn x x2
ex =1+ + + ... R
n=0
n! 1! 2!
X∞
x2n+1 x3 x 5
sin x (−1)n =x− + − ... R
n=0
(2n + 1)! 3! 5!
∞
X x2n x2 x4
cos x (−1)n =1− + − ... R
n=0
(2n)! 2! 4!
X∞
x2n+1 x3 x5
tan −1
x (−1) n
=x− − − ... [-1,1]
n=0
2n + 1 3 5
Exercises
38. Read Section 6.10 of the textbook, Taylor Series.
40. Read Section 6.11 of the textbook, Taylor Theorem.
42. Give the Maclaurin series of tan−1 x.
43. For further practice, complete the lab Taylor and Maclaurin Series in the optional Lyryx Labs.
Finishing This Unit


examination.
5. Complete the fourth assignments and submit your work to your tutor for grading.
Remember: The number of points in a question may indicate the number of steps in the solution.
6. Complete the online Lyryx Assessment quiz titled Unit 6 - II. Your grade on this quiz will count
If you have completed all the quizzes of the course, then please click the button that says, “If you
have completed all required course work, use this button to submit your grades to Athabasca
University.”

Time: 3 hours
Total points: 80
Passing grade: 55%
1. (4 points)
a. Give an example of a sequence that is eventually strictly decreasing.
b. Give an example of a sequence which is neither increasing nor decreasing.
2. (9 points) Determine whether each of the sequences below is (eventually) increasing or decreasing,
(eventually) strictly increasing or decreasing, or none of these.
a. {2n! }∞
n=0
∞
ln(n)
b.
n n=1
2 ∞
n
c.
n! + 1 n=1
3. (9 points) Determine whether the sequences in Question 2, above, are convergent or divergent.
Explain.
4. (8 points) Find the sum of each of the sums below.

∞
X e k−1
a.
k=5
π
∞
X 1
b.
k=1
(3k + 2)(3k − 1)
5. (16 points) Determine whether each of the following series is absolutely convergent, conditionally
convergent or divergent. Indicate the test you are applying to support your conclusion.
1·2 1·2·3 1·2·3·4
a. 1+ + + + ...
1·3 1·3·5 1·3·5·7
∞
X k
√ √
b. k− k−1
k=1
∞
X k!
c.
k=0
(−2)k
X∞
(−1)k
d.
k=1
(ln k)2

6. (8 points) Show that for all real number x,
∞
X x2k+1
sinh x = .
k=0
(2k + 1)!
√
7. (6 points) Use Taylor polynomials to estimate e within 0.01.
8. (4 points) Find the interval of convergence of the series

∞
X k2
(−1)k (x + 3)k .
k=0
(k + 1)!
9. (5 points) Use the Maclaurin series of tan x to find the first three terms of the Maclaurin series of
ln x cos x.
10. (5 points) Find a power series representation for the integral

Z x
tan−1 t
dt.
0 t
ex − 1
11. (6 points) Let f (x) = .
x
a. Expand f in a power series
b. Differentiate the series in part (a), and show that

∞
X n
= 1.
n=1
(n + 1)!

Appendices
Appendix A: Mathematical Induction

Properties that hold for any positive integer are proved using the Principle of Mathematical Induction.
The following are examples of properties for positive integers.
Example A1.1 The property P (n) is
X
n
n(n + 1)
1 + 2 + 3 + ··· + n = i= .
i=1
2
Hence,
5(6)
P (5) : 1 + 2 + 3 + 4 + 5 = = 15.
2
This property gives a formula for finding the sum of the first n positive integers. For example, the sum of
the first 25 integers is
25(26)
P (25) : = 325.
2
Observe that we can consider P for any representation of an integer n, such as
(n + 3)((n + 3) + 1) (n + 3)(n + 4)
P (n + 3) : =
2 2
or
n2 (n2 + 1)
P (n2 ) : 1 + 2 + · · · + n2 = .
2
Thus,
52 (52 + 1)
P (52 ) : P (25) = = 320.
2
Example A1.2. The property
P (n) : 2n < n!
is not true for n < 4; for example, P (2) does not hold because 22 = 4 > 2! = 2. However P (n) is true
for n ≥ 4. You can see, for example, that P (5) is 25 = 32 < 5! = 120.

Example A1:3. The property
d
P (n) : (f (x))n = n(f (x))n−1 f 0 (x)
dx
is the general power rule for any positive integer n. We know that this property holds for n ≥ 1.
Example A1:4. If {an } is the recurrent sequence defined as a1 = 1, a2 = 2 and an = an−1 + an−2 , then
we defined the property
n
7
P (n) : an < .
4
This property is not true for n = 2, since

2
7
a2 = 2 ≥ ;
4
but you can verify that it is true for n = 3.
The Principle of Mathematical Induction is used to prove that a property P (n) holds for an infinity
number of n ≥ n0 for some integer n0 .
Theorem A1.1. Principle of Mathematical Induction

Let P (n) be a property for a positive integer n.
A. If P (n0 ) is true for some positive integer n0 ; and
B. if whenever P (n) is true, it implies that P (n + 1), for n > n0 , is true,
then P (n) is true for any n ≥ n0 .
We will not present the proof of this theorem, which is based on the Well-ordering Principle.
Both conditions of this theorem must hold if the conclusion is to hold. Condition (B) states that if we
assume P (n) to be true, then we must prove that P (n + 1) is also true. When we assume P (n), we call
this the hypothesis of induction.
Example A1.5. Let us prove that the property in Example A1.1 on page 237
X
n
n(n + 1)
P (n) : 1 + 2 + 3 + · · · + n = i=
i=1
2
holds for all n ≥ 1.
Condition A
The property is true for n = 1 because
1(2)
P (1) : 1 = = 1.
2
Condition B

We assume that it is true that
n(n + 1)
P (n) : 1 + 2 + · · · + n = ,
2
and we must show that
(n + 1)((n + 1) + 1) (n + 1)(n + 2)
P (n + 1) : 1 + 2 + · · · + n + (n + 1) = = .
2 2
Hence
1 + 2 + 3 + . . . . + n + (n + 1) left side of P (n + 1)
= (1 + 2 + 3 + . . . . + n) + (n + 1)
n(n + 1)
= + (n + 1) since P (n) is true
2
hn i
= (n + 1) +1
2

n+2
= (n + 1)
2
(n + 1)(n + 2)
= right side of P (n + 1).
2
Conditions (A) and (B) of Theorem A1.1 hold, and the property is true for all n ≥ 1.
Example A1.6. The property in Example A1.2 on page 237, P (n) : 2n < n!, is true for n ≥ 4.
Condition A
The property is true for n = 4 because
P (4) : 24 = 16 < 24 = 4!.
Condition B
We assume that P (n) : 2n < n! for n > 4 is true, and we must show that P (n + 1) is true, that is
P (n + 1) : 2n+1 < (n + 1)!.
Hence,
2n+1 = 2n (2) < n!(2) since P (n) is true
< n!(n + 1) since n ≥ 4 > 2
= (n + 1)! right side of P (n + 1)

Both conditions of the Mathematical Principle of Induction hold; hence, the property is true for n ≥ 4.
Example A1.7. We will prove the general power rule for any positive integer n.
Condition A
We have
d
P (n) : (f (x))n = n(f (x))n−1 f 0 (x),
dx
so
d
P (1) : f (x) = (f (x))0 (f 0 (x)) = f 0 (x).
dx
Hence, the property is true for n = 1.
Condition B
We assume that
d
P (n) : (f (x))n = n(f (x))n−1 f 0 (x)
dx
is true, and we must show that
d
P (n + 1) : (f (x))n+1 = (n + 1)(f (x))n (f 0 (x)).
dx
d d
(f (x))n+1 = (f (x))n (f (x)) left side of P (n + 1)
dx dx

d d
= n
(f (x)) f (x) + (f (x)) n
f (x) by the product rule
dx dx
= (n(f (x))n−1 f 0 (x))f (x) + (f (x))n f 0 (x) since P (n) is true
= n(f (x))n f 0 (x) + (f (x))n f 0 (x)
= (n + 1)(f (x))n f 0 (x) left side of P (n + 1)
By the Principle of Mathematical Induction, we conclude that the general power rule is true for any
n ≥ 1.
There is an alternative form of the Principle of Mathematical Induction also known as the Principle of
Strong Mathematical Induction (or, alternatively, the Strong Principle of Mathematical Induction).
Theorem A1.2. Principle of Strong Mathematical Induction

Let P (n) be a property for a positive integer n.

A. If P (n0 ) is true for some positive integer n0 ; and
B. if whenever P (k) is true for all n0 ≤ k ≤ n, it implies that P (n + 1) is true,
then P (n) is true for any n ≥ n0 .
Observe the difference of the second condition of Theorem A1.1 on page 238 and Theorem A1.2, above.
The hypothesis of induction of Theorem A1.2 states that the property is true for all integers k between n0
and an integer n.
Example A1.8. Using Theorem A1.2, let us prove that the property of Example A1.4 on page 238 is true
for n ≥ 3.
Condition A
We have a1 = 1, a2 = 2 and an = an−1 + an−2 . The property P is

n
7
P (n) : an ≥ .
4
3
343 7
P (3) : a3 = a2 + a1 = 2 + 1 = 3 < = .
64 4
Thus, P (3) is true.
Condition B
We assume that
k
7
P (k) : ak <
4
is true for any integer 3 ≤ k ≤ n and we must show that

n+1
7
P (n + 1) : an+1 <
4
is true.

Hence,
an+1 = an + an−1 left side of P (n)

n n−1
7 7
< + since P (k) is true for k ≤ n
4 4
n −1 n
7 7 7 4
= [1 + ]= 1+
4 4 4 7
n n
7 11 7 7
= <
4 7 4 4
n+1
7
= right side of P (n + 1).
4
By the Principle of Strong Mathematical Induction, the property holds for

all n ≥ 3.
Recommendations
Mathematical induction is a powerful tool, and you must learn to apply it properly.
1. Always state the property P (n), to be clear about what you want to prove.
2. State the hypothesis of induction, so that you know what you are assuming to be true.
3. State the property for P (n + 1). If this property involves an equality or inequality, start on either the
left side or the right, and always keep in mind what you want to prove—that is, what you want to
conclude.
Exercises
Use Theorems A1.1 and A1.2 to prove each of the statements below. Note that answers to these exercises
are given in Appendix C of this Study Guide.
1. For any n ≥ 1,
Xn k n
1 2 1 1
− = + − .
k=0
2 3 3 2
2. For any n ≥ 1,
X
n
n(n + 1)(2n + 1)
k2 = .
k=0
6

3. For any n ≥ 1,
X
n
k(2k ) = 2 + (n − 1)2n+1 .
k=0
4. For any n ≥ 1,
2n > n.
5. For any n ≥ 2,
n! < nn .
6. The expression x2 − 1 is divisible by 8 when x is any positive odd integer.
Hint: x = 2n + 1, and the induction is over n.
7. The sequence {an } is defined recursively as a0 = 1, a1 = 1, a2 = 1, and for n ≥ 3, we define

an = an−1 + an−3 .
For all n ≥ 1, prove that

√
an+2 ≥ ( 2)n .

Appendix B: Notation
R the set of all real numbers
R−C the set of all real numbers excluding the set C
R − {0} the set of all nonzero real numbers
{an } = {an }∞
n=1 see Example 6.2 on page 158
X ∞
X
an = an see Definition 6.24 on page 192
n n=1
∞
X
cn (x − a)n see Definition 6.43 on page 222
n=0
Tn (x) see page 229
Rn (x) see Definition 6.47 on page 230

Appendix C: Answers
Answers to some of the questions assigned from the textbook are provided in the section Selected
Exercises Answers at the end of the textbook.
Unit 1
page 14
2. Graph of the inverse of f (x)
Figure C.1: Graph of the function f −1 (x)
b. The Domain of f (x) is R and the range is the interval (0, ∞). The domain of f −1 (x) is the
interval (0, ∞) and its range is R.
3. A linear function is of the form L(x) = mx + b. Its inverse is obtained by solving for x in
x−b x b 1
y = mx + b. Hence, f −1 (x) = = − this is a linear function with slope .
m m m m
4. The derivative of g(x) is g 0 (x) = 3x2 + x − 2 = (3x − 2)(x + 1). So the function is decreasing on
the interval [−1, 2/3], hence one to one, and increasing on the interval [2/3, ∞). Thus, in the
interval [0, 2] the function has a minimum (at 2/3), so it is not one to one.
5. Since f (1) = 2, we have that f −1 (2) = 1.And f 0 (x) = 5x4 − 3x2 + 2; thus, f 0 (1) = 4.
d −1 1 1 1
f (2) = 0 −1 = 0 = .
dt f (f (2)) f (1) 4
6. Two possible intervals are [0, π] and [π, 3π/2]. We see that the derivative of cosine is − sin x and
− sin x < 0 on [0, π]; hence, cosine is decreasing on this interval. Also, − sin x > 0 on the interval
[π, 3π/2]; hence, cosine is increasing on this interval. This can also be seen from the graph of
cosine.

π π
7. From the graph of the tangent function we see that tangent is increasing on − , ; hence, it is
2 2
one to one. Also, we see that the derivative of tan x is sec2 x > 0 for all x on this interval.
Therefore, the tangent function is increasing.
8. a. f 0 (x) = 1 + 2 sec2 (2x) ≥ 0 for all x ∈ R. So the function is increasing everywhere and
therefore one to one.
1
b. It is not one to one because f (1/2) = = f (2).
2
c. It is one to one because the function is decreasing everywhere as we see from its graph.
Figure C.2: Graph of the piecewise function f (x)
page 17
10. a. We need to find all t such that 1 − 4t > 0. That is, 4t < 1, from the graph of this exponential
function we see that the inequality is satisfied for all t < 0. Hence, the domain is (−∞, 0).
1
b. cos(e−t ) = cos e1t . The domain of t is all real numbers since et > 0 for all t. The domain of
e
cosine is also all real numbers. Hence, the domain is R.
3ex − e2x
11. a. We divide the numerator and denominator of the function by e3x .
3 − e3x

e−3x 3ex − e2x 3e−2x − e−x
=
e−3x 3 − e3x 3e−3x − 1
Hence,
3ex − e2x 3e−2x − e−x

lim = lim = 0.
x→∞ 3 − e3x x→∞ 3e−3x − 1

1 1
b. We have lim− = ∞. Let t = ; thus,
t→2 2−x 2−x
1
lim− e 2−x = lim et = ∞.
x→2 t→∞
1
c. The function 10−x = is exponential with base 0.1 < 1. Hence, it is decreasing and
10x
lim 10−x = 0.
x→∞
page 22
14. a. By the Chain Rule

d 3/x 3
e = − 2 e3/x .
dx x
b. By the Quotient Rule
d ex ex (1 − e3x ) − ex (−3e3x ) ex + 2e4x

= = .
dx 1 − e3x (1 − e3x )2 (1 − e3x )2
c. By the Chain Rule

d
cos(3ex − 4x ) = − sin(3ex − 4x )(3ex − m4 (4x )).
dx
d. By the Product Rule

d √ 2x e2x √ √ 1
xe = √ + 2 xe2x = xe2x +2 .
dx 2 x 2x
e. By the Chain Rule

d
exp(sin(3 − 2x)) = −2 cos(3 − 2x) exp(sin(3 − 2x)).
dx
15. We apply the Close Interval Method.

d 2 x
x e = xex (2 + x) = 0 for x = 0, −2.
dx
Both are in the interval [−3, 0] and
f (0) = 0, f (−2) = 4e−2 , f (−3) = 9e−3 .
Since 4e−2 > 9e−3 . The values are 4e−2 max and 0 min.

Z Z 2
−2x2 1 −2x2 e−2x
16. a. xe dx = − −4xe dx = − +C
4 4
Z Z
1 4−3x
b. 4−3x dx = − (−3)4−3x
dx = − +C
3 3m4
Z
c. ex cos(ex ) dx = sin(ex ) + C
Z √ Z
√
1 − e−t 2
d. dt =e−t 1 − e−t = − √ +C
et 1 − e−t
Z Z x
2x 2 m2
e. 2
dx = m2 (3 − 2x )−2 dx = −
(3 − 2 )
x m2 3 − 2x
page 27
18. Summary
1. A logarithm function base a is denoted by loga x and it is the number b = loga x we have to
rise a to obtain x. That is, xb = a.
2. The natural logarithm function ln x is the logarithmic function base e.
3. The domain of any logarithmic function is all x > 0.
4. Logarithmic functions are increasing and concave down.
5. loga a = 1 and loga (1) = 0.

ln x
6. loga x =
ln a
7. lim ln x = ∞ and lim+ ln x = −∞
x→∞ x→0
d d g 0 (x)
8. ln x = x1 and ln(g(x)) =
dx dx g(x)
Z 0
g (x)
9. = ln |g(x)| + C
g(x)
10. ln(xy) = ln x + ln y, ln(x/y) = ln x − ln y, ln(xa ) = a ln x
19. ex = ey if and only if x = y, because the exponential function is one-to-one.
20. loga x = loga y if and only if x = y, because the logarithmic function base a is one-to-one.
21. The function sin x is negative on the interval (−π, 0) and ln t is defined only for t > 0. However,
sin x > 0 on the interval (0, π);hence, ln(sin x) is defined on this interval.
24. The domain of ln(tan x) is all x so that tan x > 0, two such intervals are (0, π/2), (π, 3π/2). In
general for any integer k ≥ 0 the domain are the interval (2kπ, 2kπ + π/2) and
((2k + 1)π, π(2k + 3/2)).

25. a. lim tan x = 0 and tan x > 0 for x → 0+ . Let t = tan x; hence,
x→0+
lim ln(tan x) − lim+ ln t = −∞.

x→0+ t→0

3x2 − 2
b. 2
ln(3x − 2) − ln(2x + 5x ) = ln 2
and
2x + 5x2
3x2 − 2 3
lim 2
= .
x→∞ 2x + 5x 5
Hence,

2 23x2 − 2
lim ln(3x − 2) − ln(2x + 5x ) = ln lim = ln(3/5).
x→∞ x→∞ 2x + 5x2
c. The quadratic x2 + 2x − 1 is a concave up parabola. Hence,
lim x2 + 2x − 1 = ∞
x→∞
and
lim ln(x2 + 2x − 1) = lim ln t = ∞.

x→∞ t→∞
d. lim ln(cos x) = ln(0) = 1

x→0
page 33
32. a.
√ 2 √ 2
ln( x − 1ex (3x + 2)10 ) = ln x − 1) + ln(ex ) + ln((3x + 2)10 )
ln(x − 1)
= + x2 + 10 ln(3x + 2).
2
d ln(x − 1) 1 30
+ x2 + 10 ln(3x + 2) = + 2x + .
dx 2 2(x − 1) 3x + 2
Hence,

0
√ x2 1 10 30
g (x) = [ x − 1e (3x + 2) ] + 2x + .
2(x − 1) 3x + 2
b. ln(xtan x ) = tan x(ln x).
d tan x
tan x(ln x) = sec2 x ln x +
dx x

and

0 tan x 2 tan x
h (x) = (x ) sec x ln x + .
x

x2 − 2x + 1 ln(5 − sin x)
c. ln √ = ln(x2 − 2x + 1) −
5 − sin x 2
d ln(5 − sin x) 2x − 2 cos x

ln(x2 − 2x + 1) − = 2 +
dx 2 x − 2x + 1 2(5 − sin x)
Hence,

0 x2 − 2x + 1 2x − 2 cos x
g (x) = √ 2
+ .
5 − sin x x − 2x + 1 2(5 − sin x)
d. ln((sin x)x ) = x ln(sin x)
d x cos x
x ln(sin x) = ln(sin x) + .
dx sin x
Hence,
u0 (x) = [(sin x)x ][ln(sin x) + x cot x].
2 − u2 1
33. a. 3
= u−3 −
2u 2u
Z 6 Z 6
2 − u2 −3 1 1 ln |2u| 6
du = u − =− 2 −
1 2u3 1 2u 2u 2 1
1 ln(12) 1 ln(2) 35 ln(2) − ln(12)
=− − + + = +
72 2 2 2 72 2
35 ln(6)
= −
36 2
d
b. (4 − sin(2x)) = −2 cos(2x)
dx
Z 1 Z
cos(2x) 1 1 −2 cos(2x)
dx = − dx
0 4 − sin(2x) 2 0 4 − sin(2x)
ln |4 − sin(2x)| 1 ln |4 − sin(2)| ln |4|
=− =− +
2 !0 2 2
2
= ln p
4 − sin(2)

d x
c. (e + 1) = ex
dx
Z
3ex
x
dx = 3 ln |ex + 1| + C.
e +1
Z
8x tan(3x)
d. 8x − sec2 (3x) dx = − + C.
ln(8) 3
d 1
e. ln(2x) =
dx x
Z
(ln(2x))6 (ln(2x))7
dx = + C.
x 7
sin x d
f. tan x = and cos x = − sin x. Thus,
cos x dx
Z Z
− sin x
tan x dx = − dx = − ln | cos x| = ln | sec x| + C.
cos x
cos x d
g. cot x = and sin x = cos x
sin x dx
Z Z
cos x
cot x dx = dx = ln | sin x| + C.
sin x
csc x(csc x − cot x) csc2 x − csc x cot x

h. csc x = = . Hence,
csc x − cot x csc x − cot x
Z Z
csc2 x − csc x cot x
csc x dx = = ln | csc x − cot x| + C.
csc x − cot x
page 38
35.
Figure C.3: Graph of the hyperbolic cosecant function
36. Properties of the hyperbolic cosecant function
a. Domain is R − {0}.
b. Decreasing and concave down on (−∞, 0) and decreasing and concave up on (0, ∞).
c. lim csch x = −∞.

x→0−
d. lim csch x = ∞.
x→0+

e. lim csch x = 0.
x→∞
f. lim csch x = 0.
x→−∞
37.
Figure C.4: Graph of the hyperbolic secant function
38. Properties of the hyperbolic secant function
a. Domain is R.
b. sech(0) = 1
c. concave down on (−2, 2).
d. Increasing and concave up on (−∞, −2).
e. Decreasing and concave up on (2, −∞).
f. lim sech x = 0.
x→∞
g. lim secch x = 0.
x→−∞
39.

Figure C.5: Graph of the hyperbolic cotangent function
40. Properties of the hyperbolic cotangent function
a. Domain is R − {0}.
b. Decreasing and concave down on (−∞, 0). Decreasing and concave up on (0, ∞).
c. lim csch x = −∞.

x→0−
d. lim csch x = ∞.
x→0+
e. lim coth x = 1.
x→∞
f. lim coth x = 1.
x→−∞
d d 1 cosh x 1 cosh x
41. a. csch x = =− 2 =− = −csch x coth x
dx dx sinh x sinh x sinh x sinh x
d d 1 sinh x 1 sinh x
b. sech x = =− 2 =− = −sech x tanh x.
dx dx cosh x cosh x cosh x cosh x
d d cosh x sinh2 x − cosh2 x 1
c. coth x = = 2 =− = −csch2 x.
dx dx sinh x sinh x sinh2 x
44. a.
(ex + e−x )2 e2x + e−2x + 2
cosh2 x = =
4 4
e2x + e−2x 2 cosh(2x) 1
= + = +
4 4 2 2
cosh(2x) + 1
=
2
cosh(2x) + 1 cosh(2x) − 1
b. sinh2 x = cosh2 x − 1 = −1= .
2 2

Z Z
2 cosh(2x) − 1 sinh(2x) x
45. a. sinh x dx = dx = − + C.
2 4 2
Z Z
2 cosh(2x) + 1 sinh(2x) x
b. cosh x dx = dx = + + C.
2 4 2
46. (ex − e−x )(ey + e−y ) + (ey − e−y )(ex + e−x ) = 2ex eY − 2e−x e−y = 2ex+y − 2e−x−y
(ex − e−x )(ey + e−y ) + (ey − e−y )(ex + e−x )

sinh x cosh y + cosh x sinh y =
4
ex+y − e−x−y
= = sinh(x + y).
2
47. (ex + e−x )(ey + e−y ) + (ey − e−y )(ex − e−x ) = 2ex ey + 2e−x e−y = 2ex+y + 2ex+y
(ex + e−x )(ey + e−y ) + (ey − e−y )(ex − e−x )

cosh x cosh y + sinh x sinh y =
4
ex+y + e−x−y
= = cosh(x + y).
2
Unit 2
page 61
1.
Figure C.6: Graph of the cotangent function on the interval (0, π)
Properties: From the graph of cot−1 x in Figure C.1, above:
a. the domain is R; the range is (0, π).

π
b. cot−1 (0) = .
2
c. the function is decreasing on its domain.
d. lim cot−1 x = 0
x→∞

e. lim cot−1 x = π
x→−∞
Cancellation Laws:
cot(cot−1 x) = x if and only if x∈R

cot−1 (cot x) = x if and only if x ∈ (0, π)
Right Triangle: We can associate a right triangle with cot−1 x. Consider that θ = cot−1 x is an
angle in the interval (0, π); then
x
cot(θ) = cot(cot−1 x) = x = ,
1
√ opposite side is 1 and the adjacent side is x. By Pythagoras’ Theorem, the
hypotenuse is x + 1 and the graph is as shown below.
2
Figure C.7: The right triangle for the cotangent function on the
interval (0, π)
Derivative: By the cancelation law,
cot(cot−1 x) = x.

d
− csc2 (cot−1 x) cot−1 x = 1.
dx
Solving yields
d 1
cot−1 x = − 2 .
dx csc (cot−1 x)
Since θ = cot−1 x is in (0, π) and csc(θ) ≥ 0 for this value of θ, from the right triangle, we have
√
cot(θ) = cot(cot−1 x) = x2 + 1.
Hence,
d 1
cot−1 x = − 2 .
dx x +1

Integration:
Z
dx
2
= tan−1 x + C = − cot−1 x + C
x +1
and
Z
g 0 (x)
dx = − cot−1 (g(x)) + C.
(g(x))2 + 1

πi 3π
Figure C.8: Graph of the cosecant function on the interval 0, ∪ π,
2 2
2. Properties: From the graph of csc−1 x in Figure C.3, above:

π i 3π
a. the domain is (−∞, −1] ∪ [1, ∞); the range is 0, ∪ π, .
2 2
π 3π
b. csc−1 (1) = and csc−1 (−1) = .
2 2
c. the function is increasing on (−∞, −1] and decreasing on [1, ∞).
d. lim csc−1 x = 0.
x→∞
e. lim csc−1 x = π.
x→−∞
Cancellation Laws
csc(csc−1 x) = x if and only if x ∈ (−∞, −1] ∪ [1, ∞)

π i 3π
csc−1 (csc x) = x if and only if x ∈ 0, ∪ π,
2 2

Right Triangle: We can associate a right triangle with csc−1 x. Consider that θ = csc−1 x is an
angle in the interval
π i 3π
0, ∪ π, ,
2 2
then
x
csc(θ) = csc(csc−1 x) = x = .
1
The
√ hypoteneuse is x, and the opposite side is 1. By the Pythagorean Theorem, the adjacent side is
x2 − 1. The triangle is shown in Figure C.9, below.
Figure C.9: Graph of the right

trianglefor the cosecant function on the
πi 3π
interval 0, ∪ π,
2 2
Derivative: By the cancellation law,
csc(csc−1 x) = x.

d
− cot2 (csc−1 x) csc(csc−1 x) csc−1 x = 1.
dx
Solving yields
d 1
csc−1 x = − .
dx x cot(csc−1 x)
Since θ = csc−1 x is in
π i 3π
0, ∪ π,
2 2
and cot(θ) ≥ 0 for this value of θ, from the right triangle, we have
√
cot(θ) = cot(cot−1 x) = x2 − 1.

Hence,
d 1
cot−1 x = − √ .
dx x x2 − 1
Integration:
Z
dx
√ = sec−1 x + C = − csc−1 x + C
x x2 − 1
and
Z
g 0 (x)
p dx = − csc−1 (g(x)) + C.
2
g(x) g(x)) − 1
page 70
10. The domain of tan−1 t is |t| < 1, and the range of cosh x is x ≥ 1. Hence, if t = cosh x, then t > 1
and the composition tanh−1 (cosh x) = tanh−1 t is undefined for all x.
11. The domain of ln t is (0, −∞). If t = tanh−1 x, then t > 0 for 0 < x < 1 and the domain of
ln(tanh−1 x) is (0, 1).
12. a. Let y = cosh−1 x. Thus,
−1 ey + e−y
x = cosh(cosh x) = cosh y = ⇒ ey − 2x + e−y = 0.
2
We multiply it by ey and we get e2y − 2xey + 1 = 0. We set w = ey
√
2x ± 4x2 − 4 √
w2 − 2xw + 1 = 0 ⇒ w = = x ± x2 − 1.
2
√
Since w > 0, the value for w is x + x2 − 1. Thus,
√ √
ey = x + x2 − 1 ⇒ y = cosh−1 x = ln(x + x2 − 1).
b. Let y = coth−1 x. Thus,
ey + e−y
x = coth(coth−1 x) = tanh y = −y
⇒ x(ey − e−y ) = ey + e−y .
e −e
y
To solve for ey from this equation; we multiply it by ey
xe2y − x = e2y + 1 ⇒ (x − 1)e2y = x + 1.
We solve

2y x+1 x+1
e = ⇒ 2y = ln .
x−1 x−1

Therefore

−1 1 1+x
y = coth x = ln .
2 x−1
c. Let y = sech −1 x. Thus,

2
x = sech (sech −1 x) = sech y = ⇒ xey + xe−y = 2.
ey +ey
We multiply this equation by ey
xe2y + x = 2ey ⇒ xe2y − 2ey + x = 0.
We set w = ey and we get the quadratic equation xw2 − 2w + x = 0 with coefficients

a = x, b = −2, c = x. Thus,
√ √
2 ± 4 − 4x2 1 ± 1 − x2
w= = .
2x x
√
1 + 1 − x2
Since w = e > 0 the value for w is
y
. Therefore,
x
√
1 + 1 − x2
y = sech x = ln
−1
.
x
d. Let y = csch −1 x. Thus,

2
x = csch (csch −1 x) = csch y = ⇒ xey − xe−y = 2.
ey − ey
We multiply this equation by ey
xe2y − x = 2ey ⇒ xe2y − 2ey − x = 0.
We set w = ey and we get the quadratic equation xw2 − 2w − x = 0 with coefficients

a = x, b = −2, c = −x. Thus,
√ √
2 ± 4 + 4x2 1 ± 1 + x2
w= = .
2x x
See that for x > 0
√ √
1 + 1 + x2 1− 1 + x2
> 0 and < 0.
x x
And for x < 0
√ √ √
1 + 1 + x2 1− 1 + x2 1 1 + x2
< 0 and = + > 0.
x x x −x

√
1 + x2
1+
Therefore, the value of w is , since w = ey > 0. We conclude that
|x|
√
1 1 − x2
y = csch x = ln
−1
+ .
x |x|
d
13. a. csch 5 x coth x = (csch x)4 (csch x coth x) and csch x = −csch x coth x
dx
Z Z
csch 5 x
csch x coth x dx = −
5
csch 4 x(−csch x coth x) = − + C.
5
Z Z
sinh x
b. tanh x dx = dx = ln(cosh x) + C.
cosh x
sinh x sinh x d
c. 2 = 2 = sinh x(cosh x)−2 and cosh x = sinh x
sinh x + 1 cosh x dx
Z Z
sinh x 1
2 dx = sinh x(cosh x)−2 dx = − = −sech x + C.
sinh x + 1 cosh x
d
d. tanh x = sech 2 x
dx
Z
tanh3 x
sech 2 x tanh2 x dx = + C.
3
d
e. (2 − 3 coth x) = 3 csc2 x
dx
Z Z
csc2 x 1 3 csc2 x ln |2 − 3 coth x|
dx = dx = + C.
2 − 3 coth x 3 2 − 3 coth x 3
Unit 3
page 86
5. a. Let
u = cosn−1 x and dv = cos x.
Then,
du = −(n − 1) cosn−2 x sin x dx and v = sin x.

Applying parts
Z Z
n
cos x dx = cosn−1
x sin x + (n − 1) cosn−2 x sin2 x dx.
We replace sin2 x with 1 − cos2 x in the integral on the right
cosn−2 x sin2 x = cosn−2 x(1 − cos2 x) = cosn−2 x − cosn x
and
Z Z Z
n−2 2 n−2
(n − 1) cos x sin x dx = (n − 1) cos x dx − (n − 1) cosn x dx.
R
Adding cosn x dx to the left side yields
Z Z
n n−1
n cos x dx = cos x sin x + (n − 1) cosn−2 x dx.
The formula is obtained by dividing by n.
b. Let
u = secn−2 x and dv = sec2 x dx.
Then,
du = (n − 2) secn−2 x tan x dx and v = tan x.
Applying parts
Z Z
n
sec x dx = secn−2
x tan x − (n − 2) secn−2 x tan2 x dx.
We replace tan2 x with sec2 x − 1 in the integral to the right

Z Z Z
n−2 2 n−2
(n − 2) sec x tan x dx = −(n − 2) sec x dx + (n − 2) secn x dx.
R
Adding secn x dx to the left side yields
Z Z
n n−2
(n − 1) sec x dx = sec x tan x − (n − 2) secn−2 x dx.
The formula is obtained by dividing by n − 1.
c. Let
u = (ln x)n and dv = dx.

Then,
n(ln x)n−1
du = dx and v = x.
x
Applying parts we obtained the reduction formula
Z Z
(ln x) dx = x(ln x) − n (ln x)n−1 dx.
n n
d. Let
u = xn and dv = ex dx.
Then,
du = nxn−1 dx and v = ex .
Applying parts we obtain the reduction formula

Z Z
x e dx = x e − n xn−1 ex dx.
n x n x
e. Let
u = (x2 + a2 )n and dv = dx.
Then,
du = 2nx(x2 + a2 )n−1 dx and v = x.
Applying parts
Z Z
(x + a ) dx = x(x + a ) − 2n x2 (x2 + a2 )n−1 dx.
2 2 n 2 2 n
Replacing x2 with x2 + a2 − a2 in the integral on the right yields

Z Z Z
2 2 2 n−1 2 2 2 n−1
2n x (x + a ) dx = −2na (x + a ) dx + 2n (x2 + a2 )n dx.
R
Adding 2n (x2 + a2 )n dx to the left we obtain
Z Z
2 2 n 2 2 n 2
(2n + 1) (x + a ) dx = x(x + a ) + 2na (x2 + a2 )n−1 dx.
The reduction formula is obtained by dividing by 2n + 1.
6. Let u = cscn−2 x and dv = csc2 x; then,
du = (n − 2) cscn−3 x(− cot x csc x), and v = − cot x.

Z Z
n
csc x dx = − cot x csc n−2
x − (n − 2) cscn−2 x cot2 x dx.
Replacing cot2 x by csc2 −1, we get

Z Z
n−2 2
csc x cot x dx = cscn x − cscn−2 x dx,
R
and solving for cscn x dx yields
Z Z
n cot x cscn−2 x n − 2
csc x dx = − + cscn−2 x dx.
n−1 n−1
7. a. The function y = 2 ln(x + e) is the shift to the left by e units of ln x and a vertical stretch by 2
units of ln(x + e). Hence,
-e
Figure C.10: Graph of the function y = 2 ln(x + e)

2 2x
b. The slope of the line is m = and the equation of the line is y = +2
e e
c. The area is given by the integral
Z 2 2
2x x2
+ 2 − 2 ln(x + e) dx = + 2x − 2(x + e) ln |x + e| + 2(x + e)
−e e e −e
2 2
2 e
= + 4 − 2(2 + e) ln |2 + e| + 2(2 + e) − + 2e
e e
−1
= 4e + 3e − (4 + 2e) ln |x + e| + 8
page 94
9. a. If n = 2k is even, then we factor csc2 x from cscn x, and we replace
(csc2k−2 x) = (csc2 x)k−1 = (cot2 x + 1)k−1 .

The term cotm x cscn x becomes a sum of terms of products of the form cotr x csc2 x for some
power r.
b. If m = 2s is even, then we replace cot2 x = csc2 x − 1 and cot2s x = (csc2 x − 1)s and the
term cotm x cscn x becomes a sum of terms of the form cscr x for some power r. We then use
the reduction formula for csc x in each term.
10. a. If m = 2k is even, then we replace cot2 x = csc2 x − 1 and cot2s x = (csc2 x − 1)s , and the
term cotm x = (csc2 x − 1)k becomes a sum of terms of cscr x for some power r or constants.
We then use the reduction formula for cscr x.
b. If m = 2k + 1, then we factor cot x and we replace cot2 x = csc2 x − 1. Hence,

cotm x = (csc2 x − 1)k cot x becomes a sum of product of terms of the form cscm x cot x with
m even. Thus, we apply Exercise 9.
11. a. We apply the strategy indicated above. We have cot2 x = csc2 −1 and
cot2 x csc2 x = (csc2 x − 1) csc2 x = csc4 x − csc2 x.
The reduction formulas for cscn x

Z Z
4 cot x csc2 x 2
csc x dx = − + csc2 x dx
3 3
cot x csc2 x 2 cot x
=− − .
3 3
Hence,
Z
cot x csc2 x 2 cot x cot x csc2 x cot x
cot2 x csc2 x dx = − − + cot x = − + + C.
3 3 3 3
b. We apply the strategy indicated above. We have cot3 x = cot2 x cot x = (csc2 −1) cot x
Z Z
cot2 x
cot x dx = cot x csc2 x dx − cot x dx = −
3
− ln | sin x| + C.
2
c. We multiply out [sin(3x) − sin(4x)] cos(3x) = sin(3x) cos(3x) − sin(4x) cos(3x). We apply
identity 3.3 on page 93
Z
cos(6x) cos x cos(7x)
sin(3x) cos(3x) − sin(4x) cos(3x) dx = − + + .
12 2 14
page 98

√
15. x = a sec θ, dx = a tan θ sec θ dθ, and x2 − a2 = a tan θ; thus,
Z Z
a tan θ
a tan θ sec θ dθ = a tan2 θ dθ
a sec θ
Z
= a sec2 θ − 1 dθ
= a(tan θ − θ)

√ −1 |x|
= x2 − a2 − a sec + C.
a
√
16. x = a sec θ, dx = a tan θ sec θ, and x2 − a2 = a tan θ; thus,
Z Z
a tan θ tan2 θ
a tan θ sec θ dθ = dθ
a2 sec2 θ sec θ
Z Z
sin2 θ
= dθ = sec θ − cos θ dθ
cos θ
= ln | sec θ + tan θ| − sin θ
√ x √ x 2 − a2
x 2 − a2
=− + ln + + C.
x a x
The answer is different.
page 106
23. a. Let u = ln x, then du = 1

x
dx. The integral in terms of u is
Z √ Z √
1 + ln x 1+u
dx = du.
x ln x u
√ 1
Let w = 1 + u, then u = w2 − 1, dw = √ du. Hence 2w dw = du. The integral in
2 1+u
terms of w is
Z Z
w2 1 1 w − 1
2 dw = 2 1 + 2 dw = 2 w + ln
w2 − 1 w −1 2 w+1
√
w in terms of x is w = 1 + ln x. The answer in terms of x is
√ √1 + ln x − 1

2 1 + ln x + ln √ + C.
1 + ln x + 1
ln(tan x) cos x ln(tan x)

b. = = cot x ln(tan x) sec2 x.
sin x cos x sin x cos2 x
Let u = tan x, then cot x = 1/u, du = sec2 x dx. The integral in terms of u is
Z Z
ln(tan x) ln u (ln u)2
dx = du = .
sin x cos x u 2

(ln(tan x))2
+ C.
2
√ 3 u
c. Let u = 6x, then du = √ dx and du = dx. The integral in terms of u is
6x 3
Z
1
u cos u du.
3
We apply parts with w = u, dv = cos u du. Thus, dw = du and v = sin u.
Z Z
1 1 u sin u cos u
u cos u du = u sin u − sin u du = +
3 3 3 3

√ √ √
6x sin 6x cos 6x
+ + C.
3 3
t3 + 1 1 t2 + 2
d. By long division = + . By partial fractions
2t3 − t2 2 4t3 − 2t2
t2 + 2 t2 + 1 A B C
3 2
= 2
= + + 2
4t − 2t 2t (2t − 1) 2t − 1 2t t
Solving for A, B, C
9
t2 + 2 = (2A + 2B)t2 + (4C − B)t − 2C ⇒ C = −1, B = −4, A = .
2
We integrate
Z
1 9 2 1 t 9 ln |2t − 1| 1
+ − − 2 dt = + − 2 ln |t| + + C.
2 2(2t − 1) t t 2 4 t
1 du
e. Let u = ex , then du = ex dx and e−1 = , = dx. The integral in terms of u is
u u
Z Z
du 1 1 ln |u + 1| ln |u + 1|
2
= − du = − .
u −1 2(u + 1) 2(u − 1) 2 2

1 ex + 1
ln x + C.
2 e −1

1 dx 1
f. By part with u = ln(x + 1) and dv = 2
dx; thus, du = ,v = − .
x x+1 x
Z Z
ln(x + 1) ln(x + 1) 1
2
dx = − + .
x x x(x + 1)
By partial fractions
Z
x
dx = x − ln |x + 1|.
x+1
The answer is
ln(x + 1)
− + x − ln |x + 1| + C.
x
1 du
g. Let u = ex , then du = ex dx and e−1 = , = dx. The integral in terms of u is
u u
Z Z
dx u
= du.
1 + 2ex − e−x 2u2 +u−1
By partial fractions
u u A B
= = +
2u2 + u − 1 (2u − 1)(u + 1) 2u − 1 u + 1
1
solving A = B = .
3
Z
1 1 ln |2u − 1| ln |u + 1|
+ du = + .
3(2u − 1) 3(u + 1) 6 3

ln |2ex − 1| ln |ex + 1| 1 p
+ = ln[ |2ex − 1|(ex + 1)] + C.
6 3 3
page 106
30. The integral is improper of types 1 and 2. Thus,

Z ∞ Z 1 Z t
dx dx dx
√ = lim+ √ + lim √ .
0 x(1 + x) t→0 t x(1 + x) t→∞ 1 x(1 + x)
√
We consider the indefinite integral and evaluate with the substitution u = x, x = u2 and
2u du = dx.
Z Z Z
dx 2u 1
√ = du = 2 du
x(1 + x) u(1 + u2 ) 1 + u2
√
= 2 tan−1 (u) = 2 tan−1 x.

Z 1
dx −1
√ 1
lim √ = lim 2 tan x
t→0+ t x(1 + x) t→0+ t
π √ π
−1
= lim+ 2 − tan t =
t→0 4 2
And
Z t
dx √ t
lim √ = lim 2 tan−1 x
t→∞ 1 x(1 + x) t→∞ 1
√ π π π
= lim 2 tan−1 t − = π − =
t→∞ 2 2 2
Hence,
Z ∞
dx π π
√ = + = π.
0 x(1 + x) 2 2
Z ∞
31. a. x dx is the sum of two limits, one of them is
−∞
Z ∞
x2 t t2
lim x dx = lim = lim = ∞.
t→∞ 0 t→∞ 2 0 t→∞ 2
So the integral is divergent.
b.
Z t
x2 t
lim x dx = lim
x→∞ −t x→∞ 2 −t
t2 (−t)2
= lim − =0
x→∞ 2 2
Unit 4
page 128
1. The integral that gives the volume of the region bounded by y = cos x, x = 0, x = π/2 around the
y-axis is
Z 1
π (cos−1 y)2 dy.
0
2. The volume of the region bounded by y = sin x, x = 0, x = π around the line y = −1 is

Z π
π2
π (sin x + 1)2 − 1 dx = 2π + .
0 2
page 140

14. The area A of R is
Z π/4 π/4 √

A= cos x − sin x dx = [sin x + cos x] = 2 − 1.
0 0
The centroid (x̄, ȳ) is

Z π/4
1
x̄ = √ x[cos x − sin x] dx
2−1 0
Z π/4
1
= √ [x cos x − x sin x] dx by parts
2−1 0
1 π/4

= √ [x sin x + cos x + x cos x − sin x]
2−1 0
1 π/4

= √ [x(sin x + cos x) + cos x − sin x]
2−1 0
√
1 √ π−2 2
= √ [ 2 − 1] = √
2−1 4−2 2
Z π/4
1
ȳ = √ cos2 x − sin2 x dx
2( 2 − 1) 0
Z π/4
1
= √ cos(2x) dx
2 2−2 0
1 π/4 1

= √ (sin(2x)) = √
4 2−4 0 4 2−4
Unit 5
page 151
7. Let y(t) be the amount of salt (in kg) after t minutes, then y(0) = 0 and the total amount of liquid
in the tank remains constant at 1000 L.

dy y(t)
= (0.05)(5) + (0.04)(10) − 15
dt 1000
130 − 3y
= 0.25 + 0.40 − 0.015y = 0.65 − 0.015y =
200
We integrate
Z Z
dy dt ln |130 − 3y| t
= ⇒− = + C.
130 − 3y 200 3 200
ln(130)
Since y(0) = 0, we have C = − . Hence,
3
ln |130 − 3y| t ln(130) 3t
− = − ⇒ ln |130 − 3y| = − + ln(130).
3 200 3 200

Since t = ln(et ) we conclude that |130 − 3y| = 130e−3t/200 . Since y(0) = 0, y is continuous and
the right side of the equality is alwasy postive, we conclude that 130 − 3y > 0. Thus,
130 − 3y = 130e−3t/200 and
130 − 130e−3t/200
y(t) = .
3
130 − 130e−90/200
For t = 30 we have y(30) = ≈ 15.7027.
3
page 154
7. a. y = e−x sin(ex ) + Ce−x
b. y = e−x ln |1 − ex | + Ce−x
c. y = x2 (ex − e)
Unit 6
page 170
1. The sequence {1, 2, 3, 4, −1, e, e, e, e, e, . . . } is eventually constant at e. Any sequence similar to

this one is correct.
(−1)n (n + 1)
2. a. an = .
n
n+1
Since lim = 1, we have
n→∞ n
lim a2n = 1 and lim a2n+1 = −1.
n→∞ n→∞
Hence, the limit does not exist.
b. 2
π
c.
4
d. By the Squeeze Theorem, the limit is 0.
page 191
ε ε
6. a. Let ε > 0. For > 0, there is N > 0 such that |an | < for n > N .
|c| |c|
Hence, |can | = |c||an | < ε for n > N .
b. If lim bn = M , then for 1 there is N > 0 such that |bn − M | < 1 for n > N . Hence,
n→∞
|bn | < 1 + |M | for n > N . Then an < bn ≤ |bn | < 1 + |M | for n > N .

Hence, the sequence {an } is eventually increasing and bounded; therefore, it is convergent and
lim an = L ≤ lim bn = M.
n→∞ n→∞
7. a. For any M < 0, there is N > 0 such that an < M for n > N .
b. If lim an = −∞, then, for any M < 0, there is N0 > 0 such that an < M for all n > N0 . If
n→∞
we take N = max(N0 , k), then bn < an < M for all n > N .
Hence, lim bn = −∞.

n→∞
8. a. For M > 0, there are N1 > 0 and N2 > 0 such that an > M/2 for n > N1 and bn > M/2 for
n > N2 . Hence, if N = max{N1 , N2 }, for n > N . both previous statements hold and
an + b n ≥ M .
b. For M > 0 there is N > 0 such that an > M/c for n > N . Thus can > M for n > N .
page 202
11. As a partial fraction

2 1 1
= −
n2 + 4n + 3 n+1 n+3
The partial sum is

1 1 1 1 1 1 1 1
sn = − + − + − + −
2 4 3 5 4 6 5 7

1 1 1 1
+ ··· + − + −
n n+2 n+1 n+3
1 1 1 1
= + − −
2 3 n+2 n+3
The series is telescopic and
∞
X 2 1 1 1 1 1 1 5
= lim + − − = + = .
n=0
n2 + 4n + 3 n→∞ 2 3 n + 2 n + 3 2 3 6
page 170
26. Proof of the Root Test 6.41 on page 216.

p
a. Since limn→∞ n |an | = L < 1, for L < r < 1 there exist N > 0, such that
p
n
|an | < r for all n > N .

X
k
Hence, |an | < rn for all n > N . For an integer k > 0, let bk = |aN +i |. Thus,
i=i
X
k
bk < rN +i for all k.
i=1
P∞
Since r < 1, the series i=1 rN (ri ) is convergent, and
∞
X
lim bk = |aN +i | < ∞.
k→∞
i=i
Therefore,
∞
X ∞
X
|an | = |a1 | + |a2 | + . . . |aN | + |aN +i | < ∞.
n=1 i=i
b. If L > 1, we take 1 < r < L. For this positive number M = L − r, there is N > 0 such that
for n > N
p p
| n |an | − L| < M ⇒ −M < n |an | − L < M
p
Hence, r − L < n |an | − L < L − r and
pn
|an | > r for n > N .
p
The same is true if limn→∞ n |an | = ∞. We have,
|an | > rn for n > N ⇒ lim |an | =

6 0.
n→∞
By the Divergence Test, the series is divergent.

1 2 8 443
27. s8 = + + ··· + = ≈ 3.4609. The ratios
2 4 256 128

an+1 n + 1 2n 1 1
rn = = n+1 = 1+
an 2 n 2 n
form a decreasing sequence, and

9+1 10 5
r9 = = = <1
2(9) 18 9
by part a. of Proposition 6.42 on page 218, the error is

a9 9 81
R8 ≤ = 9 = ≈ 0.03955.
1 − r9 2 (1 − 5/9) 2048
page 227

35. a. Let an = n!xn . By the ratio test
(n + 1)!xn+1

lim = lim (n + 1)|x| = ∞.
n→∞ n!xn n→∞
The series diverges for x 6= 0 and converges for x = 0. This is the only value and the radius of
convergence is 0.
(−3)n xn
b. Let an = √ . By the ratio test
n+1
a (−3)n+1 xn+1 √n + 1
n+1
= √ n n

an n+2 (−3) x
r
n + 1
= − 3x
n+2
s
1 + 1/n
=3 |x|
1 + 2/n
Hence,
s
1 + 1/n
lim 3 |x| = 3|x| < 1 for |x| < 1/3.
n→∞ 1 + 2/n
The radius of convergence is 1/3.

(−1)n x2n
c. Let an = . By the ratio test
22n (n!)2
a (−1)n+1 x2(n+1) 22n (n!)2
n+1
= 2(n+1) 2 n 2n

an 2 ((n + 1)!) (−1) x
2n+2 2n
x 2 (n!)2
=
x2n 22n+2 ((n + 1)!)2
x2
=
4(n + 1)2
Hence,
x2
lim = 0 for all x
n→∞ 4(n + 1)2
the ratio of convergence is infinity.

∞
1 − cos(2x) 1 1 X (−1)n (2x)2n
36. a. 2
sin x = = −
2 2 2 n=0 (2n)!
∞
X
2 (−1)n+1 22n−1 x2n
sin x =
n=1
(2n)!

The radius of convergence is R = ∞.
∞
X (−1)n+1 22n−1 x2n
b. cos2 x = 1 − sin2 x = 1 −
n=1
(2n)!
∞
X
2 (−1)n 22n−1 x2n
cos x = 1 +
n=1
(2n)!
the radius of convergence is R = ∞.

∞
X ∞
X
−x2 (−x2 )n (−1)n x2n
37. e = =
n=0
n! n=0
n!
Z 1 Z
1
−x2 x2 x4 x6 nx
2n
e dx = 1− + − · · · + (−1) + . . . dx
0 0 1! 2! 3! n!

x3 x5 x7 x9 1
= x− + − + ...
3 5 · 2! 7 · 3! 9 · 4! 0
1 1 1 1
=1− + − + − ...
3 10 42 216
1 1 1 1
≈1− + − + ≈ 0.7475
3 10 42 216
The alternating Series Estimation Theorem show shows that the error is less than
1 1
= < 0.001
11 · 5! 1320
page 233
43. Let f (x) = tan−1 x.

1
f 0 (x) = f 0 (0) = 1
1 + x2
2x
f 00 (x) = − f 00 (0) = 0
(1 + x2 )2
−2 + 6x2
f 000 (x) = f 000 (0) = −2
(1 + x2 )3
24x(1 − x2 )
f (4) = f (4) (0) = 0
(1 + x2 )4
24[1 − 10x2 + 5x4 ]
f (5) (x) = f (5) (0) = 24
(1 + x2 )5

Hence,
f 00 (0) f 000 (0) f (4) (0) f (5) (0)

f (0) + f 0 (0) + + + + ...
2! 3! 4! 5!
X∞
x3 x5 (−1)n x2n+1
= x− + + ··· =
3 5 n=0
2n + 1
Appendix A
page 242
Xn k n
1 2 1 1
1. P (n) : − = + − .
k=0
2 3 3 2
X1 k
1 1 1 2 1 1
P (1) : − =1+− = = + − .
k=0
2 2 2 3 3 2
Xn k n
1 2 1 1
Induction hypothesis: P (n) : − = + − .
k=0
2 3 3 2
To prove:
n+1
X k n+1
1 2 1 1
P (n + 1) : − = + −
k=0
2 3 3 2
n
2 1 1 1
= + − −
3 3 2 2
n
2 1 1
= + − −
3 2 6
n+1
X k
1
− left side of P (n + 1)
k=0
2
X n k n+1
1 1
= − + −
k=0
2 2
n n+1
2 1 1 1
= + − + − since P (n) is true
3 3 2 2
n n
2 1 1 1 2 1 1
= + − − = + − − right side of P (n + 1)
3 2 3 2 3 2 6
X
n
n(n + 1)(2n + 1)
2. P (n) : k2 = .
k=0
6

1
X (1 + 1)(2 + 1)
P (1) : k2 = 0 + 1 = .
k=0
6
X
n
n(n + 1)(2n + 1)
Induction hypothesis: P (n) : k2 = .
k=0
6
To prove:
X
n+1
(n + 1)(n + 2)(2(n + 1) + 1)
P (n + 1) : k2 =
k=0
6
(n + 1)(n + 2)(2n + 3) (n + 1)(2n2 + 7n + 6)
= =
6 6
X
n+1
k2 left side of P (n + 1)
k=0
X
n
= k 2 + (n + 1)2
k=0
n(n + 1)(2n + 1)
= + (n + 1)2 since P (n) is true
6
(n + 1)[n(2n + 1) + 6(n + 1)]
=
6
(n + 1)(2n2 + 7n + 6)
= right side of P (n + 1)
6
X
n
3. P (n) : k(2k ) = 2 + (n − 1)2n+1 .
k=0
1
X
P (1) : k(2k ) = 0 + 2 = 2 + (1 − 1)21+1 .
k=0
X
n
Induction hypothesis: P (n) : k(2k ) = 2 + (n − 1)2n+1 .
k=0

X
n+1
To prove: k(2k ) = 2 + n2n+2 .
k=0
X
n+1
k(2k ) left side of P (n + 1)
k=0
X
n
= k(2k ) + (n + 1)2n+1
k=0
= 2 + (n − 1)2n+1 + (n + 1)2n+1 since P (n) is true
= 2 + 2n+1 [(n − 1) + (n + 1)] = 2 + n2n+2 right side of P (n + 1)
4. P (n) : 2n > n true for n = 1 since P (1) : 2 > 1.
Induction hypothesis: P (n) : 2n > n
To prove: P (n + 1) : 2n+1 > n + 1.
2n+1 = 2n (2) > n(2) = n + n > n + 1.
5. P (n) : n! < nn is true for n = 2, since P (2) : 2! = 2 < 4 = 22 .
Induction hypothesis: P (n) : n! < nn .
To prove: P (n + 1) : (n + 1)! < (n + 1)n+1 = (n + 1)n (n + 1).
(n + 1)! = n!(n + 1) < nn (n + 1) < (n + 1)n (n + 1).
6. P (n) : (2n + 1)2 − 1 is divisible by 8.
P (1) : 12 − 1 = 0 is divisible by 8.
Induction hypothesis: P (n) : (2n + 1)2 − 1 is divisible by 8, or (2n + 1)2 − 1 = 8k for some
integer k.
To prove: P (n + 1) = ((2(n + 1) + 1)2 − 1 = (2n + 3)2 − 1 is divisible by 8.
(2n + 3)2 − 1 left side of P (n + 1)
= 4n2 + 12n + 9 − 1 = 4n2 + 4n + 8n + 8
= (2n + 1)2 − 1 + 8n + 8 = 8k + 8n + 8 since P (n) is true
= 8(k + n + 1).

Hence, P (n + 1) is true.
7. a0 = 1, a1 = 1, a2 = 1, and for n ≥ 3 we define an = an−1 + an−3 .

√
P (n) : an+2 ≥ ( 2)n .
√
For n = 3 we have P (3) : a3 = a2 + a0 = 1 + 1 > 2.
√
Induction hypothesis: P (k) : ak+2 ≥ ( 2)k for all 1 ≤ k ≤ n.
√ √ √
To prove: P (n + 1) : an+1 ≥ ( 2)n+1 = 2( 2)n .
an+1 = an + an−2 left side of P (n + 1)

√ √
≥ ( 2)n−2 + ( 2)n−4 since P (k) is true for 1 ≤ k ≤ n

√ 1 1
= ( 2)n + .
2 4
√
3( 2)n √ √ n
= > 2( 2) right side of P (n + 1).
4

Appendix D: Practice Examination Solutions
In our solutions, we give references to the relevant sections of this Study Guide and of the textbook for
clarity. We do not expect you to quote references in the examinations.
In these solutions, we leave some steps for you to verify. Note that all steps must be included in your
answers to examination questions.
In general, the number of points in each question indicates the number of steps needed or concepts
applied to answer the question.
You must justify your work and give clear evidence of your knowledge.
Unit 1
1. (4 points) The graph of the function shown below is the simplest. Also correct is any function that
increases or decreases on (−∞, 0) and has one or more local maxima or minima in (0, ∞). See
Theorem 1.4 on page 6.
2. (4 points) By Theorem 1.4 on page 6 this question is equivalent to finding the intervals where the
function is either increasing or decreasing. Hence,
√ √
f 0 (x) = x2 − 6x − 2 = 0 if x = 3 − 11 or x = 3 + 11.
√ √
The function
√ is√increasing on (−∞, 3 − 11) and (3 + 11, ∞) and decreasing on
(3 − 11, 3 + 11). A correct answer is any two of these three intervals.
3. (4 points) Method 1. Finding the inverse function:

6x + 4
g(x) = = t.
3+x
Solving for x gives
3t − 4
6x + 4 = 3t + tx 6x − tx = 3t − 4 x= .
6−t
Hence,
3t − 4
g −1 (t) = .
6−t

The answer is yes.
Method 2. Using Definition 1.1 on page 3:

6f (t) + 4 3t − 4 6−t 7t 6−t
g(f (t)) = = 6 +4 = = t.
3 + f (t) 6−t 14 6−t 7
Similarly,

3g(x) − 4 6x + 4 3+x
f (g(x)) = = 3 −4 = x.
6 − g(x) 3+x 14
1
4. (4 points) h0 (x) = e4x−2 (4x2 + 2x) = 2xe4x−2 (2x + 1) = 0 if x = 0 or x = − .
2
1 1
Hence, h0 (x) > 0 for x < − and x > 0, also h0 (x) < 0 for − < x < 0.
2 2

1 1
The function is increasing on −∞, − and (0, ∞), and decreasing on − , 0 .
2 2
5. a. (4 points) By Equations (1.8) on page 20 and (1.13) on page 29,
d 2 2x 2
ln(x2 − 4) + e2−cos(x ) = 2 + 2x sin(x2 )e2−cos(x ) .
dx x −4
b. (8 points) By Equation (1.13) on page 29,

dp x
ln(3 − x2 ) = − p .
dx (3 − x2 ) ln(3 − x2 )
To use logarithmic differentiation, let

x x
y=− p = p .
(3 − x2 ) 2
ln(3 − x ) (x − 3) ln(3 − x2 )
2
Then
ln(ln(3 − x2 ))
ln y = ln x − ln(x2 − 3) −
2
and
y0 1 2x 2x
= + 2 + .
y x x − 3 2(3 − x ) ln(3 − x2 )
2
The answer is
!
x 1 2x 2x
p − 2 + .
(3 − x2 ) ln(3 − x2 ) x x − 3 2(3 − x ) ln(3 − x2 )
2

c. (4 points) Applying the Product Rule
d
sinh(3x − 1) tanh(ex ) = 3 cosh(3x − 1) tanh(ex ) + sinh(3x − 1)sech2 (ex )ex .
dx
6. (4 points) The domain consists of all x so that sinh x > 0. From the graph of sinh x, we see that
sinh x > 0 for all x > 0. Hence, the domain is the interval (0, ∞).
e−x + ex
7. a. (4 points) The hyperbolic cosine function is even because cosh(−x) = = cosh x. We
2
can also see this from its graph. Hence, g(−x) = ln(cosh(−x)) = ln(cosh x) = g(x). So the
function g is even.
e−x − ex
b. (4 points) The hyperbolic sine function is odd because sinh(−x) = = − sinh x. We
2
can also see this from its graph. Hence,
h(−x) = cos(sinh(−x)) = cos(− sinh x) = cos(sinh x)) = h(x). So the function h is even.
c. (4 points) The hyperbolic tangent function is odd because it is the product of an odd and an
even function. Hence, f (−x) = etanh(−x) = e− tanh x . So the function f is neither.
8. (4 points) h0 (x) = −2x sech2 (4 − x2 ) and h0 (2) = −4 sech2 (0) = −4.
Hence, the equation is
y = −4(x − 2).
9. a. (4 points) From the graph of tanh x, we see that lim tanh t = −1. Hence,
t→−∞
lim etanh x = lim ex = e−1 .

x→−∞ x→−1
b. (4 points) The hyperbolic cosine function is continuous everywhere and cosh(0) = 1. Hence,
lim ln(cosh x) = ln(1) = 0.

x→0
c. (4 points) We have that lim+ ln x = −∞. Hence,

x→0
lim tanh(ln x) = lim tanh t = −1.

x→0+ t→−∞
d x
10. a. (4 points) (e − 2x) = ex − 2.
dx
By Equation (1.13) on page 29,
Z Z x
2 − ex e −2
dx = − dx = − ln |ex − 2x| + C.
e − 2x
x ex − 2x

b. (4 points)
Z Z
1 sinh(3 − 5x)
cosh(3 − 5x) dx = − (−5) cosh(3 − 5x) dx = − + C.
5 5
sinh x d
c. (4 points) We have tanh x = and cosh x = sinh x. By Equation (1.13) on page 29,
cosh x dx
Z
tanhx dx = ln | cosh x| + C.
d. (4 points) We have
1 cosh x + sin hx
= = cosh x − sinh x.
cosh x − sinh x cosh2 x − sinh2 x
Hence,
Z Z
dx
= cosh x − sinh x dx = sinh x − cosh x.
cosh x − sinh x
Unit 2
1. a. (12 points) For the domain of cosh−1 (x2 − x), we need all x so that −1 ≤ x2 −
√ x ≤ 1. Since
1± 5
x2 − x is a concave up parabola and x2 − x 6= −1 and x2 − x = 1 for x = . We have that
2
√ √
1 1 − 5 1 + 5
− ≤ x2 − x ≤ 1 for all ≤x≤ .
4 2 2
h √ √ i
the domain of the function is the interval , 2 . The range is the interval [0, π].
1− 5 1+ 5
2

1 1
b. The domain of sin −1
is all x so that −1 ≤ ≤ 1. This gives the interval [0, 2].
x h− 1 i x−1
π π
The range is the interval − , .
2 2
c. The domain of cos−1 (ln x) is all x so that −1 ≤ ln x ≤ 1. Since the exponential function is
increasing e−1 ≤ x ≤ e; thus, the domain is the interval [e−1 , e]. The range is the interval [0, π].
eu − e−u
2. (6 points) By definition of tanh u = . If u = ln x, we have
eu + e−u
eln x − e− ln x x − x−1 x2 − 1
tan(ln x) = = = .
eln x + e− ln x x + x−1 x2 + 1
1 x2 − 1
Observe that x − x−1 = x − = .
x x

9 5
3. (6 points) We have that cosh2 x − sinh2 x = 1; thus, sinh2 x = − 1 = . Also,
4 4
2 2 4 5
1 − tanh x = sech x; thus, tanh x = 1 − = . Therefore,
2
9 9
2
• sech x = .
3
√
5
• sinh x = .
2
2
• csch x = √
5
√
5
• tanh x = .
3
3
• coth x = √ .
5
d 4x3 + sin−1 x
4. a. (4 points)
dx ln x
x ln x(12x2 + (1 − x2 )−1/2 ) − (4x3 + sin−1 x)
=
x(ln x)2
b. (5 points) By the Chain Rule
d 3
cosh−1 (3x + 1) = − .
dx 1 − (3x + 1)2
c. By the Chain Rule

d ex
tan−1 (ex ) = .
dx 1 + e2x
Thus,
d2 −1 x d ex ex (1 − e2x )
tan (e ) = = .
dx2 dx 1 + e2x (1 + e2x )2
5. a. (5 points) Completing the perfect square,

2
2 2 9 9 3 3
2x − 6x + 3 = 2 x − 3x + +3− =2 x− − .
4 2 2 2
3
Let u = x − ; then du = dx and
2
Z Z Z
dx du 1 du
√ = r =√ r .
2
2x − 6x + 3 3 2 3
2
2u − 2
u −
2 4

By Example 2.33 on page 66,
Z √
du
r = cosh−1 (2u/ 3) + C.
3
u2 −
4
The answer is

1 2x − 3
√ cosh−1 √ + C.
2 3
b. (5 points) Completing the perfect square,

2
2 2 25 25 5 29
2 − 10x − 2x = −2 x + 5x + +2+ = −2 x + + .
4 2 2 2
5
Let u = x + ; then du = dx and
2
Z Z Z
dx du 1 du
√ = r =√ r .
2 − 5x − 2x 2 29 2 29
2 2
−2u + −u
2 4
√
29
By Equation (2.2) on page 48, with a = ,
2
Z
du −1 2u
r = sin √ + C.
29 2 29
−u
4
The answer is

1 2x + 5
√ sin−1 √ + C.
2 29
6. a. (4 points) We apply the integral (2.7) on page 66 with a = 2. Hence,

Z ln(2) x ln(2) √
dx
√ = cosh = cosh(ln( 2)) − 1.
0 x2 − 4 2 0
√
√ 2 1 3
Moreover, by definition of cosh x we have that cosh(ln( 2)) = + √ = √ .
2 2 2 2 2
b. (4 points) The derivative of tanh x is sech2 x. Hence,
Z 4
1
3 tanh4 x 1 1 e2 − 1
2
tanh xsech x dx = = .
0 4 0 4 e2 + 1

(ex − e−x )2 e2x − 2 + e−2x
c. (4 points) We have that sinh2 x = = . Hence,
4 4
Z 1 Z
2 1 1 2x
sinh x dx = e − 2 + e−2x dx
−1 4
−12x
1 e e−2x 1
= − 2x −
4 2 2 −1
1 e2 − e−2 − 4
= (e2 − 4 − e−2 − (e−2 + 4 − e2 )) =
8 4
7. a. (4 points) Since
√
lim + cos−1 x − π = 0 and lim + x + 1 = 0,
x→−1 x→−1
the given limit is of type 0/0.
By l’Hopital’s Rule,
√
cos−1 x − π 2 x+1
lim √ = lim + − √
x→−1+ x+1 x→−1 1 − x2
s
x+1
= lim + −2
x→−1 (x + 1)(1 − x)
r
1 √
= lim + −2 = − 2.
x→−1 1−x
10. (4 points) The domain of cosh−1 t is [1, ∞), and x2 − x ≥ 1 if x2 − x − 1 ≥ 0.
Solving, we obtain
√ √
1− 5 1+ 5
x≤ and x ≥ .
2 2
The domain is
√ # " √ !
1− 5 1+ 5
−∞, ∪ ,∞ .
2 2
8. a. (4 points) Since
1 1
lim+ = ∞ and lim+ = ∞,
x→0 x x→0 ex −1
the given limit is of type ∞ − ∞.

1 1 ex − 1 − x ex − 1
lim+ − x = lim+ = lim
x→0 x e − 1 x→0 x(ex − 1) x→0+ ex − 1 + xex
x
e 1
= lim+ x x
= .
x→0 2e + xe 2
b. (4 points) Since

2
lim cos = 1 and lim x = ∞,
x→∞ x x→∞
the given limit is of type 1∞ .
By l’Hopital’s Rule, let

x
2
y = cos
x
and

2
ln cos
2 x
ln y = x ln cos = −1
.
x x
Hence,

2 2
ln cos 2 sin
x x
lim −1
= lim
x→∞ x x→∞ 2
− cos
x

2
= lim −2 tan = 0.
x→∞ x
So, the given limit is e0 = 1.
c. (4 points) Since
lim cos x = 1,
x→0
and since cos x < 1 for x → 0, we have cos x → 1− . Thus, for t = cos x, we have
lim tanh−1 (cos x) = lim− tanh−1 t = ∞.

x→0 t→1
See the graph of tan−1 x.
9. a. The function tanh(x − 1) is the shift of the function tanh x to the right by 1 unit. Hence, the
area under the curve is shown on Figure D.1 below.

1
Figure D.1: Area under the curve tanh(x − 1) in the interval [0,2]
b. Since tanh x is odd and a shift does not change the area, the area on the interval [0,2] is given by
the integral
Z 1 1

2 tanh x dx = 2(ln | cosh x|) = 2(ln | cosh(1)| − ln |1|)
0 0
−1

e + e
= 2 ln = 2 ln(e + e−1 ) − ln(4)
2
You can verify that the area is also equal to the integral
Z 0 Z 1
− tanh x dx + tanh x dx.
−1 0
Unit 3
√
1. a. (4 points) u = x, dv = a + bx; thus,
2(a + bx)3/2
du = dx and v = .
3b
Z √ Z
2x(a + bx)3/2 2
x a + bx dx = − (a + bx)3/2 dx
3b 3b
2x(a + bx)3/2 4 (a + bx)5/2
= − 2
3b 3b 5
3/2
2x(a + bx) 4(a + bx)5/2
= − + C.
3b 15b2

A simplification (not required) gives
2(3bx − 2a)(a + bx)3/2

+ C.
15b2
(u − a)
b. (4 points) By substitution, u = a + bx; thus du = b dx and x = ,
b
Z √ Z Z
1 √ 1
x a + bx dx = 2 (u − a) u du = 2 u3/2 − au1/2 du
b b
5/2 3/2

1 2u 2au 2u3/2 (3u − 5a)
= 2 − = .
b 5 3 15b2
Hence, the answer is
2(a + bx)3/2 (3bx − 2a)

+ C.
15b2
2
2x 1 1 1 1
2. a. (5 points) 3x + 2x = 3 x +
2 2
+ − =3 x+ − .
3 9 3 3 3
1
Let u = x + ; thus, du = dx and
3
Z r r
1 √ Z 1
3u2 − dx = 3 u2 − du
3 9
By trigonometric substitution, or Integral 68 from the table of integrals of the textbook, with
1
a= ,
3
r
1
Z r u u2 − r
1 9 1 1
u2 − du = − ln u + u2 − .
9 2 18 9
The answer is
 r 
1 2x
x+ x2 + r
√  3 3 1 1 2x 
3 − ln x + 2
+ x +  + C.
 2 18 3 3 
b. (5 points) x2 + 2x + 5 = (x + 1)2 + 4,

Let u = x + 1 and dx = du, then
Z √
(u − 1)2 u2 + 4 du
Z √ Z √ Z √
2
= u u2 + 4 du − 2u u2 + 4 du + u2 + 4 du.
By trigonometric substitution, or Integral 69 from the table of integrals of the textbook, with
a = 2,
Z √ √ √
u
u2 u2 + 4 du = (4 + 2u2 ) 4 + u2 − 2 ln(u + 4 + u2 ).
8
By Equation (3.1) on page 80,
Z √ 2(u2 + 4)3/2
2u u2 + 4 du = .
3
By trigonometric substitution, or Integral 68 from the able of integrals of the textbook, with
a = 2,
Z √
u√ 2 √
u2 + 4 du = u + 4 + 2 ln(u + u2 + 4)
2
Thus,
u √ √ 2(u2 + 4)3/2
(4 + 2u2 ) 4 + u2 − 2 ln(u + 4 + u2 ) − +
8 3
u√ 2 √
u + 4 + 2 ln(u + u2 + 4)
2
√ u 2 u 2 2
2
= u +4 (4 + 2u ) + − (u + 4)
8 2 3
√
u2 + 4 (3u3 − 8u2 + 12u − 32)
= .
12
The answer is
√
x2 + 2x + 5(3(x + 1)3 − 8(x + 1)2 + 12(x + 1) − 32)
.
12
A further simplification is optional.
Z Z
3. a. (5 points) sin x cos(2x) dx = sin3 x(1 − 2 sin2 x) dx
3
Z Z
3
= sin x dx − 2 sin5 x dx.
By Integral 28 from the table of integrals of the textbook, with n = 5,

Z Z
5 sin4 x cos x 4
sin x dx = − + sin3 x dx
5 5

and
Z Z
sin x dx − 2 sin5 x dx
3
Z Z
3 2 sin4 x cos x 8
= sin x dx + − sin3 x dx
5 5
Z
2 sin4 x cos x 3
= − sin3 x dx.
5 5
Again by Integral 28, but with n = 3
Z Z
3 sin2 x cos x 2 sin2 x cos x 2 cos x
sin x dx = − + sin x dx = − −
3 3 3 3
The answer is

2 sin4 x cos x 3 sin2 x cos x 2 cos x
− − −
5 5 3 3
cos x
= (2 sin4 x + sin2 x + 2)
5
Z Z
b. (5 points) 5 3
tan x sec x dx = tan4 x sec2 x(sec x tan x) dx
Z
= (1 − sec2 x)2 (sec x tan x) dx
Z Z
3
= sec x tan x dx − 2 sec4 x(sec x tan x) dx
Z
+ sec6 x(sec x tan x) dx
sec3 x 2 sec5 x sec7 x

= − + + C.
3 5 7
4. a. (5 points) Let u = x1/6 . Thus,
1 1
du = 5/6
dx = 5 dx,
6x 6u
6u5 du = dx,
x1/2 = u3 , and x1/3 = u2 .

Hence,
Z √ Z
x u3
√ dx = 6u5 du
1+ 3x 1 + u2
Z
6u8
= du
1 + u2
Z
6
= 6u6 − 6u4 + 6u2 − du
1 + u2
6u7 6u5 6u3
= − + − 6 tan−1 u.
7 5 3
The answer is
6x7/6 6x5/6 √
− + 2x3/2 − 6 tan−1 3 x + C.
7 5
sec x cos(2x) cos(2x) cos2 x − sin2 x

b. (5 points) = = .
sin x + sec x sin x cos x + 1 sin x cos x + 1
We have
d
sin x cos x = cos2 x − sin2 x.
dx
Thus,
Z Z
sec x cos(2x) cos2 x − sin2 x
dx = dx = ln | sin x cos x + 1| + C.
sin x + sec x sin x cos x + 1
π
5. (6 points) The differential is ∆x = , and by Simpson’s Rule,
6
Z π/2 p
1 + sin2 x dx
−π/2
r π r π
π 2
≈ 1 + sin − + 4 1 + sin2 −
18 2 3
r π p
+ 2 1 + sin2 − + 4 1 + sin2 0
6
r π r
2 2 π
+ 2 1 + sin + 4 1 + sin
6 3
r π
+ 1 + sin2
2
" r r !
π √ 3 1
= 2 2+8 1+ +4 1+ +4
18 4 4
π √ √ √
= 2 2+4 7+2 5+4 .
18

1
6. a. (3 points) The function is discontinuous at 0 in the interval [−1, 2].
x3
Z 2 Z 0 Z 2
dx dx dx
b. (3 points) = + .
−1 x3 −1 x
3
0 x
3
The first integral is divergent, since

Z 0 Z t
dx dx 1
3
= lim− 3
= lim− − 2 = −∞.
−1 x t→0 −1 x t→0 2t
Hence, the given integral is divergent.
7. (5 points) For all x in the interval [1, ∞), we have

1 1
√ ≤ 2,
1+x 4 x
and
Z Z
∞
dx t
dx 1 t 1
= lim = lim − = lim = − + 1 = 1.
1 x2 t→∞ 1 x 2 t→∞ x 1 t→∞ t
By the Comparison Theorem, the given integral is convergent.
8. (4 points) The error in the Trapezoidal Rule decreases by a factor of about 4 when we double the
value of n, and for Simpson’s Rule, the error decrease by a factor of about 16 when we double n.
Hence, the error in Simpson’s Rule is smaller than that in the Trapezoidal Rule.
9. We consider the indefinite integral and apply parts with u = ln x, dv = x−3 dx. Thus,
du = dx
x
, v = − 2x12 .
Z Z
ln x ln x 1 1
3
dx = − 2 + dx
x 2x 2 x3
ln x 1
=− 2 − 2
2x 4x
Z ∞ Z t
ln x ln x
dx = lim dx
1 x3 t→∞ 1 x
3
ln x 1 t ln t 1 1
= lim − 2 − 2 = lim − − +
t→∞ 2x 4x 1 t→∞ 2t2 4t2 4
By l’Hopital’s Rule
ln t 1
lim − 2
= lim − 2 = 0.
t→∞ 2t t→∞ 4t
Hence,
Z ∞
ln x 1
3
dx = .
1 x 4

Unit 4
1. a. (4 points) See the figure below.
b. i. (3 points) Washer method

Z π/2 " 2 2 #
π cos x π 4
π − x− dx
0 4 2
ii. (3 points) Shell method

Z π/2 2
π cos x π 2
2πx − x− dx
0 4 2
iii. (3 points) Washer method

Z π/2 " 2 2 2 #
π cos x π 2
π +1 − x− +1 dx
0 4 2
iv. (3 points) Shell method

Z π/2 2
3π π cos x π 2
2π −x − x− dx
0 4 4 2
2. a. (3 points) See the figure below.

b. (5 points) If a triangle is equilateral with base b, then its height is
√
3
b
2
and its area is
√ 2
3b
.
4
The base of the triangle is
√ √
36 − 4x2 4 9 − x2
2 = ,
3 3
and its area is

√ √
16 3(9 − x2 ) 4 3(9 − x)2
= .
36 9
Hence, using symmetry, we find that the integral that gives the volume is
√ Z √
3 3
√
4 3 3 4 3 x 4 3 √
9 − x2 dx = 9x − = (27 − 9) = 8 3.
9 0 9 3 0 9
3. (5 points) The volume is given by the integral

Z ∞ Z t
−2x
πe dx = lim π e−2x dx
0 t→∞ 0
πe−2x t π π
=− = lim [−e−2t + 1] = .
2 0 t→∞ 2 2
d 1
4. a. (4 points) cosh−1 x = √ . Thus,
dx x2 − 1
r
p 1 x
1 + y 02 = 1 + 2 =√ .
x −1 2
x −1
Since the domain of cosh−1 x is [1, ∞), the arc length function is
Z t
x
s(t) = √ dx
2
x −1
1
This is an improper integral, since

x
√
x2 − 1
is discontinuous at 1.

Z 4 Z 4
x x
b. (2 points) s(4) = √ dx = lim+ √ dx
2
x −1 t→1 x2−1
1 t
√ 4 √ √ √

= lim+ x2 − 1 = lim+ 15 − t2 − 1 = 15
t→1 t t→1
5. (4 points) The derivatives of sin−1 x and cos−1 x are related as

d d
sin−1 x = − cos−1 x,
dx dx
and these two functions have the same domain: [−1, 1].
Hence, their arc length functions are equal on the interval [−1, 1]. That is
s 2 s 2
Z 1 Z 1
d −1 d −1
1+ sin x dx = 1+ cos x dx.
−1 dx −1 dx
6. (6 points) Solving for x, we have
y 4 + 48
x= for y 6= 0,
7y
and
dx 3(y 4 − 16)
= .
dy 7y 2
Thus,
s 2 s p
dx 9(y 4 − 16)2 49y 4 + 9(y 4 − 16)2
1+ = 1+ = ,
dy 49y 4 7y 2
and the arc length function is

Z tp 4
49y + 9(y 4 − 16)2
s(t) = dy.
7 7y 2
√
7. (6 points) Let y = r2 − x2 be the upper half of a circle with radius r.
The surface rotated by this curve for 0 ≤ x ≤ r around the y-axis is half of a sphere. We find its
surface as follows:
dy x
= −√ ;
dx r 2 − x2
thus,
s 2
dy r
1+ =√
dx r2 − x2

and
Z r √ r
r
2πx √ dx = −2πr r2 − x2 = 2πr2 .
2
r −x 2 0
0
The total surface area of the sphere is 4πr2 .
8. (4 points) A cone is generated by rotating the line

h
y = − (x − r)
r
around the y-axis. Thus.
dy h
=−
dx r
and
s 2 √
dy r2 + h2
1+ = .
dx r
So the surface area is

Z r √ √
r2 + h2 r2 + h2 x2 r √
2πx dx = 2π = πr r2 + h2 .
0 r r 2 0
9. (5 points) The points of intersection are
x2 = −x2 + 2 ⇒ x = ±1.
The are of the region is

Z 1 1
2x3 8
−x2 + 2 − x2 dx = − + 2x = .
−1 3 −1 3
The centroid is
Z Z 4
3 1 2 3 1 3 x 2
1
x̄ x[−2x + 2] dx = −2x3 + 2x] dx = − +x =0
8 −1 8 −1 8 2 −1

Z 1
3
ȳ = (−x2 + 2)2 − x4 dx
16 −1
Z 1
3
= −4x2 + 4 dx
16 −1

3 4x3 1
= − + 4x = 1
16 3 −1
We could have found this answer by sketching the region,
Unit 5
1. (4 points) y 00 = −ab2 sin(bx), and we want
−ab2 sin(bx) + 4a sin(bx) = 0;
thus,
a sin(bx)(−b2 + 4) = 0.
Since a and b are nonzero, any value for a and b = ±2.
2. (5 points) y10 = −e−x + 6e3x and y100 = e−x + 18e3x ,
We want
y 00 − 2y 0 − 3y = e−x + 18e3x − 2(−e−x + 6e3x ) − 3(e−x + 2e3x ) = 0 6= 7e3x .
Hence, the function y1 does not satisfy the equation.
21 3x 7 3x 7e3x
y20 = xe + e = (3x + 1),
4 4 4
and
21e3x 21e3x 21e3x
y200 = + (3x + 1) = [3x + 2].
4 4 4
Thus,

00 0 21e3x 7e3x 7 3x
y − 2y − 3y = [3x + 2] − 2 (3x + 1) − 3 xe = 7e3x ,
4 4 4
and the function y2 satisfies the equation.
3. (5 points) We have x0 = 0 and y0 = 2.

Since the step size is 0.1, the x-values at which the approximate values will be obtained are x1 = 0.1,
x2 = 0.2, x3 = 0.3, x4 = 0.4 and x5 = 0.5.
The first approximations are
y1 = y0 + f (x0 , y0 )∆x = 2 + (2 − 0)(0.1) = 2.2,
y2 = y1 + f (x1 , y − 1)∆x = 2.2 + (2.2 − 0.1)(0.1) = 2.41.
For all approximations, we have
n xn yn f (xn , yn )∆x yn+1 = yn + f (xn , yn )∆x

0 0 2 0.2 2.2
1 0.1 2.2 0.21 2.41
2 0.2 2.41 0.221 2.631
3 0.3 2.631 0.23331 2.8641
4 0.4 2.841 0.24641 3.11051
5 0.5 3.11051 0.26105 3.37156
dy (x2 − 1)y
4. (6 points) = .
dx y+1
Thus, for y 6= 0
y+1
dy = (x2 − 1)dx
y
Z Z
y+1
dy = (x2 − 1)dx
y
x3
y + ln |y| = − x + C.
3
Since y(3) = 1, we have 1 + ln |1| = 9 − 3 + C and C = −5.
Thus,
x3
y + ln |y| = − x − 5.
3
5. (5 points) By Newton’s Law of Cooling, the temperature T at any time t is given by
T (t) = 32 + [72 − 32]e−kt .

1
We can determine k by applying the condition T = 50◦ .
2
18 9
50 = 32 + 40e−k/2 e−k/2 = =
40 20
k
− = ln(9/20) k = −2 ln(9/20) ≈ 1.5970.
2
Therefore, T (t) ≈ 32 + 40e−1.5970t . Now, T (1) ≈ 32 + 40e−1.5970 ≈ 40.1000; that is, the
temperature after 1 minute is approximately 40.10◦ . To find how long it will take for the temperature
to reach 35◦ , we solve
32 + 40e−1.5970t = 35.
40e−1.5970t = 3 − 1.5970t = ln(3/40)
and
ln(3) − ln(40
t= ≈ 1.62.
−1.5970
It will take approximately 1.62 minutes.
6. (5 points) Let P = 10 be the initial amount.
The amount of the sample y at time t is y = P ekt for some constant k.
We know that y(5) = 4.5; thus,
ln .45
y(5) = P e5k = 4.5 and k= ≈ −0.1597.
5
The half life of the substance is the time t such that
y(t) = 5.
Solving
10e−0.1597t = 5,
we have

1
ln
2
t= ≈ 4.3403
−0.1597
in about 4.34 days.

Unit 6
1. a. (2 points) The sequence must be an > an+1 for n > k for some k.

5
An example is .
n!
This sequence is strictly decreasing, and hence, eventually strictly decreasing.

4 4 4 4
Another example is 1, 2, 3, , , , , . . . .
3 4 5 6
b. (2 points) {cos(nπ)} = {(−1)n }. [Why?]
If {an } is increasing or decreasing, then the alternating sequence {(−1)n an } is neither

increasing nor decreasing.
2. a. (3 points) an+1 − an = 2(n+1)! − 2n! = 2n!n · 2n! − 2n! = 2n! [2n!n − 1] > 0 for n ≥ 1.
Hence, the sequence is increasing.
b. (3 points) Let
ln x
f (x) = .
x
Thus,
1 − ln x
f 0 (x) = < 0 for x > e.
x2
Hence, the sequence is strictly decreasing for n > 3, and it is eventually strictly decreasing.
c. (3 points) Let
n2
an = .
n! + 1
Thus, we consider
n2 (n + 1)2
an − an+1 = − .
n! + 1 (n + 1)! + 1
Doing the operations, we have
((n + 1)! + 1)n2 − (n! + 1)(n + 1)2

an − an+1 = .
(n! + 1)((n + 1)! + 1)
Thus, an − an+1 > 0 if the numerator is positive.

We have
((n + 1)! + 1)n2 − (n! + 1)(n + 1)2

= (n + 1)!n2 − n!(n + 1)2 + n2 − (n + 1)2
= (n + 1)!n2 − n!(n + 1)(n + 1) − (2n + 1)
= (n + 1)!n2 − (n + 1)!(n + 1) − (2n + 1)
= (n + 1)![n2 − n − 1] − (2n + 1)
For n ≥ 1,
(n + 1)!(n2 − n − 1) > n2 − n − 1.
n2 − n − 1 > 2n + 1 for n ≥ 4.
Thus, an − an+1 > 0 for n ≥ 4, and the sequence is eventually strictly decreasing.
3. a. (3 points) We have
2n ≤ 2n! for n ≥ 0 and 2n → ∞.
Hence, by Theorem 6.15 on page 176, the given sequence is divergent.
b. (3 points) By l’Hopital’s Rule.

ln x
lim = 0,
x→∞ x
and by Proposition 6.3 on page 161, the given sequence is convergent.
c. (3 points) By the previous question, the given sequence is eventually decreasing and is bounded
below by 0; hence, by the Monotonic Theorem (Theorem 6.14 on page 175), the sequence is
convergent.
e
4. a. (4 points) We have < 1, and the series is geometric with
π
e 4 e
a= and r = ,
π π

since
∞
X e 4 e 5 e 6
e k−1
= + +
+ ...
k=5
π π π π
e 4 e 4 e e 4 e 2
= + + + ...
π π π π π
X∞
e 4 e k
=
k=0
π π
e4 π e4
= = .
π4 π − e π 3 (π − e)
b. (4 points) We have
∞
X ∞
1 1X 1 1
= − .
k=1
(3k + 2)(3k − 1) 3 k=1 3k − 1 3k + 2
This is a telescoping series, and

X
n
1 1 1 1 1 1 1 1
− = − + − + − + ....
k=1
3k − 1 3k + 2 2 5 5 8 8 11

1 1 1 1
− + −
3n − 4 3n − 1 3n − 1 3n + 2
1 1
= −
2 3n + 2
Hence,
∞
X
1 1 1 1 1
= lim − = .
k=1
(3k + 2)(3k − 1) n→∞ 3 2 3n + 2 6
X ∞
1·2 1·2·3 1·2·3·4 n!
5. a. (4 points) 1 + + + + ··· =
1·3 1·3·5 1·3·5·7 n=1
1 · 3 · 5 . . . (2n − 1)
Let
n!
an = .
1 · 3 · 5 . . . (2n − 1)
We apply the Ratio Test, and we have
an+1 (n + 1)! 1 · 3 · 5 . . . (2n − 1) n+1

= = .
an 1 · 3 · 5 . . . (2n + 1) n! 2n + 1
Thus,
an+1 n+1 1
lim = lim = < 1.
n→∞ an n→∞ 2n + 1 2

Hence, the series is convergent.
b. (4 points) We apply the Root Test:

k 1/k
√ √ √ √
lim k− k−1 = lim k − k − 1
k→∞ k→∞
1
= lim √ √ = 0 < 1.
k→∞ k+ k−1
Thus, the series is convergent.
c. (4 points) We have
k! 1 · 2 · 3 · 4...k k
= > ,
2k 2 · 2 · 2 · 2...2 4
and by Theorem 6.15 on page 176 the sequence

k!
2k
is divergent. Hence, the alternating sequence

k k! k!
(−1) k =
2 (−2)k
is divergent, and by the Divergence Test, the given series is divergent.

1
d. (4 points) Let ak = .
ln k
1
Thus, lim = 0, and you can verify that the sequence {ak } is decreasing. Hence, by the
k→∞ ln k
Alternating Test, the series is convergent. We apply the Limit Comparison Test with the
X1
convergent p series .
k
x x 1
lim = lim = lim = lim x = ∞.
x→∞ (ln x)2 n→∞ 2 ln x x→∞ 2/x x→∞
So the given sequence is conditionally convergent.
6. (8 points) Let f (x) = sinh x. Hence,
f (x) = sinh x f (0) = 0

f 0 (x) = cosh x f 0 (0) = 1
f 00 (x) = sinh x f 00 (0) = 0
f 000 (x) = cosh x f 000 (0) = 1

We conclude that f (2n+1) (0) = 1, and f (2n) (0) = 0. Hence, the Maclaurin series is
X∞
x(2n+1)
n=0
(2n + 1)!
7. (6 points) The Taylor polynomial of degree n is

X
n
xk
Pn (x) =
k=0
k!
and

1/2 1
e ≈ Pn .
2
The remainder is
|1/2|n+1 |1/2|n+1
|Rn (1/2)| ≤ e1/2 <2 ,
(n + 1)! (n + 1)!
since
|1/2|4 1
2 = < 0.01.
4! 8(24)
We take n = 3 and
1 1
1 1 79
e1/2 ≈ P3 =1+ + 4 + 8 = ,
2 2 2 6 48
and we conclude that e1/2 ≈ 1.6872.
8. (4 points) Let
k2
bk = |x + 3|k .
(k + 1)!
By the Ratio Test,
bk+1 (k + 1)2 |x + 3|k+1 (k + 1)! (k + 1)2

= = |x + 3|
bk (k + 2)! k 2 |x + 3|k k 2 (k + 2)
and
(k + 1)2
lim |x + 3| = 0.
k→∞ k 2 (k + 2)
Hence, the series converges for all x, and the interval of convergence is (−∞, ∞).

Z Z
x3 2x5 17x7
9. (5 points) ln | cos x| = − tan x dx = − x+ + + + . . . dx
3 15 315
x2 x4 2x6 17x8
=C− − − − − ...
2 12 90 2520
Z Z x ∞
!
x
tan−1 t 1 X (−1)k t2k+1
10. (5 points) dt = dt
0 t 0 t k=0
2k + 1
Z ∞
!
x X (−1)k t2k
= dt
0 k=0
2k + 1
X∞ Z X∞ 2k+1
(−1)k x 2k (−1)k x
= t dt =
k=0
2k + 1 0 k=0
2k + 1 2k + 1
X∞
(−1)k
= 2
x2k+1
k=0
(2k + 1)
∞
X xk
11. a. (6 points) e − 1 =
x
; thus,
k=1
k!
∞
ex − 1 X xk−1
=
x k=1
k!
d ex − 1 ex (x − 1) + 1 ex (x − 1) + 1
b. = and 1= . Hence,
dx x x2 x2 x=1
d X xk−1 d xk−1
∞ ∞
X
1= = ,
dx k=1 k! x=1 k=1 dx k! x=1
d x x2 x3
1+ + + + ...
dx 2 3! 4!
1 2x 3x2

= + + + . . . .
2 3! 4! x=1
X∞
1 2 3 k
= + + + .... =
2 3! 4! k=1
(k + 1)!

Index
alternating harmonic series, 212 integrating factor, 152

arc length, 132 integration by parts, 79
area of a surface of revolution, 134
L’Hopital’s Rule, 71
brace notation, 158 limit
type 0∞, 73
Comparison Test, 207 type 00 , 75
differential equation type 1∞ , 75
order of, 146 type ∞/∞, 71
solution of, 146 type ∞0 , 75
Direct Fields, 147 type 0/0, 71
divergence test, 198 Limit Comparison Test, 209
limit of a sequence
Euler’s Method, 147 formal definition, 183
limit of sequence
Fibonacci sequence, 169
informal definition, 160
function
linear differential equation, 152
injective, 5
logarithmic differentiation, 31
invertible, 3
left invertible, 3 Maclaurin series, 229
logarithmic, 23 Midpoint Rule, 108
one-to-one, 5 error in, 108
rational, 98 mixing problem, 150
improper, 101 Monotonic Sequence Theorem, 174
proper, 101 monotonicity
right invertible, 3 test of, 172
hyperbolic functions, 34 natural logarithm, 26
hypothesis of induction, 238
power series, 222
improper integral radius of convergence, 222
comparison test, 116 coefficients of, 222
infinite limit of a sequence Principle of Mathematical Induction, 238
informal definition, 161 Principle of Strong Mathematical Induction, 240
initial condition, 146
initial-value problem, 146 Ratio Test, 216
Integral Root Test, 216
improper, 110
separable equation, 148
integral
sequence
improper Type 1, 111
of partial sums, 191
improper type 2, 112
strictly increasing, 171
proper, 110

strictly monotonic, 171
bounded, 174
bounded above, 173
bounded below, 173
convergent, 160
decreasing, 171
divergent, 161
eventual property, 163
geometric, 178
harmonic, 179
increasing, 171
index, 158
limit does not exist, 162
monotonic, 171
recursive, 168
strictly decreasing, 171
terms of, 157
sequences equal, 159
series, 192
-p, 204
absolutely convergent, 214
alternating, 211
conditionally convergent, 214
convergent, 193
divergent, 193
geometric, 195
telescoping, 197
Simpson’s Rule, 109
error in, 109
Squeeze Theorem for Sequences, 167
sum of the series, 194
Taylor polynomial, 229

Taylor series, 229
remainder of, 230
Taylor’s Inequality, 231
Trapezoid Rule
error in, 108
Trapezoidal Rule, 108
volume of a solid, 124

Math 266 Introduction To Calculus II Study Guide

Uploaded by

Math 266 Introduction To Calculus II Study Guide

Uploaded by

Mathematics 266

Unit 1 Inverse Functions: Exponential, Logarithmic and Hyperbolic Functions 1

Unit 2 Inverse Trigonometric and Hyperbolic Functions; L’Hopital’s Rule 43

Unit 3 Techniques of Integration 79

Unit 4 Applications of Integration 123

Unit 5 Differential Equations 145

Unit 6 Infinite Sequences and Series 157

Recommendations for Learning Mathematics

Mathematics 266: Introduction to Calculus II Study Guide i

◦ How would you describe the general appearance of the examination?

ii Study Guide Mathematics 266: Introduction to Calculus II

References in This Study Guide

Mathematics 266: Introduction to Calculus II Study Guide iii

You may also want to look into the websites below.

Labs and Quizzes

iv Study Guide Mathematics 266: Introduction to Calculus II

Inverse Functions: Exponential, Logarithmic and

1. determine whether a function is invertible.

2. find the inverse of an invertible function.

3. apply the properties of the exponential and logarithmic functions.

4. differentiate exponential functions.

5. differentiate logarithmic functions.

6. differentiate functions using logarithmic differentiation.

7. define hyperbolic functions.

8. differentiate hyperbolic functions.

• determine the domain and range of a function.

Mathematics 266: Introduction to Calculus II Study Guide 1

To solve the equation

we could proceed as follows:

2 Study Guide Mathematics 266: Introduction to Calculus II

Then, to solve Equation (1.2), we apply the function f and we get

Let us consider some easy examples.

Thus, x = (3 + 5)2 = 82 = 64.

a. A function f : D −→ R is left invertible on its domain D if there is a function gl : R −→ D

b. A function f : D −→ R is right invertible on its range R if there is a function gr : R −→ D such

Mathematics 266: Introduction to Calculus II Study Guide 3

(gl ◦ f )(gr (t)) = gr (t).

gl (t) = gl ((f ◦ gr )(t)) = gl (f (gr (t))) = (gl ◦ f )(gr (t)) = gr (t)

We then conclude that if f is invertible, there is one function g such that

(f ◦ g)(t) = t and (g ◦ f )(x) = x.

This function g is called the inverse of f , and is denoted by f −1 . Therefore,

(f ◦ f −1 )(t) = t and (f −1 ◦ f )(x) = x.

a = f −1 (f (a)) = f −1 (b) and f −1 (b) = a.

4 Study Guide Mathematics 266: Introduction to Calculus II

Figure 1.4: Functions f and g, where g = f −1

A function f is injective or one-to-one if, whenever f (a) = f (c), then a = c.

f (a) 6= f (c) whenever a 6= c.

Mathematics 266: Introduction to Calculus II Study Guide 5

The following theorem indicates that invertible functions are one-to-one.

Proof: If f is invertible, then there is a function f −1 such that

(f ◦ f −1 )(t) = t and (f ◦ f −1 )(x) = x.

If f (a) = f (c) and we apply the function f −1 , we obtain

and therefore a = c. So f is one-to-one by Definition 1.2 on the preceding page.

Let us define the function g as g(y) = x; then, g : R −→ D and g(f (x)) = x.

Therefore, g = f −1 , and f is invertible. Q.E.D.1

If a function f : I −→ R with range R is increasing or decreasing on I, then f is one-to-one, and

Proof: Let a 6= b in I. We can assume, without loss of generality, that a < b.

f (b) − f (a) = f 0 (c)(b − a) 6= 0.

6 Study Guide Mathematics 266: Introduction to Calculus II

Hence, by Definition 1.2 on page 5, the function f is one-to-one. Q.E.D.

Now let us look at the domain and range of invertible functions.

If D and R are the domain and range of the invertible function f

the domain of f = the range of f −1 ;

the range of f = the domain of f −1 .

The graphs of these functions are shown in Figure 1.5, below.

Mathematics 266: Introduction to Calculus II Study Guide 7

Figure 1.7: Inverse of the sine function on the interval [−1, 1]

Example 1.6. The domain of the function

8 Study Guide Mathematics 266: Introduction to Calculus II

and we can verify that

We then have f 0 (x) = 3x2 + 1 and f 0 (f −1 (1)) = f 0 (0) = 1. Thus,

Mathematics 266: Introduction to Calculus II Study Guide 9

Pay attention to this result.

10 Study Guide Mathematics 266: Introduction to Calculus II

Example 1.8. Sketch the graph of the function

• the domain of f is the interval (4, ∞)