APEX Calculus1 Version 4

APEX C I
Version 4.0
Gregory Hartman, Ph.D.

Department of Applied Mathema cs
Virginia Military Ins tute
Contribu ng Authors
Troy Siemers, Ph.D.
Brian Heinold, Ph.D.

Department of Mathema cs and Computer Science
Mount Saint Mary’s University
Dimplekumar Chalishajar, Ph.D.

Editor
Jennifer Bowen, Ph.D.
Department of Mathema cs and Computer Science
The College of Wooster

Copyright © 2018 Gregory Hartman
Licensed to the public under Crea ve Commons
A ribu on-Noncommercial 4.0 Interna onal Public
License
Contents
Table of Contents iii
Preface v
1 Limits 1
1.1 An Introduc on To Limits . . . . . . . . . . . . . . . . . . . . . 1
1.2 Epsilon-Delta Defini on of a Limit . . . . . . . . . . . . . . . . 9
1.3 Finding Limits Analy cally . . . . . . . . . . . . . . . . . . . . . 18
1.4 One Sided Limits . . . . . . . . . . . . . . . . . . . . . . . . . 30
1.5 Con nuity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
1.6 Limits Involving Infinity . . . . . . . . . . . . . . . . . . . . . . 46
2 Deriva ves 59
2.1 Instantaneous Rates of Change: The Deriva ve . . . . . . . . . 59
2.2 Interpreta ons of the Deriva ve . . . . . . . . . . . . . . . . . 75
2.3 Basic Differen a on Rules . . . . . . . . . . . . . . . . . . . . 82
2.4 The Product and Quo ent Rules . . . . . . . . . . . . . . . . . 89
2.5 The Chain Rule . . . . . . . . . . . . . . . . . . . . . . . . . . 100
2.6 Implicit Differen a on . . . . . . . . . . . . . . . . . . . . . . 111
2.7 Deriva ves of Inverse Func ons . . . . . . . . . . . . . . . . . 122
3 The Graphical Behavior of Func ons 129

3.1 Extreme Values . . . . . . . . . . . . . . . . . . . . . . . . . . 129
3.2 The Mean Value Theorem . . . . . . . . . . . . . . . . . . . . . 137
3.3 Increasing and Decreasing Func ons . . . . . . . . . . . . . . . 142
3.4 Concavity and the Second Deriva ve . . . . . . . . . . . . . . . 151
3.5 Curve Sketching . . . . . . . . . . . . . . . . . . . . . . . . . . 159
4 Applica ons of the Deriva ve 167

4.1 Newton’s Method . . . . . . . . . . . . . . . . . . . . . . . . . 167
4.2 Related Rates . . . . . . . . . . . . . . . . . . . . . . . . . . . 174
4.3 Op miza on . . . . . . . . . . . . . . . . . . . . . . . . . . . . 181
4.4 Differen als . . . . . . . . . . . . . . . . . . . . . . . . . . . . 188
5 Integra on 197
5.1 An deriva ves and Indefinite Integra on . . . . . . . . . . . . 197
5.2 The Definite Integral . . . . . . . . . . . . . . . . . . . . . . . 207
5.3 Riemann Sums . . . . . . . . . . . . . . . . . . . . . . . . . . 218
5.4 The Fundamental Theorem of Calculus . . . . . . . . . . . . . . 236
5.5 Numerical Integra on . . . . . . . . . . . . . . . . . . . . . . . 248
6 Techniques of An differen a on 263

6.1 Subs tu on . . . . . . . . . . . . . . . . . . . . . . . . . . . . 263
A Solu ons To Selected Problems A.1
Index A.11
P
A Note on Using this Text
Thank you for reading this short preface. Allow us to share a few key points
about the text so that you may be er understand what you will find beyond this
page.
This text is Part I of a three–text series on Calculus. The first part covers
material taught in many “Calc 1” courses: limits, deriva ves, and the basics of
integra on, found in Chapters 1 through 6.1. The second text covers material
o en taught in “Calc 2:” integra on and its applica ons, along with an introduc-
on to sequences, series and Taylor Polynomials, found in Chapters 5 through
8. The third text covers topics common in “Calc 3” or “mul variable calc:” para-
metric equa ons, polar coordinates, vector–valued func ons, and func ons of
more than one variable, found in Chapters 9 through 14. All three are available
separately for free at www.apexcalculus.com. These three texts are intended
to work together and make one cohesive text, APEX Calculus, which can also be
downloaded from the website.
Prin ng the en re text as one volume makes for a large, heavy, cumbersome
book. One can certainly only print the pages they currently need, but some
prefer to have a nice, bound copy of the text. Therefore this text has been split
into these three manageable parts, each of which can be purchased for about
$15 at Amazon.com.
A result of this spli ng is that some mes a concept is said to be explored in
a “later sec on,” though that sec on does not actually appear in this par cular
text. Also, the index makes reference to topics and page numbers that do not
appear in this text. This is done inten onally to show the reader what topics are
available for study. Downloading the .pdf of APEX Calculus will ensure that you
have all the content.
For Students: How to Read this Text
Mathema cs textbooks have a reputa on for being hard to read. High–level

mathema cal wri ng o en seeks to say much with few words, and this style
o en seeps into texts of lower–level topics. This book was wri en with the goal
of being easier to read than many other calculus textbooks, without becoming
too verbose.
Each chapter and sec on starts with an introduc on of the coming material,
hopefully se ng the stage for “why you should care,” and ends with a look ahead
to see how the just–learned material helps address future problems.
Please read the text; it is wri en to explain the concepts of Calculus. There
are numerous examples to demonstrate the meaning of defini ons, the truth
of theorems, and the applica on of mathema cal techniques. When you en-
counter a sentence you don’t understand, read it again. If it s ll doesn’t make
sense, read on anyway, as some mes confusing sentences are explained by later
sentences.
You don’t have to read every equa on. The examples generally show “all”
the steps needed to solve a problem. Some mes reading through each step is
helpful; some mes it is confusing. When the steps are illustra ng a new tech-
nique, one probably should follow each step closely to learn the new technique.
When the steps are showing the mathema cs needed to find a number to be
used later, one can usually skip ahead and see how that number is being used,
instead of ge ng bogged down in reading how the number was found.
Most proofs have been omi ed. In mathema cs, proving something is al-
ways true is extremely important, and entails much more than tes ng to see if
it works twice. However, students o en are confused by the details of a proof,
or become concerned that they should have been able to construct this proof
on their own. To alleviate this poten al problem, we do not include the proofs
to most theorems in the text. The interested reader is highly encouraged to find
proofs online or from their instructor. In most cases, one is very capable of un-
derstanding what a theorem means and how to apply it without knowing fully
why it is true.
Interac ve, 3D Graphics
New to Version 3.0 was the addi on of interac ve, 3D graphics in the .pdf
version. Nearly all graphs of objects in space can be rotated, shi ed, and zoomed
in/out so the reader can be er understand the object illustrated.
As of this wri ng, the only pdf viewers that support these 3D graphics are
Adobe Reader & Acrobat (and only the versions for PC/Mac/Unix/Linux com-
puters, not tablets or smartphones). To ac vate the interac ve mode, click on
the image. Once ac vated, one can click/drag to rotate the object and use the
scroll wheel on a mouse to zoom in/out. (A great way to inves gate an image
is to first zoom in on the page of the pdf viewer so the graphic itself takes up
much of the screen, then zoom inside the graphic itself.) A CTRL-click/drag pans
the object le /right or up/down. By right-clicking on the graph one can access
a menu of other op ons, such as changing the ligh ng scheme or perspec ve.
One can also revert the graph back to its default view. If you wish to deac vate
the interac vity, one can right-click and choose the “Disable Content” op on.
Thanks
There are many people who deserve recogni on for the important role they
have played in the development of this text. First, I thank Michelle for her sup-
port and encouragement, even as this “project from work” occupied my me
and a en on at home. Many thanks to Troy Siemers, whose most important
contribu ons extend far beyond the sec ons he wrote or the 227 figures he
coded in Asymptote for 3D interac on. He provided incredible support, advice
and encouragement for which I am very grateful. My thanks to Brian Heinold
and Dimplekumar Chalishajar for their contribu ons and to Jennifer Bowen for
reading through so much material and providing great feedback early on. Thanks
to Troy, Lee Dewald, Dan Joseph, Meagan Herald, Bill Lowe, John David, Vonda
Walsh, Geoff Cox, Jessica Liber ni and other faculty of VMI who have given me
numerous sugges ons and correc ons based on their experience with teaching
from the text. (Special thanks to Troy, Lee & Dan for their pa ence in teaching
Calc III while I was s ll wri ng the Calc III material.) Thanks to Randy Cone for
encouraging his tutors of VMI’s Open Math Lab to read through the text and
check the solu ons, and thanks to the tutors for spending their me doing so.
A very special thanks to Kris Brown and Paul Janiczek who took this opportu-
nity far above & beyond what I expected, me culously checking every solu on
and carefully reading every example. Their comments have been extraordinarily
helpful. I am also thankful for the support provided by Wane Schneiter, who as
my Dean provided me with extra me to work on this project. Finally, a huge
heap of thanks is to be bestowed on the numerous people I do not know who
took the me to email me correc ons and sugges ons. I am blessed to have so
many people give of their me to make this book be er.
APEX – Affordable Print and Electronic teXts
APEX is a consor um of authors who collaborate to produce high–quality,
low–cost textbooks. The current textbook–wri ng paradigm is facing a poten-
al revolu on as desktop publishing and electronic formats increase in popular-
ity. However, wri ng a good textbook is no easy task, as the me requirements
alone are substan al. It takes countless hours of work to produce text, write
examples and exercises, edit and publish. Through collabora on, however, the
cost to any individual can be lessened, allowing us to create texts that we freely
distribute electronically and sell in printed form for an incredibly low cost. Hav-
ing said that, nothing is en rely free; someone always bears some cost. This text
“cost” the authors of this book their me, and that was not enough. APEX Cal-
culus would not exist had not the Virginia Military Ins tute, through a generous
Jackson–Hope grant, given the lead author significant me away from teaching
so he could focus on this text.
Each text is available as a free .pdf, protected by a Crea ve Commons At-
tribu on - Noncommercial 4.0 copyright. That means you can give the .pdf to
anyone you like, print it in any form you like, and even edit the original content
and redistribute it. If you do the la er, you must clearly reference this work and
you cannot sell your edited work for money.
We encourage others to adapt this work to fit their own needs. One might
add sec ons that are “missing” or remove sec ons that your students won’t
need. The source files can be found at github.com/APEXCalculus.
You can learn more at www.vmi.edu/APEX.
Version 4.0
Key changes from Version 3.0 to 4.0:
• Numerous typographical and “small” mathema cal correc ons (again, thanks
to all my close readers!).
• “Large” mathema cal correc ons and adjustments. There were a number
of places in Version 3.0 where a defini on/theorem was not correct as
stated. See www.apexcalculus.com for more informa on.
• More useful numbering of Examples, Theorems, etc. “Defini on 11.4.2”

refers to the second defini on of Chapter 11, Sec on 4.
• The addi on of Sec on 13.7: Triple Integra on with Cylindrical and Spher-
ical Coordinates
• The addi on of Chapter 14: Vector Analysis.
1: L
Calculus means “a method of calcula on or reasoning.” When one computes
the sales tax on a purchase, one employs a simple calculus. When one finds the
area of a polygonal shape by breaking it up into a set of triangles, one is using
another calculus. Proving a theorem in geometry employs yet another calculus.
Despite the wonderful advances in mathema cs that had taken place into
the first half of the 17th century, mathema cians and scien sts were keenly
aware of what they could not do. (This is true even today.) In par cular, two
important concepts eluded mastery by the great thinkers of that me: area and
rates of change.
Area seems innocuous enough; areas of circles, rectangles, parallelograms,
etc., are standard topics of study for students today just as they were then. How-
ever, the areas of arbitrary shapes could not be computed, even if the boundary
of the shape could be described exactly.
Rates of change were also important. When an object moves at a constant
rate of change, then “distance = rate × me.” But what if the rate is not constant
– can distance s ll be computed? Or, if distance is known, can we discover the
rate of change?
It turns out that these two concepts were related. Two mathema cians, Sir
Isaac Newton and Go ried Leibniz, are credited with independently formula ng
a system of compu ng that solved the above problems and showed how they
were connected. Their system of reasoning was “a” calculus. However, as the
power and importance of their discovery took hold, it became known to many
as “the” calculus. Today, we generally shorten this to discuss “calculus.”
The founda on of “the calculus” is the limit. It is a tool to describe a par-
cular behavior of a func on. This chapter begins our study of the limit by ap-
proxima ng its value graphically and numerically. A er a formal defini on of
the limit, proper es are established that make “finding limits” tractable. Once
the limit is understood, then the problems of area and rates of change can be
approached.
1.1 An Introduc on To Limits

We begin our study of limits by considering examples that demonstrate key con-
cepts that will be explained as we progress.
Consider the func on y = sinx x . When x is near the value 1, what value (if
any) is y near?
While our ques on is not precisely formed (what cons tutes “near the value
Chapter 1 Limits
y
1”?), the answer does not seem difficult to find. One might think first to look at a
graph of this func on to approximate the appropriate y values. Consider Figure
1 1.1.1, where y = sinx x is graphed. For values of x near 1, it seems that y takes on
values near 0.85. In fact, when x = 1, then y = sin1 1 ≈ 0.84, so it makes sense
that when x is “near” 1, y will be “near” 0.84.
0.8 Consider this again at a different value for x. When x is near 0, what value
(if any) is y near? By considering Figure 1.1.2, one can see that it seems that y
0.6 takes on values near 1. But what happens when x = 0? We have
sin 0 “0”
. x
y→ → .
0 0
0.5 1 1.5
The expression “0/0” has no value; it is indeterminate. Such an expression gives
Figure 1.1.1: sin(x)/x near x = 1. no informa on about what is going on with the func on nearby. We cannot find
out how y behaves near x = 0 for this func on simply by le ng x = 0.
Finding a limit entails understanding how a func on behaves near a par cu-
lar value of x. Before con nuing, it will be useful to establish some nota on. Let
y
y = f(x); that is, let y be a func on of x for some func on f. The expression “the
limit of y as x approaches 1” describes a number, o en referred to as L, that y
1
nears as x nears 1. We write all this as
lim y = lim f(x) = L.

x→1 x→1
0.9
This is not a complete defini on (that will come in the next sec on); this is a
pseudo-defini on that will allow us to explore the idea of a limit.
0.8 Above, where f(x) = sin(x)/x, we approximated
x sin x sin x
−1 1 lim ≈ 0.84 and lim ≈ 1.
x→1 x x→0 x
Figure 1.1.2: sin(x)/x near x = 0.
(We approximated these limits, hence used the “≈” symbol, since we are work-
ing with the pseudo-defini on of a limit, not the actual defini on.)
Once we have the true defini on of a limit, we will find limits analy cally;
that is, exactly using a variety of mathema cal tools. For now, we will approx-
imate limits both graphically and numerically. Graphing a func on can provide
a good approxima on, though o en not very precise. Numerical methods can
x sin(x)/x provide a more accurate approxima on. We have already approximated limits
0.9 0.870363 graphically, so we now turn our a en on to numerical approxima ons.
0.99 0.844471
Consider again limx→1 sin(x)/x. To approximate this limit numerically, we
0.999 0.841772
1 0.841471 can create a table of x and f(x) values where x is “near” 1. This is done in Figure
1.001 0.84117 1.1.3.
1.01 0.838447 No ce that for values of x near 1, we have sin(x)/x near 0.841. The x = 1 row
1.1 0.810189 is in bold to highlight the fact that when considering limits, we are not concerned
Figure 1.1.3: Values of sin(x)/x with x

near 1.
Notes:
2
with the value of the func on at that par cular x value; we are only concerned
with the values of the func on when x is near 1.
Now approximate limx→0 sin(x)/x numerically. We already approximated x sin(x)/x
the value of this limit as 1 graphically in Figure 1.1.2. The table in Figure 1.1.4 -0.1 0.9983341665
shows the value of sin(x)/x for values of x near 0. Ten places a er the decimal -0.01 0.9999833334
-0.001 0.9999998333
point are shown to highlight how close to 1 the value of sin(x)/x gets as x takes
0 not defined
on values very near 0. We include the x = 0 row in bold again to stress that we
0.001 0.9999998333
are not concerned with the value of our func on at x = 0, only on the behavior 0.01 0.9999833334
of the func on near 0. 0.1 0.9983341665
This numerical method gives confidence to say that 1 is a good approxima-
on of limx→0 sin(x)/x; that is, Figure 1.1.4: Values of sin(x)/x with x
near 0.
lim sin(x)/x ≈ 1.
x→0
Later we will be able to prove that the limit is exactly 1.

We now consider several examples that allow us explore different aspects
of the limit concept.
y
Example 1.1.1 Approxima ng the value of a limit 0.34
Use graphical and numerical methods to approximate
0.32
x2 − x − 6
lim .
x→3 6x2 − 19x + 3 0.3
0.28
S To graphically approximate the limit, graph
0.26
y = (x2 − x − 6)/(6x2 − 19x + 3) x
2.5 3 3.5
on a small interval that contains 3. To numerically approximate the limit, create
a table of values where the x values are near 3. This is done in Figures 1.1.5 and Figure 1.1.5: Graphically approxima ng a
1.1.6, respec vely. limit in Example 1.1.1.
The graph shows that when x is near 3, the value of y is very near 0.3. By
considering values of x near 3, we see that y = 0.294 is a be er approxima on. x x2 −x−6
6x2 −19x+3
The graph and the table imply that
2.9 0.29878
x2 − x − 6 2.99 0.294569
lim 2 ≈ 0.294. 2.999 0.294163
x→3 6x − 19x + 3
3 not defined
3.001 0.294073
This example may bring up a few ques ons about approxima ng limits (and 3.01 0.293669
the nature of limits themselves). 3.1 0.289773
1. If a graph does not produce as good an approxima on as a table, why Figure 1.1.6: Numerically approxima ng
bother with it? a limit in Example 1.1.1.
2. How many values of x in a table are “enough?” In the previous example,
could we have just used x = 3.001 and found a fine approxima on?
Notes:
3
Chapter 1 Limits
Graphs are useful since they give a visual understanding concerning the be-
havior of a func on. Some mes a func on may act “erra cally” near certain
x values which is hard to discern numerically but very plain graphically. Since
graphing u li es are very accessible, it makes sense to make proper use of them.
Since tables and graphs are used only to approximate the value of a limit,
there is not a firm answer to how many data points are “enough.” Include
enough so that a trend is clear, and use values (when possible) both less than
and greater than the value in ques on. In Example 1.1.1, we used both values
less than and greater than 3. Had we used just x = 3.001, we might have been
tempted to conclude that the limit had a value of 0.3. While this is not far off,
y we could do be er. Using values “on both sides of 3” helps us iden fy trends.
1
Example 1.1.2 Approxima ng the value of a limit
Graphically and numerically approximate the limit of f(x) as x approaches 0,
where {
x+1 x<0
0.5 f(x) = .
−x2 + 1 x > 0
S Again we graph f(x) and create a table of its values near x =

x
−1 −0.5 0.5 1 0 to approximate the limit. Note that this is a piecewise defined func on, so it
behaves differently on either side of 0. Figure 1.1.7 shows a graph of f(x), and
Figure 1.1.7: Graphically approxima ng a on either side of 0 it seems the y values approach 1. Note that f(0) is not actually
limit in Example 1.1.2. defined, as indicated in the graph with the open circle.
The table shown in Figure 1.1.8 shows values of f(x) for values of x near 0.
It is clear that as x takes on values very near 0, f(x) takes on values very near 1.
x f(x) It turns out that if we let x = 0 for either “piece” of f(x), 1 is returned; this is
-0.1 0.9 significant and we’ll return to this idea later.
-0.01 0.99 The graph and table allow us to say that limx→0 f(x) ≈ 1; in fact, we are
-0.001 0.999 probably very sure it equals 1.
0.001 0.999999
0.01 0.9999
0.1 0.99 Iden fying When Limits Do Not Exist
Figure 1.1.8: Numerically approxima ng
a limit in Example 1.1.2. A func on may not have a limit for all values of x. That is, we cannot say
limx→c f(x) = L for some numbers L for all values of c, for there may not be a
number that f(x) is approaching. There are three common ways in which a limit
may fail to exist.
1. The func on f(x) may approach different values on either side of c.
2. The func on may grow without upper or lower bound as x approaches c.
3. The func on may oscillate as x approaches c without approaching a spe-

cific value.
Notes:
4
We’ll explore each of these in turn. y

3
Example 1.1.3 Different Values Approached From Le and Right

Explore why lim f(x) does not exist, where 2
x→1
{ 2
x − 2x + 3 x ≤ 1
f(x) = . 1
x x>1
S A graph of f(x) around x = 1 and a table are given in Figures . x

0.5 1 1.5 2
1.1.9 and 1.1.10, respec vely. It is clear that as x approaches 1, f(x) does not
seem to approach a single number. Instead, it seems as though f(x) approaches Figure 1.1.9: Observing no limit as x → 1
two different numbers. When considering values of x less than 1 (approaching in Example 1.1.3.
1 from the le ), it seems that f(x) is approaching 2; when considering values of
x greater than 1 (approaching 1 from the right), it seems that f(x) is approach-
ing 1. Recognizing this behavior is important; we’ll study this in greater depth x f(x)
later. Right now, it suffices to say that the limit does not exist since f(x) is not 0.9 2.01
approaching one value as x approaches 1. 0.99 2.0001
0.999 2.000001
Example 1.1.4 The Func on Grows Without Bound 1.001 1.001
Explore why lim 1/(x − 1)2 does not exist. 1.01 1.01
x→1 1.1 1.1
S A graph and table of f(x) = 1/(x − 1)2 are given in Figures Figure 1.1.10: Values of f(x) near x = 1 in
1.1.11 and 1.1.12, respec vely. Both show that as x approaches 1, f(x) grows Example 1.1.3.
larger and larger.
We can deduce this on our own, without the aid of the graph and table. If x
is near 1, then (x − 1)2 is very small, and: y
1 100
= very large number.
very small number
Since f(x) is not approaching a single number, we conclude that
1 50
lim
x→1 (x − 1)2
does not exist.
Example 1.1.5 The Func on Oscillates . x

0.5 1 1.5 2
Explore why lim sin(1/x) does not exist.
x→0
Figure 1.1.11: Observing no limit as x →
S Two graphs of f(x) = sin(1/x) are given in Figures 1.1.13. 1 in Example 1.1.4.
Figure 1.1.13(a) shows f(x) on the interval [−1, 1]; no ce how f(x) seems to os-
cillate near x = 0. One might think that despite the oscilla on, as x approaches
x f(x)
0.9 100.
0.99 10000.
0.999 1. × 106
Notes: 1.001 1. × 106
1.01 10000.
1.1 100.
Figure 1.1.12: Values of f(x) near x = 1 in

Example 1.1.4.
5
Chapter 1 Limits
0, f(x) approaches 0. However, Figure 1.1.13(b) zooms in on sin(1/x), on the

interval [−0.1, 0.1]. Here the oscilla on is even more pronounced. Finally, in
the table in Figure 1.1.13(c), we see sin(x)/x evaluated for values of x near 0. As
x approaches 0, f(x) does not appear to approach any value.
It can be shown that in reality, as x approaches 0, sin(1/x) takes on all values
between −1 and 1 infinitely many mes! Because of this oscilla on,
lim sin(1/x) does not exist.
x→0
y y
1 1
x sin(1/x)
0.1 −0.544021
0.5 0.5
0.01 −0.506366
0.001 0.82688
x x
−1 −0.5 −0.1 −5 · 10−2
0.0001 −0.305614
0.5 1 5 · 10−2 0.1
1. × 10−5 0.0357488
−0.5 −0.5
1. × 10−6 −0.349994
1. × 10−7 0.420548
. −1 . −1
(a) (b) (c)
Figure 1.1.13: Observing that f(x) = sin(1/x) has no limit as x → 0 in Example 1.1.5.
Limits of Difference Quo ents

We have approximated limits of func ons as x approached a par cular num-
ber. We will consider another important kind of limit a er explaining a few key
ideas.
Let f(x) represent the posi on func on, in feet, of some par cle that is mov-
ing in a straight line, where x is measured in seconds. Let’s say that when x = 1,
f the par cle is at posi on 10 ., and when x = 5, the par cle is at 20 . Another
way of expressing this is to say
20
f(1) = 10 and f(5) = 20.
Since the par cle traveled 10 feet in 4 seconds, we can say the par cle’s average
10
velocity was 2.5 /s. We write this calcula on using a “quo ent of differences,”
or, a difference quo ent:
f(5) − f(1) 10
= = 2.5 /s.
. x 5−1 4
2 4 6
This difference quo ent can be thought of as the familiar “rise over run” used
Figure 1.1.14: Interpre ng a difference to compute the slopes of lines. In fact, that is essen ally what we are doing:
quo ent as the slope of a secant line.
Notes:
6
given two points on the graph of f, we are finding the slope of the secant line f
through those two points. See Figure 1.1.14.

Now consider finding the average speed on another me interval. We again 20
start at x = 1, but consider the posi on of the par cle h seconds later. That is,
consider the posi ons of the par cle when x = 1 and when x = 1 + h. The
difference quo ent is now
10
f(1 + h) − f(1) f(1 + h) − f(1)

= .
(1 + h) − 1 h
. x
Let f(x) = −1.5x2 + 11.5x; note that f(1) = 10 and f(5) = 20, as in our 2 4 6
discussion. We can compute this difference quo ent for all values of h (even (a)
nega ve values!) except h = 0, for then we get “0/0,” the indeterminate form f
introduced earlier. For all values h ̸= 0, the difference quo ent computes the
average velocity of the par cle over an interval of me of length h star ng at 20
x = 1.
For small values of h, i.e., values of h close to 0, we get average veloci es
over very short me periods and compute secant lines over small intervals. See
Figure 1.1.15. This leads us to wonder what the limit of the difference quo ent 10
is as h approaches 0. That is,
f(1 + h) − f(1)
lim = ? . x
h→0 h 2 4 6
(b)
f
As we do not yet have a true defini on of a limit nor an exact method for
compu ng it, we se le for approxima ng the value. While we could graph the
difference quo ent (where the x-axis would represent h values and the y-axis 20
would represent values of the difference quo ent) we se le for making a table.
See Figure 1.1.16. The table gives us reason to assume the value of the limit is
about 8.5. 10
Proper understanding of limits is key to understanding calculus. With limits,

we can accomplish seemingly impossible mathema cal things, like adding up an
. x
infinite number of numbers (and not get infinity) and finding the slope of a line 2 4 6
between two points, where the “two points” are actually the same point. These
(c)
are not just mathema cal curiosi es; they allow us to link posi on, velocity and
accelera on together, connect cross-sec onal areas to volume, find the work Figure 1.1.15: Secant lines of f(x) at x = 1
done by a variable force, and much more. and x = 1 + h, for shrinking values of h
In the next sec on we give the formal defini on of the limit and begin our (i.e., h → 0).
study of finding limits analy cally. In the following exercises, we con nue our
introduc on and approximate the value of limits.
f(1+h)−f(1)
h h
−0.5 9.25
−0.1 8.65
−0.01 8.515
Notes: 0.01 8.485
0.1 8.35
0.5 7.75
Figure 1.1.16: The difference quo ent

evaluated at values of h near 0.
7
Exercises 1.1
Terms and Concepts 13. lim f(x), where
x→2
{
x+2 x≤2
1. In your own words, what does it mean to “find the limit of f(x) = .
3x − 5 x>2
f(x) as x approaches 3”?
14. lim f(x), where
0 x→3
2. An expression of the form 0
is called . { 2
x −x+1 x≤3
f(x) = .
2x + 1 x>3
3. T/F: The limit of f(x) as x approaches 5 is f(5).
15. lim f(x), where

4. Describe three situa ons where lim f(x) does not exist. x→0
x→c {
cos x x≤0
f(x) = .
5. In your own words, what is a difference quo ent? x2 + 3x + 1 x>0
sin x 16. lim f(x), where

6. When x is near 0, is near what value? x→π/2
{
x
sin x x ≤ π/2
f(x) = .
cos x x > π/2
Problems In Exercises 17 – 24, a func on f and a value a are

given. Approximate the limit of the difference quo ent,
In Exercises 7 – 16, approximate the given limits both numer- f(a + h) − f(a)
lim , using h = ±0.1, ±0.01.
ically and graphically. h→0 h
7. lim x2 + 3x − 5 17. f(x) = −7x + 2, a = 3

x→1
18. f(x) = 9x + 0.06, a = −1

8. lim x3 − 3x2 + x − 5
x→0
19. f(x) = x2 + 3x − 7, a = 1
x+1
9. lim 1
x→0 x2 + 3x 20. f(x) = , a=2
x+1
x2 − 2x − 3
10. lim 21. f(x) = −4x2 + 5x − 1, a = −3
x→3 x2 − 4x + 3
x2 + 8x + 7 22. f(x) = ln x, a = 5
11. lim
x→−1 x2 + 6x + 5
23. f(x) = sin x, a = π
2
x + 7x + 10
12. lim 24. f(x) = cos x, a = π
x→2 x2 − 4x + 4
8
1.2 Epsilon-Delta Defini on of a Limit

This sec on introduces the formal defini on of a limit. Many refer to this as “the Note: the common phrase “the ε-δ defi-
epsilon-delta,” defini on, referring to the le ers ε and δ of the Greek alphabet. ni on” is read aloud as “the epsilon delta
defini on.” The hyphen between ε and δ
Before we give the actual defini on, let’s consider a few informal ways of is not a minus sign.
describing a limit. Given a func on y = f(x) and an x-value, c, we say that “the
limit of the func on f, as x approaches c, is a value L”:
1. if “y tends to L” as “x tends to c.”
2. if “y approaches L” as “x approaches c.”
3. if “y is near L” whenever “x is near c.”
The problem with these defini ons is that the words “tends,” “approach,”
and especially “near” are not exact. In what way does the variable x tend to, or
approach, c? How near do x and y have to be to c and L, respec vely?
The defini on we describe in this sec on comes from formalizing 3. A quick

restatement gets us closer to what we want:
3′ . If x is within a certain tolerance level of c, then the corresponding value y =

f(x) is within a certain tolerance level of L.
The tradi onal nota on for the x-tolerance is the lowercase Greek le er
delta, or δ, and the y-tolerance is denoted by lowercase epsilon, or ε. One more
rephrasing of 3′ nearly gets us to the actual defini on:
3′′ . If x is within δ units of c, then the corresponding value of y is within ε units

of L.
We can write “x is within δ units of c” mathema cally as
|x − c| < δ, which is equivalent to c − δ < x < c + δ.
Le ng the symbol “−→” represent the word “implies,” we can rewrite 3′′ as
|x − c| < δ −→ |y − L| < ε or c − δ < x < c + δ −→ L − ε < y < L + ε.
The point is that δ and ε, being tolerances, can be any posi ve (but typically
small) values. Finally, we have the formal defini on of the limit with the nota on
seen in the previous sec on.
Notes:
9
Chapter 1 Limits
Defini on 1.2.1 The Limit of a Func on f

Let I be an open interval containing c, and let f be a func on defined on
I, except possibly at c. The limit of f(x), as x approaches c, is L, denoted
by
lim f(x) = L,
x→c
means that given any ε > 0, there exists δ > 0 such that for all x in I,
where x ̸= c, if |x − c| < δ, then |f(x) − L| < ε.
(Mathema cians o en enjoy wri ng ideas without using any words. Here is
the wordless defini on of the limit:
lim f(x) = L ⇐⇒ ∀ ε > 0, ∃ δ > 0 s.t. 0 < |x − c| < δ −→ |f(x) − L| < ε.)
x→c
Note the order in which ε and δ are given. In the defini on, the y-tolerance
ε is given first and then the limit will exist if we can find an x-tolerance δ that
works.
An example will help us understand this defini on. Note that the explana-
on is long, but it will take one through all steps necessary to understand the
y
ideas.
}
ε = .5
Example 1.2.1 √ Evalua ng a limit using the defini on
2 } Show that lim x = 2.
ε = .5
x→4
S Before we use the formal defini on, let’s try some numerical
1 Choose ε > 0. Then ...
tolerances. What if the y tolerance is 0.5, or ε = 0.5? How close to 4 does x
have to be so that y is within 0.5 units of 2, i.e., 1.5 < y < 2.5? In this case, we
can proceed as follows:
. x
2 4 6
y 1.5 < y < 2.5
√
1.5 < x < 2.5
}
ε = .5
1.52 < x < 2.52
2 } 2.25 < x < 6.25.
ε = .5
... choose δ smaller
than each of these So, what is the desired x tolerance? Remember, we want to find a symmetric
1
interval of x values, namely 4 − δ < x < 4 + δ. The lower bound of 2.25 is 1.75
width width
z
= 1.75
}| {z
= 2.25
}| {
units from 4; the upper bound of 6.25 is 2.25 units from 4. We need the smaller
. x of these two distances; we must have δ < 1.75. See Figure 1.2.1.
2 4 6
With ε = 0.5, we pick any δ < 1.75.
Figure 1.2.1: Illustra ng the ε−δ process.

Notes:
10
Given the y tolerance ε = 0.5, we have found an x tolerance, δ < 1.75, such
that whenever x is within δ units of 4, then y is within ε units of 2. That’s what
we were trying to find.
Let’s try another value of ε.
What if the y tolerance is 0.01, i.e., ε = 0.01? How close to 4 does x have to
be in order for y to be within 0.01 units of 2 (or 1.99 < y < 2.01)? Again, we
just square these values to get 1.992 < x < 2.012 , or
3.9601 < x < 4.0401.
What is the desired x tolerance? In this case we must have δ < 0.0399, which
is the minimum distance from 4 of the two bounds given above.
What we have so far: if ε = 0.5, then δ < 1.75 and if ε = 0.01, then δ <
0.0399. A pa ern is not easy to see, so we√switch to general ε try to determine
δ symbolically. We start by assuming y = x is within ε units of 2:
|y − 2| < ε
−ε < y − 2 < ε (Defini on of absolute value)
√ √
−ε <
x−2< ε (y = x)
√
2−ε< x<2+ε (Add 2)
(2 − ε) < x < (2 + ε)
2 2
(Square all)
4 − 4ε + ε < x < 4 + 4ε + ε
2 2
(Expand)
4 − (4ε − ε2 ) < x < 4 + (4ε + ε2 ). (Rewrite in the desired form)
The “desired form” in the last step is “4 − something < x < 4 + something.”
Since we want this last interval to describe an x tolerance around 4, we have that
either δ < 4ε − ε2 or δ < 4ε + ε2 , whichever is smaller:
δ < min{4ε − ε2 , 4ε + ε2 }.
Since ε > 0, the minimum is δ < 4ε − ε2 . That’s the formula: given an ε, set
δ < 4ε − ε2 .
We can check this for our previous values. If ε = 0.5, the formula gives
δ < 4(0.5)−(0.5)2 = 1.75 and when ε = 0.01, the formula gives δ < 4(0.01)−
(0.01)2 = 0.399.
So given any ε > 0, set δ < 4ε − ε2 . Then if |x − 4| < δ (and x ̸= 4), then
|f(x) − 2| < ε, sa sfying
√ the defini on of the limit. We have shown formally
(and finally!) that lim x = 2.
x→4
Notes:
11
Chapter 1 Limits
The previous example was a li le long in that we sampled a few specific cases
of ε before handling the general case. Normally
√ this is not done. The previous
example is also a bit unsa sfying in that 4 = 2; why work so hard to prove
something so obvious? Many ε-δ proofs are long and difficult to do. In this sec-
on, we will focus on examples where the answer is, frankly, obvious, because
the non–obvious examples are even harder. In the next sec on we will learn
some theorems that allow us to evaluate limits analy cally, that is, without us-
ing the ε-δ defini on.
Example 1.2.2 Evalua ng a limit using the defini on

Show that lim x2 = 4.
x→2
S Let’s do this example symbolically from the start. Let ε > 0

be given; we want |y − 4| < ε, i.e., |x2 − 4| < ε. How do we find δ such that
when |x − 2| < δ, we are guaranteed that |x2 − 4| < ε?
This is a bit trickier than the previous example, but let’s start by no cing that
|x2 − 4| = |x − 2| · |x + 2|. Consider:
ε
|x2 − 4| < ε −→ |x − 2| · |x + 2| < ε −→ |x − 2| < . (1.1)
|x + 2|
ε
Could we not set δ = ?
|x + 2|
We are close to an answer, but the catch is that δ must be a constant value (so
it can’t contain x). There is a way to work around this, but we do have to make an
assump on. Remember that ε is supposed to be a small number, which implies
that δ will also be a small value. In par cular, we can (probably) assume that
δ < 1. If this is true, then |x − 2| < δ would imply that |x − 2| < 1, giving
1 < x < 3.
ε
Now, back to the frac on . If 1 < x < 3, then 3 < x + 2 < 5 (add 2
|x + 2|
to all terms in the inequality). Taking reciprocals, we have
1 1 1
< < which implies
5 |x + 2| 3
1 1
< which implies
5 |x + 2|
ε ε
< . (1.2)
5 |x + 2|
ε
This suggests that we set δ < . To see why, let consider what follows when
5
we assume |x − 2| < δ:
Notes:
12
|x − 2| < δ
ε }
|x − 2| < (Our choice of δ) length of ε
5 ε
ε
|x − 2| · |x + 2| < |x + 2| · (Mul ply by |x + 2|) 4 length of
5 δ = ε/5
ε
|x2 − 4| < |x + 2| · (Combine le side)
5
ε ε
|x2 − 4| < |x + 2| · < |x + 2| · =ε (Using (1.2) as long as δ < 1)
5 |x + 2| δ
. z }| {
x
2
We have arrived at |x2 − 4| < ε as desired. Note again, in order to make this
happen we needed δ to first be less than 1. That is a safe assump on; we want Figure 1.2.2: Choosing δ = ε/5 in Exam-
ε to be arbitrarily small, forcing δ to also be small. ple 1.2.2.
We have also picked δ to be smaller than “necessary.” We could get . by with
a slightly larger δ, as shown in Figure 1.2.2. The dashed outer lines show the
boundaries defined by our choice of ε. The do ed inner lines show the bound-
aries defined by se ng δ = ε/5. Note how these do ed lines are within the
dashed lines. That is perfectly fine; by choosing x within the do ed lines we are
guaranteed that f(x) will be within ε of 4.
In summary, given ε > 0, set δ = ε/5. Then |x − 2| < δ implies |x2 − 4| < ε
(i.e. |y − 4| < ε) as desired. This shows that lim x2 = 4. Figure 1.2.2 gives a
x→2
visualiza on of this; by restric ng x to values within δ = ε/5 of 2, we see that
f(x) is within ε of 4.
Make note of the general pa ern exhibited in these last two examples. In
some sense, each starts out “backwards.” That is, while we want to
1. start with |x − c| < δ and conclude that
2. |f(x) − L| < ε,
we actually start by assuming
1. |f(x) − L| < ε, then perform some algebraic manipula ons to give an

inequality of the form
2. |x − c| < something.
When we have properly done this, the something on the “greater than” side of
the inequality becomes our δ. We can refer to this as the “scratch–work” phase
of our proof. Once we have δ, we can formally start with |x − c| < δ and use
algebraic manipula ons to conclude that |f(x) − L| < ε, usually by using the
same steps of our “scratch–work” in reverse order.
Notes:
13
Chapter 1 Limits
We highlight this process in the following example.

Prove that lim (x3 − 2x) = −1.
x→1
S We start our scratch–work by considering |f(x)−(−1)| < ε:
|f(x) − (−1)| < ε

|x3 − 2x + 1| < ε (Now factor)
|(x − 1)(x + x − 1)| < ε
2
ε
|x − 1| < . (1.3)
|x2 + x − 1|
We are at the phase of saying that |x − 1| < something, where something=

ε/|x2 + x − 1|. We want to turn that something into δ.
Since x is approaching 1, we are safe to assume that x is between 0 and 2.
So
0<x<2
0 < x2 < 4. (squared each term)
Since 0 < x < 2, we can add 0, x and 2, respec vely, to each part of the inequal-
ity and maintain the inequality.
0 < x2 + x < 6
−1 < x2 + x − 1 < 5. (subtracted 1 from each part)
In Equa on (1.3), we wanted |x − 1| < ε/|x2 + x − 1|. The above shows that
given any x in [0, 2], we know that
x2 + x − 1 < 5 which implies that

1 1
< 2 which implies that
5 x +x−1
ε ε
< 2 . (1.4)
5 x +x−1
So we set δ < ε/5. This ends our scratch–work, and we begin the formal proof
(which also helps us understand why this was a good choice of δ).
Given ε, let δ < ε/5. We want to show that when |x − 1| < δ, then |(x3 −
Notes:
14
2x) − (−1)| < ε. We start with |x − 1| < δ:
|x − 1| < δ
ε
|x − 1| <
5
ε ε
|x − 1| < < 2 (for x near 1, from Equa on (1.4))
5 |x + x − 1|
|x − 1| · |x2 + x − 1| < ε
|x3 − 2x + 1| < ε
|(x3 − 2x) − (−1)| < ε,
which is what we wanted to show. Thus lim (x3 − 2x) = −1.

x→1
We illustrate evalua ng limits once more.

Prove that lim ex = 1.
x→0
S Symbolically, we want to take the equa on |ex − 1| < ε and

unravel it to the form |x − 0| < δ. Here is our scratch–work:
|ex − 1| < ε
−ε < ex − 1 < ε (Defini on of absolute value)
1−ε<e <1+ε
x
(Add 1)
ln(1 − ε) < x < ln(1 + ε) (Take natural logs) Note: Recall ln 1 = 0 and ln x < 0 when
0 < x < 1. So ln(1 − ε) < 0, hence we
consider its absolute value.
Making the safe assump on that ε < 1 ensures the last inequality is valid (i.e.,
so that ln(1 − ε) is defined). We can then set δ to be the minimum of | ln(1 − ε)|
and ln(1 + ε); i.e.,
δ = min{| ln(1 − ε)|, ln(1 + ε)} = ln(1 + ε).
Now, we work through the actual the proof:
|x − 0| < δ
−δ < x < δ (Defini on of absolute value)
− ln(1 + ε) < x < ln(1 + ε).
ln(1 − ε) < x < ln(1 + ε). (since ln(1 − ε) < − ln(1 + ε))
Notes:
15
Chapter 1 Limits
The above line is true by our choice of δ and by the fact that since | ln(1 − ε)| >
ln(1 + ε) and ln(1 − ε) < 0, we know ln(1 − ε) < − ln(1 + ε).
1 − ε < ex < 1 + ε (Exponen ate)

−ε < ex − 1 < ε (Subtract 1)
In summary, given ε > 0, let δ = ln(1 + ε). Then |x − 0| < δ implies

|ex − 1| < ε as desired. We have shown that lim ex = 1.
x→0
We note that we could actually show that limx→c ex = ec for any constant c.
We do this by factoring out ec from both sides, leaving us to show limx→c ex−c =
1 instead. By using the subs tu on u = x−c, this reduces to showing limu→0 eu =
1 which we just did in the last example. As an added benefit, this shows that in
fact the func on f(x) = ex is con nuous at all values of x, an important concept
we will define in Sec on 1.5.
This formal defini on of the limit is not an easy concept grasp. Our examples
are actually “easy” examples, using “simple” func ons like polynomials, square–
roots and exponen als. It is very difficult to prove, using the techniques given
above, that lim (sin x)/x = 1, as we approximated in the previous sec on.
x→0
There is hope. The next sec on shows how one can evaluate complicated
limits using certain basic limits as building blocks. While limits are an incredibly
important part of calculus (and hence much of higher mathema cs), rarely are
limits evaluated using the defini on. Rather, the techniques of the following
sec on are employed.
Notes:
16
Exercises 1.2
Terms and Concepts 6. lim (3 − x) = −2
x→5
1. What is wrong with the following “defini on” of a limit? ( )

7. lim x2 − 3 = 6
x→3
“The limit of f(x), as x approaches a, is K”
means that given any δ > 0 there exists ε > 0 ( )
8. lim x2 + x − 5 = 15
such that whenever |f(x) − K| < ε, we have x→4
|x − a| < δ. ( )
9. lim 2x2 + 3x + 1 = 6
x→1
2. Which is given first in establishing a limit, the x–tolerance
or the y–tolerance? ( )
10. lim x3 − 1 = 7
x→2
3. T/F: ε must always be posi ve.

11. lim 5 = 5
x→2
4. T/F: δ must always be posi ve. ( )
12. lim e2x − 1 = 0
x→0
Problems 13. lim

1
=1
x→1 x
In Exercises 5 – 14, prove the given limit using an ε − δ proof.
14. lim sin x = 0 (Hint: use the fact that | sin x| ≤ |x|, with
x→0
5. lim (2x + 5) = 13 equality only when x = 0.)
x→4
17
Chapter 1 Limits
1.3 Finding Limits Analy cally

In Sec on 1.1 we explored the concept of the limit without a strict defini on,
meaning we could only make approxima ons. In the previous sec on we gave
the defini on of the limit and demonstrated how to use it to verify our approxi-
ma ons were correct. Thus far, our method of finding a limit is 1) make a really
good approxima on either graphically or numerically, and 2) verify our approx-
ima on is correct using a ε-δ proof.
Recognizing that ε-δ proofs are cumbersome, this sec on gives a series of
theorems which allow us to find limits much more quickly and intui vely.
Suppose that limx→2 f(x) = 2 and limx→2 g(x) = 3. What is limx→2 (f(x) +
g(x))? Intui on tells us that the limit should be 5, as we expect limits to behave
in a nice way. The following theorem states that already established limits do
behave nicely.
Theorem 1.3.1 Basic Limit Proper es

Let b, c, L and K be real numbers, let n be a posi ve integer, and let f and g be
func ons defined on an open interval I containing c with the following limits:
lim f(x) = L and lim g(x) = K.

x→c x→c
The following limits hold.

1. Constants: lim b = b
x→c
2. Iden ty lim x = c
x→c
3. Sums/Differences: lim (f(x) ± g(x)) = L ± K

x→c
4. Scalar Mul ples: lim b · f(x) = bL

x→c
5. Products: lim f(x) · g(x) = LK

x→c
6. Quo ents: lim f(x)/g(x) = L/K, (K ̸= 0)

x→c
7. Powers: lim f(x)n = Ln

x→c
√ √ n
8. Roots: lim n f(x) = L
x→c
(If n is even then require f(x) ≥ 0 on I.)
9. Composi ons: Adjust our previously given limit situa on to:
lim f(x) = L, lim g(x) = K and g(L) = K.

x→c x→L
Then lim g(f(x)) = K.

x→c
Notes:
18
We make a note about Property #8: when n is even, L must be greater than
0. If n is odd, then the statement is true for all L.
We apply the theorem to an example.
Example 1.3.1 Using basic limit proper es

Let
lim f(x) = 2, lim g(x) = 3 and p(x) = 3x2 − 5x + 7.
x→2 x→2
Find the following limits:

( )
1. lim f(x) + g(x) 3. lim p(x)
x→2 x→2
( 2
)
2. lim 5f(x) + g(x)
x→2
S
( )
1. Using the Sum/Difference rule, we know that lim f(x) + g(x) = 2 + 3 =
x→2
5.
(
2. Using the Scalar Mul ple and Sum/Difference rules, we find that lim 5f(x)+
) x→2
g(x)2 = 5 · 2 + 32 = 19.
3. Here we combine the Power, Scalar Mul ple, Sum/Difference and Con-
stant Rules. We show quite a few steps, but in general these can be omit-
ted:
lim p(x) = lim (3x2 − 5x + 7)

x→2 x→2
= lim 3x2 − lim 5x + lim 7
x→2 x→2 x→2
=3·2 −5·2+7
2
=9
Part 3 of the previous example demonstrates how the limit of a quadra c

polynomial can be determined using the proper es of Theorem 1.3.1. Not only
that, recognize that
lim p(x) = 9 = p(2);
x→2
i.e., the limit at 2 was found just by plugging 2 into the func on. This holds
true for all polynomials, and also for ra onal func ons (which are quo ents of
polynomials), as stated in the following theorem.
Notes:
19
Chapter 1 Limits
Theorem 1.3.2 Limits of Polynomial and Ra onal Func ons

Let p(x) and q(x) be polynomials and c a real number. Then:
1. lim p(x) = p(c)

x→c
p(x) p(c)
2. lim = , where q(c) ̸= 0.
x→c q(x) q(c)
Example 1.3.2 Finding a limit of a ra onal func on

Using Theorem 1.3.2, find
3x2 − 5x + 1
lim .
x→−1 x4 − x2 + 3
S Using Theorem 1.3.2, we can quickly state that

3x2 − 5x + 1 3(−1)2 − 5(−1) + 1
lim =
x→−1 x4 − x2 + 3 (−1)4 − (−1)2 + 3
9
= = 3.
3
It was likely frustra ng in Sec on 1.2 to do a lot of work to prove that
lim x2 = 4
x→2
as it seemed fairly obvious. The previous theorems state that many func ons
behave in such an “obvious” fashion, as demonstrated by the ra onal func on
in Example 1.3.2.
Polynomial and ra onal func ons are not the only func ons to behave in
such a predictable way. The following theorem gives a list of func ons whose
behavior is par cularly “nice” in terms of limits. In the next sec on, we will give
a formal name to these func ons that behave “nicely.”
Theorem 1.3.3 Special Limits

Let c be a real number in the domain of the given func on and let n be a posi ve integer. The
following limits hold:
1. lim sin x = sin c 4. lim csc x = csc c 7. lim ax = ac (a > 0)

x→c x→c x→c
2. lim cos x = cos c 5. lim sec x = sec c 8. lim ln x = ln c

x→c x→c x→c
√ √
3. lim tan x = tan c 6. lim cot x = cot c 9. lim n
x= n
c
x→c x→c x→c
Notes:
20
Example 1.3.3 Evalua ng limits analy cally

Evaluate the following limits.
1. lim cos x 4. lim eln x

x→π x→1
2. lim (sec x − tan x)

2 2
sin x
x→3 5. lim
x→0 x
3. lim cos x sin x
x→π/2
1. This is a straigh orward applica on of Theorem 1.3.3. lim cos x = cos π =

x→π
−1.
2. We can approach this in at least two ways. First, by directly applying The-
orem 1.3.3, we have:
lim (sec2 x − tan2 x) = sec2 3 − tan2 3.

x→3
Using the Pythagorean Theorem, this last expression is 1; therefore
lim (sec2 x − tan2 x) = 1.

x→3
We can also use the Pythagorean Theorem from the start.
lim (sec2 x − tan2 x) = lim 1 = 1,

x→3 x→3
using the Constant limit rule. Either way, we find the limit is 1.
3. Applying the Product limit rule of Theorem 1.3.1 and Theorem 1.3.3 gives
lim cos x sin x = cos(π/2) sin(π/2) = 0 · 1 = 0.

x→π/2
4. Again, we can approach this in two ways. First, we can use the exponen-
al/logarithmic iden ty that eln x = x and evaluate lim eln x = lim x = 1.
x→1 x→1
We can also use the limit Composi on Rule of Theorem 1.3.1. Using The-
orem 1.3.3, we have lim ln x = ln 1 = 0 and lim ex = e0 = 1, sa sfying
x→1 x→0
the condi ons of the Composi on Rule. Applying this rule,
lim eln x = lim ex = e0 = 1.

x→1 x→0
Both approaches are valid, giving the same result.
Notes:
21
Chapter 1 Limits
5. We encountered this limit in Sec on 1.1. Applying our theorems, we at-

tempt to find the limit as
sin x sin 0 “0”
lim → → .
x→0 x 0 0
This, of course, violates a condi on of Theorem 1.3.1, as the limit of the
denominator is not allowed to be 0. Therefore, we are s ll unable to eval-
uate this limit with tools we currently have at hand.
The sec on could have been tled “Using Known Limits to Find Unknown
Limits.” By knowing certain limits of func ons, we can find limits involving sums,
products, powers, etc., of these func ons. We further the development of such
compara ve tools with the Squeeze Theorem, a clever and intui ve way to find
the value of some limits.
Before sta ng this theorem formally, suppose we have func ons f, g and h
where g always takes on values between f and h; that is, for all x in an interval,
f(x) ≤ g(x) ≤ h(x).
If f and h have the same limit at c, and g is always “squeezed” between them,
then g must have the same limit as well. That is what the Squeeze Theorem
states.
Theorem 1.3.4 Squeeze Theorem

Let f, g and h be func ons on an open interval I containing c such that
for all x in I,
f(x) ≤ g(x) ≤ h(x).
If
lim f(x) = L = lim h(x),
x→c x→c
then
lim g(x) = L.
x→c
It can take some work to figure out appropriate func ons by which to “squeeze”
a given func on. However, that is generally the only place where work is neces-
sary; the theorem makes the “evalua ng the limit part” very simple.
We use the Squeeze Theorem in the following example to finally prove that
sin x
lim = 1.
x→0 x
Notes:
22
Example 1.3.4 Using the Squeeze Theorem

Use the Squeeze Theorem to show that
sin x
lim = 1.
x→0 x
S We begin by considering the unit circle. Each point on the

(1, tan θ)
unit circle has coordinates (cos θ, sin θ) for some angle θ as shown in Figure
1.3.1. Using similar triangles, we can extend the line from the origin through the (cos θ, sin θ)
point to the point (1, tan θ), as shown. (Here we are assuming that 0 ≤ θ ≤ π/2.
Later we will show that we can also consider θ ≤ 0.)
Figure 1.3.1 shows three regions have been constructed in the first quadrant, . θ
two triangles and a sector of a circle, which are also drawn below. The area of (1, 0)
the large triangle is 12 tan θ; the area of the sector is θ/2; the area of the triangle
contained inside the sector is 12 sin θ. It is then clear from the diagram that
Figure 1.3.1: The unit circle and related

tan θ triangles.
sin θ
. θ . θ . θ
1 1 1
tan θ θ sin θ
≥ ≥
2 2 2
2
Mul ply all terms by , giving
sin θ
1 θ
≥ ≥ 1.
cos θ sin θ
Taking reciprocals reverses the inequali es, giving
sin θ
cos θ ≤ ≤ 1.
θ
(These inequali es hold for all values of θ near 0, even nega ve values, since
cos(−θ) = cos θ and sin(−θ) = − sin θ.)
Now take limits.
sin θ
lim cos θ ≤ lim ≤ lim 1
θ→0 θ→0 θ θ→0
Notes:
23
Chapter 1 Limits
sin θ
cos 0 ≤ lim ≤1
θ→0 θ
sin θ
1 ≤ lim ≤1
θ→0 θ
sin θ
Clearly this means that lim = 1.
θ→0 θ
Two notes about the previous example are worth men oning. First, one
might be discouraged by this applica on, thinking “I would never have come up
with that on my own. This is too hard!” Don’t be discouraged; within this text we
will guide you in your use of the Squeeze Theorem. As one gains mathema cal
maturity, clever proofs like this are easier and easier to create.
Second, this limit tells us more than just that as x approaches 0, sin(x)/x
approaches 1. Both x and sin x are approaching 0, but the ra o of x and sin x
approaches 1, meaning that they are approaching 0 in essen ally the same way.
Another way of viewing this is: for small x, the func ons y = x and y = sin x are
essen ally indis nguishable.
We include this special limit, along with three others, in the following theo-
rem.
Theorem 1.3.5 Special Limits
sin x 1
3. lim (1 + x) x = e
1. lim =1 x→0
x→0 x
cos x − 1 ex − 1
2. lim =0 4. lim =1
x→0 x x→0 x
A short word on how to interpret the la er three limits. We know that as

x goes to 0, cos x goes to 1. So, in the second limit, both the numerator and
denominator are approaching 0. However, since the limit is 0, we can interpret
this as saying that “cos x is approaching 1 faster than x is approaching 0.”
In the third limit, inside the parentheses we have an expression that is ap-
proaching 1 (though never equaling 1), and we know that 1 raised to any power
is s ll 1. At the same me, the power is growing toward infinity. What happens
to a number near 1 raised to a very large power? In this par cular case, the
result approaches Euler’s number, e, approximately 2.718.
In the fourth limit, we see that as x → 0, ex approaches 1 “just as fast” as
x → 0, resul ng in a limit of 1.
Notes:
24
Our final theorem for this sec on will be mo vated by the following exam-
ple.
Example 1.3.5 Using algebra to evaluate a limit

Evaluate the following limit:
x2 − 1
lim .
x→1 x − 1
S We begin by a emp ng to apply Theorem 1.3.2 and subs -

tu ng 1 for x in the quo ent. This gives:
x2 − 1 12 − 1 “ 0 ” y
lim = = ,
x→1 x − 1 1−1 0 3
an indeterminate form. We cannot apply the theorem.

2
By graphing the func on, as in Figure 1.3.2, we see that the func on seems
to be linear, implying that the limit should be easy to evaluate. Recognize that
the numerator of our quo ent can be factored: 1
x2 − 1 (x − 1)(x + 1)
= .
x−1 x−1 . x
1 2
The func on is not defined when x = 1, but for all other x,

Figure 1.3.2: Graphing f in Example 1.3.5
to understand a limit.
x2 − 1 (x − 1)(x + 1) (x − 1)(x + 1)
= = = x + 1.
x−1 x−1 x−1
Clearly lim x + 1 = 2. Recall that when considering limits, we are not concerned
x→1
with the value of the func on at 1, only the value the func on approaches as x
approaches 1. Since (x2 − 1)/(x − 1) and x + 1 are the same at all points except
x = 1, they both approach the same value as x approaches 1. Therefore we can
conclude that
x2 − 1
lim = 2.
x→1 x − 1
The key to the above example is that the func ons y = (x2 − 1)/(x − 1) and
y = x + 1 are iden cal except at x = 1. Since limits describe a value the func on
is approaching, not the value the func on actually a ains, the limits of the two
func ons are always equal.
Notes:
25
Chapter 1 Limits
Theorem 1.3.6 Limits of Func ons Equal At All But One Point
Let g(x) = f(x) for all x in an open interval, except possibly at c, and let
lim g(x) = L for some real number L. Then
x→c
lim f(x) = L.
x→c
The Fundamental Theorem of Algebra tells us that when dealing with a ra-
g(x)
onal func on of the form g(x)/f(x) and directly evalua ng the limit lim
x→c f(x)
returns “0/0”, then (x − c) is a factor of both g(x) and f(x). One can then use
algebra to factor this term out, cancel, then apply Theorem 1.3.6. We demon-
strate this once more.
Example 1.3.6 Evalua ng a limit using Theorem 1.3.6

x3 − 2x2 − 5x + 6
Evaluate lim 3 .
x→3 2x + 3x2 − 32x + 15
S We a empt to apply Theorem 1.3.2 by subs tu ng 3 for x.

This returns the familiar indeterminate form of “0/0”. Since the numerator and
denominator are each polynomials, we know that (x−3) is factor of each. Using
whatever method is most comfortable to you, factor out (x−3) from each (using
polynomial division, synthe c division, a computer algebra system, etc.). We
find that
x3 − 2x2 − 5x + 6 (x − 3)(x2 + x − 2)
= .
2x3 + 3x2 − 32x + 15 (x − 3)(2x2 + 9x − 5)
We can cancel the (x − 3) terms as long as x ̸= 3. Using Theorem 1.3.6 we
conclude:
x3 − 2x2 − 5x + 6 (x − 3)(x2 + x − 2)
lim = lim
x→3 2x3 + 3x2 − 32x + 15 x→3 (x − 3)(2x2 + 9x − 5)
(x2 + x − 2)
= lim
x→3 (2x2 + 9x − 5)
10 1
= = .
40 4
We end this sec on by revisi ng a limit first seen in Sec on 1.1, a limit of
a difference quo ent. Let f(x) = −1.5x2 + 11.5x; we approximated the limit
f(1 + h) − f(1)
lim ≈ 8.5. We formally evaluate this limit in the following ex-
h→0 h
ample.
Notes:
26
Example 1.3.7 Evalua ng the limit of a difference quo ent

f(1 + h) − f(1)
Let f(x) = −1.5x + 11.5x; find lim
2
.
h→0 h
S Since f is a polynomial, our first a empt should be to em-

ploy Theorem 1.3.2 and subs tute 0 for h. However, we see that this gives us
“0/0.” Knowing that we have a ra onal func on hints that some algebra will
help. Consider the following steps:
( )
f(1 + h) − f(1) −1.5(1 + h)2 + 11.5(1 + h) − −1.5(1)2 + 11.5(1)
lim = lim
h→0 h h→0 h
−1.5(1 + 2h + h2 ) + 11.5 + 11.5h − 10
= lim
h→0 h
−1.5h + 8.5h
2
= lim
h→0 h
h(−1.5h + 8.5)
= lim
h→0 h
= lim (−1.5h + 8.5) (using Theorem 1.3.6, as h ̸= 0)
h→0
= 8.5 (using Theorem 1.3.3)
This matches our previous approxima on.
This sec on contains several valuable tools for evalua ng limits. One of the
main results of this sec on is Theorem 1.3.3; it states that many func ons that
we use regularly behave in a very nice, predictable way. In Sec on 1.5 we give a
name to this nice behavior; we label such func ons as con nuous. Defining that
term will require us to look again at what a limit is and what causes limits to not
exist.
Notes:
27
Exercises 1.3
Terms and Concepts In Exercises 15 – 18, use the following informa on to eval-
uate the given limit, when possible. If it is not possible to
1. Explain in your own words, without using ε-δ formality, why determine the limit, state why not.
lim b = b. • lim f(x) = 2, lim f(x) = 1, f(1) = 1/5
x→c x→1 x→10
• lim g(x) = 0, lim g(x) = π, g(10) = π
2. Explain in your own words, without using ε-δ formality, why x→1 x→10
lim x = c.
x→c
15. lim f(x)g(x)
x→1
3. What does the text mean when it says that certain func- ( )
ons’ “behavior is ‘nice’ in terms of limits”? What, in par- 16. lim cos g(x)
x→10
cular, is “nice”?
17. lim f(x)g(x)
x→1
4. Sketch a graph that visually demonstrates the Squeeze The-
orem. ( )
18. lim g 5f(x)
x→1
5. You are given the following informa on:
In Exercises 19 – 34, evaluate the given limit.
(a) lim f(x) = 0
x→1
19. lim x2 − 3x + 7
(b) lim g(x) = 0 x→3
x→1
(c) lim f(x)/g(x) = 2 ( )7

x→1 x−3
20. lim
x→π x−5
What can be said about the rela ve sizes of f(x) and g(x)
as x approaches 1?
21. lim cos x sin x
x→π/4
6. T/F: lim ln x = 0. Use a theorem to defend your answer.
x→1
2x − 2
22. lim
x→1 x+4
Problems 23. lim ln x

x→0
In Exercises 7 – 14, use the following informa on to evaluate 3

−8x
the given limit, when possible. If it is not possible to deter- 24. lim 4x
x→3
mine the limit, state why not.
25. lim csc x
• lim f(x) = 6, lim f(x) = 9, f(9) = 6 x→π/6
x→9 x→6
• lim g(x) = 3, lim g(x) = 3, g(6) = 9 26. lim ln(1 + x)

x→9 x→6
x→0
7. lim (f(x) + g(x)) x2 + 3x + 5

x→9
27. lim
x→π 5x2 − 2x − 3
8. lim (3f(x)/g(x))
x→9 3x + 1
28. lim
( ) x→π 1−x
f(x) − 2g(x)
9. lim
x→9 g(x) x2 − 4x − 12
29. lim
( ) x→6 x2 − 13x + 42
f(x)
10. lim
x→6 3 − g(x) x2 + 2x
30. lim
( ) x→0 x2 − 2x
11. lim g f(x)
x→9
x2 + 6x − 16
( ) 31. lim
x→2 x2 − 3x + 2
12. lim f g(x)
x→6
( ) x2 − 10x + 16
32. lim
13. lim g f(f(x)) x→2 x2 − x − 2
x→6
x2 − 5x − 14
14. lim f(x)g(x) − f 2 (x) + g2 (x) 33. lim
x→6 x→−2 x2 + 10x + 16
28
x2 + 9x + 8 Exercises 39 – 43 challenge your understanding of limits but
34. lim
x→−1 x2 − 6x − 7 can be evaluated using the knowledge gained in this sec on.
Use the Squeeze Theorem in Exercises 35 – 38, where appro- sin 3x

priate, to evaluate the given limit. 39. lim
x→0 x
( )
1 sin 5x
35. lim x sin 40. lim
x→0 x x→0 8x
( ) ln(1 + x)
1 41. lim
36. lim sin x cos x→0 x
x→0 x2
sin x
37. lim f(x), where 3x − 2 ≤ f(x) ≤ x3 . 42. lim , where x is measured in degrees, not radians.
x→1 x→0 x
38. lim f(x), where 6x − 9 ≤ f(x) ≤ x2 . x

x→3
43. Let f(x) = 0 and g(x) = .
x
(a) Show why lim f(x) = 0.
x→2
(b) Show why lim g(x) = 1.

x→0
( )
(c) Show why lim g f(x) does not exist.
x→2
(d) Show why the answer to part (c) does not violate the
Composi on Rule of Theorem 1.3.1.
29
Chapter 1 Limits
1.4 One Sided Limits

We introduced the concept of a limit gently, approxima ng their values graphi-
cally and numerically. Next came the rigorous defini on of the limit, along with
an admi edly tedious method for evalua ng them. The previous sec on gave us
tools (which we call theorems) that allow us to compute limits with greater ease.
Chief among the results were the facts that polynomials and ra onal, trigono-
metric, exponen al and logarithmic func ons (and their sums, products, etc.) all
behave “nicely.” In this sec on we rigorously define what we mean by “nicely.”
In Sec on 1.1 we saw three ways in which limits of func ons failed to exist:
1. The func on approached different values from the le and right,
2. The func on grows without bound, and
3. The func on oscillates.
In this sec on we explore in depth the concepts behind #1 by introducing
the one-sided limit. We begin with formal defini ons that are very similar to the
defini on of the limit given in Sec on 1.2, but the nota on is slightly different
and “x ̸= c” is replaced with either “x < c” or “x > c.”
Defini on 1.4.1 One Sided Limits: Le - and Right-Hand Limits

Le -Hand Limit
Let f be a func on defined on (a, c) for some a < c and let L be a real
number.
The limit of f(x), as x approaches c from the le , is L, or, the le -hand
limit of f at c is L, denoted by
lim f(x) = L,
x→c−
means given any ε > 0, there exists δ > 0 such that for all a < x < c, if
|x − c| < δ, then |f(x) − L| < ε.
Right-Hand Limit
Let f be a func on defined on (c, b) for some b > c and let L be a real
number.
The limit of f(x), as x approaches c from the right, is L, or, the right-hand
limit of f at c is L, denoted by
lim f(x) = L,
x→c+
means given any ε > 0, there exists δ > 0 such that for all c < x < b, if
|x − c| < δ, then |f(x) − L| < ε.
Notes:
30
Prac cally speaking, when evalua ng a le -hand limit, we consider only val-
ues of x “to the le of c,” i.e., where x < c. The admi edly imperfect nota on
x → c− is used to imply that we look at values of x to the le of c. The nota-
on has nothing to do with posi ve or nega ve values of either x or c. A similar
statement holds for evalua ng right-hand limits; there we consider only values
of x to the right of c, i.e., x > c. We can use the theorems from previous sec ons
to help us evaluate these limits; we just restrict our view to one side of c.
We prac ce evalua ng le - and right-hand limits through a series of exam-
ples.
Example 1.4.1
{ Evalua ng one sided limits
x 0≤x≤1 y
Let f(x) = , as shown in Figure 1.4.1. Find each of the
3−x 1<x<2
following: 2
1. lim f(x) 5. lim f(x)

x→1− x→0+
1
2. lim f(x) 6. f(0)
x→1+

x→2− x
x→1 1 2
.
4. f(1) 8. f(2)
Figure 1.4.1: A graph of f in Example 1.4.1.
S For these problems, the visual aid of the graph is likely more
effec ve in evalua ng the limits than using f itself. Therefore we will refer o en
to the graph.
1. As x goes to 1 from the le , we see that f(x) is approaching the value of 1.

Therefore lim f(x) = 1.
x→1−
2. As x goes to 1 from the right, we see that f(x) is approaching the value of 2.
Recall that it does not ma er that there is an “open circle” there; we are
evalua ng a limit, not the value of the func on. Therefore lim f(x) = 2.
x→1+
3. The limit of f as x approaches 1 does not exist, as discussed in the first

sec on. The func on does not approach one par cular value, but two
different values from the le and the right.
4. Using the defini on and by looking at the graph we see that f(1) = 1.
5. As x goes to 0 from the right, we see that f(x) is also approaching 0. There-
fore lim f(x) = 0. Note we cannot consider a le -hand limit at 0 as f is
x→0+
not defined for values of x < 0.
Notes:
31
Chapter 1 Limits
6. Using the defini on and the graph, f(0) = 0.

7. As x goes to 2 from the le , we see that f(x) is approaching the value of
1. Therefore lim f(x) = 1.
x→2−
8. The graph and the defini on of the func on show that f(2) is not defined.
Note how the le and right-hand limits were different at x = 1. This, of

course, causes the limit to not exist. The following theorem states what is fairly
intui ve: the limit exists precisely when the le and right-hand limits are equal.
Theorem 1.4.1 Limits and One Sided Limits

Let f be a func on defined on an open interval I containing c. Then
lim f(x) = L
x→c
if, and only if,
lim f(x) = L and lim f(x) = L.

x→c− x→c+
The phrase “if, and only if” means the two statements are equivalent: they
are either both true or both false. If the limit equals L, then the le and right
hand limits both equal L. If the limit is not equal to L, then at least one of the
le and right-hand limits is not equal to L (it may not even exist).
One thing to consider in Examples 1.4.1 – 1.4.4 is that the value of the func-
on may/may not be equal to the value(s) of its le /right-hand limits, even when
these limits agree.
y
Example 1.4.2
{ Evalua ng limits of a piecewise–defined func on
2−x 0<x<1
2 Let f(x) = , as shown in Figure 1.4.2. Evaluate the
(x − 2)2 1 < x < 2
following.
1 1. lim f(x) 5. lim f(x)

x→1− x→0+
2. lim f(x) 6. f(0)

x→1+
x
.
1 2
x→2−
x→1
Figure 1.4.2: A graph of f from Example 8. f(2)
4. f(1)
1.4.2
Notes:
32
S Again we will evaluate each using both the defini on of f and

its graph.
1. As x approaches 1 from the le , we see that f(x) approaches 1. Therefore

lim f(x) = 1.
x→1−
2. As x approaches 1 from the right, we see that again f(x) approaches 1.

Therefore lim f(x) = 1.
x→1+
3. The limit of f as x approaches 1 exists and is 1, as f approaches 1 from both

the right and le . Therefore lim f(x) = 1.
x→1
4. f(1) is not defined. Note that 1 is not in the domain of f as defined by the
problem, which is indicated on the graph by an open circle when x = 1.
5. As x goes to 0 from the right, f(x) approaches 2. So lim f(x) = 2.

x→0+
6. f(0) is not defined as 0 is not in the domain of f.
7. As x goes to 2 from the le , f(x) approaches 0. So lim f(x) = 0.

x→2−
8. f(2) is not defined as 2 is not in the domain of f.
Example 1.4.3
y
(x − 1)2 0 ≤ x ≤ 2, x ̸= 1
Let f(x) = , as shown in Figure 1.4.3. Evaluate
1 x=1
the following. 1

x→1− x→1 0.5
2. lim f(x) 4. f(1)

x→1+
x
1 2
S It is clear by looking at the graph that both the le and right- .
hand limits of f, as x approaches 1, are 0. Thus it is also clear that the limit is 0;
i.e., lim f(x) = 0. It is also clearly stated that f(1) = 1. Figure 1.4.3: Graphing f in Example 1.4.3
x→1
Example 1.4.4
x2 0≤x≤1
Let f(x) = , as shown in Figure 1.4.4. Evaluate the fol-
2−x 1<x≤2
lowing.
Notes:
33
Chapter 1 Limits
y 1. lim f(x) 3. lim f(x)

x→1− x→1
2. lim f(x) 4. f(1)

1 x→1+
S It is clear from the defini on of the func on and its graph

0.5
that all of the following are equal:
x lim f(x) = lim f(x) = lim f(x) = f(1) = 1.

x→1− x→1+ x→1
1 2
.
Figure 1.4.4: Graphing f in Example 1.4.4 In Examples 1.4.1 – 1.4.4 we were asked to find both lim f(x) and f(1). Con-
x→1
sider the following table:
lim f(x) f(1)
x→1
Example 1.4.1 does not exist 1
Example 1.4.2 1 not defined
Example 1.4.3 0 1
Example 1.4.4 1 1
Only in Example 1.4.4 do both the func on and the limit exist and agree. This
seems “nice;” in fact, it seems “normal.” This is in fact an important situa on
which we explore in the next sec on, en tled “Con nuity.” In short, a con nu-
ous func on is one in which when a func on approaches a value as x → c (i.e.,
when lim f(x) = L), it actually a ains that value at c. Such func ons behave
x→c
nicely as they are very predictable.
Notes:
34
Exercises 1.4
Terms and Concepts 2
y
1.5
1. What are the three ways in which a limit may fail to exist? 7. 1
0.5
2. T/F: If lim f(x) = 5, then lim f(x) = 5
x→1− x→1
x
.
0.5 1 1.5 2
3. T/F: If lim f(x) = 5, then lim f(x) = 5 (a) lim f(x) (d) f(1)
x→1− x→1+ x→1−
(b) lim f(x) (e) lim f(x)

x→2−
x→1+
4. T/F: If lim f(x) = 5, then lim f(x) = 5 (c) lim f(x) (f) lim f(x)
x→0+
x→1 x→1− x→1
y
2
Problems 1.5
8. 1
In Exercises 5 – 12, evaluate each expression using the given
graph of f(x). 0.5
. x
0.5 1 1.5 2
y
(a) lim f(x) (c) lim f(x)
2 x→1− x→1
(b) lim f(x) (d) f(1)

1.5 x→1+
5. 1
y
0.5 2
x 1.5
.
0.5 1 1.5 2
9. 1
(a) lim f(x) (d) f(1)

x→1− 0.5
(b) lim f(x) (e) lim f(x)

x→0−
x→1+ . x
(c) lim f(x) (f) lim f(x) 0.5 1 1.5 2

x→0+
x→1
x→1− x→1
y (b) lim f(x) (d) f(1)

x→1+
2
1.5 y
4
6. 1
2
0.5
10. x
x −4 −3 −2 −1 1 2 3 4
.
0.5 1 1.5 2 −2
(a) lim f(x) (d) f(1) −4

.
x→1−
(b) lim f(x) (e) lim f(x) (a) lim f(x) (c) lim f(x)
x→2−
x→1+ x→0− x→0
(c) lim f(x) (f) lim f(x) (b) lim f(x) (d) f(0)
x→2+
x→1 x→0+
35
{
y cos x x<π
16. f(x) =
4
sin x x≥π
2 (a) lim f(x) (c) lim f(x)
x→π − x→π
11. x (b) lim f(x) (d) f(π)
−4 −3 −2 −1 1 2 3 4 x→π +
−2
{
1 − cos2 x x<a
17. f(x) = ,
. −4 sin2 x x≥a
(a) lim f(x) (e) lim f(x) where a is a real number.
x→−2− x→2−
(b) lim f(x) (f) lim f(x) (a) lim f(x) (c) lim f(x)
x→2+ x→a− x→a
x→−2+
(g) lim f(x) (b) lim f(x) (d) f(a)
(c) lim f(x) x→2 x→a+
x→−2
(d) f(−2) (h) f(2) 

 x+1 x<1
18. f(x) = 1 x=1

y
x−1 x>1
4
x→1− x→1
2
(b) lim f(x) (d) f(1)
12. x
x→1+
−4 −3 −2 −1 1 2 3 4

−2  x2 x<2
19. f(x) = x+1 x=2

. −4 −x2 + 2x + 4 x>2
Let −3 ≤ a ≤ 3 be an integer. (a) lim f(x) (c) lim f(x)
x→2− x→2
(a) lim f(x) (c) lim f(x) (b) lim f(x) (d) f(2)
x→a− x→a x→2+
(b) lim f(x) (d) f(a) {

x→a+ a(x − b)2 + c x<b
20. f(x) = ,
a(x − b) + c x≥b
where a, b and c are real numbers.
In Exercises 13 – 21, evaluate the given limits of the piecewise
defined func ons f. (a) lim f(x) (c) lim f(x)
x→b− x→b
{ (b) lim f(x) (d) f(b)
x+1 x≤1 x→b+
13. f(x) =
x2 − 5 x>1
{ |x|
(a) lim f(x) (c) lim f(x) x ̸= 0
21. f(x) = x
x→1− x→1 0 x=0
(b) lim f(x) (d) f(1) (a) lim f(x) (c) lim f(x)
x→1+ x→0− x→0
{ (b) lim f(x) (d) f(0)

2x2 + 5x − 1 x<0 x→0+
14. f(x) =
sin x x≥0
(a) lim f(x)

x→0−
(c) lim f(x)
x→0 Review
(b) lim f(x) (d) f(0) x2 + 5x + 4
x→0+ 22. Evaluate the limit: lim .
x→−1 x2 − 3x − 4
 2
 x −1 x < −1
x2 − 16
15. f(x) = x3 + 1 −1 ≤ x ≤ 1 23. Evaluate the limit: lim .
 2 x→−4 x2 − 4x − 32
x +1 x>1
(a) lim f(x) (e) lim f(x) x2 − 15x + 54
x→−1− x→1− 24. Evaluate the limit: lim .
x→−6 x2 − 6x
(b) lim f(x) (f) lim f(x)
x→−1+ x→1+
x2 − 4.4x + 1.6
(c) lim f(x) (g) lim f(x) 25. Approximate the limit numerically: lim .
x→−1 x→1 x→0.4 x2 − 0.4x
(d) f(−1) (h) f(1)
x2 + 5.8x − 1.2
26. Approximate the limit numerically: lim .
x→0.2 x2 − 4.2x + 0.8
36
1.5 Con nuity
1.5 Con nuity

As we have studied limits, we have gained the intui on that limits measure
“where a func on is heading.” That is, if lim f(x) = 3, then as x is close to 1,
x→1
f(x) is close to 3. We have seen, though, that this is not necessarily a good indi-
cator of what f(1) actually is. This can be problema c; func ons can tend to one
value but a ain another. This sec on focuses on func ons that do not exhibit
such behavior.
Defini on 1.5.1 Con nuous Func on

Let f be a func on defined on an open interval I containing c.
1. f is con nuous at c if lim f(x) = f(c).
x→c
2. f is con nuous on I if f is con nuous at c for all values of c in I. If f

is con nuous on (−∞, ∞), we say f is con nuous everywhere.
A useful way to establish whether or not a func on f is con nuous at c is to

verify the following three things:
1. lim f(x) exists,
x→c
2. f(c) is defined, and

3. lim f(x) = f(c).
x→c
Example 1.5.1 Finding intervals of con nuity

Let f be defined as shown in Figure 1.5.1. Give the interval(s) on which f is con- y
nuous. 1.5
S We proceed by examining the three criteria for con nuity. 1

1. The limits lim f(x) exists for all c between 0 and 3.
x→c
0.5
2. f(c) is defined for all c between 0 and 3, except for c = 1. We know
immediately that f cannot be con nuous at x = 1. x
1 2 3
3. The limit lim f(x) = f(c) for all c between 0 and 3, except, of course, for .
x→c
c = 1. Figure 1.5.1: A graph of f in Example 1.5.1.
We conclude that f is con nuous at every point of (0, 3) except at x = 1.
Therefore f is con nuous on (0, 1) and (1, 3).
Our defini on of con nuity (currently) only applies to open intervals. A er
Defini on 1.5.2, we’ll be able to say that f is con nuous on [0, 1) and (1, 3].
Notes:
37
Chapter 1 Limits
Example 1.5.2 Finding intervals of con nuity

The floor func on, f(x) = ⌊x⌋, returns the largest integer smaller than, or equal
to, the input x. (For example, f(π) = ⌊π⌋ = 3.) The graph of f in Figure 1.5.2
demonstrates why this is o en called a “step func on.”
Give the intervals on which f is con nuous.
y S We examine the three criteria for con nuity.

2 1. The limits limx→c f(x) do not exist at the jumps from one “step” to the
next, which occur at all integer values of c. Therefore the limits exist for
all c except when c is an integer.
x 2. The func on is defined for all values of c.
−2 2
3. The limit lim f(x) = f(c) for all values of c where the limit exist, since each
x→c
step consists of just a line.
−2
. We conclude that f is con nuous everywhere except at integer values of c. So
Figure 1.5.2: A graph of the step func on the intervals on which f is con nuous are
in Example 1.5.2.
. . . , (−2, −1), (−1, 0), (0, 1), (1, 2), . . . .
Our defini on of con nuity on an interval specifies the interval is an open

interval. We can extend the defini on of con nuity to closed intervals by con-
sidering the appropriate one-sided limits at the endpoints.
Defini on 1.5.2 Con nuity on Closed Intervals

Let f be defined on the closed interval [a, b] for some real numbers a < b.
f is con nuous on [a, b] if:
1. f is con nuous on (a, b),
2. lim f(x) = f(a) and

x→a+
3. lim f(x) = f(b).

x→b−
We can make the appropriate adjustments to talk about con nuity on half–
open intervals such as [a, b) or (a, b] if necessary.
Using this new defini on, we can adjust our answer in Example 1.5.1 by stat-
ing that f is con nuous on [0, 1) and (1, 3], as men oned in that example. We
Notes:
38
1.5 Con nuity
can also revisit Example 1.5.2 and state that the floor func on is con nuous on
the following half–open intervals
. . . , [−2, −1), [−1, 0), [0, 1), [1, 2), . . . .
This can tempt us to conclude that f is con nuous everywhere; a er all, if f is
con nuous on [0, 1) and [1, 2), isn’t f also con nuous on [0, 2)? Of course, the
answer is no, and the graph of the floor func on immediately confirms this.
Con nuous func ons are important as they behave in a predictable fashion:
func ons a ain the value they approach. Because con nuity is so important,
most of the func ons you have likely seen in the past are con nuous on their
domains. This is demonstrated in the following example where we examine the
intervals of con nuity of a variety of common func ons.
Example 1.5.3 Determining intervals on which a func on is con nuous

For each of the following func ons, give the domain of the func on and the
interval(s) on which it is con nuous.
√
1. f(x) = 1/x 4. f(x) = 1 − x2
2. f(x) = sin x 5. f(x) = |x|
√
3. f(x) = x
S We examine each in turn.

1. The domain of f(x) = 1/x is (−∞, 0) ∪ (0, ∞). As it is a ra onal func-
on, we apply Theorem 1.3.2 to recognize that f is con nuous on all of its
domain.
2. The domain of f(x) = sin x is all real numbers, or (−∞, ∞). Applying
Theorem 1.3.3 shows that sin x is con nuous everywhere.
√
√ of f(x) = x is [0, ∞). Applying Theorem 1.3.3 shows that
3. The domain
f(x) = x is con nuous on its domain of [0, ∞).
√
4. The domain of f(x) = 1 − x2 is [−1, 1]. Applying Theorems 1.3.1 and
1.3.3 shows that f is con nuous on all of its domain, [−1, 1].
5. The domain of f(x){= |x| is (−∞, ∞). We can define the absolute value
−x x < 0
func on as f(x) = . Each “piece” of this piecewise defined
x x≥0
func on is con nuous on all of its domain, giving that f is con nuous on
(−∞, 0) and [0, ∞). We cannot assume this implies that f is con nuous
on (−∞, ∞); we need to check that lim f(x) = f(0), as x = 0 is the point
x→0
where f transi ons from one “piece” of its defini on to the other. It is
easy to verify that this is indeed true, hence we conclude that f(x) = |x|
is con nuous everywhere.
Notes:
39
Chapter 1 Limits
Con nuity is inherently ed to the proper es of limits. Because of this, the

proper es of limits found in Theorems 1.3.1 and 1.3.2 apply to con nuity as well.
Further, now knowing the defini on of con nuity we can re–read Theorem 1.3.3
as giving a list of func ons that are con nuous on their domains. The following
theorem states how con nuous func ons can be combined to form other con-
nuous func ons, followed by a theorem which formally lists func ons that we
know are con nuous on their domains.
Theorem 1.5.1 Proper es of Con nuous Func ons

Let f and g be con nuous func ons on an interval I, let c be a real number
and let n be a posi ve integer. The following func ons are con nuous on
I.
1. Sums/Differences: f ± g
2. Constant Mul ples: c · f
3. Products: f·g
4. Quo ents: f/g (as long as g ̸= 0 on I)
5. Powers: fn
√
6. Roots: n
f (If n is even then require f(x) ≥ 0 on I.)
7. Composi ons: Adjust the defini ons of f and g to: Let f be

con nuous on I, where the range of f on I is J,
and let g be con nuous on J. Then g ◦ f, i.e.,
g(f(x)), is con nuous on I.
Theorem 1.5.2 Con nuous Func ons

Let n be a posi ve integer. The following func ons are con nuous on their domains.
1. f(x) = sin x 4. f(x) = csc x 7. f(x) = ax (a > 0)

2. f(x) = cos x 5. f(x) = sec x 8. f(x) = ln x
√
3. f(x) = tan x 6. f(x) = cot x 9. f(x) = n x
We apply these theorems in the following Example.
Notes:
40
1.5 Con nuity
Example 1.5.4 Determining intervals on which a func on is con nuous

State the interval(s) on which each of the following func ons is con nuous.
√ √
1. f(x) = x − 1 + 5 − x 3. f(x) = tan x
√
2. f(x) = x sin x 4. f(x) = ln x
S We examine each in turn, applying Theorems 1.5.1 and 1.5.2

as appropriate. y
3
1. The square–root terms are con nuous on the intervals [1, ∞) and (−∞, 5],
respec vely. As f is con nuous only where each term is con nuous, f is
con nuous on [1, 5], the intersec on of these two intervals. A graph of f 2
is given in Figure 1.5.3.

1
2. The func ons y = x and y = sin x are each con nuous everywhere, hence
their product is, too.
x
3. Theorem 1.5.2 states that f(x) = tan x is con nuous “on its domain.” Its . 2 4
domain includes all real numbers except odd mul ples of π/2. Thus the
intervals on which f(x) = tan x is con nuous are Figure 1.5.3: A graph of f in Example
( ) ( ( ) 1.5.4(1).
3π π π π) π 3π
... − ,− , − , , , ,...,.
2 2 2 2 2 2
√
4. The domain of y = x is [0, ∞). The range of y = ln x is (−∞, ∞), but if
we restrict its domain to [1, ∞) its range is [0, ∞). So restric ng y =√ln x
to the domain of [1, ∞) restricts its √
output is [0, ∞), on which y = x is
defined. Thus the domain of f(x) = ln x is [1, ∞).
A common way of thinking of a con nuous func on is that “its graph can
be sketched without li ing your pencil.” That is, its graph forms a “con nuous”
curve, without holes, breaks or jumps. While beyond the scope of this text,
this pseudo–defini on glosses over some of the finer points of con nuity. Very
strange func ons are con nuous that one would be hard pressed to actually
sketch by hand.
This intui ve no on of con nuity does help us understand another impor-
tant concept as follows. Suppose f is defined on [1, 2] and f(1) = −10 and
f(2) = 5. If f is con nuous on [1, 2] (i.e., its graph can be sketched as a con nu-
ous curve from (1, −10) to (2, 5)) then we know intui vely that somewhere on
[1, 2] f must be equal to −9, and −8, and −7, −6, . . . , 0, 1/2, etc. In short, f
takes on all intermediate values between −10 and 5. It may take on more val-
ues; f may actually equal 6 at some me, for instance, but we are guaranteed all
values between −10 and 5.
Notes:
41
Chapter 1 Limits
While this no on seems intui ve, it is not trivial to prove and its importance
is profound. Therefore the concept is stated in the form of a theorem.
Theorem 1.5.3 Intermediate Value Theorem

Let f be a con nuous func on on [a, b] and, without loss of generality,
let f(a) < f(b). Then for every value y, where f(a) < y < f(b), there is
at least one value c in (a, b) such that f(c) = y.
One important applica on of the Intermediate Value Theorem is root find-

ing. Given a func on f, we are o en interested in finding values of x where
f(x) = 0. These roots may be very difficult to find exactly. Good approxima-
ons can be found through successive applica ons of this theorem. Suppose
through direct computa on we find that f(a) < 0 and f(b) > 0, where a < b.
The Intermediate Value Theorem states that there is at least one c in (a, b) such
that f(c) = 0. The theorem does not give us any clue as to where to find such a
value in the interval (a, b), just that at least one such value exists.
There is a technique that produces a good approxima on of c. Let d be the
midpoint of the interval [a, b] and consider f(d). There are three possibili es:
1. f(d) = 0: We got lucky and stumbled on the actual value. We stop as we

found a root.
2. f(d) < 0: Then we know there is a root of f on the interval [d, b] – we have
halved the size of our interval, hence are closer to a good approxima on
of the root.
3. f(d) > 0: Then we know there is a root of f on the interval [a, d] – again,we
have halved the size of our interval, hence are closer to a good approxi-
ma on of the root.
Successively applying this technique is called the Bisec on Method of root

finding. We con nue un l the interval is sufficiently small. We demonstrate this
y
in the following example.
0.5 Example 1.5.5 Using the Bisec on Method

Approximate the root of f(x) = x − cos x, accurate to three places a er the
x
decimal.
0.5 1
S Consider the graph of f(x) = x−cos x, shown in Figure 1.5.4.
−0.5 It is clear that the graph crosses the x-axis somewhere near x = 0.8. To start the
Bisec on Method, pick an interval that contains 0.8. We choose [0.7, 0.9]. Note
−1 that all we care about are signs of f(x), not their actual value, so this is all we
.
display.
Figure 1.5.4: Graphing a root of f(x) =
x − cos x.
Notes:
42
1.5 Con nuity
Itera on 1: f(0.7) < 0, f(0.9) > 0, and f(0.8) > 0. So replace 0.9 with 0.8 and
repeat.
Itera on 2: f(0.7) < 0, f(0.8) > 0, and at the midpoint, 0.75, we have f(0.75) >
0. So replace 0.8 with 0.75 and repeat. Note that we don’t need to con-
nue to check the endpoints, just the midpoint. Thus we put the rest of
the itera ons in Figure 1.5.5.
Itera on # Interval Midpoint Sign
No ce that in the 12th itera on we have the endpoints of the interval each 1 [0.7, 0.9] f(0.8) > 0
star ng with 0.739. Thus we have narrowed the zero down to an accuracy of 2 [0.7, 0.8] f(0.75) > 0
the first three places a er the decimal. Using a computer, we have 3 [0.7, 0.75] f(0.725) < 0
4 [0.725, 0.75] f(0.7375) < 0
5 [0.7375, 0.75] f(0.7438) > 0
f(0.7390) = −0.00014, f(0.7391) = 0.000024. 6 [0.7375, 0.7438] f(0.7407) > 0
7 [0.7375, 0.7407] f(0.7391) > 0
Either endpoint of the interval gives a good approxima on of where f is 0. The 8 [0.7375, 0.7391] f(0.7383) < 0
Intermediate Value Theorem states that the actual zero is s ll within this interval. 9 [0.7383, 0.7391] f(0.7387) < 0
10 [0.7387, 0.7391] f(0.7389) < 0
While we do not know its exact value, we know it starts with 0.739. 11 [0.7389, 0.7391] f(0.7390) < 0
This type of exercise is rarely done by hand. Rather, it is simple to program 12 [0.7390, 0.7391]
a computer to run such an algorithm and stop when the endpoints differ by a
preset small amount. One of the authors did write such a program and found Figure 1.5.5: Itera ons of the Bisec on
the zero of f, accurate to 10 places a er the decimal, to be 0.7390851332. While Method of Root Finding
it took a few minutes to write the program, it took less than a thousandth of a
second for the program to run the necessary 35 itera ons. In less than 8 hun-
dredths of a second, the zero was calculated to 100 decimal places (with less
than 200 itera ons).
It is a simple ma er to extend the Bisec on Method to solve problems similar

to “Find x, where f(x) = 0.” For instance, we can find x, where f(x) = 1. It
actually works very well to define a new func on g where g(x) = f(x) − 1. Then
use the Bisec on Method to solve g(x) = 0.
Similarly, given two func ons f and g, we can use the Bisec on Method to
solve f(x) = g(x). Once again, create a new func on h where h(x) = f(x) − g(x)
and solve h(x) = 0.
In Sec on 4.1 another equa on solving method will be introduced, called
Newton’s Method. In many cases, Newton’s Method is much faster. It relies on
more advanced mathema cs, though, so we will wait before introducing it.
This sec on formally defined what it means to be a con nuous func on.
“Most” func ons that we deal with are con nuous, so o en it feels odd to have
to formally define this concept. Regardless, it is important, and forms the basis
of the next chapter.
In the next sec on we examine one more aspect of limits: limits that involve
infinity.
Notes:
43
Exercises 1.5
Terms and Concepts 13. a = 1
y
2
1. In your own words, describe what it means for a func on
to be con nuous. 1.5
2. In your own words, describe what the Intermediate Value 1
Theorem states.
0.5
3. What is a “root” of a func on? x

.
0.5 1 1.5 2
4. Given func ons f and g on an interval I, how can the Bisec-

on Method be used to find a value c where f(c) = g(c)? 14. a = 0
y
5. T/F: If f is defined on an open interval containing c, and 2
lim f(x) exists, then f is con nuous at c.

x→c 1.5
6. T/F: If f is con nuous at c, then lim f(x) exists. 1

x→c
0.5
7. T/F: If f is con nuous at c, then lim f(x) = f(c).
x→c+
. x
0.5 1 1.5 2
8. T/F: If f is con nuous on [a, b], then lim f(x) = f(a).
x→a−
15. a = 1
y
9. T/F: If f is con nuous on [0, 1) and [1, 2), then f is con nu-
2
ous on [0, 2).
1.5
10. T/F: The sum of con nuous func ons is also con nuous.
1
Problems 0.5
. x
In Exercises 11 – 18, a graph of a func on f is given along with 0.5 1 1.5 2
a value a. Determine if f is con nuous at a; if it is not, state

why it is not. 16. a = 4
y
11. a = 1 4
y
2 2
1.5 x
−4 −3 −2 −1 1 2 3 4
1 −2
−4
0.5 .
. x
0.5 1 1.5 2 17. (a) a = −2
(b) a = 0
12. a = 1
y
(c) a = 2
2 y
4
1.5
2
1
x
−4 −3 −2 −1 1 2 3 4
0.5
−2
. x
0.5 1 1.5 2 −4
.
44
18. a = 3π/2 31. g(s) = ln s
y
2
32. h(t) = cos t
1.5 √
33. f(k) = 1 − ek
1
34. f(x) = sin(ex + x2 )

0.5
x Exercises 35 – 38 test your understanding of the Intermediate

π/2 π 3π/2 2π Value Theorem.
In Exercises 19 – 22, determine if f is con nuous at the indi- 35. Let f be con nuous on [1, 5] where f(1) = −2 and f(5) =
cated values. If not, explain why. −10. Does a value 1 < c < 5 exist such that f(c) = −9?
Why/why not?
{
1 x=0
19. f(x) = sin x 36. Let g be con nuous on [−3, 7] where g(0) = 0 and g(2) =
x
x>0
25. Does a value −3 < c < 7 exist such that g(c) = 15?
(a) x = 0 Why/why not?
(b) x = π
37. Let f be con nuous on [−1, 1] where f(−1) = −10 and
{
x3 − x x<1 f(1) = 10. Does a value −1 < c < 1 exist such that
20. f(x) =
x−2 x≥1 f(c) = 11? Why/why not?
(a) x = 0
38. Let h be a func on on [−1, 1] where h(−1) = −10 and
(b) x = 1 h(1) = 10. Does a value −1 < c < 1 exist such that
{ h(c) = 0? Why/why not?
x2 +5x+4
x2 +3x+2
x ̸= −1
21. f(x) =
3 x = −1 In Exercises 39 – 42, use the Bisec on Method to approxi-
mate, accurate to two decimal places, the value of the root
(a) x = −1 of the given func on in the given interval.
(b) x = 10
{ 39. f(x) = x2 + 2x − 4 on [1, 1.5].
x2 −64
x2 −11x+24
x ̸= 8
22. f(x) = 40. f(x) = sin x − 1/2 on [0.5, 0.55]
5 x=8
(a) x = 0 41. f(x) = ex − 2 on [0.65, 0.7].
(b) x = 8
42. f(x) = cos x − sin x on [0.7, 0.8].
In Exercises 23 – 34, give the intervals on which the given
func on is con nuous.
Review
23. f(x) = x2 − 3x + 9 {
x2 − 5 x<5
√ 43. Let f(x) = .
24. g(x) = x2 − 4 5x x≥5
√ (a) lim f(x) (c) lim f(x)

x→5− x→5
25. g(x) = 4 − x2
(b) lim f(x) (d) f(5)
√ √ x→5+
26. h(k) = 1−k+ k+1
√ 44. Numerically approximate the following limits:
27. f(t) = 5t2 − 30
x2 − 8.2x − 7.2
1 (a) lim
28. g(t) = √ x→−4/5+ x2 + 5.8x + 4
1 − t2
x2 − 8.2x − 7.2
(b) lim
1 x→−4/5− x2 + 5.8x + 4
29. g(x) =
1 + x2
45. Give an example of func on f(x) for which lim f(x) does not
x→0
30. f(x) = ex exist.
45
Chapter 1 Limits
1.6 Limits Involving Infinity

In Defini on 1.2.1 we stated that in the equa on limx→c f(x) = L, both c and
y
L were numbers. In this sec on we relax that defini on a bit by considering
situa ons when it makes sense to let c and/or L be “infinity.”
100 As a mo va ng example, consider f(x) = 1/x2 , as shown in Figure 1.6.1.
Note how, as x approaches 0, f(x) grows very, very large – in fact, it grows without
bound. It seems appropriate, and descrip ve, to state that
50 1
lim = ∞.
x→0 x2
Also note that as x gets very large, f(x) gets very, very small. We could represent
. this concept with nota on such as
x
−1 −0.5 0.5 1
1
lim = 0.
2
Figure 1.6.1: Graphing f(x) = 1/x for x→∞ x2
values of x near 0. We explore both types of use of ∞ in turn.
Defini on 1.6.1 Limit of Infinity, ∞

Let I be an open interval containing c, and let f be a func on defined on
I, except possibly at c.
• The limit of f(x), as x approaches c, is infinity, denoted by
lim f(x) = ∞,
x→c
means that given any N > 0, there exists δ > 0 such that for all x
in I, where x ̸= c, if |x − c| < δ, then f(x) > N.
• The limit of f(x), as x approaches c, is nega ve infinity, denoted

by
lim f(x) = −∞,
x→c
means that given any N < 0, there exists δ > 0 such that for all x
in I, where x ̸= c, if |x − c| < δ, then f(x) < N.
The first defini on is similar to the ε–δ defini on from Sec on 1.2. In that
defini on, given any (small) value ε, if we let x get close enough to c (within δ
units of c) then f(x) is guaranteed to be within ε of L. Here, given any (large)
value N, if we let x get close enough to c (within δ units of c), then f(x) will be
Notes:
46
at least as large as N. In other words, if we get close enough to c, then we can

make f(x) as large as we want. We define limits equal to −∞ in a similar way.
It is important to note that by saying limx→c f(x) = ∞ we are implicitly stat-
ing that the limit of f(x), as x approaches c, does not exist. A limit only exists
when f(x) approaches an actual numeric value. We use the concept of limits
that approach infinity because it is helpful and descrip ve.
We define one-sided limits that approach infinity in a similar way.
Defini on 1.6.2 One-Sided Limits of Infinity
• Let f be a func on defined on (a, c) for some a < c.

The limit of f(x), as x approaches c from the le , is infinity, or, the
le -hand limit of f at c is infinity, denoted by
lim f(x) = ∞,
x→c−
means given any N > 0, there exists δ > 0 such that for all
a < x < c, if |x − c| < δ, then f(x) > N.
• Let f be a func on defined on (c, b) for some b > c.
The limit of f(x), as x approaches c from the right, is infinity, or,
the right-hand limit of f at c is infinity, denoted by
lim f(x) = ∞,
x→c+
means given any N > 0, there exists δ > 0 such that for all
c < x < b, if |x − c| < δ, then f(x) > N.
• The term le - (or, right-) hand limit of f at c is nega ve infinity is
defined in a manner similar to Defini on 1.6.1.
Example 1.6.1 Evalua ng limits involving infinity y

1
Find lim as shown in Figure 1.6.2. 100
x→1 (x − 1)2
S In Example 1.1.4 of Sec on 1.1, by inspec ng values of x

close to 1 we concluded that this limit does not exist. That is, it cannot equal any
50
real number. But the limit could be infinite. And in fact, we see that the func on
does appear to be growing larger and larger, as f(.99) = 104 , f(.999) = 106 ,
f(.9999) = 108 . A similar thing happens on√the other side of 1. In general,
let a “large” value N be given. Let δ = 1/ N. If x is within δ of 1, i.e., if . x
0.5 1 1.5 2
Figure 1.6.2: Observing infinite limit as

x → 1 in Example 1.6.1.
Notes:
47
Chapter 1 Limits
√
|x − 1| < 1/ N, then:
1
|x − 1| < √
N
1
(x − 1) <
2
N
1
> N,
(x − 1)2
which is what we wanted to show. So we may say lim 1/(x − 1)2 = ∞.

x→1
y
Example 1.6.2 Evalua ng limits involving infinity
50 1
Find lim , as shown in Figure 1.6.3.
x→0 x
S It is easy to see that the func on grows without bound near

x 0, but it does so in different ways on different sides of 0. Since its behavior is not
−1 −0.5 0.5 1 1
consistent, we cannot say that lim = ∞. However, we can make a statement
x→0 x
1 1
about one–sided limits. We can state that lim = ∞ and lim = −∞.
x→0+ x x→0− x
. −50
1
Figure 1.6.3: Evalua ng lim . The graphs in the two previous examples demonstrate that if a func on f has
x→0 x
a limit (or, le - or right-hand limit) of infinity at x = c, then the graph of f looks
similar to a ver cal line near x = c. This observa on leads to a defini on.
Defini on 1.6.3 Ver cal Asymptote

Let I be an interval that either contains c or has c as an endpoint, and let
f be a func on defined on I, except possibly at c.
If the limit of f(x) as x approaches c from either the le or right (or both)
is ∞ or −∞, then the line x = c is a ver cal asymptote of f.
Example 1.6.3 Finding ver cal asymptotes

3x
Find the ver cal asymptotes of f(x) = 2 .
x −4
S Ver cal asymptotes occur where the func on grows with-

out bound; this can occur at values of c where the denominator is 0. When x is
near c, the denominator is small, which in turn can make the func on take on
large values. In the case of the given func on, the denominator is 0 at x = ±2.
Subs tu ng in values of x close to 2 and −2 seems to indicate that the func on
Notes:
48
y
tends toward ∞ or −∞ at those points. We can graphically confirm this by look-
ing at Figure 1.6.4. Thus the ver cal asymptotes are at x = ±2.
10
When a ra onal func on has a ver cal asymptote at x = c, we can conclude

that the denominator is 0 at x = c. However, just because the denominator x
is 0 at a certain point does not mean there is a ver cal asymptote there. For −5 5
instance, f(x) = (x2 − 1)/(x − 1) does not have a ver cal asymptote at x = 1,
as shown in Figure 1.6.5. While the denominator does get small near x = 1, −10
the numerator gets small too, matching the denominator step for step. In fact,
.
factoring the numerator, we get
3x
(x − 1)(x + 1) Figure 1.6.4: Graphing f(x) =
x2 − 4
.
f(x) = .
x−1
Canceling the common term, we get that f(x) = x + 1 for x ̸= 1. So there is
clearly no asymptote; rather, a hole exists in the graph at x = 1.
The above example may seem a li le contrived. Another example demon- y
stra ng this important concept is f(x) = (sin x)/x. We have considered this 3
func on several mes in the previous sec ons. We found that limx→0 sinx x = 1;
i.e., there is no ver cal asymptote. No simple algebraic cancella on makes this
2
fact obvious; we used the Squeeze Theorem in Sec on 1.3 to prove this.
If the denominator is 0 at a certain point but the numerator is not, then 1

there will usually be a ver cal asymptote at that point. On the other hand, if the
numerator and denominator are both zero at that point, then there may or may
x
not be a ver cal asymptote at that point. This case where the numerator and .−1 1 2
denominator are both zero returns us to an important topic.
Figure 1.6.5: Graphically showing that
Indeterminate Forms f(x) =
x2 − 1
does not have an asymp-
x−1
tote at x = 1.
We have seen how the limits
sin x x2 − 1
lim and lim
x→0 x x→1 x − 1
each return the indeterminate form “0/0” when we blindly plug in x = 0 and
x = 1, respec vely. However, 0/0 is not a valid arithme cal expression. It gives
no indica on that the respec ve limits are 1 and 2.
With a li le cleverness, one can come up with 0/0 expressions which have
a limit of ∞, 0, or any other real number. That is why this expression is called
indeterminate.
A key concept to understand is that such limits do not really return 0/0.
Rather, keep in mind that we are taking limits. What is really happening is that
Notes:
49
Chapter 1 Limits
the numerator is shrinking to 0 while the denominator is also shrinking to 0.

The respec ve rates at which they do this are very important and determine the
actual value of the limit.
An indeterminate form indicates that one needs to do more work in order to
compute the limit. That work may be algebraic (such as factoring and canceling)
or it may require a tool such as the Squeeze Theorem. In a later sec on we will
learn a technique called l’Hôspital’s Rule that provides another way to handle
indeterminate forms.
Some other common indeterminate forms are ∞ − ∞, ∞ · 0, ∞/∞, 00 , ∞0
and 1∞ . Again, keep in mind that these are the “blind” results of evalua ng a
limit, and each, in and of itself, has no meaning. The expression ∞ − ∞ does
not really mean “subtract infinity from infinity.” Rather, it means “One quan ty
is subtracted from the other, but both are growing without bound.” What is the
result? It is possible to get every value between −∞ and ∞.
Note that 1/0 and ∞/0 are not indeterminate forms, though they are not
exactly valid mathema cal expressions, either. In each, the func on is growing
without bound, indica ng that the limit will be ∞, −∞, or simply not exist if the
le - and right-hand limits do not match.
Limits at Infinity and Horizontal Asymptotes
At the beginning of this sec on we briefly considered what happens to f(x) =

1/x2 as x grew very large. Graphically, it concerns the behavior of the func on to
the “far right” of the graph. We make this no on more explicit in the following
defini on.
Defini on 1.6.4 Limits at Infinity and Horizontal Asymptotes

Let L be a real number.
1. Let f be a func on defined on (a, ∞) for some number a. The

limit of f at infinity is L, or lim f(x) = L, means for every ε > 0
x→∞
there exists M > a such that if x > M, then |f(x) − L| < ε.
2. Let f be a func on defined on (−∞, b) for some number b. The

limit of f at nega ve infinity is L, or lim f(x) = L, means
x→−∞
for every ε > 0 there exists M < b such that if x < M, then
|f(x) − L| < ε.
3. If lim f(x) = L or lim f(x) = L, we say the line y = L is a

x→∞ x→−∞
horizontal asymptote of f.
Notes:
50
We can also define limits such as lim f(x) = ∞ by combining this defini on 1
x→∞
with Defini on 1.6.1.
Example 1.6.4 Approxima ng horizontal asymptotes 0.5

x2
Approximate the horizontal asymptote(s) of f(x) = 2 .
x +4
x
S We will approximate the horizontal asymptotes by approxi- −20 −10 10 20
ma ng the limits .
(a)
x2 x2
lim 2 and lim 2 . x f(x)
x→−∞ x + 4 x→∞ x + 4
10 0.9615
Figure 1.6.6(a) shows a sketch of f, and part (b) gives values of f(x) for large mag- 100 0.9996
nitude values of x. It seems reasonable to conclude from both of these sources 10000 0.999996
that f has a horizontal asymptote at y = 1. −10 0.9615
−100 0.9996
Later, we will show how to determine this analy cally.
−10000 0.999996
(b)
Horizontal asymptotes can take on a variety of forms. Figure 1.6.7(a) shows
that f(x) = x/(x2 + 1) has a horizontal asymptote of y = 0, where 0 is ap- Figure 1.6.6: Using a graph and a table
proached from both above and below. √ to approximate a horizontal asymptote in
Figure 1.6.7(b) shows that f(x) = x/ x2 + 1 has two horizontal asymptotes; Example 1.6.4.
one at y = 1 and the other at y = −1.
Figure 1.6.7(c) shows that f(x) = (sin x)/x has even more interes ng behav-
ior than at just x = 0; as x approaches ±∞, f(x) approaches 0, but oscillates as
it does this.
y y y
1 1
1
0.5 0.5
0.5
x x
−20 −10 10 20 −20 −10 10 20
−0.5 −0.5
x
−20 −10 10 20
. −1 . −1
.
(a) (b) (c)
Figure 1.6.7: Considering different types of horizontal asymptotes.
Notes:
51
Chapter 1 Limits
We can analy cally evaluate limits at infinity for ra onal func ons once we
understand limx→∞ 1/x. As x gets larger and larger, 1/x gets smaller and smaller,
approaching 0. We can, in fact, make 1/x as small as we want by choosing a large
enough value of x. Given ε, we can make 1/x < ε by choosing x > 1/ε. Thus
we have limx→∞ 1/x = 0.
It is now not much of a jump to conclude the following:
1 1
lim = 0 and lim =0
x→∞ xn x→−∞ xn
Now suppose we need to compute the following limit:
x3 + 2x + 1
lim .
x→∞ 4x3 − 2x2 + 9
A good way of approaching this is to divide through the numerator and denom-
inator by x3 (hence mul plying by 1), which is the largest power of x to appear
in the func on. Doing this, we get
x3 + 2x + 1 1/x3 x3 + 2x + 1
lim = lim ·
x→∞ 4x3 − 2x2 + 9 x→∞ 1/x3 4x3 − 2x2 + 9
x3 /x3 + 2x/x3 + 1/x3

= lim
x→∞ 4x3 /x3 − 2x2 /x3 + 9/x3
1 + 2/x2 + 1/x3
= lim .
x→∞ 4 − 2/x + 9/x3
Then using the rules for limits (which also hold for limits at infinity), as well as
the fact about limits of 1/xn , we see that the limit becomes
1+0+0 1
= .
4−0+0 4
This procedure works for any ra onal func on. In fact, it gives us the follow-
ing theorem.
Notes:
52
Theorem 1.6.1 Limits of Ra onal Func ons at Infinity

Let f(x) be a ra onal func on of the following form:
an xn + an−1 xn−1 + · · · + a1 x + a0
f(x) = ,
bm xm + bm−1 xm−1 + · · · + b1 x + b0
where any of the coefficients may be 0 except for an and bm .

an
1. If n = m, then lim f(x) = lim f(x) = .
x→∞ x→−∞ bm
2. If n < m, then lim f(x) = lim f(x) = 0.
x→∞ x→−∞
3. If n > m, then lim f(x) and lim f(x) are both infinite.
x→∞ x→−∞
We can see why this is true. If the highest power of x is the same in both
the numerator and denominator (i.e. n = m), we will be in a situa on like the
example above, where we will divide by xn and in the limit all the terms will
approach 0 except for an xn /xn and bm xm /xn . Since n = m, this will leave us with
the limit an /bm . If n < m, then a er dividing through by xm , all the terms in the
numerator will approach 0 in the limit, leaving us with 0/bm or 0. If n > m, and
we try dividing through by xn , we end up with all the terms in the denominator
tending toward 0, while the xn term in the numerator does not approach 0. This
is indica ve of some sort of infinite limit.
Intui vely, as x gets very large, all the terms in the numerator are small in
comparison to an xn , and likewise all the terms in the denominator are small
compared to bn xm . If n = m, looking only at these two important terms, we
have (an xn )/(bn xm ). This reduces to an /bm . If n < m, the func on behaves
like an /(bm xm−n ), which tends toward 0. If n > m, the func on behaves like
an xn−m /bm , which will tend to either ∞ or −∞ depending on the values of n,
m, an , bm and whether you are looking for limx→∞ f(x) or limx→−∞ f(x).
With care, we can quickly evaluate limits at infinity for a large number of
func ons by considering the largest powers of x. For instance, consider again
x
lim √ , graphed in Figure 1.6.7(b). When x is very large, x2 + 1 ≈ x2 .
x→±∞ x2 + 1
Thus √ √ x x
x2 + 1 ≈ x2 = |x|, and √ ≈ .
x2 + 1 |x|
This expression is 1 when x is posi ve and −1 when x is nega ve. Hence we get
asymptotes of y = 1 and y = −1, respec vely.
Notes:
53
Chapter 1 Limits
Example 1.6.5 Finding a limit of a ra onal func on

x2
Confirm analy cally that y = 1 is the horizontal asymptote of f(x) = , as
x2 +4
approximated in Example 1.6.4.
S Before using Theorem 1.6.1, let’s use the technique of eval-

ua ng limits at infinity of ra onal func ons that led to that theorem. The largest
y power of x in f is 2, so divide the numerator and denominator of f by x2 , then
0.5 take limits.
x2 x2 /x2
lim = lim
x→∞ x2 + 4 x→∞ x2 /x2 + 4/x2
x 1
−40 −30 −20 −10 = lim
x→∞ 1 + 4/x2
1
=
−0.5
1+0
. = 1.
(a) We can also use Theorem 1.6.1 directly; in this case n = m so the limit is the
y
ra o of the leading coefficients of the numerator and denominator, i.e., 1/1 = 1.
0.5
Example 1.6.6 Finding limits of ra onal func ons
Use Theorem 1.6.1 to evaluate each of the following limits.
x x2 + 2x − 1 x2 − 1
1. lim 3. lim
10 20 30 40 x→−∞ x3 + 1 x→∞ 3 − x
x2 + 2x − 1
2. lim
x→∞ 1 − x − 3x2
−0.5 .
S
(b)
1. The highest power of x is in the denominator. Therefore, the limit is 0; see
. y
Figure 1.6.8(a).
x
20 40
2. The highest power of x is x2 , which occurs in both the numerator and de-
nominator. The limit is therefore the ra o of the coefficients of x2 , which
−20 is −1/3. See Figure 1.6.8(b).
3. The highest power of x is in the numerator so the limit will be ∞ or −∞.

To see which, consider only the dominant terms from the numerator and
−40
denominator, which are x2 and −x. The expression in the limit will behave
. like x2 /(−x) = −x for large values of x. Therefore, the limit is −∞. See
Figure 1.6.8(c).
(c)
Figure 1.6.8: Visualizing the func ons in

Example 1.6.6. Notes:
54
Chapter Summary
In this chapter we:
• defined the limit,
• found accessible ways to approximate the value of limits numerically and

graphically,
• developed a not–so–easy method of proving the value of a limit (ε-δ proofs),
• explored when limits do not exist,
• defined con nuity and explored proper es of con nuous func ons, and
• considered limits that involved infinity.
Why? Mathema cs is famous for building on itself and calculus proves to
be no excep on. In the next chapter we will be interested in “dividing by 0.”
That is, we will want to divide a quan ty by smaller and smaller numbers and
see what value the quo ent approaches. In other words, we will want to find a
limit. These limits will enable us to, among other things, determine exactly how
fast something is moving when we are only given posi on informa on.
Later, we will want to add up an infinite list of numbers. We will do so by
first adding up a finite list of numbers, then take a limit as the number of things
we are adding approaches infinity. Surprisingly, this sum o en is finite; that is,
we can add up an infinite list of numbers and get, for instance, 42.
These are just two quick examples of why we are interested in limits. Many
students dislike this topic when they are first introduced to it, but over me an
apprecia on is o en formed based on the scope of its applicability.
Notes:
55
Exercises 1.6
10. f(x) = .
(x − 3)(x − 5)2
1. T/F: If lim f(x) = ∞, then we are implicitly sta ng that the (a) lim f(x) (d) lim f(x)
x→5 x→3− x→5−
limit exists. (b) lim f(x) (e) lim f(x)
x→3+ x→5+
(c) lim f(x) (f) lim f(x)

2. T/F: If lim f(x) = 5, then we are implicitly sta ng that the x→3 x→5
x→∞
limit exists. y
50
3. T/F: If lim f(x) = −∞, then lim f(x) = ∞

x→1− x→1+
x
2 4 6
4. T/F: If lim f(x) = ∞, then f has a ver cal asymptote at
x→5
x = 5.
−50 .
5. T/F: ∞/0 is not an indeterminate form.
1
11. f(x) =
6. List 5 indeterminate forms. ex + 1

x→−∞ x→0−
7. Construct a func on with a ver cal asymptote at x = 5 and
a horizontal asymptote at y = 5. (b) lim f(x) (d) lim f(x)
x→∞ x→0+
y
8. Let lim f(x) = ∞. Explain how we know that f is/is not 1
x→7
con nuous at x = 7.
0.5
.
x
−10 −5 5 10
Problems −0.5
. −1
In Exercises 9 – 14, evaluate the given limits using the graph
of the func on.
12. f(x) = x2 sin(πx)
1
9. f(x) =
(x + 1)2
(a) lim f(x)

x→−∞
(a) lim f(x)
x→−1−
(b) lim f(x)

x→∞
(b) lim f(x)
x→−1+
y
y
100
100
50
50
x
−10 −5 5 10
−50
. x
−2 −1 .. −100
56
13. f(x) = cos(x) In Exercises 19 – 24, iden fy the horizontal and ver cal
asymptotes, if any, of the given func on.
(a) lim f(x)
x→−∞ 2x2 − 2x − 4
19. f(x) =
(b) lim f(x) x2 + x − 20
x→∞
−3x2 − 9x − 6
y 20. f(x) =
5x2 − 10x − 15
1
0.5 x2 + x − 12
21. f(x) =
7x3 − 14x2 − 21x
x
−4π −3π −2π −π π 2π 3π 4π
x2 − 9
−0.5 22. f(x) =
9x − 9
. −1
x2 − 9
23. f(x) =
9x + 27
14. f(x) = 2x + 10
x2 − 1
(a) lim f(x) 24. f(x) =
x→−∞ −x2 − 1
(b) lim f(x) In Exercises 25 – 28, evaluate the given limit.

x→∞
y x3 + 2x2 + 1
25. lim
150 x→∞ x−5
100
x3 + 2x2 + 1
26. lim
x→∞ 5−x
50
x3 + 2x2 + 1
27. lim
x→−∞ x2 − 5
. x
−10 −5 5
x3 + 2x2 + 1
28. lim
x→−∞ 5 − x2
In Exercises 15 – 18, numerically approximate the following
limits:
(a) lim f(x)

Review
x→3−
29. Use an ε − δ proof to show that
(b) lim f(x)
x→3+ lim 5x − 2 = 3.
x→1
(c) lim f(x)
x→3
30. Let lim f(x) = 3 and lim g(x) = −1. Evaluate the following
x→2 x→2
limits.
x2 − 1
15. f(x) = 2
x −x−6 (a) lim (f + g)(x) (c) lim (f/g)(x)
x→2 x→2
x + 5x − 36
2
(b) lim (fg)(x) (d) lim f(x)g(x)
16. f(x) = x→2 x→2
x3 − 5x2 + 3x + 9
{
x2 − 1 x<3
x2 − 11x + 30 31. Let f(x) = .
17. f(x) = 3 x+5 x≥3
x − 4x2 − 3x + 18
Is f con nuous everywhere?
x2 − 9x + 18
18. f(x) = 32. Evaluate the limit: lim ln x.
x2 − x − 6 x→e
57
2: D
The previous chapter introduced the most fundamental of calculus topics: the
limit. This chapter introduces the second most fundamental of calculus topics:
the deriva ve. Limits describe where a func on is going; deriva ves describe
how fast the func on is going.
2.1 Instantaneous Rates of Change: The Deriva ve

A common amusement park ride li s riders to a height then allows them to
freefall a certain distance before safely stopping them. Suppose such a ride
drops riders from a height of 150 feet. Students of physics may recall that the
height (in feet) of the riders, t seconds a er freefall (and ignoring air resistance,
etc.) can be accurately modeled by f(t) = −16t2 + 150.
Using this formula, it is easy√ to verify that, without interven on, the riders
will hit the ground at t = 2.5 1.5 ≈ 3.06 seconds. Suppose the designers of
the ride decide to begin slowing the riders’ fall a er 2 seconds (corresponding
to a height of 86 .). How fast will the riders be traveling at that me?
We have been given a posi on func on, but what we want to compute is a
velocity at a specific point in me, i.e., we want an instantaneous velocity. We
do not currently know how to calculate this.
However, we do know from common experience how to calculate an average
velocity. (If we travel 60 miles in 2 hours, we know we had an average velocity
of 30 mph.) We looked at this concept in Sec on 1.1 when we introduced the
difference quo ent. We have
change in distance “ rise ”

= = average velocity.
change in me run
We can approximate the instantaneous velocity at t = 2 by considering the

average velocity over some me period containing t = 2. If we make the me
interval small, we will get a good approxima on. (This fact is commonly used.
For instance, high speed cameras are used to track fast moving objects. Dis-
tances are measured over a fixed number of frames to generate an accurate
approxima on of the velocity.)
Consider the interval from t = 2 to t = 3 (just before the riders hit the
ground). On that interval, the average velocity is
f(3) − f(2) f(3) − f(2)

= = −80 /s,
3−2 1
Chapter 2 Deriva ves
where the minus sign indicates that the riders are moving down. By narrowing
the interval we consider, we will likely get a be er approxima on of the instan-
taneous velocity. On [2, 2.5] we have
f(2.5) − f(2) f(2.5) − f(2)

= = −72 /s.
2.5 − 2 0.5
We can do this for smaller and smaller intervals of me. For instance, over
a me span of 1/10th of a second, i.e., on [2, 2.1], we have
f(2.1) − f(2) f(2.1) − f(2)

= = −65.6 /s.
2.1 − 2 0.1
Over a me span of 1/100th of a second, on [2, 2.01], the average velocity is
f(2.01) − f(2) f(2.01) − f(2)

= = −64.16 /s.
2.01 − 2 0.01
What we are really compu ng is the average velocity on the interval [2, 2+h]
for small values of h. That is, we are compu ng
f(2 + h) − f(2)
h
Average Velocity where h is small.
h /s We really want to use h = 0, but this, of course, returns the familiar “0/0”
1 −80 indeterminate form. So we employ a limit, as we did in Sec on 1.1.
0.5 −72 We can approximate the value of this limit numerically with small values of
0.1 −65.6 h as seen in Figure 2.1.1. It looks as though the velocity is approaching −64 /s.
0.01 −64.16 Compu ng the limit directly gives
0.001 −64.016
f(2 + h) − f(2) −16(2 + h)2 + 150 − (−16(2)2 + 150)
Figure 2.1.1: Approxima ng the instan- lim = lim
h→0 h h→0 h
taneous velocity with average veloci es
over a small me period h. −64h − 16h 2
= lim
h→0 h
= lim (−64 − 16h)
h→0
= −64.
Graphically, we can view the average veloci es we computed numerically as

the slopes of secant lines on the graph of f going through the points (2, f(2))
and (2 + h, f(2 + h)). In Figure 2.1.2, the secant line corresponding to h = 1 is
shown in three contexts. Figure 2.1.2(a) shows a “zoomed out” version of f with
its secant line. In (b), we zoom in around the points of intersec on between
f and the secant line. No ce how well this secant line approximates f between
Notes:
60
those two points – it is a common prac ce to approximate func ons with straight
lines.
As h → 0, these secant lines approach the tangent line, a line that goes
through the point (2, f(2)) with the special slope of −64. In parts (c) and (d) of
Figure 2.1.2, we zoom in around the point (2, 86). In (c) we see the secant line,
which approximates f well, but not as well the tangent line shown in (d).
y y
150
100
100
50
50
x
1 2 3
x
.
−50 . 2 2.5 3
(a) (b)
.
y y
100 100
50 50
x x
. 1.5 2 2.5 . 1.5 2 2.5
(c) (d)
Figure 2.1.2: Parts (a), (b) and (c) show the secant line to f(x) with h = 1, zoomed in
different amounts. Part (d) shows the tangent line to f at x = 2.
We have just introduced a number of important concepts that we will flesh

out more within this sec on. First, we formally define two of them.
Notes:
61
Defini on 2.1.1 Deriva ve at a Point

Let f be a con nuous func on on an open interval I and let c be in I. The
deriva ve of f at c, denoted f ′ (c), is
f(c + h) − f(c)
lim ,
h→0 h
provided the limit exists. If the limit exists, we say that f is differen able
at c; if the limit does not exist, then f is not differen able at c. If f is
differen able at every point in I, then f is differen able on I.
Defini on 2.1.2 Tangent Line

Let f be con nuous on an open interval I and differen able at c, for some
c in I. The line with equa on ℓ(x) = f ′ (c)(x − c) + f(c) is the tangent line
to the graph of f at c; that is, it is the line through (c, f(c)) whose slope
is the deriva ve of f at c.
Some examples will help us understand these defini ons.
Example 2.1.1 Finding deriva ves and tangent lines

Let f(x) = 3x2 + 5x − 7. Find:
1. f ′ (1) 3. f ′ (3)
2. The equa on of the tangent line 4. The equa on of the tangent line
to the graph of f at x = 1. to the graph f at x = 3.
1. We compute this directly using Defini on 2.1.1.
f(1 + h) − f(1)
f ′ (1) = lim
h→0 h
3(1 + h)2 + 5(1 + h) − 7 − (3(1)2 + 5(1) − 7)
= lim
h→0 h
3h2 + 11h
= lim
h→0 h
= lim (3h + 11) = 11.
h→0
Notes:
62
2. The tangent line at x = 1 has slope f ′ (1) and goes through the point
(1, f(1)) = (1, 1). Thus the tangent line has equa on, in point-slope form,
y = 11(x − 1) + 1. In slope-intercept form we have y = 11x − 10.
3. Again, using the defini on,
f(3 + h) − f(3)
f ′ (3) = lim
h→0 h
y
3(3 + h)2 + 5(3 + h) − 7 − (3(3)2 + 5(3) − 7)
= lim 60
h→0 h
3h2 + 23h
= lim 40
h→0 h
= lim (3h + 23)
h→0
20
= 23.
x
4. The tangent line at x = 3 has slope 23 and goes through the point (3, f(3)) = 1 2 3 4
.
(3, 35). Thus the tangent line has equa on y = 23(x−3)+35 = 23x−34.
Figure 2.1.3: A graph of f(x) = 3x2 + 5x −
A graph of f is given in Figure 2.1.3 along with the tangent lines at x = 1 and 7 and its tangent lines at x = 1 and x = 3.
x = 3.
Another important line that can be created using informa on from the deriva-
ve is the normal line. It is perpendicular to the tangent line, hence its slope is
the opposite–reciprocal of the tangent line’s slope.
Defini on 2.1.3 Normal Line

Let f be con nuous on an open interval I and differen able at c, for some
c in I. The normal line to the graph of f at c is the line with equa on
−1
n(x) = (x − c) + f(c),
f ′ (c)
where f ′((c) ̸= )0. When f ′ (c) = 0, the normal line is the ver cal line
through c, f(c) ; that is, x = c.
Example 2.1.2 Finding equa ons of normal lines

Let f(x) = 3x2 + 5x − 7, as in Example 2.1.1. Find the equa ons of the normal
lines to the graph of f at x = 1 and x = 3.
S In Example 2.1.1, we found that f ′ (1) = 11. Hence at x = 1,
Notes:
63
y the normal line will have slope −1/11. An equa on for the normal line is
−1
3 n(x) = (x − 1) + 1.
11
The normal line is plo ed with y = f(x) in Figure 2.1.4. Note how the line looks
2 perpendicular to f. (A key word here is “looks.” Mathema cally, we say that the
normal line is perpendicular to f at x = 1 as the slope of the normal line is the
1 opposite–reciprocal of the slope of the tangent line. However, normal lines may
not always look perpendicular. The aspect ra o of the picture of the graph plays
a big role in this.)
. x
We also found that f ′ (3) = 23, so the normal line to the graph of f at x = 3
1 2 3 4
will have slope −1/23. An equa on for the normal line is
Figure 2.1.4: A graph of f(x) = 3x2 + 5x − −1
7, along with its normal line at x = 1. n(x) = (x − 3) + 35.
23
Linear func ons are easy to work with; many func ons that arise in the
course of solving real problems are not easy to work with. A common prac ce
in mathema cal problem solving is to approximate difficult func ons with not–
so–difficult func ons. Lines are a common choice. It turns out that at any given
point on the graph of a differen able func on f, the best linear approxima on
to f is its tangent line. That is one reason we’ll spend considerable me finding
tangent lines to func ons.
One type of func on that does not benefit from a tangent–line approxima-
on is a line; it is rather simple to recognize that the tangent line to a line is the
line itself. We look at this in the following example.
. Example 2.1.3 Finding the deriva ve of a linear func on

Consider f(x) = 3x + 5. Find the equa on of the tangent line to f at x = 1 and
x = 7.
S We find the slope of the tangent line by using Defini on

2.1.1.
f(1 + h) − f(1)
f ′ (1) = lim
h→0 h
3(1 + h) + 5 − (3 + 5)
= lim
h→0 h
3h
= lim
h→0 h
= lim 3
h→0
= 3.
Notes:
64
We just found that f ′ (1) = 3. That is, we found the instantaneous rate of
change of f(x) = 3x + 5 is 3. This is not surprising; lines are characterized by
being the only func ons with a constant rate of change. That rate of change
is called the slope of the line. Since their rates of change are constant, their
instantaneous rates of change are always the same; they are all the slope.
So given a line f(x) = ax + b, the deriva ve at any point x will be a; that is,
f ′ (x) = a.
It is now easy to see that the tangent line to the graph of f at x = 1 is just f,
with the same being true for x = 7.
We o en desire to find the tangent line to the graph of a func on without

knowing the actual deriva ve of the func on. In these cases, the best we may
be able to do is approximate the tangent line. We demonstrate this in the next
example.
Example 2.1.4 Numerical approxima on of the tangent line

Approximate the equa on of the tangent line to the graph of f(x) = sin x at
x = 0. y
1
S In order to find the equa on of the tangent line, we need a
slope and a point. The point is given to us: (0, sin 0) = (0, 0). To compute the 0.5
slope, we need the deriva ve. This is where we will make an approxima on.
Recall that x
sin(0 + h) − sin 0 −π − π2 π π
f ′ (0) ≈ 2
h −0.5
for a small value of h. We choose (somewhat arbitrarily) to let h = 0.1. Thus
. −1
sin(0.1) − sin 0
f ′ (0) ≈ ≈ 0.9983.
0.1 Figure 2.1.5: f(x) = sin x graphed with an
approxima on to its tangent line at x = 0.
Thus our approxima on of the equa on of the tangent line is y = 0.9983(x −
0) + 0 = 0.9983x; it is graphed in Figure 2.1.5. The graph seems to imply the
approxima on is rather good.
Recall from Sec on 1.3 that limx→0 sinx x = 1, meaning for values of x near
0, sin x ≈ x. Since the slope of the line y = x is 1 at x = 0, it should seem
reasonable that “the slope of f(x) = sin x” is near 1 at x = 0. In fact, since we
approximated the value of the slope to be 0.9983, we might guess the actual
value is 1. We’ll come back to this later.
Consider again Example 2.1.1. To find the deriva ve of f at x = 1, we needed

to evaluate a limit. To find the deriva ve of f at x = 3, we needed to again
evaluate a limit. We have this process:
Notes:
65
input specific do something return

number c to f and c number f ′ (c)
This process describes a func on; given one input (the value of c), we return
exactly one output (the value of f ′ (c)). The “do something” box is where the
tedious work (taking limits) of this func on occurs.
Instead of applying this func on repeatedly for different values of c, let us
apply it just once to the variable x. We then take a limit just once. The process
now looks like:
do something return
input variable x
to f and x func on f ′ (x)
The output is the “deriva ve func on,” f ′ (x). The f ′ (x) func on will take a
number c as input and return the deriva ve of f at c. This calls for a defini on.
Defini on 2.1.4 Deriva ve Func on

Let f be a differen able func on on an open interval I. The func on
f(x + h) − f(x)
f ′ (x) = lim
h→0 h
is the deriva ve of f.
Nota on:
Let y = f(x). The following nota ons all represent the deriva ve of f:
dy df d d
f ′ (x) = y′ = = = (f) = (y).
dx dx dx dx
dy
Important: The nota on is one symbol; it is not the frac on “dy/dx”. The
dx
nota on, while somewhat confusing at first, was chosen with care. A frac on–
looking symbol was chosen because the deriva ve has many frac on–like prop-
er es. Among other places, we see these proper es at work when we talk about
the units of the deriva ve, when we discuss the Chain Rule, and when we learn
about integra on (topics that appear in later sec ons and chapters).
Examples will help us understand this defini on.
Example 2.1.5 Finding the deriva ve of a func on

Let f(x) = 3x2 + 5x − 7 as in Example 2.1.1. Find f ′ (x).
Notes:
66
S We apply Defini on 2.1.4.
f(x + h) − f(x)
f ′ (x) = lim
h→0 h
3(x + h)2 + 5(x + h) − 7 − (3x2 + 5x − 7)
= lim
h→0 h
3h2 + 6xh + 5h
= lim
h→0 h
= lim (3h + 6x + 5)
h→0
= 6x + 5
So f ′ (x) = 6x + 5. Recall earlier we found that f ′ (1) = 11 and f ′ (3) = 23. Note
our new computa on of f ′ (x) affirm these facts.

1
Let f(x) = . Find f ′ (x).
x+1
S We apply Defini on 2.1.4.
f(x + h) − f(x)
f ′ (x) = lim
h→0 h
1
x+h+1 − 1
x+1
= lim
h→0 h
Now find common denominator then subtract; pull 1/h out front to facilitate
reading.
( )
1 x+1 x+h+1
= lim · −
h→0 h (x + 1)(x + h + 1) (x + 1)(x + h + 1)
( )
1 x + 1 − (x + h + 1)
= lim ·
h→0 h (x + 1)(x + h + 1)
( )
1 −h
= lim ·
h→0 h (x + 1)(x + h + 1)
−1
= lim
h→0 (x + 1)(x + h + 1)
−1
=
(x + 1)(x + 1)
−1
= .
(x + 1)2
Notes:
67
−1
So f ′ (x) = . To prac ce using our nota on, we could also state
(x + 1)2
( )
d 1 −1
= .
dx x + 1 (x + 1)2

Find the deriva ve of f(x) = sin x.
S Before applying Defini on 2.1.4, note that once this is found,

we can find the actual tangent line to f(x) = sin x at x = 0, whereas we se led
for an approxima on in Example 2.1.4.
sin(x + h) − sin x ( )
f ′ (x) = lim Use trig iden ty
sin(x + h) = sin x cos h + cos x sin h
h→0 h
sin x cos h + cos x sin h − sin x
= lim (regroup)
h→0 h
sin x(cos h − 1) + cos x sin h
= lim (split into two frac ons)
h→0 h
( ) ( )
sin x(cos h − 1) cos x sin h cos h − 1 sin h
= lim + use lim = 0 and lim =1
h→0 h h h→0 h h→0 h
= sin x · 0 + cos x · 1
= cos x !
We have found that when f(x) = sin x, f ′ (x) = cos x. This should be somewhat
surprising; the result of a tedious limit process and the sine func on is a nice
func on. Then again, perhaps this is not en rely surprising. The sine func on
is periodic – it repeats itself on regular intervals. Therefore its rate of change
also repeats itself on the same regular intervals. We should have known the
deriva ve would be periodic; we now know exactly which periodic func on it is.
Thinking back to Example 2.1.4, we can find the slope of the tangent line to
f(x) = sin x at x = 0 using our deriva ve. We approximated the slope as 0.9983;
y
we now know the slope is exactly cos 0 = 1.
1
Example 2.1.8 Finding the deriva ve of a piecewise defined func on
Find the deriva ve of the absolute value func on,
{
0.5 −x x < 0
f(x) = |x| = .
x x≥0
See Figure 2.1.6.
x
−1 −0.5 0.5 1
. f(x + h) − f(x)
S We need to evaluate lim . As f is piecewise–
h
h→0
Figure 2.1.6: The absolute value func on, defined, we need to consider separately the limits when x < 0 and when x > 0.
f(x) = |x|. No ce how the slope of
the lines (and hence the tangent lines)
abruptly changes at x = 0.
Notes:
68
When x < 0:
d( ) −(x + h) − (−x)
− x = lim
dx h→0 h
−h
= lim
h→0 h
= lim −1
h→0
= −1.
d( )
When x > 0, a similar computa on shows that x = 1.
dx
We need to also find the deriva ve at x = 0. By the defini on of the deriva-
ve at a point, we have
f(0 + h) − f(0)
f ′ (0) = lim .
h→0 h
Since x = 0 is the point where our func on’s defini on switches from one piece
to the other, we need to consider le and right-hand limits. Consider the fol-
lowing, where we compute the le and right hand limits side by side.
f(0 + h) − f(0) f(0 + h) − f(0)

lim = lim =
h→0− h h→0+ h
−h − 0 h−0
lim = lim = y
h→0 − h h→0 + h
lim −1 = −1 lim 1 = 1 1
h→0− h→0+
The last lines of each column tell the story: the le and right hand limits are not
equal. Therefore the limit does not exist at 0, and f is not differen able at 0. So
x
we have { −1 −0.5 0.5 1
−1 x < 0
f ′ (x) = .
1 x>0
−1
At x = 0, f ′ (x) does not exist; there is a jump discon nuity at 0; see Figure 2.1.7.
.
So f(x) = |x| is differen able everywhere except at 0.
Figure 2.1.7: A graph of the deriva ve of
The point of non-differen ability came where the piecewise defined func- f(x) = |x|.
on switched from one piece to the other. Our next example shows that this
does not always cause trouble.
Example 2.1.9 Finding the deriva ve {of a piecewise defined func on

sin x x ≤ π/2
Find the deriva ve of f(x), where f(x) = . See Figure 2.1.8.
1 x > π/2
Notes:
69
S Using Example 2.1.7, we know that when x < π/2, f ′ (x) =

y
cos x. It is easy to verify that when x > π/2, f ′ (x) = 0; consider:
1
f(x + h) − f(x) 1−1
lim = lim = lim 0 = 0.
h→0 h h→0 h h→0
0.5
So far we have {
cos x x < π/2
f ′ (x) = .
x 0 x > π/2
π
2
. We s ll need to find f ′ (π/2). No ce at x = π/2 that both pieces of f ′ are 0,

meaning we can state that f ′ (π/2) = 0.
Figure 2.1.8: A graph of f(x) as defined in Being more rigorous, we can again evaluate the difference quo ent limit at
Example 2.1.9. x = π/2, u lizing again le and right–hand limits:
f(π/2 + h) − f(π/2) f(π/2 + h) − f(π/2)

lim = lim =
h→0− h h→0+ h
y sin(π/2 + h) − sin(π/2) 1−1
lim = lim =
h→0− h h→0+ h
sin( π2 ) cos(h) + sin(h) cos( π2 ) − sin( π2 ) 0
1 lim = lim =
h h→0+ h
h→0−
1 · cos(h) + sin(h) · 0 − 1 0.
lim =
0.5 h→0− h
0.
x
Since both the le and right hand limits are 0 at x = π/2, the limit exists and
π
2
f ′ (π/2) exists (and is 0). Therefore we can fully write f ′ as
. {
cos x x ≤ π/2
Figure 2.1.9: A graph of f ′ (x) in Example f ′ (x) = .
0 x > π/2
2.1.9.
See Figure 2.1.9 for a graph of this func on.
Recall we pseudo–defined a con nuous func on as one in which we could

sketch its graph without li ing our pencil. We can give a pseudo–defini on for
differen ability as well: it is a con nuous func on that does not have any “sharp
corners.” One such sharp corner is shown in Figure 2.1.6. Even though the func-
on f in Example 2.1.9 is piecewise–defined, the transi on is “smooth” hence it
is differen able. Note how in the graph of f in Figure 2.1.8 it is difficult to tell
when f switches from one piece to the other; there is no “corner.”
Notes:
70
Differen ablity on Closed Intervals
When we defined the deriva ve at a point in Defini on 2.1.1, we specified

that the interval I over which a func on f was defined needed to be an open
interval. Open intervals are required so that we can take a limit at any point c in
I, meaning we want to approach c from both the le and right.
Recall we also required open intervals in Defini on 1.5.1 when we defined
what it meant for a func on to be con nuous. Later, we used one-sided limits to
extend con nuity to closed intervals. We now extend differen ability to closed
intervals by again considering one-sided limits.
Our mo va on is three-fold. First, we consider “common sense.” In Example
2.1.5 we found that when f(x) = 3x2 +5x−7, f ′ (x) = 6x+5, and this deriva ve
is defined for all real numbers, hence f is differen able everywhere. It seems
appropriate to also conclude that f is differen able on closed intervals, like [0, 1],
as well. A er all, f ′ (x) is defined at both x = 0 and x = 1.
√
Secondly, consider f(x) = x. The domain of f is [0, ∞). Is f differen able
on its domain – specifically, is f differen able at 0? (We’ll consider this in the
next example.)
Finally, in later sec ons, having the deriva ve defined on closed intervals will
prove useful. One such place is Sec on 7.4 where the deriva ve plays a role in
measuring the length of a curve.
A er a formal defini on of differen ability on a closed interval, we explore
the concept in an example.
Defini on 2.1.5 Differen ability on a Closed Interval

Let f be con nuous on [a, b] and differen able on (a, b), and let the one-
sided limits
f(a + h) − f(a) f(b + h) − f(b)
lim and lim
h→0+ h h→0− h
exist. Then we say f is differen able on [a, b].
For all the func ons f in this text, we can determine differen ability on [a, b]
by considering the limits limx→a+ f ′ (x) and limx→b− f ′ (x). This is o en easier to
evaluate than the limit of the difference quo ent.
Example 2.1.10 √ Differen ability at an √endpoint

Consider f(x) = x = x1/2 and g(x) = x3 = x3/2 . The domain of each func-
on is [0, ∞). It can be shown that each is differen able on (0, ∞); determine
the differen ability of each at x = 0.
Notes:
71
S We start by considering f and take the right-hand limit of the

difference quo ent:
√ √
f(a + h) − f(a) 0+h− 0
lim = lim
h→0+ h h→0+ h
√
h
= lim
h→0+ h
1
= lim 1/2 = ∞.
h→0+ h
The one-sided limit of the difference quo ent does not exist at x = 0 for f;
therefore f is differen able on (0, ∞) and not( differen
√ ) able on [0, ∞).
We state (without proof) that f ′ (x) = 1/ 2 x . Note that limx→0+ f ′ (x) =
∞; this limit was easier to evaluate than the limit of the difference quo ent,
though it required us to already know the deriva ve of f.
y
Now consider g:
√ √
1 g(a + h) − g(a) (0 + h)3 − 0
lim = lim
h→0+ h h→0+ h
y = x1/2
h3/2
= lim
h→0+ h
y = x3/2
= lim h1/2 = 0.
h→0+
x
1
As the one-sided limit exists at x = 0, we conclude g is differen able on its
domain of [0, ∞). √
Figure 2.1.10: A graph of y = x1/2 and We state (without proof) that g ′ (x) = 3 x/2. Note that limx→0+ g ′ (x) = 0;
y = x3/2 in Example 2.1.10. again, this limit is easier to evaluate than the limit of the difference quo
√ ent.
The two func ons are graphed in Figure 2.1.10. Note how f(x) = x seems
to “go ver cal” as x approaches 0, implying the slopes of its tangent lines are
growing
√ toward infinity. Also note how the slopes of the tangent lines to g(x) =
3
x approach 0 as x approaches 0.
Most calculus textbooks omit this topic and simply avoid specific cases where
it could be applied. We choose in this text to not make use of the topic unless
it is “needed.” Many theorems in later sec ons require a func on f to be differ-
en able on an open interval I; we could remove the word “open” and just use
“. . . on an interval I,” but choose to not do so in keeping with the current math-
ema cal tradi on. Our first use of differen ability on closed intervals comes in
Chapter 7, where we measure the lengths of curves.
This sec on defined the deriva ve; in some sense, it answers the ques on of
“What is the deriva ve?” The next sec on addresses the ques on “What does
the deriva ve mean?”
Notes:
72
Exercises 2.1
Terms and Concepts 19. h(x) = x3 , at x = 4.
1. T/F: Let f be a posi on func on. The average rate of change 20. f(x) = 3x2 − x + 4, at x = −1.
on [a, b] is the slope of the line through the points (a, f(a))
and (b, f(b)). 1
21. r(x) = , at x = −2.
x
2. T/F: The defini on of the deriva ve of a func on at a point
1
involves taking a limit. 22. r(x) = , at x = 3.
x−2
3. In your own words, explain the difference between the av- In Exercises 23 – 26, a func on f and an x–value a are given.
erage rate of change and instantaneous rate of change. Approximate the equa on of the tangent line to the graph of
f at x = a by numerically approxima ng f ′ (a), using h = 0.1.
4. In your own words, explain the difference between Defini-
ons 2.1.1 and 2.1.4. 23. f(x) = x2 + 2x + 1, x = 3
5. Let y = f(x). Give three different nota ons equivalent to 10

24. f(x) = ,x=9
“f ′ (x).” x+1
6. If two lines are perpendicular, what is true of their slopes? 25. f(x) = ex , x = 2
26. f(x) = cos x, x = 0

Problems
27. The graph of f(x) = x2 − 1 is shown.
In Exercises 7 – 14, use the defini on of the deriva ve to com- (a) Use the graph to approximate the slope of the tan-
pute the deriva ve of the given func on. gent line to f at the following points: (−1, 0), (0, −1)
and (2, 3).
7. f(x) = 6 (b) Using the defini on, find f ′ (x).
(c) Find the slope of the tangent line at the points
8. f(x) = 2x
(−1, 0), (0, −1) and (2, 3).
9. f(t) = 4 − 3t y
10. g(x) = x2
2
3
11. h(x) = x 1
x
12. f(x) = 3x2 − x + 4 −2 −1 1 2
−1
.
1
13. r(x) =
x 1
28. The graph of f(x) = is shown.
x+1
1
14. r(s) = (a) Use the graph to approximate the slope of the tan-
s−2
gent line to f at the following points: (0, 1) and
In Exercises 15 – 22, a func on and an x–value c are given. (1, 0.5).
(Note: these func ons are the same as those given in Exer- (b) Using the defini on, find f ′ (x).
cises 7 through 14.) (c) Find the slope of the tangent line at the points (0, 1)
(a) Give the equa on of the tangent line at x = c. and (1, 0.5).
(b) Give the equa on of the normal line at x = c. y
15. f(x) = 6, at x = −2. 4
3
16. f(x) = 2x, at x = 3.
2
17. f(x) = 4 − 3x, at x = 7. 1
x
.
18. g(x) = x2 , at x = 2. −1 1 2 3
73
In Exercises 29 – 32, a graph of a func on f(x) is given. Using
the graph, sketch f ′ (x).
y
y 5
2 34. x
−2 −1 1 2
29. 1
x −5
−2 −1 1 2 3 4
−1
.
.
y In Exercises 35 – 36, a func on f(x) is given, along with its do-

main and deriva ve. Determine if f(x) is differen able on its
2 domain.
30. √ (5 − 6x)x3/2
x5 (1 − x), domain = [0, 1], f ′ (x) =
x
−6 −4 −2 2 35. f(x) = √
2 1−x
−2
(√ )
(√ ) sin x
. 36. f(x) = cos x , domain = [0, ∞), f ′ (x) = − √
2 x
y
5
Review
31. x
−2 −1 1 2 x2 + 2x − 35
37. Approximate lim .
x→5 x2 − 10.5x + 27.5
.
. −5 38. Use the Bisec on Method to approximate, accurate to two
decimal places, the root of g(x) = x3 + x2 + x − 1 on
y [0.5, 0.6].
1
0.5 39. Give intervals on which each of the following func ons are
con nuous.
32. x
−2π −π π 2π
1 √
−0.5 (a) (c) 5−x
. ex + 1
−1 1 √
. (b) 2 (d) 5 − x2
x −1
In Exercises 33 – 34, a graph of a func on g(x) is given. Using
the graph, answer the following ques ons. 40. Use the graph of f(x) provided to answer the following.
1. Where is g(x) > 0? 1. Where is g′ (x) < 0? (a) lim f(x) =? (c) lim f(x) =?
x→−3− x→−3
2. Where is g(x) < 0? 2. Where is g′ (x) > 0?
(b) lim f(x) =? (d) Where is f con nu-
3. Where is g(x) = 0? 3. Where is g′ (x) = 0? x→−3+ ous?
y
y 3
5
1
33. x
−2 −1 1 2
x
−4 −2
. −1
. −5
74
.
2.2 Interpreta ons of the Deriva ve

The previous sec on defined the deriva ve of a func on and gave examples of
how to compute it using its defini on (i.e., using limits). The sec on also started
with a brief mo va on for this defini on, that is, finding the instantaneous ve-
locity of a falling object given its posi on func on. The next sec on will give us
more accessible tools for compu ng the deriva ve, tools that are easier to use
than repeated use of limits.
This sec on falls in between the “What is the defini on of the deriva ve?”
and “How do I compute the deriva ve?” sec ons. Here we are concerned with
“What does the deriva ve mean?”, or perhaps, when read with the right em-
phasis, “What is the deriva ve?” We offer two interconnected interpreta ons
of the deriva ve, hopefully explaining why we care about it and why it is worthy
of study.
Interpreta on of the Deriva ve #1: Instantaneous Rate of Change
The previous sec on started with an example of using the posi on of an

object (in this case, a falling amusement–park rider) to find the object’s veloc-
ity. This type of example is o en used when introducing the deriva ve because
we tend to readily recognize that velocity is the instantaneous rate of change
of posi on. In general, if f is a func on of x, then f ′ (x) measures the instan-
taneous rate of change of f with respect to x. Put another way, the deriva-
ve answers “When x changes, at what rate does f change?” Thinking back to
the amusement–park ride, we asked “When me changed, at what rate did the
height change?” and found the answer to be “By −64 feet per second.”
Now imagine driving a car and looking at the speedometer, which reads “60
mph.” Five minutes later, you wonder how far you have traveled. Certainly, lots
of things could have happened in those 5 minutes; you could have inten onally
sped up significantly, you might have come to a complete stop, you might have
slowed to 20 mph as you passed through construc on. But suppose that you
know, as the driver, none of these things happened. You know you maintained
a fairly consistent speed over those 5 minutes. What is a good approxima on of
the distance traveled?
One could argue the only good approxima on, given the informa on pro-
vided, would be based on “distance = rate × me.” In this case, we assume a
constant rate of 60 mph with a me of 5/60 hours. Hence we would approxi-
mate the distance traveled as 5 miles.
Referring back to the falling amusement–park ride, knowing that at t = 2 the

velocity was −64 /s, we could reasonably assume that 1 second later the rid-
Notes:
75
ers’ height would have dropped by about 64 feet. Knowing that the riders were
accelera ng as they fell would inform us that this is an under–approxima on. If
all we knew was that f(2) = 86 and f ′ (2) = −64, we’d know that we’d have to
stop the riders quickly otherwise they would hit the ground.
Units of the Deriva ve
It is useful to recognize the units of the deriva ve func on. If y is a func on

of x, i.e., y = f(x) for some func on f, and y is measured in feet and x in seconds,
then the units of y′ = f ′ are “feet per second,” commonly wri en as “ /s.” In
general, if y is measured in units P and x is measured in units Q, then y′ will be
measured in units “P per Q”, or “P/Q.” Here we see the frac on–like behavior
of the deriva ve in the nota on:
dy units of y
the units of are .
dx units of x
Example 2.2.1 The meaning of the deriva ve: World Popula on
Let P(t) represent the world popula on t minutes a er 12:00 a.m., January 1,
2012. It is fairly accurate to say that P(0) = 7, 028, 734, 178 (www.prb.org). It
is also fairly accurate to state that P ′ (0) = 156; that is, at midnight on January 1,
2012, the popula on of the world was growing by about 156 people per minute
(note the units). Twenty days later (or, 28,800 minutes later) we could reason-
ably assume the popula on grew by about 28, 800 · 156 = 4, 492, 800 people.
Example 2.2.2 The meaning of the deriva ve: Manufacturing
The term widget is an economic term for a generic unit of manufacturing

output. Suppose a company produces widgets and knows that the market sup-
ports a price of $10 per widget. Let P(n) give the profit, in dollars, earned by
manufacturing and selling n widgets. The company likely cannot make a (pos-
i ve) profit making just one widget; the start–up costs will likely exceed $10.
Mathema cally, we would write this as P(1) < 0.
What do P(1000) = 500 and P ′ (1000) = 0.25 mean? Approximate P(1100).
S The equa on P(1000) = 500 means that selling 1,000 wid-

gets returns a profit of $500. We interpret P ′ (1000) = 0.25 as meaning that
the profit is increasing at rate of $0.25 per widget (the units are “dollars per
widget.”) Since we have no other informa on to use, our best approxima on
for P(1100) is:
P(1100) ≈ P(1000) + P ′ (1000) × 100 = $500 + 100 · 0.25 = $525.
We approximate that selling 1,100 widgets returns a profit of $525.
Notes:
76
The previous examples made use of an important approxima on tool that

we first used in our previous “driving a car at 60 mph” example at the begin-
ning of this sec on. Five minutes a er looking at the speedometer, our best
approxima on for distance traveled assumed the rate of change was constant.
In Examples 2.2.1 and 2.2.2 we made similar approxima ons. We were given
rate of change informa on which we used to approximate total change. Nota-
onally, we would say that
f(c + h) ≈ f(c) + f ′ (c) · h.
This approxima on is best when h is “small.” “Small” is a rela ve term; when
dealing with the world popula on, h = 22 days = 28,800 minutes is small in
comparison to years. When manufacturing widgets, 100 widgets is small when
one plans to manufacture thousands.
The Deriva ve and Mo on
One of the most fundamental applica ons of the deriva ve is the study of
mo on. Let s(t) be a posi on func on, where t is me and s(t) is distance. For
instance, s could measure the height of a projec le or the distance an object has
traveled.
Let’s let s(t) measure the distance traveled, in feet, of an object a er t sec-
onds of travel. Then s ′ (t) has units “feet per second,” and s ′ (t) measures the
instantaneous rate of distance change – it measures velocity.
Now consider v(t), a velocity func on. That is, at me t, v(t) gives the ve-
locity of an object. The deriva ve of v, v ′ (t), gives the instantaneous rate of
velocity change – accelera on. (We o en think of accelera on in terms of cars:
a car may “go from 0 to 60 in 4.8 seconds.” This is an average accelera on, a
measurement of how quickly the velocity changed.) If velocity is measured in
feet per second, and me is measured in seconds, then the units of accelera on
(i.e., the units of v ′ (t)) are “feet per second per second,” or ( /s)/s. We o en
shorten this to “feet per second squared,” or /s2 , but this tends to obscure the
meaning of the units.
Perhaps the most well known accelera on is that of gravity. In this text, we
use g = 32 /s2 or g = 9.8m/s2 . What do these numbers mean?
A constant accelera on of 32( /s)/s means that the velocity changes by
32 /s each second. For instance, let v(t) measures the velocity of a ball thrown
straight up into the air, where v has units /s and t is measured in seconds. The
ball will have a posi ve velocity while traveling upwards and a nega ve velocity
while falling down. The accelera on is thus −32 /s2 . If v(1) = 20 /s, then
when t = 2, the velocity will have decreased by 32 /s; that is, v(2) = −12 /s.
We can con nue: v(3) = −44 /s, and we can also figure that v(0) = 52 /s.
These ideas are so important we write them out as a Key Idea.
Notes:
77
Key Idea 2.2.1 The Deriva ve and Mo on
1. Let s(t) be the posi on func on of an object. Then s ′ (t) is the

velocity func on of the object.
2. Let v(t) be the velocity func on of an object. Then v ′ (t) is the
accelera on func on of the object.
We now consider the second interpreta on of the deriva ve given in this

sec on. This interpreta on is not independent from the first by any means;
many of the same concepts will be stressed, just from a slightly different per-
spec ve.
y
16
Interpreta on of the Deriva ve #2: The Slope of the Tangent Line

12
f(c + h) − f(c)
8
Given a func on y = f(x), the difference quo ent gives a
h
change in y values divided by a change in x values; i.e., it is a measure of the
4 “rise(over run,”
) or(“slope,” of the) line that goes through two points on the graph
of f: c, f(c) and c+h, f(c+h) . As h shrinks to 0, these two points come close
. x together; in the limit we find f ′ (c), the slope of a special line called the tangent
1 2 3 4 line that intersects f only once near x = c.
Lines have a constant rate of change, their slope. Nonlinear func ons do not
Figure 2.2.1: A graph of f(x) = x2 .
have a constant rate of change, but we can measure their instantaneous rate of
change at a given x value c by compu ng f ′ (c). We can get an idea of how f is
behaving by looking at the slopes of its tangent lines. We explore this idea in the
following example.
y
16 Example 2.2.3 Understanding the deriva ve: the rate of change
Consider f(x) = x2 as shown in Figure 2.2.1. It is clear that at x = 3 the func on
12 is growing faster than at x = 1, as it is steeper at x = 3. How much faster is it
growing?
8
S We can answer this directly a er the following sec on, where

4 we learn to quickly compute deriva ves. For now, we will answer graphically,
by considering the slopes of the respec ve tangent lines.
.. x
With prac ce, one can fairly effec vely sketch tangent lines to a curve at a
1 2 3 4
par cular point. In Figure 2.2.2, we have sketched the tangent lines to f at x = 1
Figure 2.2.2: A graph of f(x) = x2 and tan- and x = 3, along with a grid to help us measure the slopes of these lines. At
gent lines.
Notes:
78
x = 1, the slope is 2; at x = 3, the slope is 6. Thus we can say not only is f

y
growing faster at x = 3 than at x = 1, it is growing three mes as fast. f(x)
5
Example 2.2.4 Understanding the graph of the deriva ve f ′ (x)

Consider the graph of f(x) and its deriva ve, f ′ (x), in Figure 2.2.3(a). Use these
x
graphs to find the slopes of the tangent lines to the graph of f at x = 1, x = 2, 1 2 3
and x = 3.
−5
S To find the appropriate slopes of tangent lines to the graph
of f, we need to look at the corresponding values of f ′ .
The slope of the tangent line to f at x = 1 is f ′ (1); this looks to be about −1. .
(a)
The slope of the tangent line to f at x = 2 is f ′ (2); this looks to be about 4.
.
The slope of the tangent line to f at x = 3 is f ′ (3); this looks to be about 3. y
Using these slopes, the tangent lines to f are sketched in Figure 2.2.3(b). In- 5
f(x)
cluded on the graph of f ′ in this figure are filled circles where x = 1, x = 2 and
f ′ (x)
x = 3 to help be er visualize the y value of f ′ at those points.
x
Example 2.2.5 Approxima on with the deriva ve 1 2 3
Consider again the graph of f(x) and its deriva ve f ′ (x) in Example 2.2.4. Use
the tangent line to f at x = 3 to approximate the value of f(3.1). −5
S Figure 2.2.4 shows the graph of f along with its tangent line, .
zoomed in at x = 3. No ce that near x = 3, the tangent line makes an excellent (b)
approxima on of f. Since lines are easy to deal with, o en it works well to ap- .
proximate a func on with its tangent line. (This is especially true when you don’t Figure 2.2.3: Graphs of f and f ′ in Example
actually know much about the func on at hand, as we don’t in this example.) 2.2.4, along with tangent lines in (b).
While the tangent line to f was drawn in Example 2.2.4, it was not explicitly
computed. Recall that the tangent line to f at x = c is y = f ′ (c)(x − c) + f(c).
While f is not explicitly given, by the graph it looks like f(3) = 4. Recalling that
f ′ (3) = 3, we can compute the tangent line to be approximately y = 3(x−3)+4. y
It is o en useful to leave the tangent line in point–slope form.
To use the tangent line to approximate f(3.1), we simply evaluate y at 3.1
instead of f. 4
f(3.1) ≈ y(3.1) = 3(3.1 − 3) + 4 = .1 ∗ 3 + 4 = 4.3.

3
We approximate f(3.1) ≈ 4.3.
To demonstrate the accuracy of the tangent line approxima on, we now 2 . x

state that in Example 2.2.5, f(x) = −x3 + 7x2 − 12x + 4. We can evaluate 2.8 3 3.2
f(3.1) = 4.279. Had we known f all along, certainly we could have just made
this computa on. In reality, we o en only know two things: Figure 2.2.4: Zooming in on f and its tan-
gent line at x = 3 for the func on given
in Examples 2.2.4 and 2.2.5.
Notes:
79
1. what f(c) is, for some value of c, and

2. what f ′ (c) is.
For instance, we can easily observe the loca on of an object and its instan-
taneous velocity at a par cular point in me. We do not have a “func on f ”
for the loca on, just an observa on. This is enough to create an approxima ng
func on for f.
This last example has a direct connec on to our approxima on method ex-
plained above a er Example 2.2.2. We stated there that
f(c + h) ≈ f(c) + f ′ (c) · h.
If we know f(c) and f ′ (c) for some value x = c, then compu ng the tangent
line at (c, f(c)) is easy: y(x) = f ′ (c)(x − c) + f(c). In Example 2.2.5, we used
the tangent line to approximate a value of f. Let’s use the tangent line at x = c
to approximate a value of f near x = c; i.e., compute y(c + h) to approximate
f(c + h), assuming again that h is “small.” Note:
( )
y(c + h) = f ′ (c) (c + h) − c + f(c) = f ′ (c) · h + f(c).
This is the exact same approxima on method used above! Not only does it make
intui ve sense, as explained above, it makes analy cal sense, as this approxima-
on method is simply using a tangent line to approximate a func on’s value.
The importance of understanding the deriva ve cannot be understated. When

f is a func on of x, f ′ (x) measures the instantaneous rate of change of f with re-
spect to x and gives the slope of the tangent line to f at x.
Notes:
80
Exercises 2.2
Terms and Concepts In Exercises 15 – 18, graphs of func ons f(x) and g(x) are
given. Iden fy which func on is the deriva ve of the other.
1. What is the instantaneous rate of change of posi on y
called?
4 g(x)
2. Given a func on y = f(x), in your own words describe how 2 f(x)
to find the units of f ′ (x). 15. x

−4 −2 2 4
−2
3. What func ons have a constant rate of change?
−4
.
Problems y
4
g(x) f(x)
4. Given f(5) = 10 and f ′ (5) = 2, approximate f(6). 2
16. x
5. Given P(100) = −67 and P ′ (100) = 5, approximate 2
P(110). −2
−4
6. Given z(25) = 187 and z′ (25) = 17, approximate z(20). .
y
7. Knowing f(10) = 25 and f ′ (10) = 5 and the methods de- 5
scribed in this sec on, which approxima on is likely to be
most accurate: f(10.1), f(11), or f(20)? Explain your rea-
f(x)
soning. 17. . x
−5 5
8. Given f(7) = 26 and f(8) = 22, approximate f ′ (7). g(x)
−5
9. Given H(0) = 17 and H(2) = 29, approximate H ′ (2). .
y
10. Let V(x) measure the volume, in decibels, measured inside 2
a restaurant with x customers. What are the units of V ′ (x)?
f(x)
11. Let v(t) measure the velocity, in /s, of a car moving in a 18. x
straight line t seconds a er star ng. What are the units of −4 −2 2 4
v ′ (t)?
g(x)
−2
12. The height H, in feet, of a river is recorded t hours a er mid-
night, April 1. What are the units of H ′ (t)?
13. P is the profit, in thousands of dollars, of producing and sell- Review

ing c cars.
In Exercises 19 – 20, use the defini on to compute the deriva-
′ ves of the following func ons.
(a) What are the units of P (c)?
(b) What is likely true of P(0)? 19. f(x) = 5x2
20. f(x) = (x − 2)3

14. T is the temperature in degrees Fahrenheit, h hours a er
midnight on July 4 in Sidney, NE.
In Exercises 21 – 22, numerically approximate the value of
f ′ (x) at the indicated x value.
(a) What are the units of T ′ (h)?
(b) Is T ′ (8) likely greater than or less than 0? Why? 21. f(x) = cos x at x = π.
√
(c) Is T(8) likely greater than or less than 0? Why? 22. f(x) = x at x = 9.
81
2.3 Basic Differen a on Rules

The deriva ve is a powerful tool but is admi edly awkward given its reliance on
limits. Fortunately, one thing mathema cians are good at is abstrac on. For
instance, instead of con nually finding deriva ves at a point, we abstracted and
found the deriva ve func on.
Let’s prac ce abstrac on on linear func ons, y = mx + b. What is y ′ ? With-
out limits, recognize that linear func on are characterized by being func ons
with a constant rate of change (the slope). The deriva ve, y ′ , gives the instan-
taneous rate of change; with a linear func on, this is constant, m. Thus y ′ = m.
Let’s abstract once more. Let’s find the deriva ve of the general quadra c
func on, f(x) = ax2 + bx + c. Using the defini on of the deriva ve, we have:
a(x + h)2 + b(x + h) + c − (ax2 + bx + c)
f ′ (x) = lim
h→0 h
ah2 + 2ahx + bh
= lim
h→0 h
= lim ah + 2ax + b
h→0
= 2ax + b.
So if y = 6x2 + 11x − 13, we can immediately compute y ′ = 12x + 11.
In this sec on (and in some sec ons to follow) we will learn some of what
mathema cians have already discovered about the deriva ves of certain func-
ons and how deriva ves interact with arithme c opera ons. We start with a
theorem.
Theorem 2.3.1 Deriva ves of Common Func ons
1. Constant Rule: d
5. (sin x) = cos x
d( ) dx
c = 0, where c is a constant.
dx d
6. (cos x) = − sin x
2. Power Rule: dx
d ( n)
x = nxn−1 , where n is an d x
dx 7. (e ) = ex
integer, n > 0. dx
d 1
8. (ln x) =
dx x
Notes:
82
This theorem starts by sta ng an intui ve fact: constant func ons have no
rate of change as they are constant. Therefore their deriva ve is 0 (they change
at the rate of 0). The theorem then states some fairly amazing things. The Power
Rule states that the deriva ves of Power Func ons (of the form y = xn ) are very
straigh orward: mul ply by the power, then subtract 1 from the power. We see
something incredible about the func on y = ex : it is its own deriva ve. We also
see a new connec on between the sine and cosine func ons.
One special case of the Power Rule is when n = 1, i.e., when f(x) = x. What
is f ′ (x)? According to the Power Rule,
d( ) d ( 1)
f ′ (x) = x = x = 1 · x0 = 1.
dx dx
In words, we are asking “At what rate does f change with respect to x?” Since f
is x, we are asking “At what rate does x change with respect to x?” The answer
is: 1. They change at the same rate.
Let’s prac ce using this theorem.
Example 2.3.1 Using Theorem 2.3.1 to find, and use, deriva ves
Let f(x) = x3 .
y
1. Find f ′ (x). 4
2. Find the equa on of the line tangent to the graph of f at x = −1. 2

3
3. Use the tangent line to approximate (−1.1) . x
′ −2 −1 1 2
4. Sketch f, f and the found tangent line on the same axis.
−2
−4
S .
1. The Power Rule states that if f(x) = x3 , then f ′ (x) = 3x2 . Figure 2.3.1: A graph of f(x) = x3 , along
′ 2
with
. its deriva ve f (x) = 3x and its tan-
2. To find the equa on of the line tangent to the graph of f at x = −1, we gent line at x = −1.
need a point and the slope. The point is (−1, f(−1)) = (−1, −1). The
slope is f ′ (−1) = 3. Thus the tangent line has equa on y = 3(x−(−1))+
(−1) = 3x + 2.
3. We can use the tangent line to approximate (−1.1)3 as −1.1 is close to
−1. We have
(−1.1)3 ≈ 3(−1.1) + 2 = −1.3.
We can easily find the actual answer; (−1.1)3 = −1.331.
4. See Figure 2.3.1.
Notes:
83
Theorem 2.3.1 gives useful informa on, but we will need much more. For
instance, using the theorem, we can easily find the deriva ve of y = x3 , but
it does not tell how to compute the deriva ve of y = 2x3 , y = x3 + sin x nor
y = x3 sin x. The following theorem helps with the first two of these examples
(the third is answered in the next sec on).
Theorem 2.3.2 Proper es of the Deriva ve

Let f and g be differen able on an open interval I and let c be a real
number. Then:
1. Sum/Difference Rule:
d( ) d( ) d( )
f(x) ± g(x) = f(x) ± g(x) = f ′ (x) ± g ′ (x)
dx dx dx
2. Constant Mul ple Rule:
d( ) d( )
c · f(x) = c · f(x) = c · f ′ (x).
dx dx
Theorem 2.3.2 allows us to find the deriva ves of a wide variety of func ons.
It can be used in conjunc on with the Power Rule to find the deriva ves of any
polynomial. Recall in Example 2.1.5 that we found, using the limit defini on,
the deriva ve of f(x) = 3x2 + 5x − 7. We can now find its deriva ve without
expressly using limits:
d( 2 ) d ( 2) d( ) d( )
3x + 5x + 7 = 3 x +5 x + 7
dx dx dx dx
= 3 · 2x + 5 · 1 + 0
= 6x + 5.
We were a bit pedan c here, showing every step. Normally we would do all
d( 2 )
the arithme c and steps in our head and readily find 3x +5x+7 = 6x+5.
dx
Example 2.3.2 Using the tangent line to approximate a func on value
Let f(x) = sin x + 2x + 1. Approximate f(3) using an appropriate tangent line.
S This problem is inten onally ambiguous; we are to approxi-

mate using an appropriate tangent line. How good of an approxima on are we
seeking? What does appropriate mean?
In the “real world,” people solving problems deal with these issues all me.
One must make a judgment using whatever seems reasonable. In this example,
the actual answer is f(3) = sin 3 + 7, where the real problem spot is sin 3. What
is sin 3?
Notes:
84
Since 3 is close to π, we can assume sin 3 ≈ sin π = 0. Thus one guess is

f(3) ≈ 7. Can we do be er? Let’s use a tangent line as instructed and examine
the results; it seems best to find the tangent line at x = π.
Using Theorem 2.3.1 we find f ′ (x) = cos x + 2. The slope of the tangent line
is thus f ′ (π) = cos π + 2 = 1. Also, f(π) = 2π + 1 ≈ 7.28. So the tangent line
to the graph of f at x = π is y = 1(x − π) + 2π + 1 = x + π + 1 ≈ x + 4.14.
Evaluated at x = 3, our tangent line gives y = 3 + 4.14 = 7.14. Using the
tangent line, our final approxima on is that f(3) ≈ 7.14.
Using a calculator, we get an answer accurate to 4 places a er the decimal:
f(3) = 7.1411. Our ini al guess was 7; our tangent line approxima on was more
accurate, at 7.14.
The point is not “Here’s a cool way to do some math without a calculator.”
Sure, that might be handy some me, but your phone could probably give you
the answer. Rather, the point is to say that tangent lines are a good way of
approxima ng, and many scien sts, engineers and mathema cians o en face
problems too hard to solve directly. So they approximate.
Higher Order Deriva ves
The deriva ve of a func on f is itself a func on, therefore we can take its
deriva ve. The following defini on gives a name to this concept and introduces
its nota on.
Defini on 2.3.1 Higher Order Deriva ves

Let y = f(x) be a differen able func on on I. The following are defined,
provided the corresponding limits exist.
1. The second deriva ve of f is:
( )
′′ d( ′ ) d dy d 2y Note: The second deriva ve nota on
f (x) = f (x) = = 2 = y ′′ . could be wri en as
dx dx dx dx
d 2y d 2y d2 ( )
2. The third deriva ve of f is: = = y .
dx2 (dx)2 (dx)2
( )
′′′ d ( ′′ ) d d 2y d 3y That is, we take the deriva ve of y twice
f (x) = f (x) = = = y ′′′ . (hence d 2 ), both mes with respect to x
dx dx dx2 dx3
(hence (dx)2 = dx2 ).
3. The nth deriva ve of f is:

( )
(n) d ( (n−1) ) d d n−1 y d ny
f (x) = f (x) = = = y(n) .
dx dx dxn−1 dxn
In general, when finding the fourth deriva ve and on, we resort to the f (4) (x)
Notes:
85
nota on, not f ′′′′ (x); a er a while, too many cks is confusing.
Let’s prac ce using this new concept.
Example 2.3.3 Finding higher order deriva ves

Find the first four deriva ves of the following func ons:
1. f(x) = 4x2 3. f(x) = 5ex
2. f(x) = sin x
1. Using the Power and Constant Mul ple Rules, we have: f ′ (x) = 8x. Con-
nuing on, we have
d( )
f ′′ (x) = 8x = 8; f ′′′ (x) = 0; f (4) (x) = 0.
dx
No ce how all successive deriva ves will also be 0.
2. We employ Theorem 2.3.1 repeatedly.
f ′ (x) = cos x; f ′′ (x) = − sin x; f ′′′ (x) = − cos x; f (4) (x) = sin x.
Note how we have come right back to f(x) again. (Can you quickly figure
what f (23) (x) is?)
3. Employing Theorem 2.3.1 and the Constant Mul ple Rule, we can see that
f ′ (x) = f ′′ (x) = f ′′′ (x) = f (4) (x) = 5ex .
Interpre ng Higher Order Deriva ves

What do higher order deriva ves mean? What is the prac cal interpreta-
on?
Our first answer is a bit wordy, but is technically correct and beneficial to
understand. That is,
The second deriva ve of a func on f is the rate of change of the rate
of change of f.
One way to grasp this concept is to let f describe a posi on func on. Then,
as stated in Key Idea 2.2.1, f ′ describes the rate of posi on change: velocity.
We now consider f ′′ , which describes the rate of velocity change. Sports car
Notes:
86
enthusiasts talk of how fast a car can go from 0 to 60 mph; they are bragging
about the accelera on of the car.
We started this chapter with amusement–park riders free–falling with posi-
on func on f(t) = −16t2 + 150. It is easy to compute f ′ (t) = −32t /s and
f ′′ (t) = −32 ( /s)/s. We may recognize this la er constant; it is the accelera-
on due to gravity. In keeping with the unit nota on introduced in the previous
sec on, we say the units are “feet per second per second.” This is usually short-
ened to “feet per second squared,” wri en as “ /s2 .”
It can be difficult to consider the meaning of the third, and higher order,
deriva ves. The third deriva ve is “the rate of change of the rate of change of
the rate of change of f.” That is essen ally meaningless to the unini ated. In
the context of our posi on/velocity/accelera on example, the third deriva ve
is the “rate of change of accelera on,” commonly referred to as “jerk.”
Make no mistake: higher order deriva ves have great importance even if
their prac cal interpreta ons are hard (or “impossible”) to understand. The
mathema cal topic of series makes extensive use of higher order deriva ves.
Notes:
87
Exercises 2.3
Terms and Concepts 20. f(x) = ln(5x2 )
d ( n) 21. f(t) = ln(17) + e2 + sin π/2

1. What is the name of the rule which states that x =
dx
nxn−1 , where n > 0 is an integer? 22. g(t) = (1 + 3t)2
d( )
2. What is ln x ? 23. g(x) = (2x − 5)3
dx
3. Give an example of a func on f(x) where f ′ (x) = f(x). 24. f(x) = (1 − x)3
4. Give an example of a func on f(x) where f ′ (x) = 0. 25. f(x) = (2 − 3x)2
5. The deriva ve rules introduced in this sec on explain how logb x

26. A property of logarithms is that loga x = , for all
to compute the deriva ve of which of the following func- logb a
ons? bases a, b > 0, ̸= 1.
3
• f(x) = 2 • j(x) = sin x cos x (a) Rewrite this iden ty when b = e, i.e., using loge x =
x 2 ln x, with a = 10.
• g(x) = 3x2 − x + 17 • k(x) = ex
√ (b) Use part (a) to find the deriva ve of y = log10 x.
• h(x) = 5 ln x • m(x) = x
(c) Use part (b) to find the deriva ve of y = loga x, for
any a > 0, ̸= 1.
6. Explain in your own words how to find the third deriva ve
of a func on f(x).
In Exercises 27 – 32, compute the first four deriva ves of the
′ ′′ given func on.
7. Give an example of a func on where f (x) ̸= 0 and f (x) =
0.
27. f(x) = x6
8. Explain in your own words what the second deriva ve
“means.” 28. g(x) = 2 cos x
9. If f(x) describes a posi on func on, then f ′ (x) describes 29. h(t) = t2 − et
what kind of func on? What kind of func on is f ′′ (x)?
30. p(θ) = θ4 − θ3
10. Let f(x) be a func on measured in pounds, where x is mea-
sured in feet. What are the units of f ′′ (x)? 31. f(θ) = sin θ − cos θ
32. f(x) = 1, 100

Problems
In Exercises 33 – 38, find the equa ons of the tangent and
In Exercises 11 – 26, compute the deriva ve of the given func- normal lines to the graph of the func on at the given point.
on.
33. f(x) = x3 − x at x = 1
11. f(x) = 7x − 5x + 7
2
34. f(t) = et + 3 at t = 0
12. g(x) = 14x3 + 7x2 + 11x − 29
35. g(x) = ln x at x = 1
13. m(t) = 9t5 − 18 t3 + 3t − 8
36. f(x) = 4 sin x at x = π/2
14. f(θ) = 9 sin θ + 10 cos θ
37. f(x) = −2 cos x at x = π/4
15. f(r) = 6er
38. f(x) = 2x + 3 at x = 5
16. g(t) = 10t4 − cos t + 7 sin t
17. f(x) = 2 ln x − x
Review
1 4 1 3 1 2
18. p(s) = s + s + s +s+1
4 3 2 39. Given that e0 = 1, approximate the value of e0.1 using the
tangent line to f(x) = ex at x = 0.
19. h(t) = et − sin t − cos t
88
2.4 The Product and Quo ent Rules

The previous sec on showed that, in some ways, deriva ves behave nicely. The
Constant Mul ple and Sum/Difference Rules established that the deriva ve of
f(x) = 5x2 + sin x was not complicated. We neglected compu ng the deriva ve
5x2
of things like g(x) = 5x2 sin x and h(x) = sin x on purpose; their deriva ves are
not as straigh orward. (If you had to guess what their respec ve deriva ves are,
you would probably guess wrong.) For these, we need the Product and Quo ent
Rules, respec vely, which are defined in this sec on.
We begin with the Product Rule.
Theorem 2.4.1 Product Rule

Let f and g be differen able func ons on an open interval I. Then fg is a
differen able func on on I, and
d( )
f(x)g(x) = f(x)g ′ (x) + f ′ (x)g(x).
dx
d( )
Important: f(x)g(x) ̸= f ′ (x)g ′ (x)! While this answer is simpler than
dx
the Product Rule, it is wrong.
We prac ce using this new rule in an example, followed by an example that
demonstrates why this theorem is true.
Example 2.4.1 Using the Product Rule

Use the Product Rule to compute the deriva ve of y = 5x2 sin x. Evaluate the y
deriva ve at x = π/2. 20
S To make our use of the Product Rule explicit, let’s set f(x) = 15
5x2 and g(x) = sin x. We easily compute/recall that f ′ (x) = 10x and g ′ (x) =
cos x. Employing the rule, we have 10
d( 2 )
5x sin x = 5x2 cos x + 10x sin x.
dx 5
At x = π/2, we have
( π )2 (π) π (π) . π π
x
y ′ (π/2) = 5 cos + 10 sin = 5π. 2
2 2 2 2
We graph y and its tangent line at x = π/2, which has a slope of 5π, in Figure Figure 2.4.1: A graph of y = 5x2 sin x and
2.4.1. While this does not prove that the Product Rule is the correct way to han- its tangent line at x = π/2.
dle deriva ves of products, it helps validate its truth. .
Notes:
89
We now inves gate why the Product Rule is true.
Example 2.4.2 A proof of the Product Rule

Use the defini on of the deriva ve to prove Theorem 2.4.1.
S By the limit defini on, we have
d( ) f(x + h)g(x + h) − f(x)g(x)

f(x)g(x) = lim .
dx h→0 h
We now do something a bit unexpected; add 0 to the numerator (so that nothing
is changed) in the form of −f(x+h)g(x)+f(x+h)g(x), then do some regrouping
as shown.
d( ) f(x + h)g(x + h) − f(x)g(x)

f(x)g(x) = lim (now add 0 to the numerator)
dx h→0 h
f(x + h)g(x + h) − f(x + h)g(x) + f(x + h)g(x) − f(x)g(x)
= lim (regroup)
h→0 h
( ) ( )
f(x + h)g(x + h) − f(x + h)g(x) + f(x + h)g(x) − f(x)g(x)
= lim
h→0 h
f(x + h)g(x + h) − f(x + h)g(x) f(x + h)g(x) − f(x)g(x)
= lim + lim (factor)
h→0 h h→0 h
g(x + h) − g(x) f(x + h) − f(x)
= lim f(x + h) + lim g(x) (apply limits)
h→0 h h→0 h
= f(x)g ′ (x) + f ′ (x)g(x).
It is o en true that we can recognize that a theorem is true through its proof
yet somehow doubt its applicability to real problems. In the following example,
we compute the deriva ve of a product of func ons in two ways to verify that
the Product Rule is indeed “right.”
Example 2.4.3 Exploring alternate deriva ve methods

Let y = (x2 + 3x + 1)(2x2 − 3x + 1). Find y ′ two ways: first, by expanding
the given product and then taking the deriva ve, and second, by applying the
Product Rule. Verify that both methods give the same answer.
S We first expand the expression for y; a li le algebra shows

that y = 2x4 + 3x3 − 6x2 + 1. It is easy to compute y ′ :
y ′ = 8x3 + 9x2 − 12x.
Notes:
90
Now apply the Product Rule.
y ′ = (x2 + 3x + 1)(4x − 3) + (2x + 3)(2x2 − 3x + 1)

( ) ( )
= 4x3 + 9x2 − 5x − 3 + 4x3 − 7x + 3
= 8x3 + 9x2 − 12x.
The uninformed usually assume that “the deriva ve of the product is the
product of the deriva ves.” Thus we are tempted to say that y ′ = (2x + 3)(4x −
3) = 8x2 + 6x − 9. Obviously this is not correct.
Example 2.4.4 Using the Product Rule with a product of three func ons
Let y = x3 ln x cos x. Find y ′ .
S We have a product of three func ons while the Product Rule

only specifies how to handle a product of two func ons. Our method of handling
this problem
( is to) simply group the la er two func ons together, and consider
y = x3 ln x cos x . Following the Product Rule, we have
( )′ ( )
y ′ = (x3 ) ln x cos x + 3x2 ln x cos x
( )′
To evaluate ln x cos x , we apply the Product Rule again:
( 1 ) ( )
= (x3 ) ln x(− sin x) + cos x + 3x2 ln x cos x
x
31
= x ln x(− sin x) + x cos x + 3x2 ln x cos x
3
x
Recognize the pa ern in our answer above: when applying the Product Rule to
a product of three func ons, there are three terms added together in the final
deriva ve. Each term contains only one deriva ve of one of the original func-
ons, and each func on’s deriva ve shows up in only one term. It is straigh or-
ward to extend this pa ern to finding the deriva ve of a product of 4 or more
func ons.
We consider one more example before discussing another deriva ve rule.
Example 2.4.5 Using the Product Rule

Find the deriva ves of the following func ons.
1. f(x) = x ln x
2. g(x) = x ln x − x.
Notes:
91
S Recalling that the deriva ve of ln x is 1/x, we use the Product

Rule to find our answers.
d( )
1. x ln x = x · 1/x + 1 · ln x = 1 + ln x.
dx
2. Using the result from above, we compute
d( )
x ln x − x = 1 + ln x − 1 = ln x.
dx
This seems significant; if the natural log func on ln x is an important func on (it
is), it seems worthwhile to know a func on whose deriva ve is ln x. We have
found one. (We leave it to the reader to find another; a correct answer will be
very similar to this one.)
We have learned how to compute the deriva ves of sums, differences, and
products of func ons. We now learn how to find the deriva ve of a quo ent of
func ons.
Theorem 2.4.2 Quo ent Rule

Let f and g be differen able func ons defined on an open interval I,
where g(x) ̸= 0 on I. Then f/g is differen able on I, and
( )
d f(x) g(x)f ′ (x) − f(x)g ′ (x)
= .
dx g(x) g(x)2
The Quo ent Rule is not hard to use, although it might be a bit tricky to re-
member. A useful mnemonic works as follows. Consider a frac on’s numerator
and denominator as “HI” and “LO”, respec vely. Then
( )
d HI LO· dHI – HI· dLO
= ,
dx LO LOLO
read “low dee high minus high dee low, over low low.” Said fast, that phrase can
roll off the tongue, making it easy to memorize. The “dee high” and “dee low”
parts refer to the deriva ves of the numerator and denominator, respec vely.
Let’s prac ce using the Quo ent Rule.
Example 2.4.6 Using the Quo ent Rule

5x2
Let f(x) = . Find f ′ (x).
sin x
Notes:
92
S Directly applying the Quo ent Rule gives:
( )
d 5x2 sin x · 10x − 5x2 · cos x
=
dx sin x sin2 x
10x sin x − 5x2 cos x
= .
sin2 x
The Quo ent Rule allows us to fill in holes in our understanding of deriva ves
of the common trigonometric func ons. We start with finding the deriva ve of
the tangent func on.
d
( )
Example 2.4.7 Using the Quo ent Rule to find dx tan x .
Find the deriva ve of y = tan x.
S At first, one might feel unequipped to answer this ques on.

But recall that tan x = sin x/ cos x, so we can apply the Quo ent Rule.
y
( )
d( ) d sin x 10
tan x =
dx dx cos x
5
cos x cos x − sin x(− sin x)
=
cos2 x
x
cos2 x + sin2 x − π2 − π2 π π
= 4 2
cos2 x −5
1
=
cos2 x −10
.
= sec2 x.
. 2.4.2: A graph of y = tan x along
Figure
with its tangent line at x = π/4.
This is a beau ful result. To confirm its truth, we can find the equa on of the
tangent line to y = tan x at x = π/4. The slope is sec2 (π/4) = 2; y = tan x,
along with its tangent line, is graphed in Figure 2.4.2.
We include this result in the following theorem about the deriva ves of the
trigonometric func ons. Recall we found the deriva ve of y = sin x in Example
2.1.7 and stated the deriva ve of the cosine func on in Theorem 2.3.1. The
deriva ves of the cotangent, cosecant and secant func ons can all be computed
directly using Theorem 2.3.1 and the Quo ent Rule.
Notes:
93
Theorem 2.4.3 Deriva ves of Trigonometric Func ons
d( ) d( )
1. sin x = cos x 2. cos x = − sin x
dx dx
d( ) d( )
3. tan x = sec2 x 4. cot x = − csc2 x
dx dx
d( ) d( )
5. sec x = sec x tan x 6. csc x = − csc x cot x
dx dx
To remember the above, it may be helpful to keep in mind that the deriva-
ves of the trigonometric func ons that start with “c” have a minus sign in them.

5x2
In Example 2.4.6 the deriva ve of f(x) = was found using the Quo ent
sin x
2 ′
Rule. Rewri ng f as f(x) = 5x csc x, find f using Theorem 2.4.3 and verify the
two answers are the same.

S We found in Example 2.4.6 that the f ′ (x) = .
sin2 x
′ 2
We now find f using the Product Rule, considering f as f(x) = 5x csc x.
d( 2 )
f ′ (x) = 5x csc x
dx
= 5x2 (− csc x cot x) + 10x csc x (now rewrite trig func ons)
−1 cos x 10x
= 5x2 · · +
sin x sin x sin x
−5x2 cos x 10x
= 2 + (get common denominator)
sin x sin x
=
sin2 x
Finding f ′ using either method returned the same result. At first, the answers
looked different, but some algebra verified they are the same. In general, there
is not one final form that we seek; the immediate result from the Product Rule
is fine. Work to “simplify” your results into a form that is most readable and
useful to you.
The Quo ent Rule gives other useful results, as shown in the next example.
Notes:
94
Example 2.4.9 Using the Quo ent Rule to expand the Power Rule
1
1. f(x) =
x
1
2. f(x) = , where n > 0 is an integer.
xn
S We employ the Quo ent Rule.

x·0−1·1 1
1. f ′ (x) = = − 2.
x2 x
xn · 0 − 1 · nxn−1 nxn−1 n
2. f ′ (x) = n 2
= − 2n
= − n+1 .
(x ) x x
1
The deriva ve of y = n turned out to be rather nice. It gets be er. Con-
x
sider:
( )
d 1 d ( −n )
= x (apply result from Example 2.4.9)
dx xn dx
n
= − n+1 (rewrite algebraically)
x
= −nx−(n+1)
= −nx−n−1 .
This is reminiscent of the Power Rule: mul ply by the power, then subtract 1
from the power. We now add to our previous Power Rule, which had the re-
stric on of n > 0.
Theorem 2.4.4 Power Rule with Integer Exponents

Let f(x) = xn , where n ̸= 0 is an integer. Then
f ′ (x) = n · xn−1 .
Taking the deriva ve of many func ons is rela vely straigh orward. It is
clear (with prac ce) what rules apply and in what order they should be applied.
Other func ons present mul ple paths; different rules may be applied depend-
ing on how the func on is treated. One of the beau ful things about calculus
is that there is not “the” right way; each path, when applied correctly, leads to
Notes:
95
the same result, the deriva ve. We demonstrate this concept in an example.

x2 − 3x + 1
Let f(x) = . Find f ′ (x) in each of the following ways:
x
1. By applying the Quo ent Rule,
( )
2. by viewing f as f(x) = x2 − 3x + 1 · x−1 and applying the Product and
Power Rules, and
3. by “simplifying” first through division.
Verify that all three methods give the same result.
1. Applying the Quo ent Rule gives:

( ) ( )
′ x · 2x − 3 − x2 − 3x + 1 · 1 x2 − 1 1
f (x) = 2
= 2
= 1 − 2.
x x x
2. By rewri ng f, we can apply the Product and Power Rules as follows:

( ) ( )
f ′ (x) = x2 − 3x + 1 · (−1)x−2 + 2x − 3 · x−1
x2 − 3x + 1 2x − 3
=− +
x2 x
x − 3x + 1
2
2x2 − 3x
=− +
x2 x2
x2 − 1 1
= =1− ,
x2 x2
the same result as above.
1
̸ 0, we can divide through by x first, giving f(x) = x − 3 + . Now
3. As x =
x
apply the Power Rule.
1
f ′ (x) = 1 − 2 ,
x
the same result as before.
Example 2.4.10 demonstrates three methods of finding f ′ . One is hard pressed

to argue for a “best method” as all three gave the same result without too much
difficulty, although it is clear that using the Product Rule required more steps.
Ul mately, the important principle to take away from this is: reduce the answer
Notes:
96
to a form that seems “simple” and easy to interpret. In that example, we saw
different expressions for f ′ , including:
( ) ( )
1 x · 2x − 3 − x2 − 3x + 1 · 1 ( ) ( )
1− 2
= 2
= x2 − 3x + 1 · (−1)x−2 + 2x − 3 · x−1 .
x x
They are equal; they are all correct; only the first is “clear.” Work to make an-
swers clear.
In the next sec on we con nue to learn rules that allow us to more easily
compute deriva ves than using the limit defini on directly. We have to memo-
rize the deriva ves of a certain set of func ons, such as “the deriva ve of sin x
is cos x.” The Sum/Difference, Constant Mul ple, Power, Product and Quo ent
Rules show us how to find the deriva ves of certain combina ons of these func-
ons. The next sec on shows how to find the deriva ves when we compose
these func ons together.
Notes:
97
Exercises 2.4
Terms and Concepts In Exercises 15 – 36, compute the deriva ve of the given func-
on.
d( 2 )
1. T/F: The Product Rule states that x sin x = 2x cos x.
dx 15. f(x) = x sin x
( )
d x2 cos x 16. f(x) = x2 cos x
2. T/F: The Quo ent Rule states that = .
dx sin x 2x
17. f(x) = ex ln x
3. T/F: The deriva ves of the trigonometric func ons that
start with “c” have minus signs in them.
1
18. f(t) = (csc t − 4)
t2
4. What deriva ve rule is used to extend the Power Rule to
include nega ve integer exponents? x+7
19. g(x) =
x−5
5. T/F: Regardless of the func on, there is always exactly one
right way of compu ng its deriva ve. t5
20. g(t) =
cos t − 2t2
6. In your own words, explain what it means to make your an-
swers “clear.” 21. h(x) = cot x − ex
( )
22. f(x) = tan x ln x
Problems
23. h(t) = 7t2 + 6t − 2
In Exercises 7 – 10:
(a) Use the Product Rule to differen ate the func on. x4 + 2x3
24. f(x) =
x+2
(b) Manipulate the func on algebraically and differen -
ate without the Product Rule. ( )
25. f(x) = 3x2 + 8x + 7 ex
(c) Show that the answers from (a) and (b) are equivalent.
t5 − t3
7. f(x) = x(x2 + 3x) 26. g(t) =
et
8. g(x) = 2x2 (5x3 ) 7x − 1

27. f(x) = (16x3 + 24x2 + 3x)
16x3 + 24x2 + 3x
9. h(s) = (2s − 1)(s + 4)
28. f(t) = t5 (sec t + et )
10. f(x) = (x + 5)(3 − x )
2 3
sin x
29. f(x) =
In Exercises 11 – 14: cos x + 3
(a) Use the Quo ent Rule to differen ate the func on. sin θ
30. f(θ) = θ3 sin θ +
(b) Manipulate the func on algebraically and differen - θ3
ate without the Quo ent Rule.
cos x x
31. f(x) = +
(c) Show that the answers from (a) and (b) are equivalent. x tan x
( )
x2 + 3 32. g(x) = e2 sin(π/4) − 1
11. f(x) =
x
33. g(t) = 4t3 et − sin t cos t
x3 − 2x2
12. g(x) =
2x2 t2 sin t + 3
34. h(t) =
t2 cos t + 2
3
13. h(s) =
4s3
35. f(x) = x2 ex tan x
t2 − 1
14. f(t) = 36. g(x) = 2x sin x sec x
t+1
98
In Exercises 37 – 40, find the equa ons of the tangent and Review
normal lines to the graph of g at the indicated point.
In Exercises 49 – 52, use the graph of f(x) to sketch f ′ (x).
37. g(s) = es (s2 + 2) at (0, 2). y

6
4
38. g(t) = t sin t at ( 3π
2
, − 3π
2
)
2
49. x
x2 −3 −2 −1 1 2 3
39. g(x) = at (2, 4) −2
x−1
−4
. −6
cos θ − 8θ
40. g(θ) = at (0, 1)
θ+1 .
y
6
In Exercises 41 – 44, find the x–values where the graph of the 4
func on has a horizontal tangent line. 2

50. x
−3 −2 −1 1 2 3
−2
41. f(x) = 6x2 − 18x − 24
−4
. −6
42. f(x) = x sin x on [−1, 1]
y
x 6
43. f(x) =
x+1 4
2
x2 51. x
44. f(x) = −2 −1 1 2 3 4 5
x+1 −2
−4
In Exercises 45 – 48, find the requested deriva ve. . −6
y
45. f(x) = x sin x; find f ′′ (x). 10
5
46. f(x) = x sin x; find f (4) (x).
52. x
−5 5
′′
47. f(x) = csc x; find f (x). −5
−10
48. f(x) = (x3 − 5x + 2)(x2 + x − 7); find f (8) (x). .
99
2.5 The Chain Rule

We have covered almost all of the deriva ve rules that deal with combina ons
of two (or more) func ons. The opera ons of addi on, subtrac on, mul plica-
on (including by a constant) and division led to the Sum and Difference rules,
the Constant Mul ple Rule, the Power Rule, the Product Rule and the Quo ent
Rule. To complete the list of differen a on rules, we look at the last way two (or
more) func ons can be combined: the process of composi on (i.e. one func on
“inside” another).
One example of a composi on of func ons is f(x) = cos(x2 ). We currently
do not know how to compute this deriva ve. If forced to guess, one would likely
guess f ′ (x) = − sin(2x), where we recognize − sin x as the deriva ve of cos x
and 2x as the deriva ve of x2 . However, this is not the case; f ′ (x) ̸= − sin(2x).
In Example 2.5.4 we’ll see the correct answer, which employs the new rule this
sec on introduces, the Chain Rule.
Before we define this new rule, recall the nota on for composi on of func-
ons. We write (f ◦ g)(x) or f(g(x)), read as “f of g of x,” to denote composing f
with g. In shorthand, we simply write f ◦ g or f(g) and read it as “f of g.” Before
giving the corresponding differen a on rule, we note that the rule extends to
mul ple composi ons like f(g(h(x))) or f(g(h(j(x)))), etc.
To mo vate the rule, let’s look at three deriva ves we can already compute.
Example 2.5.1 Exploring similar deriva ves

Find the deriva ves of F1 (x) = (1 − x)2 , F2 (x) = (1 − x)3 , and F3 (x) = (1 −
x)4 . (We’ll see later why we are using subscripts for different func ons and an
uppercase F.)
S In order to use the rules we already have, we must first ex-

pand each func on as F1 (x) = 1 − 2x + x2 , F2 (x) = 1 − 3x + 3x2 − x3 and
F3 (x) = 1 − 4x + 6x2 − 4x3 + x4 .
It is not hard to see that:
F ′1 (x) = −2 + 2x,
F ′2 (x) = −3 + 6x − 3x2 and
F ′3 (x) = −4 + 12x − 12x2 + 4x3 .
An interes ng fact is that these can be rewri en as
F ′1 (x) = −2(1 − x), F ′2 (x) = −3(1 − x)2 and F ′3 (x) = −4(1 − x)3 .
A pa ern might jump out at you; note how the we end up mul plying by the old
power and the new power is reduced by 1. We also always mul ply by (−1).
Notes:
100
2.5 The Chain Rule
Recognize that each of these func ons is a composi on, le ng g(x) = 1−x:
F1 (x) = f1 (g(x)), where f1 (x) = x2 ,

F2 (x) = f2 (g(x)), where f2 (x) = x3 ,
F3 (x) = f3 (g(x)), where f3 (x) = x4 .
We’ll come back to this example a er giving the formal statements of the
Chain Rule; for now, we are just illustra ng a pa ern.
When composing func ons, we need to make√sure that the new func on is
actually defined. For instance, consider f(x) = x and g(x) = −x2 − 1. The
domain of f excludes all nega ve numbers,
( but) the √range of g is only nega ve
numbers. Therefore the composi on f g(x) = −x2 − 1 is not defined for
any x, and hence is not differen able.
The following defini on takes care to ensure this problem does not arise.
We’ll focus more on the deriva ve result than on the domain/range condi ons.
Theorem 2.5.1 The Chain Rule

Let g be a differen able func on on an interval I, let the range of g be a
subset of the interval J, and let f be a differen able func on on J. Then
y = f(g(x)) is a differen able func on on I, and
y ′ = f ′ (g(x)) · g ′ (x).
To help understand the Chain Rule, we return to Example 2.5.1.
Example 2.5.2 Using the Chain Rule

Use the Chain Rule to find the deriva ves of the following func ons, as given in
Example 2.5.1.
S Example 2.5.1 ended with the recogni on that each of the

given func ons was actually a composi on of func ons. To avoid confusion, we
ignore most of the subscripts here.
F1 (x) = (1 − x)2 :
We found that
y = (1 − x)2 = f(g(x)), where f(x) = x2 and g(x) = 1 − x.
To find y ′ , we apply the Chain Rule. We need f ′ (x) = 2x and g ′ (x) = −1.
Notes:
101
Part of the Chain Rule uses f ′ (g(x)). This means subs tute g(x) for x in the
equa on for f ′ (x). That is, f ′ (x) = 2(1 − x). Finishing out the Chain Rule we
have
y ′ = f ′ (g(x)) · g ′ (x) = 2(1 − x) · (−1) = −2(1 − x) = 2x − 2.
F2 (x) = (1 − x)3 :
Let y = (1 − x)3 = f(g(x)), where f(x) = x3 and g(x) = (1 − x). We have

f (x) = 3x2 , so f ′ (g(x)) = 3(1 − x)2 . The Chain Rule then states
′
y ′ = f ′ (g(x)) · g ′ (x) = 3(1 − x)2 · (−1) = −3(1 − x)2 .
F3 (x) = (1 − x)4 :
Finally, when y = (1 − x)4 , we have f(x) = x4 and g(x) = (1 − x). Thus

f (x) = 4x3 and f ′ (g(x)) = 4(1 − x)3 . Thus
′
y ′ = f ′ (g(x)) · g ′ (x) = 4(1 − x)3 · (−1) = −4(1 − x)3 .
Example 2.5.2 demonstrated a par cular pa ern: when f(x) = xn , then

y = n · (g(x))n−1 · g ′ (x). This is called the Generalized Power Rule.
′
Theorem 2.5.2 Generalized Power Rule

Let g(x) be a differen able func on and let n ̸= 0 be an integer. Then
d( ) ( )n−1 ′
g(x)n = n · g(x) · g (x).
dx
This allows us to quickly find the deriva ve of func ons like y = (3x2 − 5x +
7 + sin x)20 . While it may look in mida ng, the Generalized Power Rule states
that
y ′ = 20(3x2 − 5x + 7 + sin x)19 · (6x − 5 + cos x).
Treat the deriva ve–taking process step–by–step. In the example just given,
first mul ply by 20, then rewrite the inside of the parentheses, raising it all to
the 19th power. Then think about the deriva ve of the expression inside the
parentheses, and mul ply by that.
We now consider more examples that employ the Chain Rule.
Notes:
102
2.5 The Chain Rule
Example 2.5.3 Using the Chain Rule

Find the deriva ves of the following func ons:
1. y = sin 2x 2. y = ln(4x3 − 2x2 ) 3. y = e−x

2
S
1. Consider y = sin 2x. Recognize that this is a composi on of func ons,
where f(x) = sin x and g(x) = 2x. Thus
y ′ = f ′ (g(x)) · g ′ (x) = cos(2x) · 2 = 2 cos 2x.
2. Recognize that y = ln(4x3 − 2x2 ) is the composi on of f(x) = ln x and

g(x) = 4x3 − 2x2 . Also, recall that
d( ) 1
ln x = .
dx x
This leads us to:
1 12x2 − 4x 4x(3x − 1) 2(3x − 1)
y′ = · (12x 2
− 4x) = = = .
4x3 − 2x2 4x3 − 2x2 2x(2x2 − x) 2x2 − x
2
3. Recognize that y = e−x is the composi on of f(x) = ex and g(x) = −x2 .
Remembering that f ′ (x) = ex , we have
2 2
y ′ = e−x · (−2x) = (−2x)e−x .
Example 2.5.4 Using the Chain Rule to find a tangent line y

Let f(x) = cos x2 . Find the equa on of the line tangent to the graph of f at x = 1. 1
S The tangent line goes through the point (1, f(1)) ≈ (1, 0.54)
with slope f ′ (1). To find f ′ , we need the Chain Rule. 0.5
f ′ (x) = − sin(x2 ) · (2x) = −2x sin x2 . Evaluated at x = 1, we have f ′ (1) =

−2 sin 1 ≈ −1.68. Thus the equa on of the tangent line is x
−2 2
y = −1.68(x − 1) + 0.54.
−0.5
The tangent line is sketched along with f in Figure 2.5.1.
The Chain Rule is used o en in taking deriva ves. Because of this, one can −1
.
become familiar with the basic process and learn pa erns that facilitate finding
deriva ves quickly. For instance, Figure 2.5.1: f(x) = cos x2 sketched along
with its tangent line at x = 1.
d( ) 1 (anything)′
ln(anything) = · (anything)′ = .
dx anything anything
Notes:
103
A concrete example of this is
d( ) 45x14 + sin x + ex
ln(3x15 − cos x + ex ) = 15 .
dx 3x − cos x + ex
While the deriva ve may look in mida ng at first, look for the pa ern. The
denominator is the same as what was inside the natural log func on; the nu-
merator is simply its deriva ve.
This pa ern recogni on process can be applied to lots of func ons. In gen-
eral, instead of wri ng “anything”, we use u as a generic func on of x. We then
say
d( ) u′
ln u = .
dx u
The following is a short list of how the Chain Rule can be quickly applied to fa-
miliar func ons.
d ( n) d( )
1. u = n · un−1 · u ′ . 4. cos u = −u ′ · sin u.
dx dx
d ( u) d( )
2. e = u ′ · eu . 5. tan u = u ′ · sec2 u.
dx dx
d( )
3. sin u = u ′ · cos u.
dx
Of course, the Chain Rule can be applied in conjunc on with any of the other
rules we have already learned. We prac ce this next.
Example 2.5.5 Using the Product, Quo ent and Chain Rules
5x3
1. f(x) = x5 sin 2x3 2. f(x) = .
e−x2
S
1. We must use the Product and Chain Rules. Do not think that you must be
able to “see” the whole answer immediately; rather, just proceed step–
by–step.
( ) ( )
f ′ (x) = x5 6x2 cos 2x3 + 5x4 sin 2x3 = 6x7 cos 2x3 + 5x4 sin 2x3 .
2. We must employ the Quo ent Rule along with the Chain Rule. Again, pro-
Notes:
104
2.5 The Chain Rule
ceed step–by–step.
2( ) ( 2) 2( )
′ e−x 15x2 − 5x3 (−2x)e−x e−x 10x4 + 15x2
f (x) = ( 2 )2 =
e−x e−2x2
2( )
= ex 10x4 + 15x2 .
A key to correctly working these problems is to break the problem down

into smaller, more manageable pieces. For instance, when using the Product
and Chain Rules together, just consider the first part of the Product Rule at first:
f(x)g ′ (x). Just rewrite f(x), then find g ′ (x). Then move on to the f ′ (x)g(x) part.
Don’t a empt to figure out both parts at once.
Likewise, using the Quo ent Rule, approach the numerator in two steps and
handle the denominator a er comple ng that. Only simplify a erward.
We can also employ the Chain Rule itself several mes, as shown in the next
example.
Example 2.5.6 Using the Chain Rule mul ple mes

Find the deriva ve of y = tan5 (6x3 − 7x).
S Recognize that we have the g(x) = tan(6x3 − 7x) func on

( )5
“inside” the f(x) = x5 func on; that is, we have y = tan(6x3 −7x) . We begin
using the Generalized Power Rule; in this first step, we do not fully compute the
deriva ve. Rather, we are approaching this step–by–step.
( )4
y ′ = 5 tan(6x3 − 7x) · g ′ (x).
We now find g ′ (x). We again need the Chain Rule;
g ′ (x) = sec2 (6x3 − 7x) · (18x2 − 7).
Combine this with what we found above to give

( )4
y ′ = 5 tan(6x3 − 7x) · sec2 (6x3 − 7x) · (18x2 − 7)
= (90x2 − 35) sec2 (6x3 − 7x) tan4 (6x3 − 7x).
This func on is frankly a ridiculous func on, possessing no real prac cal
value. It is very difficult to graph, as the tangent func on has many ver cal
asymptotes and 6x3 − 7x grows so very fast. The important thing to learn from
this is that the deriva ve can be found. In fact, it is not “hard;” one can take
several simple steps and should be careful to keep track of how to apply each of
these steps.
Notes:
105
It is a tradi onal mathema cal exercise to find the deriva ves of arbitrarily
complicated func ons just to demonstrate that it can be done. Just break every-
thing down into smaller pieces.
Example 2.5.7 Using the Product, Quo ent and Chain Rules
x cos(x−2 ) − sin2 (e4x )
Find the deriva ve of f(x) = .
ln(x2 + 5x4 )
S This func on likely has no prac cal use outside of demon-

stra ng deriva ve skills. The answer is given below without simplifica on. It
employs the Quo ent Rule, the Product Rule, and the Chain Rule three mes.
f ′ (x) =
 [( ) ] 
ln(x2 + 5x4 ) · x · (− sin(x−2 )) · (−2x−3 ) + 1 · cos(x−2 ) − 2 sin(e4x ) · cos(e4x ) · (4e4x )
 ( ) 3

− x cos(x−2 ) − sin2 (e4x ) · 2x+20x
x2 +5x4
( ) .
ln(x2 + 5x4 ) 2
The reader is highly encouraged to look at each term and recognize why it
is there. (I.e., the Quo ent Rule is used; in the numerator, iden fy the “LOdHI”
term, etc.) This example demonstrates that deriva ves can be computed sys-
tema cally, no ma er how arbitrarily complicated the func on is.
The Chain Rule also has theore c value. That is, it can be used to find the
deriva ves of func ons that we have not yet learned as we do in the following
example.
Example 2.5.8 The Chain Rule and exponen al func ons

Use the Chain Rule to find the deriva ve of y = 2x .
S We only know how to find the deriva ve of one exponen al

func on, y = ex . We can accomplish our goal by rewri ng 2 in terms of e.
Recalling that ex and ln x are inverse func ons, we can write
( )x
2 = eln 2 and so y = 2x = eln 2 = ex(ln 2) .
The func on is now the composi on y = f(g(x)), with f(x) = ex and g(x) =
x(ln 2). Since f ′ (x) = ex and g ′ (x) = ln 2, the Chain Rule gives
y ′ = ex(ln 2) · ln 2.
Recall that the ex(ln 2) term on the right hand side is just 2x , our original func on.
Thus, the deriva ve contains the original func on itself. We have
y ′ = y · ln 2 = 2x · ln 2.
Notes:
106
2.5 The Chain Rule
We can extend this process to use any base a, where a > 0 and a ̸= 1. All we
need to do is replace each “2” in our work with “a.” The Chain Rule, coupled
with the deriva ve rule of ex , allows us to find the deriva ves of all exponen al
func ons.
The comment at the end of previous example is important and is restated

formally as a theorem.
Theorem 2.5.3 Deriva ves of Exponen al Func ons

Let f(x) = ax , for a > 0, a ̸= 1. Then f is differen able for all real
numbers (i.e., differen able everywhere) and
f ′ (x) = ln a · ax .
Alternate Chain Rule Nota on
It is instruc ve to understand what the Chain Rule “looks like” using “ dy

dx ” no-
ta on instead of y ′ nota on. Suppose that y = f(u) is a func on of u, where
u = g(x) is a func on of x, as stated in Theorem 2.5.1. Then, through the com-
posi on f ◦ g, we can think of y as a func on of x, as y = f(g(x)). Thus the
deriva ve of y with respect to x makes sense; we can talk about dy dx . This leads
to an interes ng progression of nota on:
y ′ = f ′ (g(x)) · g ′ (x)
dy
= y ′ (u) · u ′ (x) (since y = f(u) and u = g(x))
dx
dy dy du
= · (using “frac onal” nota on for the deriva ve)
dx du dx
Here the “frac onal” aspect of the deriva ve nota on stands out. On the
right hand side, it seems as though the “du” terms cancel out, leaving
dy dy
= .
dx dx
It is important to realize that we are not canceling these terms; the deriva ve
dy
nota on of du is one symbol. It is equally important to realize that this nota on
was chosen precisely because of this behavior. It makes applying the Chain Rule
easy with mul ple variables. For instance,
Notes:
107
.
dy dy d⃝ d△
= · · .
dt d⃝ d△ dt
where ⃝ and △ are any variables you’d like to use.
One of the most common ways of “visualizing” the Chain Rule is to consider
a set of gears, as shown in Figure 2.5.2. The gears have 36, 18, and 6 teeth,
respec vely. That means for every revolu on of the x gear, the u gear revolves
twice. That is, the rate at which the u gear makes a revolu on is twice as fast
x as the rate at which the x gear makes a revolu on. Using the terminology of
calculus, the rate of u-change, with respect to x, is du
dx = 2.
dy
Likewise, every revolu on of u causes 3 revolu ons of y: du = 3. How does
y change with respect to x? For each revolu on of x, y revolves 6 mes; that is,
dy
du =6
=2 dx dy dy du
dx = · = 2 · 3 = 6.
dx du dx
u y We can then extend the Chain Rule with more variables by adding more gears
dy to the picture.
=3
du
It is difficult to overstate the importance of the Chain Rule. So o en the
Figure 2.5.2: A series of gears to demon- func ons that we deal with are composi ons of two or more func ons, requir-
strate the Chain Rule. Note how dy dx
= ing us to use this rule to compute deriva ves. It is also o en used in real life
dy
du
· du
dx when actual func ons are unknown. Through measurement, we can calculate
dy
(or, approximate) du and du
dx . With our knowledge of the Chain Rule, we can find
dy
dx .
In the next sec on, we use the Chain Rule to jus fy another differen a on
technique. There are many curves that we can draw in the plane that fail the
“ver cal line test.” For instance, consider x2 + y2 = 1, which describes the unit
circle. We may s ll be interested in finding slopes of tangent lines to the circle at
various points. The next sec on shows how we can find dy dx without first “solving
for y.” While we can in this instance, in many other instances solving for y is
impossible. In these situa ons, implicit differen a on is indispensable.
Notes:
108
Exercises 2.5
Terms and Concepts 23. g(r) = 4r
1. T/F: The Chain Rule describes how to evaluate the deriva- 24. g(t) = 5cos t
ve of a composi on of func ons.
d( ) 25. g(t) = 152

2. T/F: The Generalized Power Rule states that g(x)n =
( )n−1 dx
n g(x) . 3w
26. m(w) =
2w
d( ) 1
3. T/F: ln(x2 ) = 2 .
dx x 2t + 3
27. h(t) =
3t + 2
d ( x)
4. T/F: 3 ≈ 1.1 · 3x .
dx 3w + 1
28. m(w) =
dx dx dt 2w
5. T/F: = ·
dy dt dy 2
3x + x
( ) 29. f(x) =
6. f(x) = ln x + x 2 3 2x2
30. f(x) = x2 sin(5x)

Problems
31. f(x) = (x2 + x)5 (3x4 + 2x)3
In Exercises 7 – 36, compute the deriva ve of the given func-
on. 32. g(t) = cos(t2 + 3t) sin(5t − 7)
7. f(x) = (4x3 − x)10

33. f(x) = sin(3x + 4) cos(5 − 2x)
8. f(t) = (3t − 2)5 2
34. g(t) = cos( 1t )e5t
9. g(θ) = (sin θ + cos θ)3
( )
sin 4x + 1
2 35. f(x) =
10. h(t) = e3t +t−1
(5x − 9)3
( )3
11. f(x) = ln x + x2 (4x + 1)2
36. f(x) =
tan(5x)
3
12. f(x) = 2x +3x
( )4 In Exercises 37 – 40, find the equa ons of tangent and normal

13. f(x) = x + 1x lines to the graph of the func on at the given point. Note: the
func ons here are the same as in Exercises 7 through 10.
14. f(x) = cos(3x)
37. f(x) = (4x3 − x)10 at x = 0
15. g(x) = tan(5x)
( )
16. h(θ) = tan θ2 + 4θ 38. f(t) = (3t − 2)5 at t = 1
( )
17. g(t) = sin t5 + 1t 39. g(θ) = (sin θ + cos θ)3 at θ = π/2
18. h(t) = sin4 (2t) 40. h(t) = e3t

2
+t−1
at t = −1
19. p(t) = cos3 (t2 + 3t + 1)

d( )
41. Compute ln(kx) two ways:
dx
20. f(x) = ln(cos x)
(a) Using the Chain Rule, and
21. f(x) = ln(x2 )
(b) by first using the logarithm rule ln(ab) = ln a + ln b,
22. f(x) = 2 ln(x) then taking the deriva ve.
109
d( )
chill factor, in degrees Fahrenheit, when it is 25◦ F outside
42. Compute ln(xk ) two ways:
dx with a wind of w mph.
(a) Using the Chain Rule, and
(b) by first using the logarithm rule ln(ap ) = p ln a, then (a) What are the units of W ′ (w)?
taking the deriva ve.
(b) What would you expect the sign of W ′ (10) to be?
Review 44. Find the deriva ves of the following func ons.
(a) f(x) = x2 ex cot x

43. The “wind chill factor” is a measurement of how cold it
“feels” during cold, windy weather. Let W(w) be the wind (b) g(x) = 2x 3x 4x
110
2.6 Implicit Differen a on

′
In the previous sec ons we learned to find the deriva ve, dy dx , or y , when y is
given explicitly as a func on of x. That is, if we know y = f(x) for some func on
f, we can find y ′ . For example, given y = 3x2 − 7, we can easily find y ′ = 6x.
(Here we explicitly state how x and y are related. Knowing x, we can directly find
y.)
Some mes the rela onship between y and x is not explicit; rather, it is im- y
plicit. For instance, we might know that x2 − y = 4. This equality defines a
rela onship between x and y; if we know x, we could figure out y. Can we s ll 2
find y ′ ? In this case, sure; we solve for y to get y = x2 − 4 (hence we now know
y explicitly) and then differen ate to get y ′ = 2x.
Some mes the implicit rela onship between x and y is complicated. Sup- x
pose we are given sin(y) + y3 = 6 − x3 . A graph of this implicit func on is given −2 2
in Figure 2.6.1. In this case there is absolutely no way to solve for y in terms of
elementary func ons. The surprising thing is, however, that we can s ll find y ′ −2
via a process known as implicit differen a on. .
Implicit differen a on is a technique based on the Chain Rule that is used to
find a deriva ve when the rela onship between the variables is given implicitly Figure 2.6.1: A graph of the implicit func-
rather than explicitly (solved for one variable in terms of the other). on sin(y) + y3 = 6 − x3 .
We begin by reviewing the Chain Rule. Let f and g be func ons of x. Then
d( )
f(g(x)) = f ′ (g(x)) · g′ (x).
dx
Suppose now that y = g(x). We can rewrite the above as
d( ) d( ) dy
f(y) = f ′ (y) · y ′ , or f(y) = f ′ (y) · . (2.1)
dx dx dx
These equa ons look strange; the key concept to learn here is that we can find
y ′ even if we don’t exactly know how y and x relate.
We demonstrate this process in the following example.
Example 2.6.1 Using Implicit Differen a on

Find y ′ given that sin(y) + y3 = 6 − x3 .
S We start by taking the deriva ve of both sides (thus main-

taining the equality.) We have :
d( ) d( )
sin(y) + y3 = 6 − x3 .
dx dx
Notes:
111
The right hand side is easy; it returns −3x2 .

The le hand side requires more considera on. We take the deriva ve term–
by–term. Using the technique derived from Equa on 2.1 above, we can see that
d( )
sin y = cos y · y ′ .
dx
We apply the same process to the y3 term.
d ( 3) d ( 3)
y = (y) = 3(y)2 · y ′ .
dx dx
Pu ng this together with the right hand side, we have
cos(y)y ′ + 3y2 y ′ = −3x2 .
Now solve for y ′ .
cos(y)y ′ + 3y2 y ′ = −3x2 .
( )
cos y + 3y2 y ′ = −3x2
−3x2
y′ =
cos y + 3y2
This equa on for y ′ probably seems unusual for it contains both x and y
terms. How is it to be used? We’ll address that next.
Implicit func ons are generally harder to deal with than explicit func ons.
With an explicit func on, given an x value, we have an explicit formula for com-
pu ng the corresponding y value. With an implicit func on, one o en has to
find x and y values at the same me that sa sfy the equa on. It is much eas-
ier to demonstrate that a given point sa sfies the equa on than to actually find
such a point. √
For instance, we can affirm easily that the point ( 3 6, 0) lies on the graph of
the implicit func on sin√y + y3 = 6 − x3 . Plugging in 0 for y, we see the le hand
side is 0. Se ng x = 3 6, we see the right hand side is also 0; the equa on is
sa sfied. The following example finds the equa on of the tangent line to this
func on at this point.
Example 2.6.2 Using Implicit Differen a on to find a tangent line

Find the equa on of the line tangent to
√ the curve of the implicitly defined func-
on sin y + y3 = 6 − x3 at the point ( 3 6, 0).
S In Example 2.6.1 we found that

−3x2
y′ = .
cos y + 3y2
Notes:
112
√ √
We find the slope of the tangent√ line at the point ( 3 6, 0) by subs tu ng 3 6 for y
x and 0 for y. Thus at the point ( 3 6, 0), we have the slope as
√ √ 2
′ −3( 3 6)2 −3 3 36
y = = ≈ −9.91.
cos 0 + 3 · 02 1
Therefore the equa on of the√tangent line to the implicitly defined func on x
−2 2
sin y + y3 = 6 − x3 at the point ( 3 6, 0) is
√ √
y = −3 36(x − 6) + 0 ≈ −9.91x + 18.
3 3
−2
The curve and this tangent line are shown in Figure 2.6.2. .
Figure 2.6.2: The func on sin y + y3 =
This suggests a general method for implicit differen a on. For the steps be- 6√− x3 and its tangent line at the point
low assume y is a func on of x. ( 3 6, 0).
1. Take the deriva ve of each term in the equa on. Treat the x terms like
normal. When taking the deriva ves of y terms, the usual rules apply
except that, because of the Chain Rule, we need to mul ply each term
by y ′ .
2. Get all the y ′ terms on one side of the equal sign and put the remaining
terms on the other side.
3. Factor out y ′ ; solve for y ′ by dividing.
Prac cal Note: When working by hand, it may be beneficial to use the symbol
dy ′ 1
dx instead of y , as the la er can be easily confused for y or y .

Given the implicitly defined func on y3 + x2 y4 = 1 + 2x, find y ′ .
S We will take the implicit deriva ves term by term. The deriva-
ve of y3 is 3y2 y ′ .
The second term, x2 y4 , is a li le tricky. It requires the Product Rule as it is the
product of two func ons of x: x2 and y4 . Its deriva ve is x2 (4y3 y ′ ) + 2xy4 . The
first part of this expression requires a y ′ because we are taking the deriva ve of a
y term. The second part does not require it because we are taking the deriva ve
of x2 .
The deriva ve of the right hand side is easily found to be 2. In all, we get:
3y2 y ′ + 4x2 y3 y ′ + 2xy4 = 2.
Move terms around so that the le side consists only of the y ′ terms and the
right side consists of all the other terms:
3y2 y ′ + 4x2 y3 y ′ = 2 − 2xy4 .
Notes:
113
y Factor out y ′ from the le side and solve to get
2 − 2xy4
x y′ = .
5 10 3y2 + 4x2 y3
To confirm the validity of our work, let’s find the equa on of a tangent line
−5 to this func on at a point. It is easy to confirm that the point (0, 1) lies on the
graph of this func on. At this point, y ′ = 2/3. So the equa on of the tangent
line is y = 2/3(x−0)+1. The func on and its tangent line are graphed in Figure
−10
2.6.3.
.
No ce how our func on looks much different than other func ons we have
Figure 2.6.3: A graph of the implicitly de- seen. For one, it fails the ver cal line test. Such func ons are important in many
fined func on y3 + x2 y4 = 1 + 2x along areas of mathema cs, so developing tools to deal with them is also important.
with its tangent line at the point (0, 1).
Given the implicitly defined func on sin(x2 y2 ) + y3 = x + y, find y ′ .
S Differen a ng term by term, we find the most difficulty in

the first term. It requires both the Chain and Product Rules.
d( ) d ( 2 2)
y sin(x2 y2 ) = cos(x2 y2 ) · x y
dx (dx )
= cos(x2 y2 ) · x2 (2yy ′ ) + 2xy2
1 = 2(x2 yy ′ + xy2 ) cos(x2 y2 ).
x We leave the deriva ves of the other terms to the reader. A er taking the
−1 1 deriva ves of both sides, we have
−1
2(x2 yy ′ + xy2 ) cos(x2 y2 ) + 3y2 y ′ = 1 + y ′ .
. We now have to be careful to properly solve for y ′ , par cularly because of
Figure 2.6.4: A graph of the implicitly de- the product on the le . It is best to mul ply out the product. Doing this, we get
fined func on sin(x2 y2 ) + y3 = x + y.
2x2 y cos(x2 y2 )y ′ + 2xy2 cos(x2 y2 ) + 3y2 y ′ = 1 + y ′ .
From here we can safely move around terms to get the following:
2x2 y cos(x2 y2 )y ′ + 3y2 y ′ − y ′ = 1 − 2xy2 cos(x2 y2 ).
Then we can solve for y ′ to get
1 − 2xy2 cos(x2 y2 )
y′ = .
+ 3y2 − 1
2x2 y cos(x2 y2 )
Notes:
114
A graph of this implicit func on is given in Figure 2.6.4. It is easy to verify y

that the points (0, 0), (0, 1) and (0, −1) all lie on the graph. We can find the
slopes of the tangent lines at each of these points using our formula for y ′ .
At (0, 0), the slope is −1. 1
At (0, 1), the slope is 1/2.

At (0, −1), the slope is also 1/2. x
−1 1
The tangent lines have been added to the graph of the func on in Figure
2.6.5. −1
Quite a few “famous” curves have equa ons that are given implicitly. We can
use implicit differen a on to find the slope at various points on those curves.
We inves gate two such curves in the next examples. Figure 2.6.5: A graph of the implicitly de-
fined func on sin(x2 y2 ) + y3 = x + y and
certain tangent lines.
Example 2.6.5 Finding slopes of tangent lines to a circle √
Find the slope of the tangent line to the circle x2 +y2 = 1 at the point (1/2, 3/2).
S Taking deriva ves, we get 2x+2yy ′ = 0. Solving for y ′ gives:

−x
y′ = .
y
y
This is a clever formula. Recall that the slope of the line through the origin and √
1 (1/2, 3/2)
the point (x, y) on the circle will be y/x. We have found that the slope of the
tangent line to the circle at that point is the opposite reciprocal of y/x, namely,
−x/y. Hence these two √ lines are always perpendicular.
At the point (1/2, 3/2), we have the tangent line’s slope as
x
−1/2 −1 −1 1
y′ = √ = √ ≈ −0.577.
3/2 3
√
A graph of the circle and its tangent line at (1/2, 3/2) is given in Figure
2.6.6, along with a thin dashed line from the origin that is perpendicular to the −1
tangent line. (It turns out that all normal lines to a circle pass through the center .
of the circle.) Figure 2.6.6: The √
unit circle with its tan-
gent line at (1/2, 3/2).
This sec on has shown how to find the deriva ves of implicitly defined func-
ons, whose graphs include a wide variety of interes ng and unusual shapes.
Implicit differen a on can also be used to further our understanding of “regu-
lar” differen a on.
One hole in our current understanding of deriva ves is this: what is the
deriva ve of the square root func on? That is,
d (√ ) d ( 1/2 )
x = x =?
dx dx
Notes:
115
We allude to a possible solu on, as we can write the square root func on as
a power func on with a ra onal (or, frac onal) power. We are then tempted to
apply the Power Rule and obtain
d ( 1/2 ) 1 −1/2 1
x = x = √ .
dx 2 2 x
The trouble with this is that the Power Rule was ini ally defined only for
posi ve integer powers, n > 0. While we did not jus fy this at the me, gen-
erally the Power Rule is proved using something called the Binomial Theorem,
which deals only with posi ve integers. The Quo ent Rule allowed us to extend
the Power Rule to nega ve integer powers. Implicit Differen a on allows us to
extend the Power Rule to ra onal powers, as shown below.
Let y = xm/n , where m and n are integers with no common factors (so m = 2
and n = 5 is fine, but m = 2 and n = 4 is not). We can rewrite this explicit
func on implicitly as yn = xm . Now apply implicit differen a on.
y = x m/n
yn = xm
d ( n) d ( m)
y = x
dx dx
n · y n−1 · y ′ = m · x m−1
m x m−1
y′ = (now subs tute x m/n for y)
n y n−1
m x m−1
= (apply lots of algebra)
n (x m/n ) n−1
m (m−n)/n
= x
n
m m/n−1
= x .
n
The above deriva on is the key to the proof extending the Power Rule to ra-
onal powers. Using limits, we can extend this once more to include all powers,
including irra onal (even transcendental!) powers, giving the following theo-
rem.
Theorem 2.6.1 Power Rule for Differen a on

Let f(x) = x , where n ̸= 0 is a real number. Then f is differen able on
n
its domain, except possibly at x = 0, and f ′ (x) = n · x n−1 .
Notes:
116
20
This theorem allows us to say the deriva ve of xπ is πxπ−1 .
We now apply this final version of the Power Rule in the next example, the
second inves ga on of a “famous” curve.
x
−20 20
Example 2.6.6 Using the Power Rule
Find the slope of x2/3 + y2/3 = 8 at the point (8, 8).
S This is a par cularly interes ng curve called an astroid. It −20

is the shape traced out by a point on the edge of a circle that is rolling around .
inside of a larger circle, as shown in Figure 2.6.7. Figure 2.6.7: An astroid, traced out by a
To find the slope of the astroid at the point (8, 8), we take the deriva ve point on the smaller circle as it rolls inside
implicitly. the larger circle.
2 −1/3 2 −1/3 ′ y
x + y y =0
3 3 20
2 −1/3 ′ 2
y y = − x−1/3
3 3 (8, 8)
x−1/3
y ′ = − −1/3
y
√ x
y1/3 y −20 20
y ′ = − 1/3 = − 3 .
x x
−20
Plugging in x = 8 and y = 8, we get a slope of −1. The astroid, with its
.
tangent line at (8, 8), is shown in Figure 2.6.8.
Figure 2.6.8: An astroid with a tangent
line.
Implicit Differen a on and the Second Deriva ve
We can use implicit differen a on to find higher order deriva ves. In theory,
this is simple: first find dy
dx , then take its deriva ve with respect to x. In prac ce,
it is not hard, but it o en requires a bit of algebra. We demonstrate this in an
example.
Example 2.6.7 Finding the second deriva ve

d 2y
Given x + y = 1, find 2 = y ′′ .
2 2
dx
S We found that y ′ = dy
dx = −x/y in Example 2.6.5. To find y ′′ ,
Notes:
117
we apply implicit differen a on to y ′ .
d ( ′)
y ′′ = y
dx ( )
d x
= − (Now use the Quo ent Rule.)
dx y
y(1) − x(y ′ )
=−
y2
replace y ′ with −x/y:
y − x(−x/y)
=−
y2
y + x2 /y
=− .
y2
While this is not a par cularly simple expression, it is usable. We can see that
y ′′ > 0 when y < 0 and y ′′ < 0 when y > 0. In Sec on 3.4, we will see how
this relates to the shape of the graph.
y
Logarithmic Differen a on
4
3 Consider the func on y = xx ; it is graphed in Figure 2.6.9. It is well–defined

for x > 0 and we might be interested in finding equa ons of lines tangent and
2 normal to its graph. How do we take its deriva ve?
The func on is not a power func on: it has a “power” of x, not a constant.
1 It is not an exponen al func on: it has a “base” of x, not a constant.
A differen a on technique known as logarithmic differen a on becomes
. x
1 2 useful here. The basic principle is this: take the natural log of both sides of an
equa on y = f(x), then use implicit differen a on to find y ′ . We demonstrate
Figure 2.6.9: A plot of y = xx . this in the following example.
Example 2.6.8 Using Logarithmic Differen a on

Given y = xx , use logarithmic differen a on to find y ′ .
S As suggested above, we start by taking the natural log of
Notes:
118
both sides then applying implicit differen a on. y

4
y = xx
ln(y) = ln(xx ) (apply logarithm rule) 3
ln(y) = x ln x (now use implicit differen a on)

d( ) d( ) 2
ln(y) = x ln x
dx dx
y′ 1 1
= ln x + x ·
y x
. x
y′ 1 2
= ln x + 1
y
( ) Figure 2.6.10: A graph of y = xx and its
y ′ = y ln x + 1 (subs tute y = xx ) tangent line at x = 1.5.
( )
y ′ = xx ln x + 1 .
To “test” our answer, let’s use it to find the equa on of the tangent line at x =
1.5. The point on the graph our tangent line must pass through is (1.5, 1.51.5 ) ≈
(1.5, 1.837). Using the equa on for y ′ , we find the slope as
( )
y ′ = 1.51.5 ln 1.5 + 1 ≈ 1.837(1.405) ≈ 2.582.
Thus the equa on of the tangent line is y = 1.6833(x − 1.5) + 1.837. Figure
2.6.10 graphs y = xx along with this tangent line.
Implicit differen a on proves to be useful as it allows us to find the instan-

taneous rates of change of a variety of func ons. In par cular, it extended the
Power Rule to ra onal exponents, which we then extended to all real numbers.
In the next sec on, implicit differen a on will be used to find the deriva ves of
inverse func ons, such as y = sin−1 x.
Notes:
119
Exercises 2.6
Terms and Concepts 21. (y2 + 2y − x)2 = 200
1. In your own words, explain the difference between implicit x2 + y

func ons and explicit func ons. 22. = 17
x + y2
2. Implicit differen a on is based on what other differen a- sin(x) + y

on rule? 23. =1
cos(y) + x
3. T/F: Implicit√differen a on can be used to find the deriva- 24. ln(x2 + y2 ) = e

ve of y = x.
25. ln(x2 + xy + y2 ) = 1
4. T/F: Implicit differen a on can be used to find the deriva-
ve of y = x3/4 . dy
26. Show that is the same for each of the following implicitly
dx
defined func ons.
Problems (a) xy = 1
In Exercises 5 – 12, compute the deriva ve of the given func- (b) x2 y2 = 1
on.
(c) sin(xy) = 1
√ 1
5. f(x) = x + √ (d) ln(xy) = 1
x
√ In Exercises 27 – 32, find the equa on of the tangent line to
6. f(x) = 3
x + x2/3
the graph of the implicitly defined func on at the indicated
√ points. As a visual aid, each func on is graphed.
7. f(t) = 1 − t2
√ 27. x2/5 + y2/5 = 1
8. g(t) = t sin t
(a) At (1, 0).
9. h(x) = x1.5
(b) At (0.1, 0.281) (which does not exactly lie on the
π
10. f(x) = x + x 1.9
+π 1.9 curve, but is very close).
y
x+7 1
11. g(x) = √
x
0.5 (0.1, 0.281)
√
12. f(t) = 5 t(sec t + et ) x
−1 −0.5 0.5 1
dy −0.5
In Exercises 13 – 25, find using implicit differen a on.
dx
. −1
13. x4 + y2 + y = 7
28. x4 + y4 = 1
14. x2/5 + y2/5 = 1
(a) At (1, 0).
15. cos(x) + sin(y) = 1 √ √
(b) At ( 0.6, 0.8).
x (c) At (0, 1).
16. = 10
y
y
y 1
√ √
17. = 10 ( 0.6, 0.8)
x
0.5
18. x2 e2 + 2y = 5 x
−1 −0.5 0.5 1
19. x2 tan y = 50 −0.5
−1
20. (3x2 + 2y3 )4 = 2 .
120
29. (x2 + y2 − 4)3 = 108y2 32. x2 + y3 + 2xy = 0
(a) At (−1, 1).
(a) At (0, 4). ( )
√ 1 √
(b) At (2, − 4 108). (b) At −1, (−1 + 5) .
2
( )
1 √
y (c) At −1, (−1 − 5) .
4 2
y
2
2
x (−1, 1)
−4 −2 2 4 ( √ )
−1+ 5
−1,
−2 2
x
√ −2 2
4
−4 (2, − 108)
. ( √ )
−1− 5
−1, 2
−2
30. (x2 + y2 + x)2 = x2 + y2
(a) At (0, 1). In Exercises 33 – 36, an implicitly defined func on is given.

( √ ) d2 y
3 3 3 Find 2 . Note: these are the same problems used in Exer-
(b) At − , . dx
4 4 cises 13 through 16.
y 33. x4 + y2 + y = 7
( √ )
− 34 , 3 3
4 1
34. x2/5 + y2/5 = 1
x 35. cos x + sin y = 1

−2 −1
x
36. = 10
−1 y
.
In Exercises 37 – 42, use logarithmic differen a on to find
dy
31. (x − 2) + (y − 3) = 9
2 2 , then find the equa on of the tangent line at the indicated
dx
( √ ) x–value.
7 6+3 3
(a) At , .
2 2 37. y = (1 + x)1/x , x = 1
( √ )
4+3 3 3 2
(b) At , . 38. y = (2x)x , x = 1
2 2
xx
y 39. y = , x=1
6
x+1
( √
6+3 3
)
3.5,
40. y = xsin(x)+2 , x = π/2
2
x+1
2 ( 4+3√3 ) 41. y = , x=1
2 , 1.5 x+2
.
x (x + 1)(x + 2)
2 4 6 42. y = , x=0
(x + 3)(x + 4)
121
2.7 Deriva ves of Inverse Func ons

Recall that a func on y = f(x) is said to be one to one if it passes the horizontal
line test; that is, for two different x values x1 and x2 , we do not have f(x1 ) = f(x2 ).
In some cases the domain of f must be restricted so that it is one to one. For
instance, consider f(x) = x2 . Clearly, f(−1) = f(1), so f is not one to one on its
regular domain, but by restric ng f to (0, ∞), f is one to one.
Now recall that one to one func ons have inverses. That is, if f is one to
one, it has an inverse func on, denoted by f −1 , such that if f(a) = b, then
f −1 (b) = a. The domain of f −1 is the range of f, and vice-versa. For ease of
nota on, we set g = f −1 and treat g as a func on of x.
Since f(a) = b implies g(b) = a, when we compose f and g we get a nice
result:
y
( )
f g(b) = f(a) = b.
(1, 1.5) ( ) ( )
In general, f g(x) = x and g f(x) = x. This gives us a convenient way to check
1 if two func ons are inverses of each other: compose them and if the result is x,
(1.5, 1)
(−0.5, 0.375)
then they are inverses (on the appropriate domains.)
When the point (a, b) lies on the graph of f, the point (b, a) lies on the graph
x
−1 1
of g. This leads us to discover that the graph of g is the reflec on of f across the
line y = x. In Figure 2.7.1 we see a func on graphed along with its inverse. See
(0.375, −0.5) how the point (1, 1.5) lies on one graph, whereas (1.5, 1) lies on the other. Be-
−1 cause of this rela onship, whatever we know about f can quickly be transferred
.
into knowledge about g.
Figure 2.7.1: A func on f along with its in- For example, consider Figure 2.7.2 where the tangent line to f at the point
verse f −1 . (Note how it does not ma er
(a, b) is drawn. That line has slope f ′ (a). Through reflec on across y = x, we
which func on we refer to as f; the other 1
is f −1 .) can see that the tangent line to g at the point (b, a) should have slope ′ .
f (a)
1
This then tells us that g ′ (b) = ′ .
f (a)
Consider:
y
Informa on about f Informa on about g = f −1
1
(−0.5, 0.375) lies on f (0.375, −0.5) lies on g
(a, b) Slope of tangent line to f Slope of tangent line to
at x = −0.5 is 3/4 g at x = 0.375 is 4/3
x
−1 1 f ′ (−0.5) = 3/4 g ′ (0.375) = 4/3
(b, a)
We have discovered a rela onship between f ′ and g ′ in a mostly graphical
−1 way. We can realize this rela onship analy cally as well. Let y = g(x), where
.
again g = f −1 . We want to find y ′ . Since y = g(x), we know that f(y) = x. Using
Figure 2.7.2: Corresponding tangent lines the Chain Rule and Implicit Differen a on, take the deriva ve of both sides of
drawn to f and f −1 .
Notes:
122
this last equality.

d( ) d( )
f(y) = x
dx dx
f ′ (y) · y ′ = 1
1
y′ = ′
f (y)
1
y′ = ′ .
f (g(x))
This leads us to the following theorem.
Theorem 2.7.1 Deriva ves of Inverse Func ons

Let f be differen able and one to one on an open interval I, where f ′ (x) ̸=
0 for all x in I, let J be the range of f on I, let g be the inverse func on of
f, and let f(a) = b for some a in I. Then g is a differen able func on on
J, and in par cular,
( )′ 1 ( )′ 1
1. f −1 (b) = g ′ (b) = ′ and 2. f −1 (x) = g ′ (x) = ′
f (a) f (g(x))
The results of Theorem 2.7.1 are not trivial; the nota on may seem confusing
at first. Careful considera on, along with examples, should earn understanding.
In the next example we apply Theorem 2.7.1 to the arcsine func on.
Example 2.7.1 Finding the deriva ve of an inverse trigonometric func on

Let y = arcsin x = sin−1 x. Find y ′ using Theorem 2.7.1.
S Adop ng our previously defined nota on, let g(x) = arcsin x

and f(x) = sin x. Thus f ′ (x) = cos x. Applying the theorem, we have
1
g ′ (x) =
f ′ (g(x))
1
= .
cos(arcsin x)
This last expression is not immediately illumina ng. Drawing a figure will
help, as shown in Figure 2.7.4. Recall that the sine func on can be viewed as
taking in an angle and returning a ra o of sides of a right triangle, specifically,
the ra o “opposite over hypotenuse.” This means that the arcsine func on takes
as input a ra o of sides and returns an angle. The equa on y = arcsin x can
be rewri en as y = arcsin(x/1); that is, consider a right triangle where the
Notes:
123
hypotenuse has length 1 and the side opposite√

of the angle with measure y has
length x. This means the final side has length 1 − x2 , using the Pythagorean
Theorem.
1
x
√ √
. y
√
Therefore cos(sin−1 x) = cos y = 1 − x2 /1 = 1 − x2 , resul ng in
1 − x2
Figure 2.7.4: A right triangle defined by d( ) 1

y = sin−1 (x/1) with the length of the arcsin x = g ′ (x) = √ .
third leg found using the Pythagorean
dx 1 − x2
Theorem.
Remember that the input x of the arcsine func on is a ra o of a side of a right

y
triangle to its hypotenuse; the absolute value of this ra o will never be greater
√
than 1. Therefore the inside of the square root will never be nega ve.
3
1 ( π3 , 2 )
In order to make y = sin x one to one, we restrict its domain to [−π/2, π/2];
on this domain, the range is [−1, 1]. Therefore the domain of y = arcsin x is
x [−1, 1] and the range is [−π/2, π/2]. When x = ±1, note how the deriva ve of
− π2 − π4 π π
4 2 the arcsine func on is undefined; this corresponds to the fact that as x → ±1,
the tangent lines to arcsine approach ver cal lines with undefined slopes.
y = sin x −1
In Figure 2.7.5 we see f(x) = sin x and f −1 (x) = sin−1 x graphed
√ on their re-
.
spec ve domains. The line tangent to sin x at the point (π/3, 3/2) has slope
−1
π
2
cos
√ π/3 = 1/2. The slope of the corresponding point on sin x, the point
√
( 2
3
, π3 ) ( 3/2, π/3), is
π
4
1 1 1 1
√ √ =√ =√ = = 2,
−1 1
1 − ( 3/2)2 1 − 3/4 1/4 1/2
y = sin−1 x
− π4
. − π2
verifying yet again that at corresponding points, a func on and its inverse have
−1 reciprocal slopes.
Figure 2.7.5: Graphs of sin x and sin x
along with corresponding tangent lines.
Using similar techniques, we can find the deriva ves of all the inverse trigono-
metric func ons. In Figure 2.7.3 we show the restric ons of the domains of the
standard trigonometric func ons that allow them to be inver ble.
Notes:
124
Inverse
Func on Domain Range Func on Domain Range
−1
sin x [−π/2, π/2] [−1, 1] sin x [−1, 1] [−π/2, π/2]
cos x [0, π] [−1, 1] cos−1 x [−1, 1] [0, π]
−1
tan x (−π/2, π/2) (−∞, ∞) tan x (−∞, ∞) (−π/2, π/2)
−1
csc x [−π/2, 0) ∪ (0, π/2] (−∞, −1] ∪ [1, ∞) csc x (−∞, −1] ∪ [1, ∞) [−π/2, 0) ∪ (0, π/2]
−1
sec x [0, π/2) ∪ (π/2, π] (−∞, −1] ∪ [1, ∞) sec x (−∞, −1] ∪ [1, ∞) [0, π/2) ∪ (π/2, π]
−1
cot x (0, π) (−∞, ∞) cot x (−∞, ∞) (0, π)
Figure 2.7.3: Domains and ranges of the trigonometric and inverse trigonometric func ons.
Theorem 2.7.2 Deriva ves of Inverse Trigonometric Func ons

The inverse trigonometric func ons are differen able on all open sets
contained in their domains (as listed in Figure 2.7.3) and their deriva ves
are as follows:
d ( −1 ) d( ) 1
1. sin x = √
1 4. cos−1 x = − √
dx 1 − x2 dx 1 − x2
d ( −1 ) d ( ) 1
2. sec x = √
1 5. csc−1 x = − √
dx |x| x2 − 1 dx |x| x2 − 1
d( ) 1 d ( −1 ) 1
3. tan−1 x = 6. cot x = −
dx 1 + x2 dx 1 + x2
Note how the last three deriva ves are merely the opposites of the first
three, respec vely. Because of this, the first three are used almost exclusively
throughout this text.
d( ) 1
In Sec on 2.3, we stated without proof or explana on that ln x = .
dx x
We can jus fy that now using Theorem 2.7.1, as shown in the example.
Example 2.7.2 Finding the deriva ve of y = ln x

d( )
Use Theorem 2.7.1 to compute ln x .
dx
S View y = ln x as the inverse of y = ex . Therefore, using our

standard nota on, let f(x) = ex and g(x) = ln x. We wish to find g ′ (x). Theorem
Notes:
125
2.7.1 gives:
1
g ′ (x) =
f ′ (g(x))
1
= ln x
e
1
= .
x
In this chapter we have defined the deriva ve, given rules to facilitate its
computa on, and given the deriva ves of a number of standard func ons. We
restate the most important of these in the following theorem, intended to be a
reference for further work.
Theorem 2.7.3 Glossary of Deriva ves of Elementary Func ons

Let u and v be differen able func ons, and let a, c and n be real
numbers, a > 0, n ̸= 0.
( ) ( )
1. d
dx cu = cu′ 2. d
dx u ± v = u′ ± v′
( ) ′ ′
( ) u′ v−uv′
dx u · v = uv + u v
d d u
3. 4. dx v = v2
d
( ) ′ ′ d
( )
5. dx u(v) = u (v)v 6. dx c = 0
d
( ) d
( n)
7. dx x = 1 8. dx x = nxn−1
d
( x) ( x)
9. dx e = ex 10. d
dx a = ln a · ax
d
( ) 1 ( )
dx loga x = ln a · x
11. d 1 1
dx ln x = x 12.
d
( ) ( )
dx cos x = − sin x
13. d
dx sin x = cos x 14.
d
( ) ( )
dx csc x = − csc x cot x
15. d
dx sec x = sec x tan x 16.
d
( ) ( )
dx cot x = − csc x
2 d 2
17. dx tan x = sec x 18.
d
( −1 ) 1
( −1
)
19. dx sin x = √1−x 2 20. d
dx cos x = − √1−x 1
2
( ) ( )
21. d
dx sec−1 x = √1
|x| x2 −1
22. d
dx csc−1 x = − |x|√1x2 −1
( ) ( )
23. d
dx tan−1 x = 1
1+x2 24. d
dx cot−1 x = − 1+x
1
2
Notes:
126
Exercises 2.7
Terms and Concepts 14. f(x) = 6e3x
Point= (0, 6)
( )′
1. T/F: Every func on has an inverse. Evaluate f−1 (6)
2. In your own words explain what it means for a func on to In Exercises 15 – 24, compute the deriva ve of the given func-
be “one to one.” on.
3. If (1, 10) lies on the graph of y = f(x), what can be said 15. h(t) = sin−1 (2t)
about the graph of y = f−1 (x)?
16. f(t) = sec−1 (2t)
′
4. If (1, 10) lies on the graph of y = f(x) and f (1) = 5, what
can be said about y = f−1 (x)? 17. g(x) = tan−1 (2x)
18. f(x) = x sin−1 x

Problems
19. g(t) = sin t cos−1 t
In Exercises 5 – 8, verify that the given func ons are inverses.
20. f(t) = ln tet
5. f(x) = 2x + 6 and g(x) = 12 x − 3
sin−1 x
21. h(x) =
6. f(x) = x + 6x + 11, x ≥ 3 and
2 cos−1 x
√
g(x) = x − 2 − 3, x ≥ 2 √
22. g(x) = tan−1 ( x)
3
7. f(x) = , x ̸= 5 and 23. f(x) = sec−1 (1/x)
x−5
3 + 5x
g(x) = , x ̸= 0 24. f(x) = sin(sin−1 x)
x
x+1 In Exercises 25 – 26, compute the deriva ve of the given func-

8. f(x) = , x ̸= 1 and g(x) = f(x) on in two ways:
x−1
(a) By simplifying first, then taking the deriva ve, and
In Exercises 9 – 14, an inver ble func on f(x) is given along
(b) by using the Chain Rule first then simplifying.
with a( point
)′ that lies on its graph. Using Theorem 2.7.1, eval-
uate f−1 (x) at the indicated value. Verify that the two answers are the same.
9. f(x) = 5x + 10 25. f(x) = sin(sin−1 x)

Point= (2, 20)
( )′ 26. f(x) = tan−1 (tan x)
Evaluate f−1 (20)
In Exercises 27 – 28, find the equa on of the line tangent to
10. f(x) = x2 − 2x + 4, x ≥ 1 the graph of f at the indicated x value.
Point= (3, 7)
( )′ √
Evaluate f−1 (7) 27. f(x) = sin−1 x at x = 2
2
√
11. f(x) = sin 2x, −π/4 ≤ x ≤ π/4 28. f(x) = cos−1 (2x) at x = 3
√ 4
Point= (π/6, 3/2)
( )′ √
Evaluate f−1 ( 3/2)
Review
12. f(x) = x3 − 6x2 + 15x − 2
Point= (1, 8) 29. Find dy
dx
, where x2 y − y2 x = 1.
( )′
Evaluate f−1 (8)
30. Find the equa on of the line tangent to the graph of x2 +
y2 + xy = 7 at the point (1, 2).
1
13. f(x) = ,x≥0
1 + x2
31. Let f(x) = x3 + x.
Point= (1, 1/2)
( )′ f(x + s) − f(x)
Evaluate f−1 (1/2) Evaluate lim .
s→0 s
127
3: T G B
F
y
Our study of limits led to con nuous func ons, a certain class of func ons that
behave in a par cularly nice way. Limits then gave us an even nicer class of
4
func ons, func ons that are differen able.
This chapter explores many of the ways we can take advantage of the infor-
ma on that con nuous and differen able func ons provide. 2
3.1 Extreme Values ( ) x

−1
.−2 1 2
Given any quan ty described by a func on, we are o en interested in the largest
(a)
and/or smallest values that quan ty a ains. For instance, if a func on describes
the speed of an object, it seems reasonable to want to know the fastest/slowest y
the object traveled. If a func on describes the value of a stock, we might want
to know the highest/lowest values the stock a ained over the past year. We call
such values extreme values. 4
Defini on 3.1.1 Extreme Values 2
Let f be defined on an interval I containing c.
1. f(c) is the minimum (also, absolute minimum) of f on I if f(c) ≤ [ ] x

−2 −1 1 2
f(x) for all x in I. .
(b)
2. f(c) is the maximum (also, absolute maximum) of f on I if f(c) ≥
f(x) for all x in I. y
The maximum and minimum values are the extreme values, or extrema,
of f on I. 4
2
Consider Figure 3.1.1. The func on displayed in (a) has a maximum, but
no minimum, as the interval over which the func on is defined is open. In (b),
the func on has a minimum, but no maximum; there is a discon nuity in the [ ] x
“natural” place for the maximum to occur. Finally, the func on shown in (c) has −2
. −1 1 2
both a maximum and a minimum; note that the func on is con nuous and the
(c)
interval on which it is defined is closed.
It is possible for discon nuous func ons defined on an open interval to have Figure 3.1.1: Graphs of func ons with and
both a maximum and minimum value, but we have just seen examples where without extreme values.
they did not. On the other hand, con nuous func ons on a closed interval al-
ways have a maximum and minimum value.
Note: The extreme values of a func on

are “y” values, values the func on a ains,
not the input values.
Chapter 3 The Graphical Behavior of Func ons
Theorem 3.1.1 The Extreme Value Theorem

Let f be a con nuous func on defined on a closed interval I. Then f has
both a maximum and minimum value on I.
This theorem states that f has extreme values, but it does not offer any ad-
vice about how/where to find these values. The process can seem to be fairly
easy, as the next example illustrates. A er the example, we will draw on lessons
learned to form a more general and powerful method for finding extreme values.
y
(5, 25)
Example 3.1.1 Approxima ng extreme values
Consider f(x) = 2x3 − 9x2 on I = [−1, 5], as graphed in Figure 3.1.2. Approxi-
20
mate the extreme values of f.
(0, 0)
x
S The graph is drawn in such a way to draw a en on to certain
−1 5 points. It certainly seems that the smallest y value is −27, found when x = 3.
(−1, −11)
It also seems that the largest y value is 25, found at the endpoint of I, x = 5.
−20 (3, −27)
We use the word seems, for by the graph alone we cannot be sure the smallest
value is not less than −27. Since the problem asks for an approxima on, we
. approximate the extreme values to be 25 and −27.
Figure 3.1.2: A graph of f(x) = 2x3 − 9x2
as in Example 3.1.1. No ce how the minimum value came at “the bo om of a hill,” and the maxi-
mum value came at an endpoint. Also note that while 0 is not an extreme value,
it would be if we narrowed our interval to [−1, 4]. The idea that the point (0, 0)
is the loca on of an extreme value for some interval is important, leading us to
a defini on of a rela ve maximum. In short, a “rela ve max” is a y-value that’s
the largest y-value “nearby.”
Defini on 3.1.2 Rela ve Minimum and Rela ve Maximum

Note: The terms local minimum and local Let f be defined on an interval I containing c.
maximum are o en used as synonyms for
rela ve minimum and rela ve maximum. 1. If there is a δ > 0 such that f(c) ≤ f(x) for all x in I where |x − c| <
δ, then f(c) is a rela ve minimum of f. We also say that f has a
As it makes intui ve sense that an ab- rela ve minimum at (c, f(c)).
solute maximum is also a rela ve max-
imum, Defini on 3.1.2 allows a rela ve 2. If there is a δ > 0 such that f(c) ≥ f(x) for all x in I where |x − c| <
maximum to occur at an interval’s end- δ, then f(c) is a rela ve maximum of f. We also say that f has a
point. rela ve maximum at (c, f(c)).
The rela ve maximum and minimum values comprise the rela ve ex-
trema of f.
Notes:
130
3.1 Extreme Values
We briefly prac ce using these defini ons.
Example 3.1.2 Approxima ng rela ve extrema

Consider f(x) = (3x4 −4x3 −12x2 +5)/5, as shown in Figure 3.1.3. Approximate y
the rela ve extrema of f. At each of these points, evaluate f ′ .
6
4
S We s ll do not have the tools to exactly find the rela ve
2
extrema, but the graph does allow us to make reasonable approxima ons. It
seems f has rela ve minima at x = −1 and x = 2, with values of f(−1) = 0 and x
−2 −1 1 2 3
f(2) = −5.4. It also seems that f has a rela ve maximum at the point (0, 1). −2
We approximate the rela ve minima to be 0 and −5.4; we approximate the −4

rela ve maximum to be 1. −6
.
It is straigh orward to evaluate f ′ (x) = 15 (12x3 − 12x2 − 24x) at x = 0, 1
and 2. In each case, f ′ (x) = 0. Figure 3.1.3: A graph of f(x) = (3x4 −
4x3 − 12x2 + 5)/5 as in Example 3.1.2.
Example 3.1.3 Approxima ng rela ve extrema

Approximate the rela ve extrema of f(x) = (x−1)2/3 +2, shown in Figure 3.1.4.
At each of these points, evaluate f ′ . y
3
S The figure implies that f does not have any rela ve maxima,
but has a rela ve minimum at (1, 2). In fact, the graph suggests that not only 2
is this point a rela ve minimum, y = f(1) = 2 is the minimum value of the
func on. 1
We compute f ′ (x) = 23 (x − 1)−1/3 . When x = 1, f ′ is undefined.
x
. 1 2
What can we learn from the previous two examples? We were able to vi-
sually approximate rela ve extrema, and at each such point, the deriva ve was Figure 3.1.4: A graph of f(x) = (x −
either 0 or it was not defined. This observa on holds for all func ons, leading 1)2/3 + 2 as in Example 3.1.3.
to a defini on and a theorem.
Defini on 3.1.3 Cri cal Numbers and Cri cal Points

Let f be defined at c. The value c is a cri cal number (or cri cal value)
of f if f ′ (c) = 0 or f ′ (c) is not defined.
If c is a cri cal number of f, then the point (c, f(c)) is a cri cal point of f.
Notes:
131
y
1
Theorem 3.1.2 Rela ve Extrema and Cri cal Points
Let a func on f be defined on an open interval I containing c, and let f
have a rela ve extremum at the point (c, f(c)). Then c is a cri cal number
of f.
x
−1 1
Be careful to understand that this theorem states “Rela ve extrema on open

intervals occur at cri cal points.” It does not say “All cri cal numbers produce
−1
. rela ve extrema.” For instance, consider f(x) = x3 . Since f ′ (x) = 3x2 , it is
straigh orward to determine that x = 0 is a cri cal number of f. However, f has
Figure 3.1.5: A graph of f(x) = x3 which
has a cri cal value of x = 0, but no rela- no rela ve extrema, as illustrated in Figure 3.1.5.
ve extrema. Theorem 3.1.1 states that a con nuous func on on a closed interval will have
both an absolute maximum and an absolute minimum. Common sense tells us
“extrema occur either at the endpoints or somewhere in between.” It is easy
to check for extrema at endpoints, but there are infinitely many points to check
that are “in between.” Our theory tells us we need only check at the cri cal
points that are in between the endpoints. We combine these concepts to offer
a strategy for finding extrema.
Key Idea 3.1.1 Finding Extrema on a Closed Interval

Let f be a con nuous func on defined on a closed interval [a, b]. To find
the maximum and minimum values of f on [a, b]:
1. Evaluate f at the endpoints a and b of the interval.

2. Find the cri cal numbers of f in [a, b].
3. Evaluate f at each cri cal number.
4. The absolute maximum of f is the largest of these values, and the

absolute minimum of f is the least of these values.
We prac ce these ideas in the next examples.
Example 3.1.4 Finding extreme values

Find the extreme values of f(x) = 2x3 + 3x2 − 12x on [0, 3], graphed in Figure
3.1.6(a).
S We follow the steps outlined in Key Idea 3.1.1. We first eval-
Notes:
132
3.1 Extreme Values
uate f at the endpoints: y
f(0) = 0 and f(3) = 45. 40
Next, we find the cri cal values of f on [0, 3]. f ′ (x) = 6x2 + 6x − 12 = 6(x +
2)(x − 1); therefore the cri cal values of f are x = −2 and x = 1. Since x = −2 20
does not lie in the interval [0, 3], we ignore it. Evalua ng f at the only cri cal
number in our interval gives: f(1) = −7.
The table in Figure 3.1.6(b) gives f evaluated at the “important” x values in
x
[0, 3]. We can easily see the maximum and minimum values of f: the maximum 1 2 3
value is 45 and the minimum value is −7. .
(a)
Note that all this was done without the aid of a graph; this work followed an
analy c algorithm and did not depend on any visualiza on. Figure 3.1.6 shows x f(x)
f and we can confirm our answer, but it is important to understand that these 0 0
answers can be found without graphical assistance. 1 −7
We prac ce again. 3 45
(b)
Find the maximum and minimum values of f on [−4, 2], where Figure 3.1.6: Finding the extreme values
of f(x) = 2x3 +3x2 −12x in Example 3.1.4.
{
(x − 1)2 x ≤ 0
f(x) = ,
x+1 x>0
y
graphed in Figure 3.1.7(a).
S Here f is piecewise–defined, but we can s ll apply Key Idea 20
3.1.1 as it is con nuous on [−4, 2] (one should check to verify that lim f(x) =
x→0
f(0)). Evalua ng f at the endpoints gives:
10
f(−4) = 25 and f(2) = 3.
We now find the cri cal numbers of f. We have to define f ′ in a piecewise

manner; it is . x
{ −4 −2 2
′ 2(x − 1) x < 0
f (x) = .
1 x>0 (a)
Note that while f is defined for all of [−4, 2], f ′ is not, as the deriva ve of f does x f(x)
not exist when x = 0. (From the le , the deriva ve approaches −2; from the −4 25
right the deriva ve is 1.) Thus one cri cal number of f is x = 0. 0 1
We now set f ′ (x) = 0. When x > 0, f ′ (x) is never 0. When x < 0, f ′ (x) is 2 3
also never 0, so we find no cri cal values from se ng f ′ (x) = 0. (b)
So we have three important x values to consider: x = −4, 2 and 0. Evaluat-
ing f at each gives, respec vely, 25, 3 and 1, shown in Figure 3.1.7(b). Thus the Figure 3.1.7: Finding the extreme values
of a piecewise–defined func on in Exam-
ple 3.1.5.
Notes:
133
y
1 absolute minimum of f is 1, the absolute maximum of f is 25, confirmed by the
graph of f.
0.5
x Find the extrema of f(x) = cos(x2 ) on [−2, 2], graphed in Figure 3.1.8(a).
−2 −1 1 2
−0.5 S We again use Key Idea 3.1.1. Evalua ng f at the endpoints of

the interval gives: f(−2) = f(2) = cos(4) ≈ −0.6536. We now find the cri cal
−1 values of f.
.
Applying the Chain Rule, we find f ′ (x) = −2x sin(x2 ). Set f ′ (x) = 0 and
(a) solve for x to find the cri cal values of f.
We have f ′ (x) = 0 when x = 0 and when sin(x2 ) = 0. In general, sin t = 0
x f(x) when t = . . . − 2π, −π, 0, π, . . . Thus sin(x2 ) = 0 when x2 = 0, π, 2π, (x2 is
−2 −0.65
√ . . .√
√ always posi ve so we ignore −π, etc.) So sin(x ) = 0 when x = 0, ± √π, ± 2π,
2
− π −1
0 1 √ The only values to fall in the given interval of [−2, 2] are 0 and ± π, where
etc.
√ π ≈ 1.77.
π −1
We again construct a table of important values√in Figure 3.1.8(b). In this
2 −0.65
(b) example we have 5 values to consider: x = 0, ±2, ± π.
From the table it is clear that the maximum value of f on [−2, 2] is 1; the
Figure 3.1.8: Finding the extrema of minimum value is −1. The graph of f confirms our results.
f(x) = cos(x2 ) in Example 3.1.6.
We consider one more example.
Example 3.1.7 Finding extreme

√ values
Find the extreme values of f(x) = 1 − x2 , graphed in Figure 3.1.9(a).
y
1
S A closed interval is not given, so we find the extreme values
of f on its domain. f is defined whenever 1 − x2 ≥ 0; thus the domain of f is
[−1, 1]. Evalua ng f at either endpoint returns 0.
−x
Using the Chain Rule, we find f ′ (x) = √ . The cri cal points of f are
x 1 − x2
′ ′
. −1 1 found when f (x) = 0 or when f is undefined. It is straigh orward to find that
f ′ (x) = 0 when x = 0, and f ′ is undefined when x = ±1, the endpoints of the
(a) interval. The table of important values is given in Figure 3.1.9(b). The maximum
value is 1, and the minimum value is 0. (See also the marginal note.)
x f(x)
−1 0 We have seen that con nuous func ons on closed intervals always have a
0 1
maximum and minimum value, and we have also developed a technique to find
1 0
these values. In the next sec on, we further our study of the informa on we can
(b)
glean from “nice” func ons with the Mean Value Theorem. On a closed interval,
Figure 3.1.9: Finding the extrema of the we can find the average rate of change of a func on (as we did at the beginning
half–circle in Example 3.1.7. of Chapter 2). We will see that differen able func ons always have a point at
which their instantaneous rate of change is same as the average rate of change.
This is surprisingly useful, as we’ll see.
Note: We implicitly found the deriva ve
of x2 + y2 = 1, the unit circle, in Ex-
ample 2.6.5 as dy dx
= −x/y. In Exam- Notes:
ple 3.1.7, half√of the unit circle is given as
y = f(x) = 1 − x2 . We found f ′ (x) =
√−x . Recognize that the denominator
2
1−x
of this frac on is y; that is, we again found
f ′ (x) = dy
dx
= −x/y.
134
Exercises 3.1
Terms and Concepts In Exercises 9 – 16, evaluate f ′ (x) at the points indicated in
the graph.
2
9. f(x) =
1. Describe what an “extreme value” of a func on is in your x2 + 1
own words. y
(0, 2)
2
2. Sketch the graph of a func on f on (−1, 1) that has both a

maximum and minimum value. 1
3. Describe the difference between absolute and rela ve x

−5 5
maxima in your own words. .
√
10. f(x) = x2 6 − x2
4. Sketch the graph of a func on f where f has a rela ve max- y
imum at x = 1 and f ′ (1) is undefined. 6 √

(2, 4 2)
4
5. T/F: If c is a cri cal value of a func on f, then f has either a
rela ve maximum or rela ve minimum at x = c. 2
x
6. Fill in the blanks: The cri cal points of a func on f are . −2 (0, 0) 2
found where f ′ (x) is equal to or where f ′ (x) is

. 11. f(x) = sin x
y
(π/2, 1)
1
Problems x
2 4 6
In Exercises 7 – 8, iden fy each of the marked points as being −1

(3π/2, −1)
an absolute maximum or minimum, a rela ve maximum or .
minimum, or none of the above.
√
12. f(x) = x2 4 − x
y
( 16 512
)
, √
10 5 25 5
y
B
D G
2
5
E
7. C
x
(4, 0)
2 4 6 x
F . −2 (0, 0) 2 4
−2 A
.
(x − 2)2/3
13. f(x) = 1 +
x
y
y
C 6
2
B D 4
8. x ( √
3
)
2
2 4 6, 1 + 3
2
E
(2, 1)
A
−2 x
. . 5 10
135
√
x − 2x + 1
3 4
14. f(x) = In Exercises 17 – 26, find the extreme values of the func on
on the given interval.
y
3 17. f(x) = x2 + x + 4 on [−1, 2].
2 9 2
18. f(x) = x3 − x − 30x + 3 on [0, 6].
2
1
19. f(x) = 3 sin x on [π/4, 2π/3].
(−1, 0) (1, 0)
x
−2 −1 1 2 √
20. f(x) = x2 4 − x2 on [−2, 2].
{ 3
x2 x≤0 21. f(x) = x + on [1, 5].
15. f(x) = x
x5 x>0
y x2
22. f(x) = on [−3, 5].
1 x2 +5
23. f(x) = ex cos x on [0, π].

0.5
24. f(x) = ex sin x on [0, π].

x
−1 −0.5 (0, 0) 0.5 1
ln x
25. f(x) = on [1, 4].
. −0.5 x
26. f(x) = x2/3 − x on [0, 2].

{
x2
x≤0
16. f(x) =
x x>0
y
Review
1
27. Find dy
dx
, where x2 y − y2 x = 1.
0.5 28. Find the equa on of the line tangent to the graph of x2 +
y2 + xy = 7 at the point (1, 2).
x
−1 −0.5 (0, 0) 0.5 1
29. Let f(x) = x3 + x.
. −0.5
f(x + s) − f(x)
Evaluate lim .
s→0 s
136
3.2 The Mean Value Theorem

We mo vate this sec on with the following ques on: Suppose you leave your
house and drive to your friend’s house in a city 100 miles away, comple ng the
trip in two hours. At any point during the trip do you necessarily have to be going
50 miles per hour?
In answering this ques on, it is clear that the average speed for the en re
trip is 50 mph (i.e. 100 miles in 2 hours), but the ques on is whether or not your
instantaneous speed is ever exactly 50 mph. More simply, does your speedome-
ter ever read exactly 50 mph?. The answer, under some very reasonable as-
sump ons, is “yes.”
Let’s now see why this situa on is in a calculus text by transla ng it into
mathema cal symbols.
First assume that the func on y = f(t) gives the distance (in miles) traveled
from your home at me t (in hours) where 0 ≤ t ≤ 2. In par cular, this gives
f(0) = 0 and f(2) = 100. The slope of the secant line connec ng the star ng
and ending points (0, f(0)) and (2, f(2)) is therefore
∆f f(2) − f(0) 100 − 0
= = = 50 mph.
∆t 2−0 2
The slope at any point on the graph itself is given by the deriva ve f ′ (t). So,
since the answer to the ques on above is “yes,” this means that at some me
during the trip, the deriva ve takes on the value of 50 mph. Symbolically,
f(2) − f(0)
f ′ (c) = = 50
2−0
for some me 0 ≤ c ≤ 2.
How about more generally? Given any func on y = f(x) and a range a ≤
x ≤ b does the value of the deriva ve at some point between a and b have to
match the slope of the secant line connec ng the points (a, f(a)) and (b, f(b))?
Or equivalently, does the equa on f ′ (c) = f(b)−f(a)
b−a have to hold for some a <
c < b?
Let’s look at two func ons in an example.
Example 3.2.1 Comparing average and instantaneous rates of change

Consider func ons
1
f1 (x) = 2 and f2 (x) = |x|
x
with a = −1 and b = 1 as shown in Figure 3.2.1(a) and (b), respec vely. Both
func ons have a value of 1 at a and b. Therefore the slope of the secant line
Notes:
137
y
connec ng the end points is 0 in each case. But if you look at the plots of each,
you can see that there are no points on either graph where the tangent lines
have slope zero. Therefore we have found that there is no c in [−1, 1] such that
f(1) − f(−1)
f ′ (c) = = 0.
2 1 − (−1)
So what went “wrong”? It may not be surprising to find that the discon nuity
x of f1 and the corner of f2 play a role. If our func ons had been con nuous and
−1 1 differen able, would we have been able to find that special value c? This is our
..
(a) mo va on for the following theorem.
y
Theorem 3.2.1 The Mean Value Theorem of Differen a on
1
Let y = f(x) be a con nuous func on on the closed interval [a, b] and
differen able on the open interval (a, b). There exists a value c, a < c <
b, such that
f(b) − f(a)
f ′ (c) =
0.5
.
b−a
That is, there is a value c in (a, b) where the instantaneous rate of change
x of f at c is equal to the average rate of change of f on [a, b].
. −1 1
(b)
Note that the reasons that the func ons in Example 3.2.1 fail are indeed that
Figure 3.2.1: A graph of f1 (x) = 1/x2 and f1 has a discon nuity on the interval [−1, 1] and f2 is not differen able at the ori-
f2 (x) = |x| in Example 3.2.1.
gin.
We will give a proof of the Mean Value Theorem below. To do so, we use a
fact, called Rolle’s Theorem, stated here.
y
Theorem 3.2.2 Rolle’s Theorem

5
Let f be con nuous on [a, b] and differen able on (a, b), where f(a) =
f(b). There is some c in (a, b) such that f ′ (c) = 0.
x
−1 a c 1 b 2
Consider Figure 3.2.2 where the graph of a func on f is given, where f(a) =
f(b). It should make intui ve sense that if f is differen able (and hence, con nu-
. −5 ous) that there would be a value c in (a, b) where f ′ (c) = 0; that is, there would
be a rela ve maximum or minimum of f in (a, b). Rolle’s Theorem guarantees at
Figure 3.2.2: A graph of f(x) = x3 − 5x2 + least one; there may be more.
3x + 5, where f(a) = f(b). Note the ex-
istence of c, where a < c < b, where
f ′ (c) = 0.
.
Notes:
138
Rolle’s Theorem is really just a special case of the Mean Value Theorem. If
f(a) = f(b), then the average rate of change on (a, b) is 0, and the theorem
guarantees some c where f ′ (c) = 0. We will prove Rolle’s Theorem, then use it
to prove the Mean Value Theorem.
Proof of Rolle’s Theorem

Let f be differen able on (a, b) where f(a) = f(b). We consider two cases.
Case 1: Consider the case when f is constant on [a, b]; that is, f(x) = f(a) = f(b)
for all x in [a, b]. Then f ′ (x) = 0 for all x in [a, b], showing there is at least one
value c in (a, b) where f ′ (c) = 0.
Case 2: Now assume that f is not constant on [a, b]. The Extreme Value Theorem
guarantees that f has a maximal and minimal value on [a, b], found either at the
endpoints or at a cri cal value in (a, b). Since f(a) = f(b) and f is not constant, it
is clear that the maximum and minimum cannot both be found at the endpoints.
Assume, without loss of generality, that the maximum of f is not found at the
endpoints. Therefore there is a c in (a, b) such that f(c) is the maximum value
of f. By Theorem 3.1.2, c must be a cri cal number of f; since f is differen able,
we have that f ′ (c) = 0, comple ng the proof of the theorem. □
We can now prove the Mean Value Theorem.
Proof of the Mean Value Theorem

Define the func on
f(b) − f(a)
g(x) = f(x) − x.
b−a
We know g is differen able on (a, b) and con nuous on [a, b] since f is. We can
show g(a) = g(b) (it is actually easier to show g(b) − g(a) = 0, which suffices).
We can then apply Rolle’s theorem to guarantee the existence of c in (a, b) such
that g ′ (c) = 0. But note that
f(b) − f(a)
0 = g ′ (c) = f ′ (c) − ;
b−a
hence
f(b) − f(a)
f ′ (c) = ,
b−a
which is what we sought to prove. □
Going back to the very beginning of the sec on, we see that the only as-
sump on we would need about our distance func on f(t) is that it be con nu-
ous and differen able for t from 0 to 2 hours (both reasonable assump ons). By
the Mean Value Theorem, we are guaranteed a me during the trip where our
Notes:
139
instantaneous speed is 50 mph. This fact is used in prac ce. Some law enforce-
ment agencies monitor traffic speeds while in aircra . They do not measure
speed with radar, but rather by ming individual cars as they pass over lines
painted on the highway whose distances apart are known. The officer is able
to measure the average speed of a car between the painted lines; if that aver-
age speed is greater than the posted speed limit, the officer is assured that the
driver exceeded the speed limit at some me.
Note that the Mean Value Theorem is an existence theorem. It states that a
special value c exists, but it does not give any indica on about how to find it. It
turns out that when we need the Mean Value Theorem, existence is all we need.
Example 3.2.2 Using the Mean Value Theorem

Consider f(x) = x3 + 5x + 5 on [−3, 3]. Find c in [−3, 3] that sa sfies the Mean
Value Theorem.
S The average rate of change of f on [−3, 3] is:
y
f(3) − f(−3) 84
= = 14.
3 − (−3) 6
40
We want to find c such that f ′ (c) = 14. We find f ′ (x) = 3x2 + 5. We set this
20 equal to 14 and solve for x.
x f ′ (x) = 14
−3 −2 −1 1 2 3
−20 3x2 + 5 = 14
x2 = 3
. −40 √
x = ± 3 ≈ ±1.732
Figure 3.2.3: Demonstra ng the Mean
Value Theorem in Example 3.2.2. We have found 2 values c in [−3, 3] where the instantaneous rate of change
is equal to the average rate of change; the Mean Value Theorem guaranteed at
least one. In Figure 3.2.3 f is graphed with a dashed√ line represen ng the aver-
age rate of change; the lines tangent to f at x = ± 3 are also given. Note how
these lines are parallel (i.e., have the same slope) with the dashed line.
While the Mean Value Theorem has prac cal use (for instance, the speed
monitoring applica on men oned before), it is mostly used to advance other
theory. We will use it in the next sec on to relate the shape of a graph to its
deriva ve.
Notes:
140
Exercises 3.2
Terms and Concepts In Exercises 11 – 20, a func on f(x) and interval [a, b] are
given. Check if the Mean Value Theorem can be applied to f
on [a, b]; if so, find a value c in [a, b] guaranteed by the Mean
1. Explain in your own words what the Mean Value Theorem Value Theorem.
states.
11. f(x) = x2 + 3x − 1 on [−2, 2].
2. Explain in your own words what Rolle’s Theorem states.
12. f(x) = 5x2 − 6x + 8 on [0, 5].
√
13. f(x) = 9 − x2 on [0, 3].
Problems
√
14. f(x) = 25 − x on [0, 9].
In Exercises 3 – 10, a func on f(x) and interval [a, b] are given.
Check if Rolle’s Theorem can be applied to f on [a, b]; if so, find x2 − 9
c in [a, b] such that f ′ (c) = 0. 15. f(x) = on [0, 2].
x2 − 1
16. f(x) = ln x on [1, 5].

3. f(x) = 6 on [−1, 1].
17. f(x) = tan x on [−π/4, π/4].
4. f(x) = 6x on [−1, 1].
18. f(x) = x3 − 2x2 + x + 1 on [−2, 2].
5. f(x) = x + x − 6 on [−3, 2].
2
19. f(x) = 2x3 − 5x2 + 6x + 1 on [−5, 2].

6. f(x) = x2 + x − 2 on [−3, 2].
20. f(x) = sin−1 x on [−1, 1].
7. f(x) = x2 + x on [−2, 2].

Review
8. f(x) = sin x on [π/6, 5π/6].
21. Find the extreme values of f(x) = x2 − 3x + 9 on [−2, 5].
9. f(x) = cos x on [0, π].
22. Describe the cri cal points of f(x) = cos x.
1
10. f(x) = on [0, 2]. 23. Describe the cri cal points of f(x) = tan x.
x2 − 2x + 1
141
3.3 Increasing and Decreasing Func ons

Our study of “nice” func ons f in this chapter has so far focused on individual
y
points: points where f is maximal/minimal, points where f ′ (x) = 0 or f ′ does
not exist, and points c where f ′ (c) is the average rate of change of f on some
interval.
4
In this sec on we begin to study how func ons behave between special
points; we begin studying in more detail the shape of their graphs.
We start with an intui ve concept. Given the graph in Figure 3.3.1, where
2 would you say the func on is increasing? Decreasing? Even though we have
not defined these terms mathema cally, one likely answered that f is increasing
when x > 1 and decreasing when x < 1. We formally define these terms here.
x
. 1 2
Defini on 3.3.1 Increasing and Decreasing Func ons
Figure 3.3.1: A graph of a func on f used
to illustrate the concepts of increasing Let f be a func on defined on an interval I.
and decreasing.
1. f is increasing on I if for every a < b in I, f(a) < f(b).
2. f is decreasing on I if for every a < b in I, f(a) > f(b).
Informally, a func on is increasing if as x gets larger (i.e., looking le to right)

f(x) gets larger.
Our interest lies in finding intervals in the domain of f on which f is either
increasing or decreasing. Such informa on should seem useful. For instance, if
f describes the speed of an object, we might want to know when the speed was
increasing or decreasing (i.e., when the object was accelera ng vs. decelerat-
ing). If f describes the popula on of a city, we should be interested in when the
popula on is growing or declining.
To find such intervals, we again consider secant lines. Let f be an increasing,
y differen able func on on an open interval I, such as the one shown in Figure
2 (b, f(b)) 3.3.2, and let a < b be given in I. The secant line on the graph of f from x = a
to x = b is drawn; it has a slope of (f(b) − f(a))/(b − a). But note:
Average rate of
f(b) − f(a) numerator > 0 slope of the
1 ⇒ ⇒ ⇒ change of f on
b−a denominator > 0 secant line > 0 [a, b] is > 0.
(a, f(a)) We have shown mathema cally what may have already been obvious: when
x f is increasing, its secant lines will have a posi ve slope. Now recall the Mean
a 1 b 2
. Value Theorem guarantees that there is a number c, where a < c < b, such that
f(b) − f(a)
Figure 3.3.2: Examining the secant line of f ′ (c) = > 0.
an increasing func on. b−a
Notes:
142
By considering all such secant lines in I, we strongly imply that f ′ (x) > 0 on I. A
similar statement can be made for decreasing func ons.
Our above logic can be summarized as “If f is increasing, then f ′ is probably
posi ve.” Theorem 3.3.1 below turns this around by sta ng “If f ′ is posi ve,
then f is increasing.” This leads us to a method for finding when func ons are
increasing and decreasing.
Theorem 3.3.1 Test For Increasing/Decreasing Func ons

Let f be a con nuous func on on [a, b] and differen able on (a, b).
Note: Parts 1 & 2 of Theorem 3.3.1 also
1. If f ′ (c) > 0 for all c in (a, b), then f is increasing on [a, b]. hold if f ′ (c) = 0 for a finite number of
values of c in I.
2. If f ′ (c) < 0 for all c in (a, b), then f is decreasing on [a, b].
3. If f ′ (c) = 0 for all c in (a, b), then f is constant on [a, b].
Let f be differen able on an interval I and let a and b be in I where f ′ (a) > 0
and f ′ (b) < 0. If f ′ is con nuous on [a, b], it follows from the Intermediate Value
Theorem that there must be some value c between a and b where f ′ (c) = 0. (It
turns out that this is s ll true even if f ′ is not con nuous on [a, b].) This leads us
to the following method for finding intervals on which a func on is increasing or
decreasing.
Key Idea 3.3.1 Finding Intervals on Which f is Increasing or

Decreasing
Let f be a differen able func on on an interval I. To find intervals on
which f is increasing and decreasing:
1. Find the cri cal values of f. That is, find all c in I where f ′ (c) = 0
or f ′ is not defined.
2. Use the cri cal values to divide I into subintervals.
3. Pick any point p in each subinterval, and find the sign of f ′ (p).
(a) If f ′ (p) > 0, then f is increasing on that subinterval.
(b) If f ′ (p) < 0, then f is decreasing on that subinterval.
We demonstrate using this process in the following example.
Notes:
143
Example 3.3.1 Finding intervals of increasing/decreasing

Let f(x) = x3 + x2 − x + 1. Find intervals on which f is increasing or decreasing.
S Using Key Idea 3.3.1, we first find the cri cal values of f. We
have f ′ (x) = 3x2 + 2x − 1 = (3x − 1)(x + 1), so f ′ (x) = 0 when x = −1 and
when x = 1/3. f ′ is never undefined.
Since an interval was not specified for us to consider, we consider the en-
re domain of f which is (−∞, ∞). We thus break the whole real line into
three subintervals based on the two cri cal values we just found: (−∞, −1),
(−1, 1/3) and (1/3, ∞). This is shown in Figure 3.3.3.
f ′ > 0 incr f ′ < 0 decr f ′ > 0 incr

.
−1 1/3
Figure 3.3.3: Number line for f in Example 3.3.1.

We now pick a value p in each subinterval and find the sign of f ′ (p). All we
care about is the sign, so we do not actually have to fully compute f ′ (p); pick
“nice” values that make this simple.
y Subinterval 1, (−∞, −1): We (arbitrarily) pick p = −2. We can compute
10 f ′ (−2) directly: f ′ (−2) = 3(−2)2 + 2(−2) − 1 = 7 > 0. We conclude that f is
increasing on (−∞, −1).
Note we can arrive at the same conclusion without computa on. For in-
f ′ (x) stance, we could choose p = −100. The first term in f ′ (−100), i.e., 3(−100)2 is
5
clearly posi ve and very large. The other terms are small in comparison, so we
f(x) know f ′ (−100) > 0. All we need is the sign.
x
−2 −1 1/3 1 2 Subinterval 2, (−1, 1/3): We pick p = 0 since that value seems easy to deal
. with. f ′ (0) = −1 < 0. We conclude f is decreasing on (−1, 1/3).
Figure 3.3.4: A graph of f(x) in Example
Subinterval 3, (1/3, ∞): Pick an arbitrarily large value for p > 1/3 and note
3.3.1, showing where f is increasing and
decreasing.
that f ′ (p) = 3p2 + 2p − 1 > 0. We conclude that f is increasing on (1/3, ∞).
We can verify our calcula ons by considering Figure 3.3.4, where f is graphed.
The graph also presents f ′ ; note how f ′ > 0 when f is increasing and f ′ < 0
when f is decreasing.
One is jus fied in wondering why so much work is done when the graph
seems to make the intervals very clear. We give three reasons why the above
work is worthwhile.
First, the points at which f switches from increasing to decreasing are not
precisely known given a graph. The graph shows us something significant hap-
pens near x = −1 and x = 0.3, but we cannot determine exactly where from
the graph.
Notes:
144
One could argue that just finding cri cal values is important; once we know
the significant points are x = −1 and x = 1/3, the graph shows the increas-
ing/decreasing traits just fine. That is true. However, the technique prescribed
here helps reinforce the rela onship between increasing/decreasing and the
sign of f ′ . Once mastery of this concept (and several others) is obtained, one
finds that either (a) just the cri cal points are computed and the graph shows
all else that is desired, or (b) a graph is never produced, because determining
increasing/decreasing using f ′ is straigh orward and the graph is unnecessary.
So our second reason why the above work is worthwhile is this: once mastery
of a subject is gained, one has op ons for finding needed informa on. We are
working to develop mastery.
Finally, our third reason: many problems we face “in the real world” are very
complex. Solu ons are tractable only through the use of computers to do many
calcula ons for us. Computers do not solve problems “on their own,” however;
they need to be taught (i.e., programmed) to do the right things. It would be
beneficial to give a func on to a computer and have it return maximum and
minimum values, intervals on which the func on is increasing and decreasing,
the loca ons of rela ve maxima, etc. The work that we are doing here is easily
programmable. It is hard to teach a computer to “look at the graph and see if it
is going up or down.” It is easy to teach a computer to “determine if a number
is greater than or less than 0.”
In Sec on 3.1 we learned the defini on of rela ve maxima and minima and
found that they occur at cri cal points. We are now learning that func ons can
switch from increasing to decreasing (and vice–versa) at cri cal points. This new
understanding of increasing and decreasing creates a great method of determin-
ing whether a cri cal point corresponds to a maximum, minimum, or neither.
Imagine a func on increasing un l a cri cal point at x = c, a er which it de-
creases. A quick sketch helps confirm that f(c) must be a rela ve maximum. A
similar statement can be made for rela ve minimums. We formalize this con-
cept in a theorem.
Theorem 3.3.2 First Deriva ve Test

Let f be differen able on an interval I and let c be a cri cal number in I.
1. If the sign of f ′ switches from posi ve to nega ve at c, then f(c) is
a rela ve maximum of f.
2. If the sign of f ′ switches from nega ve to posi ve at c, then f(c) is
a rela ve minimum of f.
3. If f ′ is posi ve (or, nega ve) before and a er c, then f(c) is not a
rela ve extrema of f.
Notes:
145
y Example 3.3.2 Using the First Deriva ve Test

20
f(x) Find the intervals on which f is increasing and decreasing, and use the First
Deriva ve Test to determine the rela ve extrema of f, where
10
x2 + 3
f ′ (x) f(x) = .
x x−1
−4 −2 2 4
−10
S We start by no ng the domain of f: (−∞, 1) ∪ (1, ∞). Key
Idea 3.3.1 describes how to find intervals where f is increasing and decreasing
. −20 when the domain of f is an interval. Since the domain of f in this example is
the union of two intervals, we apply the techniques of Key Idea 3.3.1 to both
Figure 3.3.5: A graph of f(x) in Example
intervals of the domain of f.
3.3.2, showing where f is increasing and
decreasing.
Since f is not defined at x = 1, the increasing/decreasing nature of f could
switch at this value. We do not formally consider x = 1 to be a cri cal value of
f, but we will include it in our list of cri cal values that we find next.
Using the Quo ent Rule, we find
x2 − 2x − 3
f ′ (x) = .
(x − 1)2
We need to find the cri cal values of f; we want to know when f ′ (x) = 0 and
when f ′ is not defined. That la er is straigh orward: when the denominator
of f ′ (x) is 0, f ′ is undefined. That occurs when x = 1, which we’ve already
recognized as an important value.
f ′ (x) = 0 when the numerator of f ′ (x) is 0. That occurs when x2 − 2x − 3 =
(x − 3)(x + 1) = 0; i.e., when x = −1, 3.
We have found that f has two cri cal numbers, x = −1, 3, and at x = 1
something important might also happen. These three numbers divide the real
number line into 4 subintervals:
(−∞, −1), (−1, 1), (1, 3) and (3, ∞).
Pick a number p from each subinterval and test the sign of f ′ at p to determine
whether f is increasing or decreasing on that interval. Again, we do well to avoid
complicated computa ons; no ce that the denominator of f ′ is always posi ve
so we can ignore it during our work.
Interval 1, (−∞, −1): Choosing a very small number (i.e., a nega ve number
with a large magnitude) p returns p2 − 2p − 3 in the numerator of f ′ ; that will
be posi ve. Hence f is increasing on (−∞, −1).
Interval 2, (−1, 1): Choosing 0 seems simple: f ′ (0) = −3 < 0. We conclude
f is decreasing on (−1, 1).
Interval 3, (1, 3): Choosing 2 seems simple: f ′ (2) = −3 < 0. Again, f is
decreasing.
Notes:
146
Interval 4, (3, ∞): Choosing an very large number p from this subinterval will
give a posi ve numerator and (of course) a posi ve denominator. So f is increas-
ing on (3, ∞).
In summary, f is increasing on the intervals (−∞, −1) and (3, ∞) and is de-
creasing on the intervals (−1, 1) and (1, 3). Since at x = −1, the sign of f ′
switched from posi ve to nega ve, Theorem 3.3.2 states that f(−1) is a rela ve
maximum of f. At x = 3, the sign of f ′ switched from nega ve to posi ve, mean-
ing f(3) is a rela ve minimum. At x = 1, f is not defined, so there is no rela ve
extrema at x = 1.
rel. rel.
f ′ > 0 incr max f ′ < 0 decr f ′ < 0 decr min f ′ > 0 incr
.
−1 1 3
This is summarized in the number line shown in Figure 3.3.6. Also, Figure
3.3.5 shows a graph of f, confirming our calcula ons. This figure also shows
f ′ , again demonstra ng that f is increasing when f ′ > 0 and decreasing when
f ′ < 0.
One is o en tempted to think that func ons always alternate “increasing,

decreasing, increasing, decreasing,. . .” around cri cal values. Our previous ex-
ample demonstrated that this is not always the case. While x = 1 was not
technically a cri cal value, it was an important value we needed to consider.
We found that f was decreasing on “both sides of x = 1.”
We examine one more example.
Example 3.3.3 Using the First Deriva ve Test

Find the intervals on which f(x) = x8/3 − 4x2/3 is increasing and decreasing and
iden fy the rela ve extrema.
S We again start with taking a deriva ve. Since we know we

want to solve f ′ (x) = 0, we will do some algebra a er taking the deriva ve.
8 2
f(x) = x 3 − 4x 3
8 5 8 1
f ′ (x) = x 3 − x− 3
3 3
8 −1 ( 6 )
= x 3 x3 − 1
3
Notes:
147
8 −1 2
= x 3 (x − 1)
3
8 1
= x− 3 (x − 1)(x + 1).
3
This deriva on of f ′ shows that f ′ (x) = 0 when x = ±1 and f ′ is not de-
fined when x = 0. Thus we have 3 cri cal values, breaking the number line into
4 subintervals as shown in Figure 3.3.7.
Interval 1, (∞, −1): We choose p = −2; we can easily verify that f ′ (−2) <
0. So f is decreasing on (−∞, −1).
Interval 2, (−1, 0): Choose p = −1/2. Once more we prac ce finding the sign
of f ′ (p) without compu ng an actual value. We have f ′ (p) = (8/3)p−1/3 (p −
1)(p + 1); find the sign of each of the three terms.
8 −1
f ′ (p) = · p 3 · (p − 1) (p + 1) .
3 |{z} | {z } | {z }
<0 <0 >0
We have a “nega ve × nega ve × posi ve” giving a posi ve number; f is in-

creasing on (−1, 0).
Interval 3, (0, 1): We do a similar sign analysis as before, using p in (0, 1).
8 −1
f ′ (p) = · p 3 · (p − 1) (p + 1) .
3 |{z} | {z } | {z }
>0 <0 >0
We have 2 posi ve factors and one nega ve factor; f ′ (p) < 0 and so f is de-
creasing on (0, 1).
y
Interval 4, (1, ∞): Similar work to that done for the other three intervals shows
10
that f ′ (x) > 0 on (1, ∞), so f is increasing on this interval.
rel. rel. rel.

f(x) 5 f ′ < 0 decr min f ′ > 0 incr max f ′ < 0 decr min f ′ > 0 incr
f ′ (x)
.
−1 0 1
x
−3 −2 −1 1 2 3
.
Figure 3.3.8: A graph of f(x) in Example We conclude by sta ng that f is increasing on the intervals (−1, 0) and (1, ∞)
3.3.3, showing where f is increasing and and decreasing on the intervals (−∞, −1) and (0, 1). The sign of f ′ changes
decreasing. from nega ve to posi ve around x = −1 and x = 1, meaning by Theorem 3.3.2
that f(−1) and f(1) are rela ve minima of f. As the sign of f ′ changes from pos-
i ve to nega ve at x = 0, we have a rela ve maximum at f(0). Figure 3.3.8
.
Notes:
148
shows a graph of f, confirming our result. We also graph f ′ , highligh ng once

more that f is increasing when f ′ > 0 and is decreasing when f ′ < 0.
We have seen how the first deriva ve of a func on helps determine when
the func on is going “up” or “down.” In the next sec on, we will see how the
second deriva ve helps determine how the graph of a func on curves.
Notes:
149
Exercises 3.3
14. f(x) =
x2 + 1
1. In your own words describe what it means for a func on to In Exercises 15 – 24, a func on f(x) is given.
be increasing.
(a) Give the domain of f.
2. What does a decreasing func on “look like”? (b) Find the cri cal numbers of f.
(c) Create a number line to determine the intervals on
3. Sketch a graph of a func on on [0, 2] that is increasing, which f is increasing and decreasing.
where it is increasing “quickly” near x = 0 and increasing
(d) Use the First Deriva ve Test to determine whether
“slowly” near x = 2.
each cri cal point is a rela ve maximum, minimum,
or neither.
4. Give an example of a func on describing a situa on where
it is “bad” to be increasing and “good” to be decreasing.
15. f(x) = x2 + 2x − 3
5. T/F: Func ons always switch from increasing to decreasing,

16. f(x) = x3 + 3x2 + 3
or decreasing to increasing, at cri cal points.
2 17. f(x) = 2x3 + x2 − x + 3

6. A func on f has deriva ve f ′ (x) = (sin x + 2)ex +1 , where
f ′ (x) > 1 for all x. Is f increasing, decreasing, or can we not 18. f(x) = x3 − 3x2 + 3x − 1
tell from the given informa on?
1
19. f(x) =
x2 − 2x + 2
Problems
x2 − 4
20. f(x) =
In Exercises 7 – 14, a func on f(x) is given. x2 − 1
(a) Compute f ′ (x). x
21. f(x) =
′
(b) Graph f and f on the same axes (using technology is x2 − 2x − 8
permi ed) and verify Theorem 3.3.1.
(x − 2)2/3
22. f(x) =
x
7. f(x) = 2x + 3
23. f(x) = sin x cos x on (−π, π).
8. f(x) = x2 − 3x + 5
24. f(x) = x5 − 5x
9. f(x) = cos x
10. f(x) = tan x Review

11. f(x) = x3 − 5x2 + 7x − 1 25. Consider f(x) = x2 − 3x + 5 on [−1, 2]; find c guaranteed
by the Mean Value Theorem.
12. f(x) = 2x3 − x2 + x − 1
26. Consider f(x) = sin x on [−π/2, π/2]; find c guaranteed by
13. f(x) = x4 − 5x2 + 4 the Mean Value Theorem.
150
3.4 Concavity and the Second Deriva ve
y
30
Our study of “nice” func ons con nues. The previous sec on showed how the
first deriva ve of a func on, f ′ , can relay important informa on about f. We
20
now apply the same technique to f ′ itself, and learn what this tells us about f.
The key to studying f ′ is to consider its deriva ve, namely f ′′ , which is the
second deriva ve of f. When f ′′ > 0, f ′ is increasing. When f ′′ < 0, f ′ is 10
decreasing. f ′ has rela ve maxima and minima where f ′′ = 0 or is undefined.
This sec on explores how knowing informa on about f ′′ gives informa on . x
about f. −2 2
Concavity Figure 3.4.1: A func on f with a concave

up graph. No ce how the slopes of the
tangent lines, when looking from le to
We begin with a defini on, then explore its meaning.
right, are increasing.
Defini on 3.4.1 Concave Up and Concave Down

Note: We o en state that “f is concave
Let f be differen able on an interval I. The graph of f is concave up on I
up” instead of “the graph of f is concave
if f ′ is increasing. The graph of f is concave down on I if f ′ is decreasing. up” for simplicity.
If f ′ is constant then the graph of f is said to have no concavity.
The graph of a func on f is concave up when f ′ is increasing. That means as

one looks at a concave up graph from le to right, the slopes of the tangent lines
will be increasing. Consider Figure 3.4.1, where a concave up graph is shown Note: A mnemonic for remembering
along with some tangent lines. No ce how the tangent line on the le is steep, what concave up/down means is: “Con-
downward, corresponding to a small value of f ′ . On the right, the tangent line cave up is like a cup; concave down is like
is steep, upward, corresponding to a large value of f ′ . a frown.” It is admi edly terrible, but it
If a func on is decreasing and concave up, then its rate of decrease is slow- works.
ing; it is “leveling off.” If the func on is increasing and concave up, then the rate
of increase is increasing. The func on is increasing at a faster and faster rate.
Now consider a func on which is concave down. We essen ally repeat the
above paragraphs with slight varia on. y
The graph of a func on f is concave down when f ′ is decreasing. That means
as one looks at a concave down graph from le to right, the slopes of the tangent 30
lines will be decreasing. Consider Figure 3.4.2, where a concave down graph is
shown along with some tangent lines. No ce how the tangent line on the le 20
is steep, upward, corresponding to a large value of f ′ . On the right, the tangent
line is steep, downward, corresponding to a small value of f ′ .
10
If a func on is increasing and concave down, then its rate of increase is slow-
ing; it is “leveling off.” If the func on is decreasing and concave down, then the
rate of decrease is decreasing. The func on is decreasing at a faster and faster . x
rate. −2 2
Figure 3.4.2: A func on f with a concave

down graph. No ce how the slopes of the
Notes: tangent lines, when looking from le to
right, are decreasing.
151
Our defini on of concave up and concave down is given in terms of when

the first deriva ve is increasing or decreasing. We can apply the results of the
previous sec on and to find intervals on which a graph is concave up or down.
f ′ > 0, increasing f ′ < 0, decreasing
f ′′ < 0, c. down f ′′ < 0, c. down
That is, we recognize that f ′ is increasing when f ′′ > 0, etc.
. Theorem 3.4.1 Test for Concavity

Let f be twice differen able on an interval I. The graph of f is concave up
f ′ < 0, decreasing f ′ > 0, increasing
f ′′ > 0, c. up f ′′ > 0, c. up
if f ′′ > 0 on I, and is concave down if f ′′ < 0 on I.
If knowing where a graph is concave up/down is important, it makes sense

Figure 3.4.3: Demonstra ng the 4 ways that the places where the graph changes from one to the other is also important.
that concavity interacts with increas- This leads us to a defini on.
ing/decreasing, along with the rela on-
ships with the first and second deriva-
ves. Defini on 3.4.2 Point of Inflec on
A point of inflec on is a point on the graph of f at which the concavity
of f changes.
Note: Geometrically speaking, a func on

is concave up if its graph lies above its tan- Figure 3.4.4 shows a graph of a func on with inflec on points labeled.
gent lines. A func on is concave down if If the concavity of f changes at a point (c, f(c)), then f ′ is changing from
its graph lies below its tangent lines. increasing to decreasing (or, decreasing to increasing) at x = c. That means that
the sign of f ′′ is changing from posi ve to nega ve (or, nega ve to posi ve) at
x = c. This leads to the following theorem.
y
15 Theorem 3.4.2 Points of Inflec on
If (c, f(c)) is a point of inflec on on the graph of f, then either f ′′ (c) = 0
10
or f ′′ is not defined at c.
f ′′ > 0 f ′′ > 0
c. up c. up
5 f ′′ < 0 We have iden fied the concepts of concavity and points of inflec on. It is
c. down
now me to prac ce using these concepts; given a func on, we should be able
x
to find its points of inflec on and iden fy intervals on which it is concave up or
. 1 2 3 4 down. We do so in the following examples.
Figure 3.4.4: A graph of a func on with Example 3.4.1 Finding intervals of concave up/down, inflec on points
its inflec on points marked. The inter- Let f(x) = x3 − 3x + 1. Find the inflec on points of f and the intervals on which
vals where concave up/down are also in-
it is concave up/down.
dicated.
Notes:
152
S We start by finding f ′ (x) = 3x2 − 3 and f ′′ (x) = 6x. To find

f ′′ < 0 c. down f ′′ > 0 c. up
the inflec on points, we use Theorem 3.4.2 and find where f ′′ (x) = 0 or where
f ′′ is undefined. We find f ′′ is always defined, and is 0 only when x = 0. So the .
point (0, 1) is the only possible point of inflec on. 0
This possible inflec on point divides the real line into two intervals, (−∞, 0)
and (0, ∞). We use a process similar to the one used in the previous sec on to Figure 3.4.5: A number line determining
the concavity of f in Example 3.4.1.
determine increasing/decreasing. Pick any c < 0; f ′′ (c) < 0 so f is concave
down on (−∞, 0). Pick any c > 0; f ′′ (c) > 0 so f is concave up on (0, ∞). Since
the concavity changes at x = 0, the point (0, 1) is an inflec on point.
The number line in Figure 3.4.5 illustrates the process of determining con- y
cavity; Figure 3.4.6 shows a graph of f and f ′′ , confirming our results. No ce how
f is concave down precisely when f ′′ (x) < 0 and concave up when f ′′ (x) > 0.
f(x) 2 f ′′ (x)
Example 3.4.2 Finding intervals of concave up/down, inflec on points

Let f(x) = x/(x2 − 1). Find the inflec on points of f and the intervals on which x
−2 −1 1 2
it is concave up/down.
−2
S We need to find f ′ and f ′′ . Using the Quo ent Rule and sim-
plifying, we find .
−(1 + x2 ) 2x(x2 + 3) Figure 3.4.6: A graph of f(x) used in Ex-

f ′ (x) = and f ′′ (x) = . ample 3.4.1.
(x2 − 1)2 (x2 − 1)3
To find the possible points of inflec on, we seek to find where f ′′ (x) = 0 and
where f ′′ is not defined. Solving f ′′ (x) = 0 reduces to solving 2x(x2 + 3) = 0;
we find x = 0. We find that f ′′ is not defined when x = ±1, for then the
denominator of f ′′ is 0. We also note that f itself is not defined at x = ±1,
having a domain of (−∞, −1) ∪ (−1, 1) ∪ (1, ∞). Since the domain of f is the
union of three intervals, it makes sense that the concavity of f could switch across
intervals. We technically cannot say that f has a point of inflec on at x = ±1 as
they are not part of the domain, but we must s ll consider these x-values to be
important and will include them in our number line.
The important x-values at which concavity might switch are x = −1, x = 0
and x = 1, which split the number line into four intervals as shown in Figure
3.4.7. We determine the concavity on each. Keep in mind that all we are con-
cerned with is the sign of f ′′ on the interval.
Interval 1, (−∞, −1): Select a number c in this interval with a large magnitude
(for instance, c = −100). The denominator of f ′′ (x) will be posi ve. In the
numerator, the (c2 + 3) will be posi ve and the 2c term will be nega ve. Thus
the numerator is nega ve and f ′′ (c) is nega ve. We conclude f is concave down
on (−∞, −1).
Notes:
153
Interval 2, (−1, 0): For any number c in this interval, the term 2c in the numer-
y
ator will be nega ve, the term (c2 + 3) in the numerator will be posi ve, and
10
the term (c2 − 1)3 in the denominator will be nega ve. Thus f ′′ (c) > 0 and f is
concave up on this interval.
5 f(x) f ′′ (x)
Interval 3, (0, 1): Any number c in this interval will be posi ve and “small.” Thus
x
the numerator is posi ve while the denominator is nega ve. Thus f ′′ (c) < 0
−2 2 and f is concave down on this interval.
−5 Interval 4, (1, ∞): Choose a large value for c. It is evident that f ′′ (c) > 0, so we
conclude that f is concave up on (1, ∞).
. −10
f ′′ < 0 c. down f ′′ > 0 c. up f ′′ < 0 c. down f ′′ > 0 c. up
′′ .
Figure 3.4.8: A graph of f(x) and f (x) in
Example 3.4.2. −1 0 1

We conclude that f is concave up on (−1, 0) and (1, ∞) and concave down
on (−∞, −1) and (0, 1). There is only one point of inflec on, (0, 0), as f is not
defined at x = ±1. Our work is confirmed by the graph of f in Figure 3.4.8. No-
ce how f is concave up whenever f ′′ is posi ve, and concave down when f ′′ is
nega ve.
Recall that rela ve maxima and minima of f are found at cri cal points of
f; that is, they are found when f ′ (x) = 0 or when f ′ is undefined. Likewise,
the rela ve maxima and minima of f ′ are found when f ′′ (x) = 0 or when f ′′ is
undefined; note that these are the inflec on points of f.
What does a “rela ve maximum of f ′ ” mean? The deriva ve measures the
rate of change of f; maximizing f ′ means finding where f is increasing the most –
where f has the steepest tangent line. A similar statement can be made for min-
imizing f ′ ; it corresponds to where f has the steepest nega vely–sloped tangent
y
line.
20
We u lize this concept in the next example.
S(t)
15
Example 3.4.3 Understanding inflec on points
10 The sales of a certain product over a three-year span are modeled by S(t) =
t4 − 8t2 + 20, where t is the me in years, shown in Figure 3.4.9. Over the first
5 two years, sales are decreasing. Find the point at which sales are decreasing at
their greatest rate.
. t
1 2 3 S We want to maximize the rate of decrease, which is to say,
we want to find where S ′ has a minimum. To do this, we find where S ′′ is 0. We
Figure 3.4.9: A graph of S(t) in Example
find S ′ (t)√= 4t3 − 16t and S ′′ (t) = 12t2 − 16. Se ng S ′′ (t) = 0 and solving, we
3.4.3, modeling the sale of a product over
get t = 4/3 ≈ 1.16 (we ignore the nega ve value of t since it does not lie in
me.
Notes:
154
the domain of our func on S).

y
This is both the inflec on point and the point of maximum decrease. This
20
is the point at which things first start looking up for the company. A er the S(t)
inflec on point, it will s ll take some me before sales start to increase, but at
least sales are not decreasing quite as quickly as they had been. 10
A graph of S(t) and S ′ (t) is given in Figure 3.4.10. When S ′ (t) < 0, sales are
decreasing; note how at t ≈ 1.16, S ′ (t) is minimized. That is, sales are decreas- t
1 2 3
ing at the fastest rate at t ≈ 1.16. On the interval of (1.16, 2), S is decreasing
but concave up, so the decline in sales is “leveling off.” −10 S ′ (t)
.
Not every cri cal point corresponds to a rela ve extrema; f(x) = x3 has a
cri cal point at (0, 0) but no rela ve maximum or minimum. Likewise, just be- Figure 3.4.10: A graph of S(t) in Example
cause f ′′ (x) = 0 we cannot conclude concavity changes at that point. We were 3.4.3 along with S ′ (t).
careful before to use terminology “possible point of inflec on” since we needed
to check to see if the concavity changed. The canonical example of f ′′ (x) = 0
without concavity changing is f(x) = x4 . At x = 0, f ′′ (x) = 0 but f is always y
concave up, as shown in Figure 3.4.11.
1
The Second Deriva ve Test
The first deriva ve of a func on gave us a test to find if a cri cal value cor- 0.5
responded to a rela ve maximum, minimum, or neither. The second deriva ve
gives us another way to test if a cri cal point is a local maximum or minimum.
The following theorem officially states something that is intui ve: if a cri cal
x
value occurs in a region where a func on f is concave up, then that cri cal value .−1 −0.5 0.5 1
must correspond to a rela ve minimum of f, etc. See Figure 3.4.12 for a visual-
iza on of this. Figure 3.4.11: A graph of f(x) = x4 .
Clearly f is always concave up, despite the
fact that f ′′ (x) = 0 when x = 0. It this
Theorem 3.4.3 The Second Deriva ve Test example, the possible point of inflec on
(0, 0) is not a point of inflec on.
Let c be a cri cal value of f where f ′′ (c) is defined.
1. If f ′′ (c) > 0, then f has a local minimum at (c, f(c)). y
′′ 10
2. If f (c) < 0, then f has a local maximum at (c, f(c)).
c. down
⇒ rel. max 5
The Second Deriva ve Test relates to the First Deriva ve Test in the following x
way. If f ′′ (c) > 0, then the graph is concave up at a cri cal point c and f ′ itself −2 −1 1 2
is growing. Since f ′ (c) = 0 and f ′ is growing at c, then it must go from nega ve −5 c. up
to posi ve at c. This means the func on goes from decreasing to increasing, in- ⇒ rel. min
dica ng a local minimum at c. −10

.
Figure 3.4.12: Demonstra ng the fact
that rela ve maxima occur when the
graph is concave down and rela ve min-
Notes:
ima occur when the graph is concave up.
155
40 Example 3.4.4 Using the Second Deriva ve Test

f ′′ (10) > 0 Let f(x) = 100/x + x. Find the cri cal points of f and use the Second Deriva ve
20 Test to label them as rela ve maxima or minima.
x
−20 −10 10 20 S We find f ′ (x) = −100/x2 + 1 and f ′′ (x) = 200/x3 . We set
−20 f (x) = 0 and solve for x to find the cri cal values (note that f ′ is not defined at
′
f ′′ (−10) < 0 x = 0, but neither is f so this is not a cri cal value.) We find the cri cal values
−40 are x = ±10. Evalua ng f ′′ at x = 10 gives 0.1 > 0, so there is a local minimum
. at x = 10. Evalua ng f ′′ (−10) = −0.1 < 0, determining a rela ve maximum
Figure 3.4.13: A graph of f(x) in Example at x = −10. These results are confirmed in Figure 3.4.13.
3.4.4. The second deriva ve is evaluated
at each cri cal point. When the graph is We have been learning how the first and second deriva ves of a func on
concave up, the cri cal point represents relate informa on about the graph of that func on. We have found intervals of
a local minimum; when the graph is con- increasing and decreasing, intervals where the graph is concave up and down,
cave down, the cri cal point represents a along with the loca ons of rela ve extrema and inflec on points. In Chapter 1
local maximum. we saw how limits explained asympto c behavior. In the next sec on we com-
bine all of this informa on to produce accurate sketches of func ons.
Notes:
156
Exercises 3.4
Terms and Concepts In Exercises 15 – 28, a func on f(x) is given.
(a) Find the possible points of inflec on of f.
1. Sketch a graph of a func on f(x) that is concave up on (0, 1) (b) Create a number line to determine the intervals on
and is concave down on (1, 2). which f is concave up or concave down.
2. Sketch a graph of a func on f(x) that is: 15. f(x) = x2 − 2x + 1
(a) Increasing, concave up on (0, 1), 16. f(x) = −x2 − 5x + 7
(b) increasing, concave down on (1, 2), 17. f(x) = x3 − x + 1
(c) decreasing, concave down on (2, 3) and 18. f(x) = 2x3 − 3x2 + 9x + 5
(d) increasing, concave down on (3, 4).
x4 x3
19. f(x) = + − 2x + 3
4 3
3. Is is possible for a func on to be increasing and concave
down on (0, ∞) with a horizontal asymptote of y = 1? If 20. f(x) = −3x4 + 8x3 + 6x2 − 24x + 2
so, give a sketch of such a func on.
21. f(x) = x4 − 4x3 + 6x2 − 4x + 1
4. Is is possible for a func on to be increasing and concave up
on (0, ∞) with a horizontal asymptote of y = 1? If so, give 22. f(x) = sec x on (−3π/2, 3π/2)
a sketch of such a func on.
1
23. f(x) =
x2 + 1
Problems x
24. f(x) =
x2 − 1
In Exercises 5 – 14, a func on f(x) is given.
25. f(x) = sin x + cos x on (−π, π)
(a) Compute f ′′ (x).
26. f(x) = x2 ex
(b) Graph f and f ′′ on the same axes (using technology is
permi ed) and verify Theorem 3.4.1. 27. f(x) = x2 ln x
2
5. f(x) = −7x + 3 28. f(x) = e−x
6. f(x) = −4x2 + 3x − 8 In Exercises 29 – 42, a func on f(x) is given. Find the cri cal
points of f and use the Second Deriva ve Test, when possi-
ble, to determine the rela ve extrema. (Note: these are the
7. f(x) = 4x2 + 3x − 8
same func ons as in Exercises 15 – 28.)
8. f(x) = x3 − 3x2 + x − 1 29. f(x) = x2 − 2x + 1
9. f(x) = −x3 + x2 − 2x + 5 30. f(x) = −x2 − 5x + 7
10. f(x) = sin x 31. f(x) = x3 − x + 1
11. f(x) = tan x 32. f(x) = 2x3 − 3x2 + 9x + 5
x4 x3
1 33. f(x) = + − 2x + 3
12. f(x) = 2 4 3
x +1
34. f(x) = −3x4 + 8x3 + 6x2 − 24x + 2
1
13. f(x) =
x
35. f(x) = x4 − 4x3 + 6x2 − 4x + 1
1
14. f(x) = 36. f(x) = sec x on (−3π/2, 3π/2)
x2
157
1 x4 x3
37. f(x) = 47. f(x) = + − 2x + 3
x2 + 1 4 3
x
38. f(x) = 48. f(x) = −3x4 + 8x3 + 6x2 − 24x + 2
x2 − 1
39. f(x) = sin x + cos x on (−π, π) 49. f(x) = x4 − 4x3 + 6x2 − 4x + 1
40. f(x) = x2 ex
50. f(x) = sec x on (−3π/2, 3π/2)
41. f(x) = x2 ln x
1
51. f(x) =
42. f(x) = e−x
2
x2 + 1
x
In Exercises 43 – 56, a func on f(x) is given. Find the x val- 52. f(x) =
ues where f ′ (x) has a rela ve maximum or minimum. (Note: x2 − 1
these are the same func ons as in Exercises 15 – 28.)
53. f(x) = sin x + cos x on (−π, π)
43. f(x) = x2 − 2x + 1
54. f(x) = x2 ex
44. f(x) = −x2 − 5x + 7
55. f(x) = x2 ln x
45. f(x) = x3 − x + 1
2
46. f(x) = 2x3 − 3x2 + 9x + 5 56. f(x) = e−x
158
3.5 Curve Sketching
3.5 Curve Sketching

We have been learning how we can understand the behavior of a func on based
on its first and second deriva ves. While we have been trea ng the proper es
of a func on separately (increasing and decreasing, concave up and concave
down, etc.), we combine them here to produce an accurate graph of the func on
without plo ng lots of extraneous points.
Why bother? Graphing u li es are very accessible, whether on a computer,
a hand–held calculator, or a smartphone. These resources are usually very fast
and accurate. We will see that our method is not par cularly fast – it will require
me (but it is not hard). So again: why bother?
We are a emp ng to understand the behavior of a func on f based on the
informa on given by its deriva ves. While all of a func on’s deriva ves relay
informa on about it, it turns out that “most” of the behavior we care about is
explained by f ′ and f ′′ . Understanding the interac ons between the graph of f
and f ′ and f ′′ is important. To gain this understanding, one might argue that all
that is needed is to look at lots of graphs. This is true to a point, but is somewhat
similar to sta ng that one understands how an engine works a er looking only at
pictures. It is true that the basic ideas will be conveyed, but “hands–on” access
increases understanding.
The following Key Idea summarizes what we have learned so far that is ap-
plicable to sketching graphs of func ons and gives a framework for pu ng that
informa on together. It is followed by several examples.
Key Idea 3.5.1 Curve Sketching

To produce an accurate sketch a given func on f, consider the following
steps.
1. Find the domain of f. Generally, we assume that the domain is the
en re real line then find restric ons, such as where a denominator
is 0 or where nega ves appear under the radical.
2. Find the cri cal values of f.
3. Find the possible points of inflec on of f.
4. Find the loca on of any ver cal asymptotes of f (usually done in

conjunc on with item 1 above).
5. Consider the limits lim f(x) and lim f(x) to determine the end
x→−∞ x→∞
behavior of the func on.
(con nued)
Notes:
159
Key Idea 3.5.1 Curve Sketching – Con nued
6. Create a number line that includes all cri cal points, possible
points of inflec on, and loca ons of ver cal asymptotes. For each
interval created, determine whether f is increasing or decreasing,
concave up or down.
7. Evaluate f at each cri cal point and possible point of inflec on.
Plot these points on a set of axes. Connect these points with curves
exhibi ng the proper concavity. Sketch asymptotes and x and y
intercepts where applicable.
Example 3.5.1 Curve sketching

Use Key Idea 3.5.1 to sketch f(x) = 3x3 − 10x2 + 7x + 5.
S We follow the steps outlined in the Key Idea.
1. The domain of f is the en re real line; there are no values x for which f(x)
is not defined.
2. Find the cri cal values of f. We compute f ′ (x) = 9x2 − 20x + 7. Use the
Quadra c Formula to find the roots of f ′ :
√
20 ± (−20)2 − 4(9)(7) 1( √ )
x= = 10 ± 37 ⇒ x ≈ 0.435, 1.787.
2(9) 9
3. Find the possible points of inflec on of f. Compute f ′′ (x) = 18x − 20. We

have
f ′′ (x) = 0 ⇒ x = 10/9 ≈ 1.111.
4. There are no ver cal asymptotes.
5. We determine the end behavior using limits as x approaches ±infinity.
lim f(x) = −∞ lim f(x) = ∞.

x→−∞ x→∞
We do not have any horizontal asymptotes.

√
6. We place the values x = (10 ± 37)/9 and x = 10/9 on a number
line, as shown in Figure 3.5.1. We mark each subinterval as increasing or
Notes:
160
3.5 Curve Sketching
decreasing, concave up or down, using the techniques used in Sec ons

3.3 and 3.4.
y
10
f ′ > 0 incr f ′ < 0 decr f ′ < 0 decr f ′ > 0 incr
f ′′ < 0 c. down f ′′ < 0 c. down f ′′ > 0 c. up f ′′ < 0 c. up
.
5
√ √
9 (10 −
1
37) 10
9 ≈ 1.111 1
9 (10 + 37)
≈ 0.435 ≈ 1.787
x
Figure 3.5.1: Number line for f in Example 3.5.1. −1 1 2 3
7. We plot the appropriate points on axes as shown in Figure 3.5.2(a) and . −5

connect the points with straight lines. In Figure 3.5.2(b) we adjust these (a)
lines to demonstrate the proper concavity. Our curve crosses the y axis at
y = 5 and crosses the x axis near x = −0.424. In Figure 3.5.2(c) we show y
a graph of f drawn with a computer program, verifying the accuracy of our 10
sketch.
5
x2 − x − 2
Sketch f(x) = 2 .
x −x−6 x
−1 1 2 3
S We again follow the steps outlined in Key Idea 3.5.1.

. −5
1. In determining the domain, we assume it is all real numbers and look for
(b)
restric ons. We find that at x = −2 and x = 3, f(x) is not defined. So the
domain of f is D = {real numbers x | x ̸= −2, 3}. y
2. To find the cri cal values of f, we first find f ′ (x). Using the Quo ent Rule, 10
we find
−8x + 4 −8x + 4
f ′ (x) = 2 = . 5
(x + x − 6) 2 (x − 3)2 (x + 2)2
f ′ (x) = 0 when x = 1/2, and f ′ is undefined when x = −2, 3. Since f ′
x
is undefined only when f is, these are not cri cal values. The only cri cal −1 1 2 3
value is x = 1/2.
3. To find the possible points of inflec on, we find f ′′ (x), again employing . −5
the Quo ent Rule: (c)

24x2 − 24x + 56
f ′′ (x) = . Figure 3.5.2: Sketching f in Example 3.5.1.
(x − 3)3 (x + 2)3
We find that f ′′ (x) is never 0 (se ng the numerator equal to 0 and solving
for x, we find the only roots to this quadra c are imaginary) and f ′′ is
Notes:
161
y
undefined when x = −2, 3. Thus concavity will possibly only change at
5
x = −2 and x = 3.
4. The ver cal asymptotes of f are at x = −2 and x = 3, the places where f
is undefined.
x 5. There is a horizontal asymptote of y = 1, as lim f(x) = 1 and lim f(x) =

x→−∞ x→∞
−4 −2 2 4
1.
6. We place the values x = 1/2, x = −2 and x = 3 on a number line as
shown in Figure 3.5.3. We mark in each interval whether f is increasing or
. −5
decreasing, concave up or down. We see that f has a rela ve maximum at
x = 1/2; concavity changes only at the ver cal asymptotes.
(a)
y f ′ > 0 incr f ′ > 0 incr f ′ < 0 decr f ′ < 0 decr

5 f ′′ > 0 c. up f ′′ < 0 c. down f ′′ < 0 c. down f ′′ > 0 c. up
.
−2 1 3
2
x Figure 3.5.3: Number line for f in Example 3.5.2.

−4 −2 2 4
7. In Figure 3.5.4(a), we plot the points from the number line on a set of
axes and connect the points with straight lines to get a general idea of
.. −5 what the func on looks like (these lines effec vely only convey increas-
ing/decreasing informa on). In Figure 3.5.4(b), we adjust the graph with
(b) the appropriate concavity. We also show f crossing the x axis at x = −1
and x = 2.
y
5
Figure 3.5.4(c) shows a computer generated graph of f, which verifies the accu-
racy of our sketch.

5(x − 2)(x + 1)
x Sketch f(x) = .
−4 −2 2 4 x2 + 2x + 4
S We again follow Key Idea 3.5.1.

−5 1. We assume that the domain of f is all real numbers and consider restric-
.
ons. The only restric ons come when the denominator is 0, but this
. never occurs. Therefore the domain of f is all real numbers, R.
(c)
2. We find the cri cal values of f by se ng f ′ (x) = 0 and solving for x. We
Figure 3.5.4: Sketching f in Example 3.5.2.
find
15x(x + 4)
f ′ (x) = 2 ⇒ f ′ (x) = 0 when x = −4, 0.
(x + 2x + 4)2
Notes:
162
3.5 Curve Sketching
3. We find the possible points of inflec on by solving f ′′ (x) = 0 for x. We

find
30x3 + 180x2 − 240
f ′′ (x) = − . y
(x2 + 2x + 4)3
The cubic in the numerator does not factor very “nicely.” We instead ap-
proximate the roots at x = −5.759, x = −1.305 and x = 1.064. 5
4. There are no ver cal asymptotes.
5. We have a horizontal asymptote of y = 5, as lim f(x) = lim f(x) = 5. x

x→−∞ x→∞
−5 5
6. We place the cri cal points and possible points on a number line as shown
.
in Figure 3.5.5 and mark each interval as increasing/decreasing, concave
up/down appropriately. (a)
y
f ′ > 0 incr f ′ > 0 incr f ′ < 0 decr f ′ < 0 decr f ′ > 0 incr f ′ > 0 decr
f ′′ > 0 c. up f ′′ < 0 c. down f ′′ < 0 c. down f ′′ > 0 c. up f ′′ > 0 c. up f ′′ < 0 c. down
.
−5.579 −4 −1.305 0 1.064 5
7. In Figure 3.5.6(a) we plot the significant points from the number line as x
well as the two roots of f, x = −1 and x = 2, and connect the points −5 5
with straight lines to get a general impression about the graph. In Figure .
3.5.6(b), we add concavity. Figure 3.5.6(c) shows a computer generated
graph of f, affirming our results. (b)
y
In each of our examples, we found a few, significant points on the graph of
f that corresponded to changes in increasing/decreasing or concavity. We con-
nected these points with straight lines, then adjusted for concavity, and finished 5
by showing a very accurate, computer generated graph.
Why are computer graphics so good? It is not because computers are “smart-
er” than we are. Rather, it is largely because computers are much faster at com-
pu ng than we are. In general, computers graph func ons much like most stu- x
−5 5
dents do when first learning to draw graphs: they plot equally spaced points,
then connect the dots using lines. By using lots of points, the connec ng lines .
are short and the graph looks smooth.
(c)
This does a fine job of graphing in most cases (in fact, this is the method
used for many graphs in this text). However, in regions where the graph is very
Figure 3.5.6: Sketching f in Example 3.5.3.
“curvy,” this can generate no ceable sharp edges on the graph unless a large
number of points are used. High quality computer algebra systems, such as
Notes:
163
Mathema ca, use special algorithms to plot lots of points only where the graph
is “curvy.”
In Figure 3.5.7, a graph of y = sin x is given, generated by Mathema ca.
The small points represent each of the places Mathema ca sampled the func-
on. No ce how at the “bends” of sin x, lots of points are used; where sin x is
rela vely straight, fewer points are used. (Many points are also used at the end-
points to ensure the “end behavior” is accurate.) In fact, in the interval of length
0.2 centered around π/2, Mathema ca plots 72 of the 431 points plo ed; that
is, it plots about 17% of its points in a subinterval that accounts for about 3% of
the total interval length.
1.0
0.5
1 2 3 4 5 6
-0.5
-1.0
Figure 3.5.7: A graph of y = sin x generated by Mathema ca.
How does Mathema ca know where the graph is “curvy”? Calculus. When
we study curvature in a later chapter, we will see how the first and second
deriva ves of a func on work together to provide a measurement of “curvi-
ness.” Mathema ca employs algorithms to determine regions of “high curva-
ture” and plots extra points there.
Again, the goal of this sec on is not “How to graph a func on when there
is no computer to help.” Rather, the goal is “Understand that the shape of the
graph of a func on is largely determined by understanding the behavior of the
func on at a few key places.” In Example 3.5.3, we were able to accurately sketch
a complicated graph using only 5 points and knowledge of asymptotes!
There are many applica ons of our understanding of deriva ves beyond curve
sketching. The next chapter explores some of these applica ons, demonstrat-
ing just a few kinds of problems that can be solved with a basic knowledge of
differen a on.
Notes:
164
Exercises 3.5
Terms and Concepts 15. f(x) = x3 + 3x2 + 3x + 1
1. Why is sketching curves by hand beneficial even though 16. f(x) = x3 − x2 − x + 1

technology is ubiquitous?
17. f(x) = (x − 2) ln(x − 2)
2. What does “ubiquitous” mean?
18. f(x) = (x − 2)2 ln(x − 2)
3. T/F: When sketching graphs of func ons, it is useful to find
the cri cal points.
x2 − 4
19. f(x) =
x2
the possible points of inflec on. x2 − 4x + 3
20. f(x) =
x2 − 6x + 8
the horizontal and ver cal asymptotes.
x2 − 2x + 1
21. f(x) =
x2 − 6x + 8
6. T/F: When sketching graphs of func ons, one need not plot
any points at all. √
22. f(x) = x x + 1
23. f(x) = x2 ex
Problems
In Exercises 7 – 12, prac ce using Key Idea 3.5.1 by applying 24. f(x) = sin x cos x on [−π, π]
the principles to the given func ons with familiar graphs.
25. f(x) = (x − 3)2/3 + 2
7. f(x) = 2x + 4
(x − 1)2/3
26. f(x) =
8. f(x) = −x2 + 1 x
9. f(x) = sin x In Exercises 27 – 30, a func on with the parameters a and b

are given. Describe the cri cal points and possible points of
10. f(x) = ex inflec on of f in terms of a and b.
1 a
11. f(x) = 27. f(x) =
x x2 + b 2
1 28. f(x) = ax2 + bx + 1

12. f(x) =
x2
29. f(x) = sin(ax + b)
In Exercises 13 – 26, sketch a graph of the given func on us-
ing Key Idea 3.5.1. Show all work; check your answer with
technology. 30. f(x) = (x − a)(x − b)
dy
13. f(x) = x3 − 2x2 + 4x + 1 31. Given x2 + y2 = 1, use implicit differen a on to find dx
2
and ddx2y . Use this informa on to jus fy the sketch of the
3 2
14. f(x) = −x + 5x − 3x + 2 unit circle.
165
4: A
D
In Chapter 3, we learned how the first and second deriva ves of a func on influ-
ence its graph. In this chapter we explore other applica ons of the deriva ve.
y
1
4.1 Newton’s Method
0.5
Solving equa ons is one of the most important things we do in mathema cs,
yet we are surprisingly limited in what we can solve analy cally. For instance,
x
equa ons as simple as x5 + x + 1 = 0 or cos x = x cannot be solved by algebraic x0 x1
methods in terms of familiar func ons. Fortunately, there are methods that
−0.5
can give us approximate solu ons to equa ons like these. These methods can
usually give an approxima on correct to as many decimal places as we like. In
−1
.
Sec on 1.5 we learned about the Bisec on Method. This sec on focuses on
another technique (which generally works faster), called Newton’s Method. (a)
y
Newton’s Method is built around tangent lines. The main idea is that if x is
1
sufficiently close to a root of f(x), then the tangent line to the graph at (x, f(x))
will cross the x-axis at a point closer to the root than x.
0.5
We start Newton’s Method with an ini al guess about roughly where the
root is. Call this x0 . (See Figure 4.1.1(a).) Draw the tangent line to the graph at
x
(x0 , f(x0 )) and see where it meets the x-axis. Call this point x1 . Then repeat the x0 x2 x1
process – draw the tangent line to the graph at (x1 , f(x1 )) and see where it meets
−0.5
the x-axis. (See Figure 4.1.1(b).) Call this point x2 . Repeat the process again to
get x3 , x4 , etc. This sequence of points will o en converge rather quickly to a
−1
.
root of f.
We can use this geometric process to create an algebraic process. Let’s look (b)
y
at how we found x1 . We started with the tangent line to the graph at (x0 , f(x0 )).
The slope of this tangent line is f ′ (x0 ) and the equa on of the line is 1
y = f ′ (x0 )(x − x0 ) + f(x0 ). 0.5
This line crosses the x-axis when y = 0, and the x–value where it crosses is what x
x0 x2 x3 x1
we called x1 . So let y = 0 and replace x with x1 , giving the equa on:
−0.5
0 = f ′ (x0 )(x1 − x0 ) + f(x0 ).
−1
.
Now solve for x1 :
f(x0 ) (c)
x1 = x0 − .
f ′ (x0 ) Figure 4.1.1: Demonstra ng the geo-
metric concept behind Newton’s Method.
Note how x3 is very close to a solu on to
f(x) = 0.
Chapter 4 Applica ons of the Deriva ve
Since we repeat the same geometric process to find x2 from x1 , we have
f(x1 )
x2 = x1 − .
f ′ (x1 )
In general, given an approxima on xn , we can find the next approxima on, xn+1
as follows:
f(xn )
xn+1 = xn − .
f ′ (xn )
We summarize this process as follows.
Key Idea 4.1.1 Newton’s Method

Let f be a differen able func on on an interval I with a root in I. To ap-
proximate the value of the root, accurate to d decimal places:
1. Choose a value x0 as an ini al approxima on of the root. (This is

o en done by looking at a graph of f.)
Note: Newton’s Method is not infalli-
2. Create successive approxima ons itera vely; given an approxima-
ble. The sequence of approximate values
on xn , compute the next approxima on xn+1 as
may not converge, or it may converge so
slowly that one is “tricked” into thinking a
f(xn )
certain approxima on is be er than it ac- xn+1 = xn − .
tually is. These issues will be discussed at f ′ (xn )
the end of the sec on.
3. Stop the itera ons when successive approxima ons do not differ
in the first d places a er the decimal point.
Let’s prac ce Newton’s Method with a concrete example.
Example 4.1.1 Using Newton’s Method

Approximate the real root of x3 − x2 − 1 = 0, accurate to the first 3 places a er
the decimal, using Newton’s Method and an ini al approxima on of x0 = 1.
S To begin, we compute f ′ (x) = 3x2 − 2x. Then we apply the
Notes:
168
Newton’s Method algorithm, outlined in Key Idea 4.1.1.
f(1) 13 − 12 − 1
x1 = 1 − = 1 − = 2,
f ′ (1) 3 · 12 − 2 · 1
f(2) 23 − 22 − 1
x2 =2− ′ =2− = 1.625,
f (2) 3 · 22 − 2 · 2
f(1.625) 1.6253 − 1.6252 − 1
x3 = 1.625 − ′ = 1.625 − ≈ 1.48579. y
f (1.625) 3 · 1.6252 − 2 · 1.625 0.5
f(1.48579)
x4 = 1.48579 − ′ ≈ 1.46596
f (1.48579) x
0.5 1 1.5
f(1.46596)
x5 = 1.46596 − ′ ≈ 1.46557
f (1.46596) −0.5
We performed 5 itera ons of Newton’s Method to find a root accurate to the

−1
first 3 places a er the decimal; our final approxima on is 1.465. The exact value
of the root, to six decimal places, is 1.465571; It turns out that our x5 is accurate
−1.5.
to more than just 3 decimal places.
A graph of f(x) is given in Figure 4.1.2. We can see from the graph that our Figure 4.1.2: A graph of f(x) = x3 − x2 − 1
ini al approxima on of x0 = 1 was not par cularly accurate; a closer guess in Example 4.1.1.
would have been x0 = 1.5. Our choice was based on ease of ini al calcula on,
and shows that Newton’s Method can be robust enough that we do not have to
make a very accurate ini al approxima on.
We can automate this process on a calculator that has an Ans key that re-
turns the result of the previous calcula on. Start by pressing 1 and then Enter.
(We have just entered our ini al guess, x0 = 1.) Now compute
f(Ans)
Ans −
f ′ (Ans)
by entering the following and repeatedly press the Enter key:
Ans-(Ans^3-Ans^2-1)/(3*Ans^2-2*Ans)
Each me we press the Enter key, we are finding the successive approxima ons,
x1 , x2 , …, and each one is ge ng closer to the root. In fact, once we get past
around x7 or so, the approxima ons don’t appear to be changing. They actually
are changing, but the change is far enough to the right of the decimal point that
it doesn’t show up on the calculator’s display. When this happens, we can be
pre y confident that we have found an accurate approxima on.
Using a calculator in this manner makes the calcula ons simple; many iter-
a ons can be computed very quickly.
Notes:
169
Example 4.1.2 Using Newton’s Method to find where func ons intersect
Use Newton’s Method to approximate a solu on to cos x = x, accurate to 5
places a er the decimal.
S Newton’s Method provides a method of solving f(x) = 0; it

is not (directly) a method for solving equa ons like f(x) = g(x). However, this is
not a problem; we can rewrite the la er equa on as f(x) − g(x) = 0 and then
y
use Newton’s Method.
1
So we rewrite cos x = x as cos x − x = 0. Wri en this way, we are finding
a root of f(x) = cos x − x. We compute f ′ (x) = − sin x − 1. Next we need a
0.5
star ng value, x0 . Consider Figure 4.1.3, where f(x) = cos x − x is graphed. It
seems that x0 = 0.75 is pre y close to the root, so we will use that as our x0 .
x
−1 −0.5 0.5 1 (The figure also shows the graphs of y = cos x and y = x, drawn with dashed
lines. Note how they intersect at the same x value as when f(x) = 0.)
−0.5
We now compute x1 , x2 , etc. The formula for x1 is
. −1 cos(0.75) − 0.75
x1 = 0.75 − ≈ 0.7391111388.
− sin(0.75) − 1
Figure 4.1.3: A graph of f(x) = cos x − x
used to find an ini al approxima on of its Apply Newton’s Method again to find x2 :
root.
cos(0.7391111388) − 0.7391111388
x2 = 0.7391111388 − ≈ 0.7390851334.
− sin(0.7391111388) − 1
We can con nue this way, but it is really best to automate this process. On a cal-
culator with an Ans key, we would start by pressing 0.75, then Enter, inpu ng
our ini al approxima on. We then enter:
Ans - (cos(Ans)-Ans)/(-sin(Ans)-1).
Repeatedly pressing the Enter key gives successive approxima ons. We

quickly find:
x3 = 0.7390851332
x4 = 0.7390851332.
Our approxima ons x2 and x3 did not differ for at least the first 5 places a er the
decimal, so we could have stopped. However, using our calculator in the man-
ner described is easy, so finding x4 was not hard. It is interes ng to see how we
found an approxima on, accurate to as many decimal places as our calculator
displays, in just 4 itera ons.
If you know how to program, you can translate the following pseudocode
into your favorite language to perform the computa on in this problem.
Notes:
170
x = .75 y
while true 0.5
oldx = x
x = x - (cos(x)-x)/(-sin(x)-1) x
−0.5 0.5 1 1.5
print x
if abs(x-oldx) < .0000000001 −0.5
break
−1
This code calculates x1 , x2 , etc., storing each result in the variable x. The pre-
vious approxima on is stored in the variable oldx. We con nue looping un l . −1.5
the difference between two successive approxima ons, abs(x-oldx), is less
than some small tolerance, in this case, .0000000001. Figure 4.1.4: A graph of f(x) = x3 −x2 −1,
showing why an ini al approxima on of
Convergence of Newton’s Method x0 = 0 with Newton’s Method fails.
What should one use for the ini al guess, x0 ? Generally, the closer to the y
actual root the ini al guess is, the be er. However, some ini al guesses should
be avoided. For instance, consider Example 4.1.1 where we sought the root to
1
f(x) = x3 − x2 − 1. Choosing x0 = 0 would have been a par cularly poor choice.
Consider Figure 4.1.4, where f(x) is graphed along with its tangent line at x = 0.
Since f ′ (0) = 0, the tangent line is horizontal and does not intersect the x–axis. x
Graphically, we see that Newton’s Method fails. −1 x0 1
We can also see analy cally that it fails. Since
−1
f(0)
x1 = 0 − ′ .
f (0)
(a)
′ y
and f (0) = 0, we see that x1 is not well defined.
This problem can also occur if, for instance, it turns out that f ′ (x5 ) = 0.
Adjus ng the ini al approxima on x0 by a very small amount will likely fix the 1
problem.
It is also possible for Newton’s Method to not converge while each successive
x
approxima on is well defined. Consider f(x) = x1/3 , as shown in Figure 4.1.5. It −1 x1 x0 1
is clear that the root is x = 0, but let’s approximate this with x0 = 0.1. Figure
4.1.5(a) shows graphically the calcula on of x1 ; no ce how it is farther from the −1
root than x0 . Figures 4.1.5(b) and (c) show the calcula on of x2 and x3 , which are
even farther away; our successive approxima ons are ge ng worse. (It turns .
out that in this par cular example, each successive approxima on is twice as far (b)
y
from the true answer as the previous approxima on.)
There is no “fix” to this problem; Newton’s Method simply will not work and
another method must be used. 1
While Newton’s Method does not always work, it does work “most of the
me,” and it is generally very fast. Once the approxima ons get close to the root,
x
−1x3 x1 x0 x2 1
−1
Notes: .
(c)
Figure 4.1.5: Newton’s Method fails to

find a root of f(x) = x1/3 , regardless of
the choice of x0 .
171
Newton’s Method can as much as double the number of correct decimal places
with each successive approxima on. A course in Numerical Analysis will intro-
duce the reader to more itera ve root finding methods, as well as give greater
detail about the strengths and weaknesses of Newton’s Method.
Notes:
172
Exercises 4.1
Terms and Concepts imal. If an interval is given, find only the roots that lie in
that interval. Use technology to obtain good ini al approx-
1. T/F: Given a func on f(x), Newton’s Method produces an ima ons.
exact solu on to f(x) = 0.
9. f(x) = x3 + 5x2 − x − 1
2. T/F: In order to get a solu on to f(x) = 0 accurate to d
places a er the decimal, at least d + 1 itera ons of New- 10. f(x) = x4 + 2x3 − 7x2 − x + 5
tons’ Method must be used.
11. f(x) = x17 − 2x13 − 10x8 + 10 on (−2, 2)
Problems 12. f(x) = x2 cos x + (x − 1) sin x on (−3, 3)
In Exercises 3 – 8, the roots of f(x) are known or are easily

In Exercises 13 – 16, use Newton’s Method to approximate
found. Use 5 itera ons of Newton’s Method with the given
when the given func ons are equal, accurate to 3 places af-
ini al approxima on to approximate the root. Compare it to
ter the decimal. Use technology to obtain good ini al approx-
the known value of the root.
ima ons.
3. f(x) = cos x, x0 = 1.5
13. f(x) = x2 , g(x) = cos x
4. f(x) = sin x, x0 = 1
14. f(x) = x2 − 1, g(x) = sin x
5. f(x) = x2 + x − 2, x0 = 0 2
15. f(x) = ex , g(x) = cos x
6. f(x) = x2 − 2, x0 = 1.5
16. f(x) = x, g(x) = tan x on [−6, 6]
7. f(x) = ln x, x0 = 2
17. Why does Newton’s Method fail in finding a root of f(x) =
8. f(x) = x3 − x2 + x − 1, x0 = 1 x3 − 3x2 + x + 3 when x0 = 1?
In Exercises 9 – 12, use Newton’s Method to approximate all 18. Why does Newton’s Method fail in finding a root of f(x) =
roots of the given func ons accurate to 3 places a er the dec- −17x4 + 130x3 − 301x2 + 156x + 156 when x0 = 1?
173
4.2 Related Rates

When two quan es are related by an equa on, knowing the value of one quan-
ty can determine the value of the other. For instance, the circumference and
radius of a circle are related by C = 2πr; knowing that C = 6πin determines the
radius must be 3in.
Note: This sec on relies heavily on im-
plicit differen a on, so referring back to The topic of related rates takes this one step further: knowing the rate
Sec on 2.6 may help. at which one quan ty is changing can determine the rate at which another
changes.
We demonstrate the concepts of related rates through examples.
Example 4.2.1 Understanding related rates

The radius of a circle is growing at a rate of 5in/hr. At what rate is the circumfer-
ence growing?
S The circumference and radius of a circle are related by C =

2πr. We are given informa on about how the length of r changes with respect
to me; that is, we are told dr
dt = 5in/hr. We want to know how the length of C
changes with respect to me, i.e., we want to know dCdt .
Implicitly differen ate both sides of C = 2πr with respect to t:
C = 2πr
d( ) d( )
C = 2πr
dt dt
dC dr
= 2π .
dt dt
dr
As we know dt = 5in/hr, we know
dC
= 2π5 = 10π ≈ 31.4in/hr.
dt
Consider another, similar example.
Example 4.2.2 Finding related rates

Water streams out of a faucet at a rate of 2in3 /s onto a flat surface at a constant
rate, forming a circular puddle that is 1/8in deep.
1. At what rate is the area of the puddle growing?
2. At what rate is the radius of the circle growing?
Notes:
174
4.2 Related Rates
1. We can answer this ques on two ways: using “common sense” or related
rates. The common sense method states that the volume of the puddle is
growing by 2in3 /s, where
volume of puddle = area of circle × depth.
Since the depth is constant at 1/8in, the area must be growing by 16in2 /s.
This approach reveals the underlying related–rates principle. Let V and A
represent the Volume and Area of the puddle. We know V = A × 81 . Take
the deriva ve of both sides with respect to t, employing implicit differen-
a on.
1
V=A
8 ( )
d( ) d 1
V = A
dt dt 8
dV 1 dA
=
dt 8 dt
As dV
dt = 2, we know 2 =
1 dA
8 dt , and hence dA
dt = 16. Thus the area is
growing by 16in2 /s.
2. To start, we need an equa on that relates what we know to the radius.

We just learned something about the surface area of the circular puddle,
and we know A = πr2 . We should be able to learn about the rate at which
the radius is growing with this informa on.
Implicitly derive both sides of A = πr2 with respect to t:
A = πr2
d( ) d ( 2)
A = πr
dt dt
dA dr
= 2πr
dt dt
dA dr
Our work above told us that dt = 16in2 /s. Solving for dt , we have
dr 8
= .
dt πr
Note how our answer is not a number, but rather a func on of r. In other
words, the rate at which the radius is growing depends on how big the
Notes:
175
circle already is. If the circle is very large, adding 2in3 of water will not
make the circle much bigger at all. If the circle is dime–sized, adding the
same amount of water will make a radical change in the radius of the circle.
In some ways, our problem was (inten onally) ill–posed. We need to spec-
ify a current radius in order to know a rate of change. When the puddle
has a radius of 10in, the radius is growing at a rate of
dr 8 4
= = ≈ 0.25in/s.
dt 10π 5π
Example 4.2.3 Studying related rates

Radar guns measure the rate of distance change between the gun and the object
it is measuring. For instance, a reading of “55mph” means the object is moving
away from the gun at a rate of 55 miles per hour, whereas a measurement of
“−25mph” would mean that the object is approaching the gun at a rate of 25
miles per hour.
If the radar gun is moving (say, a ached to a police car) then radar readouts
are only immediately understandable if the gun and the object are moving along
the same line. If a police officer is traveling 60mph and gets a readout of 15mph,
he knows that the car ahead of him is moving away at a rate of 15 miles an hour,
meaning the car is traveling 75mph. (This straight–line principle is one reason
officers park on the side of the highway and try to shoot straight back down the
road. It gives the most accurate reading.)
Suppose an officer is driving due north at 30 mph and sees a car moving due
east, as shown in Figure 4.2.1. Using his radar gun, he measures a reading of
20mph. By using landmarks, he believes both he and the other car are about
1/2 mile from the intersec on of their two roads.
N
B = 1/2 If the speed limit on the other road is 55mph, is the other driver speeding?
. E
Car
S Using the diagram in Figure 4.2.1, let’s label what we know
about the situa on. As both the police officer and other driver are 1/2 mile from
A = 1/2
C the intersec on, √

we have A = 1/2, B = 1/2, and through the Pythagorean
Theorem, C = 1/ 2 ≈ 0.707.
dt = −30. The
We know the police officer is traveling at 30mph; that is, dA
Officer
reason this rate of change is nega ve is that A is ge ng smaller; the distance
between the officer and the intersec on is shrinking. The radar measurement
is dC dB
dt = 20. We want to find dt .
Figure 4.2.1: A sketch of a police car
We need an equa on that relates B to A and/or C. The Pythagorean Theorem
(at bo om) a emp ng to measure the
speed of a car (at right) in Example 4.2.3.
Notes:
176
4.2 Related Rates
is a good choice: A2 + B2 = C2 . Differen ate both sides with respect to t:

Note: Example 4.2.3 is both interes ng
2
A +B =C2 2 and imprac cal. It highlights the difficulty
in using radar in a non–linear fashion, and
d ( 2 ) d ( 2) explains why “in real life” the police offi-
A + B2 = C
dt dt cer would follow the other driver to de-
dA dB dC termine their speed, and not pull out pen-
2A + 2B = 2C
dt dt dt cil and paper.
The principles here are important,
dB though. Many automated vehicles make
We have values for everything except dt . Solving for this we have judgments about other moving objects
based on perceived distances, radar–like
dB C dC − A dA measurements and the concepts of
= dt dt
≈ 58.28mph. related rates.
dt B
The other driver appears to be speeding slightly.
Example 4.2.4 Studying related rates

A camera is placed on a tripod 10 from the side of a road. The camera is to turn
to track a car that is to drive by at 100mph for a promo onal video. The video’s
planners want to know what kind of motor the tripod should be equipped with
in order to properly track the car as it passes by. Figure 4.2.2 shows the proposed
setup.
100mph
How fast must the camera be able to turn to track the car? .
x
S We seek informa on about how fast the camera is to turn;

therefore, we need an equa on that will relate an angle θ to the posi on of the 10
camera and the speed and posi on of the car. θ

Figure 4.2.2 suggests we use a trigonometric equa on. Le ng x represent
the distance the car is from the point on the road directly in front of the camera, Figure 4.2.2: Tracking a speeding car (at
we have le ) with a rota ng camera.
x
tan θ = . (4.1)
10
dt = −100mph (as in the last example,

As the car is moving at 100mph, we have dx
since x is ge ng smaller as the car travels, dx
dt is nega ve). We need to convert
the measurements so they use the same units; rewrite −100mph in terms of
/s:
dx m m 1 hr
= −100 = −100 · 5280 · = −146.6 /s.
dt hr hr m 3600 s
Now take the deriva ve of both sides of Equa on (4.1) using implicit differen -
Notes:
177
a on:
x
tan θ =
10
d( ) d ( x )
tan θ =
dt dt 10
dθ 1 dx
sec2 θ =
dt 10 dt
dθ cos2 θ dx
= (4.2)
dt 10 dt
We want to know the fastest the camera has to turn. Common sense tells us this
is when the car is directly in front of the camera (i.e., when θ = 0). Our mathe-
ma cs bears this out. In Equa on (4.2) we see this is when cos2 θ is largest; this
is when cos θ = 1, or when θ = 0.
dt ≈ −146.67 /s, we have
With dx
dθ 1rad
=− 146.67 /s = −14.667radians/s.
dt 10
We find that dθ dt is nega ve; this matches our diagram in Figure 4.2.2 for θ is
ge ng smaller as the car approaches the camera.
What is the prac cal meaning of −14.667radians/s? Recall that 1 circular
revolu on goes through 2π radians, thus 14.667rad/s means 14.667/(2π) ≈
2.33 revolu ons per second. The nega ve sign indicates the camera is rota ng
in a clockwise fashion.
We introduced the deriva ve as a func on that gives the slopes of tangent

lines of func ons. This chapter emphasizes using the deriva ve in other ways.
Newton’s Method uses the deriva ve to approximate roots of func ons; this
sec on stresses the “rate of change” aspect of the deriva ve to find a rela on-
ship between the rates of change of two related quan es.
In the next sec on we use Extreme Value concepts to op mize quan es.
Notes:
178
Exercises 4.2
Terms and Concepts (a) 1 mile away?
(b) 1/5 mile away?
1. T/F: Implicit differen a on is o en used when solving “re-
lated rates” type problems. (c) Directly overhead?
2. T/F: A study of related rates is part of the standard police 8. An F-22 aircra is flying at 500mph with an eleva on of
officer training. 100 on a straight–line path that will take it directly over
an an –aircra gun as in Exercise 7 (note the lower eleva-
on here).
Problems How fast must the gun be able to turn to accurately track
the aircra when the plane is:
3. Water flows onto a flat surface at a rate of 5cm3 /s forming a (a) 1000 feet away?
circular puddle 10mm deep. How fast is the radius growing
when the radius is: (b) 100 feet away?
(a) 1 cm? (c) Directly overhead?
(b) 10 cm?
9. A 24 . ladder is leaning against a house while the base is
(c) 100 cm? pulled away at a constant rate of 1 /s.
4. A circular balloon is inflated with air flowing at a rate of
24
.
10cm3 /s. How fast is the radius of the balloon increasing 1 /s
when the radius is: .
At what rate is the top of the ladder sliding down the side
(a) 1 cm? of the house when the base is:
(b) 10 cm?
(a) 1 foot from the house?
(c) 100 cm?
(b) 10 feet from the house?
5. Consider the traffic situa on introduced in Example 4.2.3. (c) 23 feet from the house?
How fast is the “other car” traveling if the officer and the
(d) 24 feet from the house?
other car are each 1/2 mile from the intersec on, the other
car is traveling due west, the officer is traveling north at
50mph, and the radar reading is −80mph? 10. A boat is being pulled into a dock at a constant rate of
30 /min by a winch located 10 above the deck of the
boat.
6. Consider the traffic situa on introduced in Example 4.2.3.
Calculate how fast the “other car” is traveling in each of the
10
following situa ons. .
(a) The officer is traveling due north at 50mph and is . .
1/2 mile from the intersec on, while the other car At what rate is the boat approaching the dock when the
is 1 mile from the intersec on traveling west and the boat is:
radar reading is −80mph?
(b) The officer is traveling due north at 50mph and is (a) 50 feet out?
1 mile from the intersec on, while the other car is (b) 15 feet out?
1/2 mile from the intersec on traveling west and the
(c) 1 foot from the dock?
radar reading is −80mph?
(d) What happens when the length of rope pulling in the
7. An F-22 aircra is flying at 500mph with an eleva on of boat is less than 10 feet long?
10,000 on a straight–line path that will take it directly over
an an –aircra gun. 11. An inverted cylindrical cone, 20 deep and 10 across at
the top, is being filled with water at a rate of 10 3 /min. At
what rate is the water rising in the tank when the depth of
. 10,000
the water is:
(a) 1 foot?
. θ
.
(b) 10 feet?
x
(c) 19 feet?
How fast must the gun be able to turn to accurately track
the aircra when the plane is: How long will the tank take to fill when star ng at empty?
179
12. A rope, a ached to a weight, goes up through a pulley at (b) How fast is the weight rising a er the man has walked
the ceiling and back down to a worker. The man holds the 10 feet?
rope at the same height as the connec on point between
(c) How fast is the weight rising a er the man has walked
rope and weight.
30 feet?
(d) How far must the man walk to raise the weight all the
way to the pulley?
30
2 /s
. 14. A hot air balloon li s off from ground rising ver cally. From
Suppose the man stands directly next to the weight (i.e., a 100 feet away, a 5’ woman tracks the path of the balloon.
total rope length of 60 ) and begins to walk away at a rate When her sightline with the balloon makes a 45◦ angle with
of 2 /s. How fast is the weight rising when the man has the horizontal, she notes the angle is increasing at about
walked: 5◦ /min.
(a) 10 feet? (a) What is the eleva on of the balloon?
(b) 40 feet?
(b) How fast is it rising?
How far must the man walk to raise the weight all the way
to the pulley? 15. A company that produces landscaping materials is dumping
sand into a conical pile. The sand is being poured at a rate
13. Consider the situa on described in Exercise 12. Suppose of 5 3 /sec; the physical proper es of the sand, in conjunc-
the man starts 40 from the weight and begins to walk on with gravity, ensure that the cone’s height is roughly
away at a rate of 2 /s. 2/3 the length of the diameter of the circular base.
(a) How long is the rope? How fast is the cone rising when it has a height of 30 feet?
180
4.3 Op miza on
4.3 Op miza on
In Sec on 3.1 we learned about extreme values – the largest and smallest values
a func on a ains on an interval. We mo vated our interest in such values by
discussing how it made sense to want to know the highest/lowest values of a
stock, or the fastest/slowest an object was moving. In this sec on we apply
the concepts of extreme values to solve “word problems,” i.e., problems stated
in terms of situa ons that require us to create the appropriate mathema cal
framework in which to solve the problem.
We start with a classic example which is followed by a discussion of the topic
of op miza on.
Example 4.3.1 Op miza on: perimeter and area

A man has 100 feet of fencing, a large yard, and a small dog. He wants to create
a rectangular enclosure for his dog with the fencing that provides the maximal
area. What dimensions provide the maximal area?
S One can likely guess the correct answer – that is great. We

will proceed to show how calculus can provide this answer in a context that
proves this answer is correct.
.
.
.
.
.
It helps to make a sketch of the situa on. Our enclosure is sketched twice
.
.
.
in Figure 4.3.1, either with green grass and nice fence boards or as a simple
.
.
.
rectangle. Either way, drawing a rectangle forces us to realize that we need to . y
.
.
.
.
.
know the dimensions of this rectangle so we can create an area func on – a er
all, we are trying to maximize the area. .
We let x and y denote the lengths of the sides of the rectangle. Clearly, x
Area = xy.
We do not yet know how to handle func ons with 2 variables; we need to
y
reduce this down to a single variable. We know more about the situa on: the
man has 100 feet of fencing. By knowing the perimeter of the rectangle must
be 100, we can create another equa on:
x
Perimeter = 100 = 2x + 2y.
Figure 4.3.1: A sketch of the enclosure in
We now have 2 equa ons and 2 unknowns. In the la er equa on, we solve Example 4.3.1.
for y:
y = 50 − x.
Now subs tute this expression for y in the area equa on:
Area = A(x) = x(50 − x).
Note we now have an equa on of one variable; we can truly call the Area a
func on of x.
Notes:
181
This func on only makes sense when 0 ≤ x ≤ 50, otherwise we get nega ve
values of area. So we find the extreme values of A(x) on the interval [0, 50].
To find the cri cal points, we take the deriva ve of A(x) and set it equal to
0, then solve for x.
A(x) = x(50 − x)
= 50x − x2
A′ (x) = 50 − 2x
We solve 50 − 2x = 0 to find x = 25; this is the only cri cal point. We evaluate
A(x) at the endpoints of our interval and at this cri cal point to find the extreme
values; in this case, all we care about is the maximum.
Clearly A(0) = 0 and A(50) = 0, whereas A(25) = 625 2 . This is the max-
imum. Since we earlier found y = 50 − x, we find that y is also 25. Thus the
dimensions of the rectangular enclosure with perimeter of 100 . with maxi-
mum area is a square, with sides of length 25 .
This example is very simplis c and a bit contrived. (A er all, most people
create a design then buy fencing to meet their needs, and not buy fencing and
plan later.) But it models well the necessary process: create equa ons that de-
scribe a situa on, reduce an equa on to a single variable, then find the needed
extreme value.
“In real life,” problems are much more complex. The equa ons are o en
not reducible to a single variable (hence mul –variable calculus is needed) and
the equa ons themselves may be difficult to form. Understanding the princi-
ples here will provide a good founda on for the mathema cs you will likely en-
counter later.
We outline here the basic process of solving these op miza on problems.
Key Idea 4.3.1 Solving Op miza on Problems
1. Understand the problem. Clearly iden fy what quan ty is to be

maximized or minimized. Make a sketch if helpful.
2. Create equa ons relevant to the context of the problem, using the
informa on given. (One of these should describe the quan ty to
be op mized. We’ll call this the fundamental equa on.)
3. If the fundamental equa on defines the quan ty to be op mized
as a func on of more than one variable, reduce it to a single vari-
able func on using subs tu ons derived from the other equa-
ons.
(con nued). . .
Notes:
182
4.3 Op miza on
Key Idea 4.3.1 Solving Op miza on Problems – Con nued
4. Iden fy the domain of this func on, keeping in mind the context
of the problem.
5. Find the extreme values of this func on on the determined do-
main.
6. Iden fy the values of all relevant quan es of the problem.
We will use Key Idea 4.3.1 in a variety of examples.
Example 4.3.2 Op miza on: perimeter and area

Here is another classic calculus problem: A woman has a 100 feet of fencing, a
small dog, and a large yard that contains a stream (that is mostly straight). She
wants to create a rectangular enclosure with maximal area that uses the stream
as one side. (Apparently her dog won’t swim away.) What dimensions provide
the maximal area?
S We will follow the steps outlined by Key Idea 4.3.1.

1. We are maximizing area. A sketch of the region will help; Figure 4.3.2
gives two sketches of the proposed enclosed area. A key feature of the
sketches is to acknowledge that one side is not fenced.
.
.
.
.
.
2. We want to maximize the area; as in the example before,
.
.
.
.
.
Area = xy.
.
. y
.
.
.
This is our fundamental equa on. This defines area as a func on of two
.
variables, so we need another equa on to reduce it to one variable. .
We again appeal to the perimeter; here the perimeter is

x
Perimeter = 100 = x + 2y.
x
Note how this is different than in our previous example.
3. We now reduce the fundamental equa on to a single variable. In the y
perimeter equa on, solve for y: y = 50 − x/2. We can now write Area as
1
Area = A(x) = x(50 − x/2) = 50x − x2 .
2 Figure 4.3.2: A sketch of the enclosure in
Area is now defined as a func on of one variable. Example 4.3.2.
Notes:
183
4. We want the area to be nonnega ve. Since A(x) = x(50 − x/2), we want
x ≥ 0 and 50 − x/2 ≥ 0. The la er inequality implies that x ≤ 100, so
0 ≤ x ≤ 100.
5. We now find the extreme values. At the endpoints, the minimum is found,
giving an area of 0.
Find the cri cal points. We have A′ (x) = 50 − x; se ng this equal to 0
and solving for x returns x = 50. This gives an area of
A(50) = 50(25) = 1250.
6. We earlier set y = 50 − x/2; thus y = 25. Thus our rectangle will have
two sides of length 25 and one side of length 50, with a total area of 1250
2
.
Keep in mind as we do these problems that we are prac cing a process; that
is, we are learning to turn a situa on into a system of equa ons. These equa-
ons allow us to write a certain quan ty as a func on of one variable, which we
then op mize.
Example 4.3.3 Op miza on: minimizing cost

A power line needs to be run from a power sta on located on the beach to an
offshore facility. Figure 4.3.3 shows the distances between the power sta on to
. 1000
the facility.
It costs $50/ . to run a power line along the land, and $130/ . to run a
. power line under water. How much of the power line should be run along the
.
land to minimize the overall cost? What is the minimal cost?
5000
Figure 4.3.3: Running a power line from

S We will follow the strategy of Key Idea 4.3.1 implicitly, with-
the power sta on to an offshore facility out specifically numbering steps.
with minimal cost in Example 4.3.3. There are two immediate solu ons that we could consider, each of which we
will reject through “common sense.” First, we could minimize the distance by
directly connec ng the two loca ons with a straight line. However, this requires
that all the wire be laid underwater, the most costly op on. Second, we could
minimize the underwater length by running a wire all 5000 . along the beach,
directly across from the offshore facility. This has the undesired effect of having
the longest distance of all, probably ensuring a non–minimal cost.
. √ x2 +
10
00
2
1000
The op mal solu on likely has the line being run along the ground for a
while, then underwater, as the figure implies. We need to label our unknown
. distances – the distance run along the ground and the distance run underwater.
.
5000 − x x Recognizing that the underwater distance can be measured as the hypotenuse
of a right triangle, we choose to label the distances as shown in Figure 4.3.4.
Figure 4.3.4: Labeling unknown distances
in Example 4.3.3.
Notes:
184
4.3 Op miza on
By choosing x as we did, we make the expression under the square root sim-
ple. We now create the cost func on.
Cost = land cost + water cost

$50 × land distance + $130 ×√water distance
50(5000 − x) + 130 x2 + 10002 .
√
So we have c(x) = 50(5000 − x) + 130 x2 + 10002 . This func on only
makes sense on the interval [0, 5000]. While we are fairly certain the endpoints
will not give a minimal cost, we s ll evaluate c(x) at each to verify.
c(0) = 380, 000 c(5000) ≈ 662, 873.
We now find the cri cal values of c(x). We compute c ′ (x) as
130x
c ′ (x) = −50 + √ .
x2 + 10002
Recognize that this is never undefined. Se ng c ′ (x) = 0 and solving for x,

we have:
130x
−50 + √ =0
x2 + 10002
130x
√ = 50
x + 10002
2
1302 x2
= 502
x2 + 10002
1302 x2 = 502 (x2 + 10002 )
1302 x2 − 502 x2 = 502 · 10002
(1302 − 502 )x2 = 50, 0002
50, 0002
x2 =
1302 − 502
50, 000
x= √
1302 − 502
50, 000 1250
x= = ≈ 416.67.
120 3
Evalua ng c(x) at x = 416.67 gives a cost of about $370,000. The distance
√ along land is 5000 − 416.67 = 4583.33 ., and the under-
the power line is laid
water distance is 416.672 + 10002 ≈ 1083 .
Notes:
185
In the exercises you will see a variety of situa ons that require you to com-
bine problem–solving skills with calculus. Focus on the process; learn how to
form equa ons from situa ons that can be manipulated into what you need.
Eschew memorizing how to do “this kind of problem” as opposed to “that kind
of problem.” Learning a process will benefit one far longer than memorizing a
specific technique.
The next sec on introduces our final applica on of the deriva ve: differen-
als. Given y = f(x), they offer a method of approxima ng the change in y a er
x changes by a small amount.
Notes:
186
Exercises 4.3
Terms and Concepts What is the maximum volume of a package with a square
cross sec on (w = h) that does not exceed the 108” stan-
1. T/F: An “op miza on problem” is essen ally an “extreme dard?
values” problem in a “story problem” se ng.
12. The strength S of a wooden beam is directly propor onal
2. T/F: This sec on teaches one to find the extreme values of to its cross sec onal width w and the square of its height h;
a func on that has more than one variable. that is, S = kwh2 for some constant k.
w
Problems 12 h
3. Find the maximum product of two numbers (not necessar-

ily integers) that have a sum of 100. Given a circular log with diameter of 12 inches, what sized
beam can be cut from the log with maximum strength?
4. Find the minimum sum of two posi ve numbers whose
product is 500. 13. A power line is to be run to an offshore facility in the man-
ner described in Example 4.3.3. The offshore facility is 2
5. Find the maximum sum of two posi ve numbers whose miles at sea and 5 miles along the shoreline from the power
product is 500. plant. It costs $50,000 per mile to lay a power line under-
ground and $80,000 to run the line underwater.
6. Find the maximum sum of two numbers, each of which is How much of the power line should be run underground to
in [0, 300] whose product is 500. minimize the overall costs?
7. Find the maximal area of a right triangle with hypotenuse 14. A power line is to be run to an offshore facility in the man-
of length 1. ner described in Example 4.3.3. The offshore facility is 5
miles at sea and 2 miles along the shoreline from the power
8. A rancher has 1000 feet of fencing in which to construct plant. It costs $50,000 per mile to lay a power line under-
adjacent, equally sized rectangular pens. What dimensions ground and $80,000 to run the line underwater.
should these pens have to maximize the enclosed area? How much of the power line should be run underground to
minimize the overall costs?
15. A woman throws a s ck into a lake for her dog to fetch;

the s ck is 20 feet down the shore line and 15 feet into the
water from there. The dog may jump directly into the wa-
ter and swim, or run along the shore line to get closer to
the s ck before swimming. The dog runs about 22 /s and
9. A standard soda can is roughly cylindrical and holds 355cm3 swims about 1.5 /s.
of liquid. What dimensions should the cylinder be to min- How far along the shore should the dog run to minimize
imize the material needed to produce the can? Based on the me it takes to get to the s ck? (Hint: the figure from
your dimensions, determine whether or not the standard Example 4.3.3 can be useful.)
can is produced to minimize the material costs.
16. A woman throws a s ck into a lake for her dog to fetch;
10. Find the dimensions of a cylindrical can with a volume of the s ck is 15 feet down the shore line and 30 feet into the
206in3 that minimizes the surface area. water from there. The dog may jump directly into the wa-
The “#10 can”is a standard sized can used by the restau- ter and swim, or run along the shore line to get closer to
rant industry that holds about 206in3 with a diameter of 6 the s ck before swimming. The dog runs about 22 /s and
2/16in and height of 7in. Does it seem these dimensions swims about 1.5 /s.
were chosen with minimiza on in mind? How far along the shore should the dog run to minimize the
me it takes to get to the s ck? (Google “calculus dog” to learn
11. The United States Postal Service charges more for boxes more about a dog’s ability to minimize mes.)
whose combined length and girth exceeds 108” (the
“length” of a package is the length of its longest side; the 17. What are the dimensions of the rectangle with largest area
girth is the perimeter of the cross sec on, i.e., 2w + 2h). that can be drawn inside the unit circle?
187
4.4 Differen als

In Sec on 2.2 we explored the meaning and use of the deriva ve. This sec on
starts by revisi ng some of those ideas.
Recall that the deriva ve of a func on f can be used to find the slopes of
lines tangent to the graph of f. At x = c, the tangent line to the graph of f has
equa on
y = f ′ (c)(x − c) + f(c).
The tangent line can be used to find good approxima ons of f(x) for values of x
near c.
For instance, we can approximate sin 1.1 using the tangent
√ line to the graph
y of f(x) = sin x at x = π/3 ≈ 1.05. Recall that sin(π/3) = 3/2 ≈ 0.866, and
1 cos(π/3) = 1/2. Thus the tangent line to f(x) = sin x at x = π/3 is:
√
3
2
1
ℓ(x) = (x − π/3) + 0.866.
2
0.5
In Figure 4.4.1(a), we see a graph of f(x) = sin x graphed along with its tan-
gent line at x = π/3. The small rectangle shows the region that is displayed in
Figure 4.4.1(b). In this figure, we see how we are approxima ng sin 1.1 with the
x tangent line, evaluated at 1.1. Together, the two figures show how close these
. π
3 values are.
Using this line to approximate sin 1.1, we have:
(a)
1
y
ℓ(1.1) = (1.1 − π/3) + 0.866
2
1
= (0.053) + 0.866 = 0.8925.
ℓ(1.1) ≈ sin 1.1 2
0.89
(We leave it to the reader to see how good of an approxima on this is.)
0.88 sin 1.1
We now generalize this concept. Given f(x) and an x–value c, the tangent
line is ℓ(x) = f ′ (c)(x − c) + f(c). Clearly, f(c) = ℓ(c). Let ∆x be a small number,
0.87
√
3 represen ng a small change in x value. We assert that:
2
.
π 1.1
x
f(c + ∆x) ≈ ℓ(c + ∆x),
3
since the tangent line to a func on approximates well the values of that func on
(b) near x = c.
As the x-value changes from c to c + ∆x, the y-value of f changes from f(c)
Figure 4.4.1: Graphing f(x) = sin x and its
to f(c + ∆x). We call this change of y value ∆y. That is:
tangent line at x = π/3 in order to es -
mate sin 1.1.
∆y = f(c + ∆x) − f(c).
Notes:
188
4.4 Differen als
Replacing f(c + ∆x) with its tangent line approxima on, we have
∆y ≈ ℓ(c + ∆x) − f(c)

( )
= f ′ (c) (c + ∆x) − c + f(c) − f(c)
= f ′ (c)∆x (4.3)
This final equa on is important; it becomes the basis of the upcoming Def-
ini on and Key Idea. In short, it says that when the x-value changes from c to
c + ∆x, the y value of a func on f changes by about f ′ (c)∆x.
We introduce two new variables, dx and dy in the context of a formal defini-
on.
Defini on 4.4.1 Differen als of x and y.

Let y = f(x) be differen able. The differen al of x, denoted dx, is any
nonzero real number (usually taken to be a small number). The differ-
en al of y, denoted dy, is
dy = f ′ (x)dx.
We can solve for f ′ (x) in the above equa on: f ′ (x) = dy/dx. This states that
the deriva ve of f with respect to x is the differen al of y divided by the differ-
en al of x; this is not the alternate nota on for the deriva ve, dy dx . This la er
nota on was chosen because of the frac on–like quali es of the deriva ve, but
again, it is one symbol and not a frac on.
It is helpful to organize our new concepts and nota ons in one place.
Key Idea 4.4.1 Differen al Nota on

Let y = f(x) be a differen able func on.
1. Let ∆x represent a small, nonzero change in x value.
2. Let dx represent a small, nonzero change in x value (i.e., ∆x = dx).
3. Let ∆y be the change in y value as x changes by ∆x; hence
∆y = f(x + ∆x) − f(x).
4. Let dy = f ′ (x)dx which, by Equa on (4.3), is an approxima on of

the change in y value as x changes by ∆x; dy ≈ ∆y.
Notes:
189
What is the value of differen als? Like many mathema cal concepts, differ-
en als provide both prac cal and theore cal benefits. We explore both here.
Example 4.4.1 Finding and using differen als

Consider f(x) = x2 . Knowing f(3) = 9, approximate f(3.1).
S The x value is changing from x = 3 to x = 3.1; therefore, we

see that dx = 0.1. If we know how much the y value changes from f(3) to f(3.1)
(i.e., if we know ∆y), we will know exactly what f(3.1) is (since we already know
f(3)). We can approximate ∆y with dy.
∆y ≈ dy
= f ′ (3)dx
= 2 · 3 · 0.1 = 0.6.
We expect the y value to change by about 0.6, so we approximate f(3.1) ≈

9.6.
We leave it to the reader to verify this, but the preceding discussion links the
differen al to the tangent line of f(x) at x = 3. One can verify that the tangent
line, evaluated at x = 3.1, also gives y = 9.6.
Of course, it is easy to compute the actual answer (by hand or with a calcula-
tor): 3.12 = 9.61. (Before we get too cynical and say “Then why bother?”, note
our approxima on is really good!)
So why bother?
In “most” real life situa ons, we do not know the func on that describes
a par cular behavior. Instead, we can only take measurements of how things
change – measurements of the deriva ve.
Imagine water flowing down a winding channel. It is easy to measure the
speed and direc on (i.e., the velocity) of water at any loca on. It is very hard
to create a func on that describes the overall flow, hence it is hard to predict
where a floa ng object placed at the beginning of the channel will end up. How-
ever, we can approximate the path of an object using differen als. Over small
intervals, the path taken by a floa ng object is essen ally linear. Differen als
allow us to approximate the true path by piecing together lots of short, linear
paths. This technique is called Euler’s Method, studied in introductory Differen-
al Equa ons courses.
We use differen als once more to approximate the value of a func on. Even
though calculators are very accessible, it is neat to see how these techniques can
some mes be used to easily compute something that looks rather hard.
Notes:
190
4.4 Differen als
Example 4.4.2√ Using differen als to approximate a func on value

Approximate 4.5.
√ √
S We expect 4.5 ≈ 2, yet we can do be er.√Let f(x) = x,
and let c = 4. Thus f(4) = 2. We can compute f ′ (x) = 1/(2 x), so f ′ (4) =
1/4.
We approximate the difference between f(4.5) and f(4) using differen als,
with dx = 0.5:
f(4.5) − f(4) = ∆y ≈ dy = f ′ (4) · dx = 1/4 · 1/2 = 1/8 = 0.125.
The
√ approximate change in f from x = 4 to x = 4.5 is 0.125, so we approximate
4.5 ≈ 2.125.
Differen als are important when we discuss integra on. When we study
that topic, we will use nota on such as
∫
f(x) dx
quite o en. While we don’t discuss here what all of that nota on means, note
the existence of the differen al dx. Proper handling of integrals comes with
proper handling of differen als.
In light of that, we prac ce finding differen als in general.
Example 4.4.3 Finding differen als

In each of the following, find the differen al dy.
√
1. y = sin x 2. y = ex (x2 + 2) 3. y = x2 + 3x − 1
1. y = sin x: As f(x) = sin x, f ′ (x) = cos x. Thus
dy = cos(x)dx.
2. y = ex (x2 + 2): Let f(x) = ex (x2 + 2). We need f ′ (x), requiring the
Product Rule.
We have f ′ (x) = ex (x2 + 2) + 2xex , so
( )
dy = ex (x2 + 2) + 2xex dx.
Notes:
191
√ √
3. y = x2 + 3x − 1: Let f(x) = x2 + 3x − 1; we need f ′ (x), requiring
the Chain Rule.
1 2x + 3
We have f ′ (x) = (x2 + 3x − 1)− 2 (2x + 3) = √
1
. Thus
2 2 x2 + 3x − 1
(2x + 3)dx
dy = √ .
2 x2 + 3x − 1
Finding the differen al dy of y = f(x) is really no harder than finding the

deriva ve of f; we just mul ply f ′ (x) by dx. It is important to remember that we
are not simply adding the symbol “dx” at the end.
We have seen a prac cal use of differen als as they offer a good method of
making certain approxima ons. Another use is error propaga on. Suppose a
length is measured to be x, although the actual value is x + ∆x (where ∆x is the
error, which we hope is small). This measurement of x may be used to compute
some other value; we can think of this la er value as f(x) for some func on f.
As the true length is x + ∆x, one really should have computed f(x + ∆x). The
difference between f(x) and f(x + ∆x) is the propagated error.
How close are f(x) and f(x + ∆x)? This is a difference in “y” values:
f(x + ∆x) − f(x) = ∆y ≈ dy.
We can approximate the propagated error using differen als.
Example 4.4.4 Using differen als to approximate propagated error

A steel ball bearing is to be manufactured with a diameter of 2cm. The manu-
facturing process has a tolerance of ±0.1mm in the diameter. Given that the
density of steel is about 7.85g/cm3 , es mate the propagated error in the mass
of the ball bearing.
S The mass of a ball bearing is found using the equa on “mass

= volume × density.” In this situa on the mass func on is a product of the radius
of the ball bearing, hence it is m = 7.85 34 πr3 . The differen al of the mass is
dm = 31.4πr2 dr.
The radius is to be 1cm; the manufacturing tolerance in the radius is ±0.05mm,
or ±0.005cm. The propagated error is approximately:
∆m ≈ dm
= 31.4π(1)2 (±0.005)
= ±0.493g
Notes:
192
4.4 Differen als
Is this error significant? It certainly depends on the applica on, but we can get
an idea by compu ng the rela ve error. The ra o between amount of error to
the total mass is
dm 0.493
=±
m 7.85 34 π
0.493
=±
32.88
= ±0.015,
or ±1.5%.
We leave it to the reader to confirm this, but if the diameter of the ball was
supposed to be 10cm, the same manufacturing tolerance would give a propa-
gated error in mass of ±12.33g, which corresponds to a percent error of ±0.188%.
While the amount of error is much greater (12.33 > 0.493), the percent error
is much lower.
We first learned of the deriva ve in the context of instantaneous rates of

change and slopes of tangent lines. We furthered our understanding of the
power of the deriva ve by studying how it relates to the graph of a func on
(leading to ideas of increasing/decreasing and concavity). This chapter has put
the deriva ve to yet more uses:
• Equa on solving (Newton’s Method),

• Related Rates (furthering our use of the deriva ve to find instantaneous
rates of change),
• Op miza on (applied extreme values), and
• Differen als (useful for various approxima ons and for something called
integra on).
In the next chapters, we will consider the “reverse” problem to compu ng
the deriva ve: given a func on f, can we find a func on whose deriva ve is f?
Being able to do so opens up an incredible world of mathema cs and applica-
ons.
Notes:
193
Exercises 4.4
22. y =
x4
1. T/F: Given a differen able func on y = f(x), we are gen-
2x
erally free to choose a value for dx, which then determines 23. y =
tan x + 1
the value of dy.
24. y = ln(5x)
2. T/F: The symbols “dx” and “∆x” represent the same con-
cept.
25. y = ex sin x
3. T/F: The symbols “dy” and “∆y” represent the same con-
cept. 26. y = cos(sin x)
4. T/F: Differen als are important in the study of integra on. x+1
27. y =
x+2
5. How are differen als and tangent lines related?
28. y = 3x ln x
6. T/F: In real life, differen als are used to approximate func-
on values when the func on itself is not known. 29. y = x ln x − x
( )
30. f(x) = ln sec x
Problems
Exercises 31 – 34 use differen als to approximate propagated
In Exercises 7 – 16, use differen als to approximate the given
error.
value by hand.
7. 2.052 31. A set of plas c spheres are to be made with a diameter

of 1cm. If the manufacturing process is accurate to 1mm,
8. 5.932 what is the propagated error in volume of the spheres?
9. 5.13 32. The distance, in feet, a stone drops in t seconds is given by

d(t) = 16t2 . The depth of a hole is to be approximated by
10. 6.83 dropping a rock and listening for it to hit the bo om. What
is the propagated error if the me measurement is accurate
√
11. 16.5 to 2/10ths of a second and the measured me is:
√ (a) 2 seconds?
12. 24
√
3
(b) 5 seconds?
13. 63
√
3
33. What is the propagated error in the measurement of the
14. 8.5 cross sec onal area of a circular log if the diameter is mea-
sured at 15′′ , accurate to 1/4′′ ?
15. sin 3
34. A wall is to be painted that is 8′ high and is measured to
16. e0.1
be 10′ , 7′′ long. Find the propagated error in the measure-
ment of the wall’s surface area if the measurement is accu-
In Exercises 17 – 30, compute the differen al dy. rate to 1/2′′ .
17. y = x2 + 3x − 5
Exercises 35 – 39 explore some issues related to surveying in
which distances are approximated using other measured dis-
18. y = x7 − x5
tances and measured angles. (Hint: Convert all angles to ra-
1 dians before compu ng.)
19. y =
4x2
35. The length l of a long wall is to be approximated. The angle
20. y = (2x + sin x)2 θ, as shown in the diagram (not to scale), is measured to be
85.2◦ , accurate to 1◦ . Assume that the triangle formed is a
21. y = x2 e3x right triangle.
194
l =? θ 50′ l =?
θ
25′
(a) What is the measured length l of the wall? (a) What is the measured length of the wall?
(b) What is the propagated error? (b) What is the propagated error?
(c) What is the percent error?
(c) What is the percent error?
36. Answer the ques ons of Exercise 35, but with a measured
angle of 71.5◦ , accurate to 1◦ , measured from a point 100′ 38. The length of the walls in Exercises 35 – 37 are essen ally
from the wall. the same. Which setup gives the most accurate result?
37. The length l of a long wall is to be calculated by measuring 39. Consider the setup in Exercise 37. This me, assume the
the angle θ shown in the diagram (not to scale). Assume angle measurement of 143◦ is exact but the measured 50′
the formed triangle is an isosceles triangle. The measured from the wall is accurate to 6′′ . What is the approximate
angle is 143◦ , accurate to 1◦ . percent error?
195
5: I
We have spent considerable me considering the deriva ves of a func on and
their applica ons. In the following chapters, we are going to star ng thinking
in “the other direc on.” That is, given a func on f(x), we are going to consider
func ons F(x) such that F ′ (x) = f(x). There are numerous reasons this will
prove to be useful: these func ons will help us compute area, volume, mass,
force, pressure, work, and much more.
5.1 An deriva ves and Indefinite Integra on

Given a func on y = f(x), a differen al equa on is one that incorporates y, x,
and the deriva ves of y. For instance, a simple differen al equa on is:
y ′ = 2x.
Solving a differen al equa on amounts to finding a func on y that sa sfies

the given equa on. Take a moment and consider that equa on; can you find a
func on y such that y ′ = 2x?
Can you find another?
And yet another?
Hopefully one was able to come up with at least one solu on: y = x2 . “Find-
ing another” may have seemed impossible un l one realizes that a func on like
y = x2 + 1 also has a deriva ve of 2x. Once that discovery is made, finding “yet
another” is not difficult; the func on y = x2 + 123, 456, 789 also has a deriva-
ve of 2x. The differen al equa on y ′ = 2x has many solu ons. This leads us
to some defini ons.
Defini on 5.1.1 An deriva ves and Indefinite Integrals

Let a func on f(x) be given. An an deriva ve of f(x) is a func on F(x)
such that F ′ (x) = f(x).
The set of all an deriva ves of f(x) is the indefinite integral of f, denoted
by ∫
f(x) dx.
Make a note about our defini on: we refer to an an deriva ve of f, as op-

posed to the an deriva ve of f, since there is always an infinite number of them.
Chapter 5 Integra on
We o en use upper-case le ers to denote an deriva ves.

Knowing one an deriva ve of f allows us to find infinitely more, simply by
adding a constant. Not only does this give us more an deriva ves, it gives us all
of them.
Theorem 5.1.1 An deriva ve Forms

Let F(x) and G(x) be an deriva ves of f(x) on an interval I. Then there
exists a constant C such that, on I,
G(x) = F(x) + C.
Given a func on f defined on an interval I and one of its an deriva ves F,

we know all an deriva ves of f on I have the form F(x) + C for some constant
C. Using Defini on 5.1.1, we can say that
∫
f(x) dx = F(x) + C.
Let’s analyze this indefinite integral nota on.
Integra on Differen al Constant of

symbol of x integra on
∫
f(x) dx = F(x) + C
Integrand One
an deriva ve
Figure 5.1.1: Understanding

. the indefinite integral nota on.
Figure 5.1.1 shows

∫ the typical nota on of the indefinite integral. The inte-
gra on symbol, , is in reality an “elongated S,” represen ng “take the sum.”
We will later see how sums and an deriva ves are related.
The func on we want to find an an deriva ve of is called the integrand. ∫It
contains the differen al of the variable we are integra ng with respect to. The
symbol and the differen al dx∫are not “bookends” with a func on sandwiched in
between; rather, the symbol means “find all an deriva ves of what follows,”
and the func on f(x) and dx are mul plied together; the dx does not “just sit
there.”
Let’s prac ce using this nota on.
Notes:
198
Example ∫5.1.1 Evalua ng indefinite integrals

Evaluate sin x dx.
S We are asked to find all func ons F(x) such that F ′ (x) =
sin x. Some thought will lead us to one solu on: F(x) = − cos x, because dx d
(− cos x) =
sin x.
The indefinite integral of sin x is thus − cos x, plus a constant of integra on.
So: ∫
sin x dx = − cos x + C.
A commonly asked ques on is “What happened to the dx?” The unenlight-

ened response is “Don’t worry about it. It just goes away.” A full understanding
includes the following.
This process of an differen a on is really solving a differen al ques on. The
integral ∫
sin x dx
presents us with a differen al, dy = sin x dx. It is asking: “What is y?” We found
lots of solu ons, all of the form y = − cos x + C.
Le ng dy = sin x dx, rewrite
∫ ∫
sin x dx as dy.
This is asking: “What func ons have a differen al of the form dy?” The answer
is “Func ons of the form y + C, where C is a constant.” What is y? We have lots
of choices, all differing by a constant; the simplest choice is y = − cos x.
Understanding all of this is more important later as we try to find an deriva-
ves of more complicated func ons. In this sec on, we will simply explore the
rules of indefinite integra on, and one can succeed for now with answering
“What happened to the dx?” with “It went away.”
Let’s prac ce once more before sta ng integra on rules.
Example ∫5.1.2 Evalua ng indefinite integrals

2
Evaluate (3x + 4x + 5) dx.
S We seek a func on F(x) whose deriva ve is 3x2 + 4x + 5.

When taking deriva ves, we can consider func ons term–by–term, so we can
likely do that here.
What func ons have a deriva ve of 3x2 ? Some thought will lead us to a
cubic, specifically x3 + C1 , where C1 is a constant.
Notes:
199
What func ons have a deriva ve of 4x? Here the x term is raised to the first
power, so we likely seek a quadra c. Some thought should lead us to 2x2 + C2 ,
where C2 is a constant.
Finally, what func ons have a deriva ve of 5? Func ons of the form 5x + C3 ,
where C3 is a constant.
Our answer appears to be
∫
(3x2 + 4x + 5) dx = x3 + C1 + 2x2 + C2 + 5x + C3 .
We do not need three separate constants of integra on; combine them as one
constant, giving the final answer of
∫
(3x2 + 4x + 5) dx = x3 + 2x2 + 5x + C.
It is easy to verify our answer; take the deriva ve of x3 + 2x3 + 5x + C and

see we indeed get 3x2 + 4x + 5.
This final step of “verifying our answer” is important both prac cally and
theore cally. In general, taking deriva ves is easier than finding an deriva ves
so checking our work is easy and vital as we learn.
We also see that taking the deriva ve of our answer returns the func on in
the integrand. Thus we can say that:
(∫ )
d
f(x) dx = f(x).
dx
Differen a on “undoes” the work done by an differen a on.

Theorem 2.7.3 gave a list of the deriva ves of common func ons we had
learned at that point. We restate part of that list here to stress the rela onship
between deriva ves and an deriva ves. This list will also be useful as a glossary
of common an deriva ves as we learn.
Notes:
200
Theorem 5.1.2 Deriva ves and An deriva ves

Common Differen a on Rules Common Indefinite Integral Rules
( ) ∫ ∫
d
1. dx cf(x) = c · f ′ (x) 1. c · f(x) dx = c · f(x) dx
( ) ∫( )
d
2. dx f(x) ± g(x) = 2. f(x) ± g(x) dx =
∫ ∫
f ′ (x) ± g′ (x) f(x) dx ± g(x) dx
d
( ) ∫
3. dx C =0 3. 0 dx = C
d
( ) ∫ ∫
4. dx x =1 4. 1 dx = dx = x + C
( n) ∫
d
5. dx x = n · xn−1 5. xn dx = n+1 1 n+1
x + C (n ̸= −1)
d
( ) ∫
6. dx sin x = cos x 6. cos x dx = sin x + C
( ) ∫
d
7. dx cos x = − sin x 7. sin x dx = − cos x + C
d
( ) ∫
8. dx tan x = sec2 x 8. sec2 x dx = tan x + C
( ) ∫
d
9. dx csc x = − csc x cot x 9. csc x cot x dx = − csc x + C
d
( ) ∫
10. dx sec x = sec x tan x 10. sec x tan x dx = sec x + C
( ) ∫
d
11. dx cot x = − csc2 x 11. csc2 x dx = − cot x + C
d
( x) ∫
12. dx e = ex 12. ex dx = ex + C
( x) ∫
d
13. dx a = ln a · ax 13. ax dx = ln1a · ax + C
( ) ∫
d
14. dx ln x = 1x 14. 1x dx = ln |x| + C
We highlight a few important points from Theorem 5.1.2:

∫ ∫
• Rule #1 states c · f(x) dx = c · f(x) dx. This is the Constant Mul ple
Rule: we can temporarily ignore constants when finding
( 2 ) an deriva ves,
d
just as we did when compu ng deriva ves (i.e., dx 3x is just as easy to
d
( 2)
compute as dx x ). An example:
∫ ∫
5 cos x dx = 5 · cos x dx = 5 · (sin x + C) = 5 sin x + C.
In the last step we can consider the constant as also being mul plied by
Notes:
201
5, but “5 mes a constant” is s ll a constant, so we just write “C ”.

• Rule #2 is the Sum/Difference Rule: we can split integrals apart when the
integrand contains terms that are added/subtracted, as we did in Example
5.1.2. So:
∫ ∫ ∫ ∫
(3x2 + 4x + 5) dx = 3x2 dx + 4x dx + 5 dx
∫ ∫ ∫
= 3 x2 dx + 4 x dx + 5 dx
1 1
= 3 · x3 + 4 · x2 + 5x + C
3 2
= x3 + 2x2 + 5x + C
In prac ce we generally do not write out all these steps, but we demon-
strate them here for completeness.
• Rule #5 is the Power Rule of indefinite integra on. There are two impor-
tant things to keep in mind:
∫
1. No ce the restric on that n ̸= −1. This is important: 1x dx ̸=
“ 01 x0 + C”; rather, see Rule #14.
2. We are presen ng an differen a on as the “inverse opera on” of
differen a on. Here is a useful quote to remember:
“Inverse opera ons do the opposite things in the opposite
order.”
When taking a deriva ve using the Power Rule, we first mul ply by
the power, then second subtract 1 from the power. To find the an-
deriva ve, do the opposite things in the opposite order: first add
one to the power, then second divide by the power.
• Note that Rule #14 incorporates the absolute value of x. The exercises will
work the reader through why this is the case; for now, know the absolute
value is important and cannot be ignored.
Ini al Value Problems

In Sec on 2.3 we saw that the deriva ve of a posi on func on gave a velocity
func on, and the deriva ve of a velocity func on describes accelera on. We
can now go “the other way:” the an deriva ve of an accelera on func on gives
a velocity func on, etc. While there is just one deriva ve of a given func on,
there are infinitely many an deriva ves. Therefore we cannot ask “What is the
velocity of an object whose accelera on is −32 /s2 ?”, since there is more than
one answer.
Notes:
202
We can find the answer if we provide more informa on with the ques on,
as done in the following example. O en the addi onal informa on comes in the
form of an ini al value, a value of the func on that one knows beforehand.
Example 5.1.3 Solving ini al value problems

The accelera on due to gravity of a falling object is −32 /s2 . At me t = 3,
a falling object had a velocity of −10 /s. Find the equa on of the object’s
velocity.
S We want to know a velocity func on, v(t). We know two

things:
• The accelera on, i.e., v ′ (t) = −32, and
• the velocity at a specific me, i.e., v(3) = −10.
Using the first piece of informa on, we know that v(t) is an an deriva ve of
v ′ (t) = −32. So we begin by finding the indefinite integral of −32:
∫
(−32) dt = −32t + C = v(t).
Now we use the fact that v(3) = −10 to find C:

v(t) = −32t + C
v(3) = −10
−32(3) + C = −10
C = 86
Thus v(t) = −32t + 86. We can use this equa on to understand the mo on
of the object: when t = 0, the object had a velocity of v(0) = 86 /s. Since the
velocity is posi ve, the object was moving upward.
When did the object begin moving down? Immediately a er v(t) = 0:
43
−32t + 86 = 0 ⇒ t= ≈ 2.69s.
16
Recognize that we are able to determine quite a bit about the path of the object
knowing just its accelera on and its velocity at a single point in me.
Example 5.1.4 Solving ini al value problems

Find f(t), given that f ′′ (t) = cos t, f ′ (0) = 3 and f(0) = 5.
S We start by finding f ′ (t), which is an an deriva ve of f ′′ (t):

∫ ∫
f (t) dt = cos t dt = sin t + C = f ′ (t).
′′
Notes:
203
So f ′ (t) = sin t + C for the correct value of C. We are given that f ′ (0) = 3,
so:
f ′ (0) = 3 ⇒ sin 0 + C = 3 ⇒ C = 3.
Using the ini al value, we have found f ′ (t) = sin t + 3.
We now find f(t) by integra ng again.
∫ ∫
f(t) = f ′ (t) dt = (sin t + 3) dt = − cos t + 3t + C.
We are given that f(0) = 5, so
− cos 0 + 3(0) + C = 5
−1 + C = 5
C=6
Thus f(t) = − cos t + 3t + 6.
This sec on introduced an deriva ves and the indefinite integral. We found
they are needed when finding a func on given informa on about its deriva-
ve(s). For instance, we found a velocity func on given an accelera on func-
on.
In the next sec on, we will see how posi on and velocity are unexpectedly
related by the areas of certain regions on a graph of the velocity func on. Then,
in Sec on 5.4, we will see how areas and an deriva ves are closely ed together.
This connec on is incredibly important, as indicated by the name of the theorem
that describes it: The Fundamental Theorem of Calculus.
Notes:
204
Exercises 5.1 ∫
Terms and Concepts 19. (sec x tan x + csc x cot x) dx
1. Define the term “an deriva ve” in your own words. ∫

20. 5eθ dθ
2. Is it more accurate to refer to “the” an deriva ve of f(x) or
“an” an deriva ve of f(x)? ∫
21. 3t dt
3. Use your own words to define the indefinite integral of
f(x). ∫
5t
22. dt
2
4. Fill in the blanks: “Inverse opera ons do the
things in the order.” ∫
23. (2t + 3)2 dt
5. What is an “ini al value problem”?
∫
6. The deriva ve of a posi on func on is a func- 24. (t2 + 3)(t3 − 2t) dt
on.
∫
7. The an deriva ve of an accelera on func on is a 25. x2 x3 dx
func on.
∫
26. eπ dx
8. If F(x) is an an deriva ve of f(x), and G(x) is an an deriva-
ve of g(x), give an an deriva ve of f(x) + g(x).
∫
27. a dx
Problems
28. ∫
This problem inves gates why Theorem 5.1.2 states that
1
In Exercises 9 – 27, evaluate the given indefinite integral. dx = ln |x| + C.
x
∫
(a) What is the domain of y = ln x?
9. 3x3 dx ( )
(b) Find dxd ln x .
∫ (c) What is the domain of y = ln(−x)?
10. x8 dx d
( )
(d) Finddx
ln(−x) .
∫ (e) You should find that 1/x has two types of an deriva-
11. (10x2 − 2) dx ves, depending on whether x >∫ 0 or x < 0. In
1
one expression, give a formula for dx that takes
∫ x
12. dt these different domains into account, and explain
your answer.
∫
13. 1 ds In Exercises 29 – 39, find f(x) described by the given ini al
value problem.
∫
1 29. f ′ (x) = sin x and f(0) = 2
14. dt
3t2
∫ 30. f ′ (x) = 5ex and f(0) = 10
3
15. dt
t2
31. f ′ (x) = 4x3 − 3x2 and f(−1) = 9
∫
1
16. √ dx 32. f ′ (x) = sec2 x and f(π/4) = 5
x
∫ 33. f ′ (x) = 7x and f(2) = 1
2
17. sec θ dθ
34. f ′′ (x) = 5 and f ′ (0) = 7, f(0) = 3
∫
18. sin θ dθ 35. f ′′ (x) = 7x and f ′ (1) = −1, f(1) = 10
205
36. f ′′ (x) = 5ex and f ′ (0) = 3, f(0) = 5 Review
37. f ′′ (θ) = sin θ and f ′ (π) = 2, f(π) = 4 40. Use informa on gained from the first and second deriva-
1
ves to sketch f(x) = x .
38. f ′′ (x) = 24x2 + 2x − cos x and f ′ (0) = 5, f(0) = 0 e +1
39. f ′′ (x) = 0 and f ′ (1) = 3, f(1) = 1 41. Given y = x2 ex cos x, find dy.
206
5.2 The Definite Integral
5.2 The Definite Integral y ( /s)
We start with an easy problem. An object travels in a straight line at a constant 5

velocity of 5 /s for 10 seconds. How far away from its star ng point is the ob-
ject?
We approach this problem with the familiar “Distance = Rate × Time” equa-
on. In this case, Distance = 5 /s × 10s = 50 feet.
It is interes ng to note that this solu on of 50 feet can be represented graph-
ically. Consider Figure 5.2.1, where the constant velocity of 5 /s is graphed on
t (s)
the axes. Shading the area under the line from t = 0 to t = 10 gives a rectangle 5 10
.
with an area of 50 square units; when one considers the units of the axes, we
can say this area represents 50 . Figure 5.2.1: The area under a constant
Now consider a slightly harder situa on (and not par cularly realis c): an velocity func on corresponds to distance
object travels in a straight line with a constant velocity of 5 /s for 10 seconds, traveled.
then instantly reverses course at a rate of 2 /s for 4 seconds. (Since the object
is traveling in the opposite direc on when reversing course, we say the velocity
is a constant −2 /s.) How far away from the star ng point is the object – what
is its displacement? y ( /s)
Here we use “Distance = Rate1 × Time1 + Rate2 × Time2 ,” which is 5
Distance = 5 · 10 + (−2) · 4 = 42 .
Hence the object is 42 feet from its star ng loca on.
We can again depict this situa on graphically. In Figure 5.2.2 we have the
veloci es graphed as straight lines on [0, 10] and [10, 14], respec vely. The dis- t (s)
5 10 15
placement of the object is
−2
“Area above the t–axis − Area below the t–axis,” .
which is easy to calculate as 50 − 8 = 42 feet. Figure 5.2.2: The total displacement is the
Now consider a more difficult problem. area above the t–axis minus the area be-
low the t–axis.
Example 5.2.1 Finding posi on using velocity
The velocity of an object moving straight up/down under the accelera on of
gravity is given as v(t) = −32t+48, where me t is given in seconds and velocity
is in /s. When t = 0, the object had a height of 0 .
1. What was the ini al velocity of the object?
2. What was the maximum height of the object?
3. What was the height of the object at me t = 2?
S It is straigh orward to find the ini al velocity; at me t = 0,

v(0) = −32 · 0 + 48 = 48 /s.
Notes:
207
To answer ques ons about the height of the object, we need to find the
object’s posi on func on s(t). This is an ini al value problem, which we studied
in the previous sec on. We are told the ini al height is 0, i.e., s(0) = 0. We
know s ′ (t) = v(t) = −32t + 48. To find s, we find the indefinite integral of v(t):
∫ ∫
v(t) dt = (−32t + 48) dt = −16t2 + 48t + C = s(t).
Since s(0) = 0, we conclude that C = 0 and s(t) = −16t2 + 48t.

To find the maximum height of the object, we need to find the maximum of
s. Recalling our work finding extreme values, we find the cri cal points of s by
se ng its deriva ve equal to 0 and solving for t:
s ′ (t) = −32t + 48 = 0 ⇒ t = 48/32 = 1.5s.
(No ce how we ended up just finding when the velocity was 0 /s!) The first
deriva ve test shows this is a maximum, so the maximum height of the object
is found at
s(1.5) = −16(1.5)2 + 48(1.5) = 36 .
The height at me t = 2 is now straigh orward to compute: it is s(2) = 32 .
y ( /s)
50
While we have answered all three ques ons, let’s look at them again graph-
ically, using the concepts of area that we explored earlier.
Figure 5.2.3 shows a graph of v(t) on axes from t = 0 to t = 3. It is again
straigh orward to find v(0). How can we use the graph to find the maximum
t (s)
height of the object?
1 2 3 Recall how in our previous work that the displacement of the object (in this
case, its height) was found as the area under the velocity curve, as shaded in the
figure. Moreover, the area between the curve and the t–axis that is below the
−50
t–axis counted as “nega ve” area. That is, it represents the object coming back
. toward its star ng posi on. So to find the maximum distance from the star ng
Figure 5.2.3: A graph of v(t) = −32t + point – the maximum height – we find the area under the velocity line that is
48; the shaded areas help determine dis- above the t–axis, i.e., from t = 0 to t = 1.5. This region is a triangle; its area is
placement.
1 1
Area = Base × Height = × 1.5s × 48 /s = 36 ,
2 2
which matches our previous calcula on of the maximum height.
Finally, to find the height of the object at me t = 2 we calculate the total
“signed area” (where some area is nega ve) under the velocity func on from
t = 0 to t = 2. This signed area is equal to s(2), the displacement (i.e., signed
distance) from the star ng posi on at t = 0 to the posi on at me t = 2. That
is,
Displacement = Area above the t–axis − Area below t–axis.
Notes:
208
The regions are triangles, and we find

1 1
Displacement = (1.5s)(48 /s) − (.5s)(16 /s) = 32 .
2 2
This also matches our previous calcula on of the height at t = 2.
No ce how we answered each ques on in this example in two ways. Our first
method was to manipulate equa ons using our understanding of an deriva ves
and deriva ves. Our second method was geometric: we answered ques ons
looking at a graph and finding the areas of certain regions of this graph.
The above example does not prove a rela onship between area under a ve-
locity func on and displacement, but it does imply a rela onship exists. Sec on
5.4 will fully establish fact that the area under a velocity func on is displace-
ment.
Given a graph of a func on y = f(x), we will find that there is great use in
compu ng the area between the curve y = f(x) and the x-axis. Because of this,
we need to define some terms.
Defini on 5.2.1 The Definite Integral, Total Signed Area

Let y = f(x) be defined on a closed interval [a, b]. The total signed area
from x = a to x = b under f is:
(area under f and above the x–axis on [a, b]) − (area above f and under
the x–axis on [a, b]).
The definite integral of f on [a, b] is the total signed area of f on [a, b],
denoted ∫ b
f(x) dx,
a
where a and b are the bounds of integra on.
By our defini on, the definite integral gives the “signed area under f.” We
usually drop the word “signed” when talking about the definite integral, and
simply say the definite integral gives “the area under f ” or, more commonly,
“the area under the curve.”
The previous sec on introduced the indefinite integral, which related to an-
deriva ves. We have now defined the definite integral, which relates to areas
under a func on. The two are very much related, as we’ll see when we learn
the Fundamental
∫ Theorem of Calculus in Sec on 5.4. Recall that earlier we said
that the “ ” symbol was an “elongated S” that represented finding a “sum.” In
the context of the definite integral, this nota on makes a bit more sense, as we
are adding up areas under the func on f.
Notes:
209
We prac ce using this nota on.

y
1
Example 5.2.2 Evalua ng definite integrals
Consider the func on f given in Figure 5.2.4.
Find:
x
1 2 3 4 5 ∫ 3 ∫ 3
1. f(x) dx 4. 5f(x) dx
0 0
−1
∫ 5 ∫ 1
.
2. f(x) dx 5. f(x) dx
Figure 5.2.4: A graph of f(x) in Example 3 1
5.2.2. ∫ 5
3. f(x) dx
0
∫3
1. f(x) dx is the area under f on the interval [0, 3]. This region is a triangle,
0 ∫3
y so the area is 0 f(x) dx = 12 (3)(1) = 1.5.
5
∫5
2. 3
f(x) dx represents the area of the triangle found under the x–axis on
[3, 5]. The area is 21 (2)(1) = 1; since it is found under the x–axis, this is
∫5
x “nega ve area.” Therefore 3 f(x) dx = −1.
1 2 3 4 5
∫5
3. 0
f(x) dx is the total signed area under f on [0, 5]. This is 1.5+(−1) = 0.5.
−5
.
Figure 5.2.5: A graph of 5f in Example ∫3
4. 0
5f(x) dx is the area under 5f on [0, 3]. This is sketched in Figure 5.2.5.
5.2.2. (Yes, it looks just like the graph of
Again, the region is a triangle, with∫ height 5 mes that of the height of the
f in Figure 5.2.4, just with a different y- 3
scale.)
original triangle. Thus the area is 0 5f(x) dx = 12 (15)(1) = 7.5.
∫1
5. 1
f(x) dx is the area under f on the “interval” [1, 1]. This describes a line
segment, not a region; it has no width. Therefore the area is 0.
This example illustrates some of the proper es of the definite integral, given
here.
Notes:
210
Theorem 5.2.1 Proper es of the Definite Integral

Let f and g be defined on a closed interval I that contains the values a, b
and c, and let k be a constant. The following hold:
∫ a
1. f(x) dx = 0
a
∫ b ∫ c ∫ c
2. f(x) dx + f(x) dx = f(x) dx
a b a
∫ b ∫ a
3. f(x) dx = − f(x) dx
a b
∫ ∫ ∫
b ( ) b b
4. f(x) ± g(x) dx = f(x) dx ± g(x) dx
a a a
∫ b ∫ b
5. k · f(x) dx = k · f(x) dx
a a
We give a brief jus fica on of Theorem 5.2.1 here.

1. As demonstrated in Example 5.2.2, there is no “area under the curve”
when the region has no width; hence this definite integral is 0.
2. This states that total area is the sum of the areas of subregions. It is easily
considered when we let a < b < c. We can break the interval [a, c] into
two subintervals, [a, b] and [b, c]. The total area over [a, c] is the area over
[a, b] plus the area over [b, c].
It is important to note that this s ll holds true even if a < b < c is not
true. We discuss this in the next point.
3. This property can be viewed a merely a conven on to make other proper-
es work well. (Later we will see how this property has a jus fica on all its
own, not necessarily in support of other proper es.) Suppose b < a < c.
The discussion from the previous point clearly jus fies
∫ a ∫ c ∫ c
f(x) dx + f(x) dx = f(x) dx. (5.1)
b a b
However, we s ll claim that, as originally stated,

∫ b ∫ c ∫ c
f(x) dx + f(x) dx = f(x) dx. (5.2)
a b a
Notes:
211
How do Equa ons (5.1) and (5.2) relate? Start with Equa on (5.1):
∫ a ∫ c ∫ c
f(x) dx + f(x) dx = f(x) dx
b
∫a c b
∫ a ∫ c
f(x) dx = − f(x) dx + f(x) dx
a b b
Property (3) jus ∫

fies changing the sign and switching the bounds of inte-
a
gra on on the − f(x) dx term; when this is done, Equa ons (5.1) and
b
(5.2) are equivalent.
The conclusion is this: by adop ng the conven on of Property (3), Prop-
erty (2) holds no ma er the order of a, b and c. Again, in the next sec on
we will see another jus fica on for this property.
4,5. Each of these may be non–intui ve. Property (5) states that when one
scales a func on by, for instance, 7, the area of the enclosed region also
is scaled by a factor of 7. Both Proper es (4) and (5) can be proved using
geometry. The details are not complicated but are not discussed here.
Example 5.2.3 Evalua ng definite integrals using Theorem 5.2.1.
y Consider the graph of a func on f(x) shown in Figure 5.2.6. Answer the follow-
ing:
∫ b ∫ c
1. Which value is greater: f(x) dx or f(x) dx?
a b
x
∫ c
a b c 2. Is f(x) dx greater or less than 0?
a
∫ b ∫ b
3. Which value is greater: f(x) dx or f(x) dx?
. a c
Figure 5.2.6: A graph of a func on in Ex-

S
ample 5.2.3. ∫b
1. f(x) dx has a posi ve value (since the area is above the x–axis) whereas
∫ac ∫b
b
f(x) dx has a nega ve value. Hence a f(x) dx is bigger.
∫c
2. a f(x) dx is the total signed area under f between x = a and x = c. Since
the region below the x–axis looks to be larger than the region above, we
conclude that the definite integral has a value less than 0.
∫b
3. Note how the second integral has the bounds “reversed.” Therefore c f(x) dx
represents a posi ve number, greater than the area described by the first
∫b
definite integral. Hence c f(x) dx is greater.
Notes:
212
The area defini on of the definite integral allows us to use geometry to com- y
pute the definite integral of some simple func ons. 10
(5, 6)
Example 5.2.4 Evalua ng definite integrals using geometry 5

Evaluate the following definite integrals: R2
x
∫ 5 ∫ 3√ −2 R1 2 5
1. (2x − 4) dx 2. 9 − x2 dx.
−2 −3 −5
(−2, −8) −10

S .
(a)
1. It is useful to sketch the func on in the integrand, as shown in Figure
y
5.2.7(a). We see we need to compute the areas of two regions, which
we have labeled R1 and R2 . Both are triangles, so the area computa on is 5
straigh orward:
1 1
R1 : (4)(8) = 16 R2 : (3)6 = 9.
2 2
Region R1 lies under the x–axis, hence it is counted as nega ve area (we
can think of the triangle’s height as being “−8”), so
∫ 5 x
. −3 3
(2x − 4) dx = −16 + 9 = −7.
−2 (b)
2. Recognize that the integrand of this definite integral describes a half circle, graph of f(x) = 2x − 4 in
Figure 5.2.7: A√
as sketched in Figure 5.2.7(b), with radius 3. Thus the area is: (a) and f(x) = 9 − x2 in (b), from Exam-
∫ 3√ ple 5.2.4.
1 9
9 − x2 dx = πr2 = π.
−3 2 2
y ( /s)
Example 5.2.5 Understanding mo on given velocity 15
Consider the graph of a velocity func on of an object moving in a straight line,
given in Figure 5.2.8, where the numbers in the given regions gives the area of 10
38
that region. Assume that the definite integral of a velocity func on gives dis-
placement. Find the maximum speed of the object and its maximum displace- 5
ment from its star ng posi on. t (s)

a b c
S Since the graph gives velocity, finding the maximum speed 11 11
−5
is simple: it looks to be 15 /s. .
At me t = 0, the displacement is 0; the object is at its star ng posi on. At
me t = a, the object has moved backward 11 feet. Between mes t = a and Figure 5.2.8: A graph of a velocity in Ex-
ample 5.2.5.
Notes:
213
y t = b, the object moves forward 38 feet, bringing it into a posi on 27 feet for-
10 ward of its star ng posi on. From t = b to t = c the object is moving backwards
again, hence its maximum displacement is 27 feet from its star ng posi on.
In our examples, we have either found the areas of regions that have nice
5 geometric shapes (such as rectangles, triangles and circles) or the areas were
given to us. Consider Figure 5.2.9, where a region below y = x2 is shaded. What
is its area? The func on y = x2 is rela vely simple, yet the shape it defines has
an area that is not simple to find geometrically.
x
. 1 2 3
In the next sec on we will explore how to find the areas of such regions.
Figure 5.2.9: What is the area below y =
x2 on [0, 3]? The region is not a usual ge-
ometric shape.
Notes:
214
Exercises 5.2
Terms and Concepts y
1. What is “total signed area”? y = f(x)

7. 2
2. What is “displacement”?
x
∫ 3 . 1 2 3 4
3. What is sin x dx? ∫ 2 ∫ 1
3 (a) f(x) dx (d) 4x dx
0 0
∫ 4 ∫ 3
4. Give a single definite integral that has the same value as (b) f(x) dx (e) (2x − 4) dx
∫ 1 ∫ 2 2 2
(2x + 3) dx + (2x + 3) dx. ∫ 4 ∫ 3
0 1 (c) 2f(x) dx (f) (4x − 8) dx
2 2
Problems
In Exercises 5 – 10, a graph of a func on f(x) is given. Using y
the geometry of the graph, evaluate the definite integrals. 3

y=x−1
2
y 8. 1
4
y = −2x + 4 x
2 1 2 3 4
−1
5. x .
2 4
−2
−4
.
∫ ∫ ∫ 1 ∫ 3
1 3
(a) (−2x + 4) dx (d) (−2x + 4) dx (a) (x − 1) dx (d) (x − 1) dx
0 2
∫
0 1
∫ ∫ 2 ∫ 4
2 4
(b) (−2x + 4) dx (e) (−2x + 4) dx (b) (x − 1) dx (e) (x − 1) dx
0 1
∫
0 2
∫ ∫ ∫
3 1
3 4 ( )
(c) (−2x + 4) dx (f) (−6x + 12) dx (c) (x − 1) dx (f) (x − 1) + 1 dx
0 1
0 0
y
2
1
y
6. x 3
1 2 3 4 5 √
f(x) = 4 − (x − 2)2
−1
2
y = f(x) 9.
. −2
∫ 2 ∫ 5
1
(a) f(x) dx (d) f(x) dx

0 2 x
∫ ∫ . 1 2 3 4
3 3
∫ 2 ∫ 4
(b) f(x) dx (e) f(x) dx
0 5 (a) f(x) dx (c) f(x) dx
∫ ∫ 0 0
5 3
∫ ∫
(c) f(x) dx (f) −2f(x) dx 4 4
0 0 (b) f(x) dx (d) 5f(x) dx

2 0
215
y
10 f(x) = 3x2 − 3
y
f(x) = 3 5
3 13.
4 4
x
2 4
−2 −1 1 2
10.
1 −5
∫ −1 ∫ 1
x (a) f(x) dx (c) f(x) dx
5 10 −2 −1
∫ 5 ∫ 0 ∫ 2 ∫ 1
(a) f(x) dx (c) f(x) dx (b) f(x) dx (d) f(x) dx
0 0 1 0
∫ 7 ∫ b
(b) f(x) dx (d) f(x) dx, where y
3 a 4
0 ≤ a ≤ b ≤ 10
3 f(x) = x2
In Exercises 11 – 14, a graph of a func on f(x) is given; the 14. 2
numbers inside the shaded regions give the area of that re-
1
gion. Evaluate the definite integrals using this area informa-
on. .
1/3 7/3
x
1 2
∫ 2 ∫ 3
y (a) 5x2 dx (c) (x − 1)2 dx
50 0 1
∫ ∫
y = f(x) 2 4 ( )
11 21 (b) (x2 + 3) dx (d) (x − 2)2 + 5 dx
x 0 2
59 1 2 3
11.
−50 In Exercises 15 – 16, a graph of the velocity func on of an ob-
ject moving in a straight line is given. Answer the ques ons
based on that graph.
. −100
∫ 1 ∫ 3
y ( /s)
(a) f(x) dx (c) f(x) dx
0 0 2
∫ 2 ∫ 2
(b) f(x) dx (d) −3f(x) dx
1
0 1
15.
y t (s)
1 2 3
1
f(x) = sin(πx/2) −1.
4/π
12. (a) What is the object’s maximum velocity?

x
1 2 3 4 (b) What is the object’s maximum displacement?
4/π
(c) What is the object’s total displacement on [0, 3]?
−1
.
∫ 2 ∫ 4
y ( /s)
3
(a) f(x) dx (c) f(x) dx
0 0
∫ 4 ∫ 1 2
(b) f(x) dx (d) f(x) dx 16.
2 0
1
t (s)
1 2 3 4 5
.
(a) What is the object’s maximum velocity?
(b) What is the object’s maximum displacement?
(c) What is the object’s total displacement on [0, 5]?
216
17. An object is thrown straight up with a velocity, in /s, given In Exercises 23 – 26, let
by v(t) = −32t + 64, where t is in seconds, from a height ∫ 3
of 48 feet. • s(t) dt = 10,
0
(a) What is the object’s maximum velocity? ∫ 5
(b) What is the object’s maximum displacement? • s(t) dt = 8,
3
(c) When does the maximum displacement occur? ∫ 5
(d) When will the object reach a height of 0? (Hint: find • r(t) dt = −1, and
when the displacement is −48 .) 3
∫ 5
18. An object is thrown straight up with a velocity, in /s, given • r(t) dt = 11.
0
by v(t) = −32t + 96, where t is in seconds, from a height
of 64 feet. Use these values to evaluate the given definite integrals.
(a) What is the object’s ini al velocity? ∫ 3 ( )

(b) When is the object’s displacement 0? 23. s(t) + r(t) dt
0
(c) How long does it take for the object to return to its
∫
ini al height? 0 ( )
24. s(t) − r(t) dt
(d) When will the object reach a height of 210 feet? 5
∫
In Exercises 19 – 22, let 3 ( )
∫ 2 25. πs(t) − 7r(t) dt
• f(x) dx = 5, 3
0
∫ 3
26. Find nonzero values for a and b such that
• f(x) dx = 7, ∫ 5
( )
0
∫ 2 ar(t) + bs(t) dt = 0
0
• g(x) dx = −3, and
0
∫ 3
• g(x) dx = 5.
2
Review
Use these values to evaluate the given definite integrals.
In Exercises 27 – 30, evaluate the given indefinite integral.
∫ 2 ( ) ∫
19. f(x) + g(x) dx ( 3 )
0 27. x − 2x2 + 7x − 9 dx
∫ 3 ( ) ∫
20. f(x) − g(x) dx ( )
0 28. sin x − cos x + sec2 x dx
∫ 3 ( ) ∫
21. 3f(x) + 2g(x) dx (√ 1 )
2 29. 3
t + 2 + 2t dt
t
22. ∫
Find nonzero values for a and b such that ∫ ( )
3( ) 1
af(x) + bg(x) dx = 0 30. − csc x cot x dx
0 x
217
5.3 Riemann Sums

In the previous sec on we defined the definite integral of a func on on [a, b] to
be the signed area between the curve and the x–axis. Some areas were simple
y to compute; we ended the sec on with a region whose area was not simple to
compute. In this sec on we develop a technique to find such areas.
4 A fundamental calculus technique is to first answer a given problem with an
3
approxima on, then refine that approxima on to make it be er, then use limits
in the refining process to find the exact answer. That is what we will do here.
2 Consider the region given in Figure 5.3.1, which is the area∫under y = 4x−x2
4
1
on [0, 4]. What is the signed area of this region – i.e., what is 0 (4x − x2 ) dx?
We start by approxima ng. We can surround the region with a rectangle
x
1 2 3 4
with height and width of 4 and find the area is approximately 16 square units.
. This is obviously an over–approxima on; we are including area in the rectangle
that is not under the parabola.
Figure 5.3.1: A graph of f(x) = 4x − x2 .
What is the area of the shaded region?
We have an approxima on of the area, using one rectangle. How can we
refine our approxima on to make it be er? The key to this sec on is this answer:
use more rectangles.
Let’s use 4 rectangles with an equal width of 1. This par ons the interval
y
[0, 4] into 4 subintervals, [0, 1], [1, 2], [2, 3] and [3, 4]. On each subinterval we
will draw a rectangle.
4
There are three common ways to determine the height of these rectangles:
the Le Hand Rule, the Right Hand Rule, and the Midpoint Rule. The Le Hand
3 Rule says to evaluate the func on at the le –hand endpoint of the subinterval
2
and make the rectangle that height. In Figure 5.3.2, the rectangle drawn on the
interval [2, 3] has height determined by the Le Hand Rule; it has a height of
1 f(2). (The rectangle is labeled “LHR.”)
RHR MPR LHR other The Right Hand Rule says the opposite: on each subinterval, evaluate the
x
1 2 3 4 func on at the right endpoint and make the rectangle that height. In the figure,
.
the rectangle drawn on [0, 1] is drawn using f(1) as its height; this rectangle is
∫4
Figure 5.3.2: Approxima ng 0 (4x − labeled “RHR.”.
2
x ) dx using rectangles. The heights of the The Midpoint Rule says that on each subinterval, evaluate the func on at
rectangles are determined using different the midpoint and make the rectangle that height. The rectangle drawn on [1, 2]
rules. was made using the Midpoint Rule, with a height of f(1.5). That rectangle is
labeled “MPR.”
These are the three most common rules for determining the heights of ap-
proxima ng rectangles, but one is not forced to use one of these three methods.
The rectangle on [3, 4] has a height of approximately f(3.53), very close to the
Midpoint Rule. It was chosen so that the area of the rectangle is exactly the area
of the region under f on [3, 4]. (Later you’ll be able to figure how to do this, too.)
∫4
The following example will approximate the value of 0 (4x − x2 ) dx using
these rules.
Notes:
218
5.3 Riemann Sums
Using the y
Example 5.3.1 ∫ 4 Le Hand, Right Hand and Midpoint Rules
Approximate the value of 0 (4x − x2 ) dx using the Le Hand Rule, the Right
Hand Rule, and the Midpoint Rule, using 4 equally spaced subintervals. 4
3
S We break the interval [0, 4] into four subintervals as before.
In Figure 5.3.3(a) we see 4 rectangles drawn on f(x) = 4x − x2 using the Le 2
Hand Rule. (The areas of the rectangles are given in each figure.) 1
Note how in the first subinterval, [0, 1], the rectangle has height f(0) = 0. 0 3 4 3 x
We add up the areas of each rectangle (height× width) for our Le Hand Rule 1 2 3 4
approxima on: .
(a)
f(0) · 1 + f(1) · 1 + f(2) · 1 + f(3) · 1 = y
0 + 3 + 4 + 3 = 10.
4
Figure 5.3.3(b) shows 4 rectangles drawn under f using the Right Hand Rule; 3
note how the [3, 4] subinterval has a rectangle of height 0.
In this example, these rectangle seem to be the mirror image of those found 2
in part (a) of the Figure. This is because of the symmetry of our shaded region. 1
Our approxima on gives the same answer as before, though calculated a differ- 3 4 3 0 x
ent way: 1 2 3 4
.
f(1) · 1 + f(2) · 1 + f(3) · 1 + f(4) · 1 = (b)
y
3 + 4 + 3 + 0 = 10.
Figure 5.3.3(c) shows 4 rectangles

∫4 drawn under f using the Midpoint Rule. 4
This gives an approxima on of 0 (4x − x2 ) dx as: 3
f(0.5) · 1 + f(1.5) · 1 + f(2.5) · 1 + f(3.5) · 1 = 2
1.75 + 3.75 + 3.75 + 1.75 = 11. 1
∫4 1.75 3.75 3.75 1.75 x

Our three methods provide two approxima ons of 0
(4x − x2 ) dx: 10 and 11. 1 2 3 4
.
Summa on Nota on (c)
∫4
It is hard to tell at this moment which is a be er approxima on: 10 or 11? Figure 5.3.3: Approxima ng 0 (4x −
x2 ) dx in Example 5.3.1. In (a), the Le
We can con nue to refine our approxima on by using more rectangles. The
Hand Rule is used; in (b), the Right Hand
nota on can become unwieldy, though, as we add up longer and longer lists of
Rule is used; in (c), the Midpoint Rule is
numbers. We introduce summa on nota on to ameliorate this problem. used.
Notes:
219
Suppose we wish to add up a list of numbers a1 , a2 , a3 , …, a9 . Instead of

wri ng
a 1 + a2 + a 3 + a4 + a5 + a6 + a7 + a 8 + a9 ,
we use summa on nota on and write
upper
bound summand
∑
9
ai .
i=1
.
i=index lower
of summa on bound
Figure 5.3.4: Understanding summa on nota on.
The upper case sigma represents the term “sum.” The index of summa on
in this example is i; any symbol can be used. By conven on, the index takes on
only the integer values between (and including) the lower and upper bounds.
Let’s prac ce using this nota on.
Example 5.3.2 Using summa on nota on

Let the numbers {ai } be defined as ai = 2i − 1 for integers i, where i ≥ 1. So
a1 = 1, a2 = 3, a3 = 5, etc. (The output is the posi ve odd integers). Evaluate
the following summa ons:
∑
6 ∑
7 ∑
4
1. ai 2. (3ai − 4) 3. (ai )2
i=1 i=3 i=1
S
∑
6
1. ai = a 1 + a2 + a 3 + a4 + a5 + a6
i=1
= 1 + 3 + 5 + 7 + 9 + 11
= 36.
2. Note the star ng value is different than 1:
∑
7
(3ai − 4) = (3a3 − 4) + (3a4 − 4) + (3a5 − 4) + (3a6 − 4) + (3a7 − 4)
i=3
= 11 + 17 + 23 + 29 + 35
= 115.
Notes:
220
5.3 Riemann Sums
3.
∑
4
(ai )2 = (a1 )2 + (a2 )2 + (a3 )2 + (a4 )2
i=1
= 12 + 32 + 52 + 72
= 84.
It might seem odd to stress a new, concise way of wri ng summa ons only
to write each term out as we add them up. It is. The following theorem gives
some of the proper es of summa ons that allow us to work with them without
wri ng individual terms. Examples will follow.
Theorem 5.3.1 Proper es of Summa ons
∑
n ∑
n
n(n + 1)
1. c = c · n, where c is a constant. 5. i=
i=1 i=1
2
∑
n ∑
n ∑
n ∑
n
n(n + 1)(2n + 1)
2. (ai ± bi ) = ai ± bi 6. i2 =
i=m i=m i=m i=1
6
∑
n ∑
n ∑
n ( )2
n(n + 1)
3. c · ai = c · ai 7. i3 =
i=m i=m i=1
2
∑
j ∑
n ∑
n
4. ai + ai = ai
i=m i=j+1 i=m
Example 5.3.3 Evalua ng summa ons using Theorem 5.3.1

Revisit Example 5.3.2 and, using Theorem 5.3.1, evaluate
∑
6 ∑
6
ai = (2i − 1).
i=1 i=1
Notes:
221
∑
6 ∑
6 ∑
6
(2i − 1) = 2i − (1)
i=1 i=1 i=1
( )
∑
6
= 2 i −6
i=1
6(6 + 1)
=2 −6
2
= 42 − 6 = 36
We obtained the same answer without wri ng out all six terms. When dealing
with small sizes of n, it may be faster to write the terms out by hand. However,
Theorem 5.3.1 is incredibly important when dealing with large sums as we’ll
soon see.
Riemann Sums
∫4
Consider again 0 (4x − x2 ) dx. We will approximate this definite integral
using 16 equally spaced subintervals and the Right Hand Rule in Example 5.3.4.
Before doing so, it will pay to do some careful prepara on.
Figure 5.3.5 shows a number line of [0, 4] divided, or par oned, into 16
equally spaced subintervals. We denote 0 as x1 ; we have marked the values of x5 ,
x9 , x13 and x17 . We could mark them all, but the figure would get crowded. While
.
0 1 2 3 4 it is easy to figure that x10 = 2.25, in general, we want a method of determining
x1 x5 x9 x13 x17 the value of xi without consul ng the figure. Consider:
Figure 5.3.5: Dividing [0, 4] into 16 number of

equally spaced subintervals. subintervals
between x1 and xi
xi = x1 + (i − 1)∆x
. star ng subinterval
value size
So x10 = x1 + 9(4/16) = 2.25.

If we had par oned [0, 4] into 100 equally spaced subintervals, each subin-
terval would have length ∆x = 4/100 = 0.04. We could compute x32 as
x32 = x1 + 31(4/100) = 1.24.
(That was far faster than crea ng a sketch first.)
Notes:
222
5.3 Riemann Sums
Given any subdivision of [0, 4], the first subinterval is [x1 , x2 ]; the second is
[x2 , x3 ]; the i th subinterval is [xi , xi+1 ].
When using the Le Hand Rule, the height of the i th rectangle will be f(xi ).
When using the Right Hand Rule, the height of the i th rectangle will be f(xi+1 ).
( )
th xi + xi+1
When using the Midpoint Rule, the height of the i rectangle will be f .
2
∫4
Thus approxima ng 0 (4x − x2 ) dx with 16 equally spaced subintervals can
be expressed as follows, where ∆x = 4/16 = 1/4:
∑
16
Le Hand Rule: f(xi )∆x
i=1
∑
16
Right Hand Rule: f(xi+1 )∆x
i=1
∑16 ( )
xi + xi+1
Midpoint Rule: f ∆x
i=1
2
We use these formulas in the next two examples. The following example lets
us prac ce using the Right Hand Rule and the summa on formulas introduced
in Theorem 5.3.1.
Example 5.3.4∫ Approxima ng definite integrals using sums

4
Approximate 0 (4x − x2 ) dx using the Right Hand Rule and summa on formulas
with 16 and 1000 equally spaced intervals.
S Using the formula derived before, using 16 equally spaced

intervals and the Right Hand Rule, we can approximate the definite integral as
∑
16
f(xi+1 )∆x.
i=1
We have ∆x = 4/16 = 0.25. Since xi = 0 + (i − 1)∆x, we have
( )
xi+1 = 0 + (i + 1) − 1 ∆x
= i∆x
Notes:
223
Using the summa on formulas, consider:

∫ 4 ∑
16
(4x − x2 ) dx ≈ f(xi+1 )∆x
0 i=1
∑
16
= f(i∆x)∆x
i=1
∑
16
( )
= 4i∆x − (i∆x)2 ∆x
i=1
∑
16
= (4i∆x2 − i2 ∆x3 )
i=1
∑
16 ∑
16
= (4∆x2 ) i − ∆x3 i2 (5.3)
i=1 i=1
16 · 17 16(17)(33)
= (4∆x2 ) − ∆x3 (∆x = 0.25)
2 6
= 10.625
y We were able to sum up the areas of 16 rectangles with very li le computa on.
In Figure 5.3.6 the func on and the 16 rectangles are graphed. While some
4 rectangles over–approximate the area, other under–approximate the area (by
about the same amount). Thus our approximate area of 10.625 is likely a fairly
3
good approxima on.
2 No ce Equa on (5.3); by changing the 16’s to 1,000’s (and appropriately
changing the value of ∆x), we can use that equa on to sum up 1000 rectan-
1
gles! We do so here, skipping from the original summand to the equivalent of
x Equa on (5.3) to save space. Note that ∆x = 4/1000 = 0.004.
1 2 3 4
. ∫ 4 ∑
1000
∫4 (4x − x2 ) dx ≈ f(xi+1 )∆x
Figure 5.3.6: Approxima ng 0 (4x −
0 i=1
x2 ) dx with the Right Hand Rule and 16
evenly spaced subintervals. ∑
1000 ∑
1000
= (4∆x2 ) i − ∆x3 i2
i=1 i=1
1000 · 1001 1000(1001)(2001)
= (4∆x2 ) − ∆x3
2 6
= 10.666656
∫Using
4
many, many rectangles, we have a likely good approxima on of
0
(4x − x2 )∆x. That is,
∫ 4
(4x − x2 ) dx ≈ 10.666656.
0
Notes:
224
5.3 Riemann Sums
Before the above example, we stated what the summa ons for the Le Hand,
Right Hand and Midpoint Rules looked like. Each had the same basic structure,
which was:
1. each rectangle has the same width, which we referred to as ∆x, and
2. each rectangle’s height is determined by evalua ng f at a par cular point
in each subinterval. For instance, the Le Hand Rule states that each rect-
angle’s height is determined by evalua ng f at the le hand endpoint of
the subinterval the rectangle lives on.
One could par on an interval [a, b] with subintervals that do not have the same
size. We refer to the length of the i th subinterval as ∆xi . Also, one could deter-
mine each rectangle’s height by evalua ng f at any point ci in the i th subinterval.
Thus the height of the i th subinterval would be f(ci ), and the area of the i th rect-
angle would be f(ci )∆xi . These ideas are formally defined below.
Defini on 5.3.1 Par on

A par on ∆x of a closed interval [a, b] is a set of numbers x1 , x2 , . . .
xn+1 where
a = x1 < x2 < . . . < xn < xn+1 = b.
The length of the i th subinterval, [xi , xi+1 ], is ∆xi = xi+1 − xi . If [a, b] is
par oned into subintervals of equal length, we let ∆x represent the
length of each subinterval.
The size of the par on, denoted ||∆x||, is the length of the largest
subinterval of the par on.
Summa ons of rectangles with area f(ci )∆xi are named a er mathema cian
Georg Friedrich Bernhard Riemann, as given in the following defini on.
Defini on 5.3.2 Riemann Sum

Let f be defined on a closed interval [a, b], let ∆x be a par on of [a, b]
and let ci denote any value in the i th subinterval.
The sum
∑n
f(ci )∆xi
i=1
is a Riemann sum of f on [a, b].
Notes:
225
y ∫4
Figure 5.3.7 shows the approxima ng rectangles of a Riemann sum of 0 (4x−
4
x2 ) dx. While the rectangles in this example do not approximate well the shaded
area, they demonstrate that the subinterval widths may vary and the heights of
3 the rectangles can be determined without following a par cular rule.
2 “Usually” Riemann sums are calculated using one of the three methods we
have introduced. The uniformity of construc on makes computa ons easier.
1
Before working another example, let’s summarize some of what we have learned
x in a convenient way.
1 2 3 4
.
Figure 5.3.7: An example of∫a general Rie- Key Idea 5.3.1 Riemann Sum Concepts
4
mann sum to approximate 0 (4x−x2 ) dx. ∫ b ∑
n
Consider f(x) dx ≈ f(ci )∆xi .
a i=1
b−a
1. When the n subintervals have equal length, ∆xi = ∆x = .
n
2. The i th term of an equally spaced par on is xi = a + (i − 1)∆x.
(Thus x1 = a and xn+1 = b.)
∑
n
3. The Le Hand Rule summa on is: f(xi )∆x.
i=1
∑
n
4. The Right Hand Rule summa on is: f(xi+1 )∆x.
i=1
∑n ( )
xi + xi+1
5. The Midpoint Rule summa on is: f ∆x.
i=1
2
Let’s do another example.
Example 5.3.5∫ Approxima ng definite integrals with sums

3
Approximate −2 (5x + 2) dx using the Midpoint Rule and 10 equally spaced
intervals.
S Following Key Idea 5.3.1, we have
3 − (−2)
∆x = = 1/2 and xi = (−2) + (1/2)(i − 1) = i/2 − 5/2.
10
Notes:
226
5.3 Riemann Sums
xi + xi+1
As we are using the Midpoint Rule, we will also need xi+1 and . Since
2
xi = i/2 − 5/2, xi+1 = (i + 1)/2 − 5/2 = i/2 − 2. This gives
xi + xi+1 (i/2 − 5/2) + (i/2 − 2) i − 9/2
= = = i/2 − 9/4.
2 2 2
We now construct the Riemann sum and compute its value using summa on
formulas.
∫ 3 ∑10 ( )
xi + xi+1
(5x + 2) dx ≈ f ∆x
−2 i=1
2
∑
10
= f(i/2 − 9/4)∆x
i=1
∑
10
( )
= 5(i/2 − 9/4) + 2 ∆x
i=1
10 [( )
∑ ] y
5 37
= ∆x i− 17
i=1
2 4
( 10 ( ))
5∑ ∑
10
37 10
= ∆x (i) −
2 i=1 i=1
4
( )
1 5 10(11) 37
= · − 10 · x
2 2 2 4 −2 −1 1 2 3
45
= = 22.5 −8
2 .
∫3
Note the graph of f(x) = 5x + 2 in Figure 5.3.8. The regions whose area is Figure 5.3.8: Approxima ng −2 (5x +
2) dx using the Midpoint Rule and 10
computed by the definite integral are triangles, meaning we can find the exact
evenly spaced subintervals in Example
answer without summa on techniques. We find that the exact answer is indeed
5.3.5.
22.5. One of the strengths of the Midpoint Rule is that o en each rectangle
includes area that should not be counted, but misses other area that should.
When the par on size is small, these two amounts are about equal and these
errors almost “cancel each other out.” In this example, since our func on is a
line, these errors are exactly equal and they do cancel each other out, giving us
the exact answer.
Note too that when the func on is nega ve, the rectangles have a “nega ve”
height. When we compute the area of the rectangle, we use f(ci )∆x; when f is
nega ve, the area is counted as nega ve.
No ce in the previous example that while we used 10 equally spaced inter-

vals, the number “10” didn’t play a big role in the calcula ons un l the very end.
Notes:
227
Mathema cians love to abstract ideas; let’s approximate the area of another re-
gion using n subintervals, where we do not specify a value of n un l the very end.
Example∫ 45.3.6 Approxima ng definite integrals with a formula, using sums

Revisit 0 (4x−x2 ) dx yet again. Approximate this definite integral using the Right
Hand Rule with n equally spaced subintervals.
S Using Key Idea 5.3.1, we know ∆x = 4−0 n = 4/n. We also

find xi = 0 + ∆x(i − 1) = 4(i − 1)/n. The Right Hand Rule uses xi+1 , which is
xi+1 = 4i/n.
We construct the Right Hand Rule Riemann sum as follows. Be sure to fol-
low each step carefully. If you get stuck, and do not understand how one line
proceeds to the next, you may skip to the result and consider how this result
is used. You should come back, though, and work through each step for full
understanding.
∫ 4 ∑
n
(4x − x2 ) dx ≈ f(xi+1 )∆x
0 i=1
∑n ( )
4i
= f ∆x
i=1
n
[ ( )2 ]
∑n
4i 4i
= 4 − ∆x
i=1
n n
∑n ( ) ∑ n ( )
16∆x 16∆x 2
= i− i
i=1
n i=1
n2
( ) n ( ) n
16∆x ∑ 16∆x ∑ 2
= i− i
n i=1
n2 i=1
( ) ( ) ( )
16∆x n(n + 1) 16∆x n(n + 1)(2n + 1) recall
= · − ∆x = 4/n
n 2 n2 6
32(n + 1) 32(n + 1)(2n + 1)
= − (now simplify)
(n ) 3n2
32 1
= 1− 2
3 n
The result is an amazing, easy to use formula. To approximate the definite
integral with 10 equally spaced subintervals and the Right Hand Rule, set n = 10
and compute
∫ 4 ( )
32 1
(4x − x2 ) dx ≈ 1 − 2 = 10.56.
0 3 10
Notes:
228
5.3 Riemann Sums
Recall how earlier we approximated the definite integral with 4 subintervals;

with n = 4, the formula gives 10, our answer as before.
It is now easy to approximate the integral with 1,000,000 subintervals! Hand-
held calculators will round off the answer a bit prematurely giving an answer of
10.66666667. (The actual answer is 10.666666666656.)
We now take an important leap. Up to this point, our mathema cs has been
limited to geometry and algebra (finding areas and manipula ng expressions).
Now we apply calculus. For any finite n, we know that
∫ 4 ( )
32 1
(4x − x ) dx ≈
2
1− 2 .
0 3 n
Both common sense and high–level mathema cs tell us that as n gets large, the
∫ 4 In fact, if we take the limit as n → ∞, we get the
approxima on gets be er.
exact area described by 0 (4x − x2 ) dx. That is,
∫ 4 ( )
32 1
(4x − x ) dx = lim
2
1− 2
0 n→∞ 3 n
32
= (1 − 0)
3
32
= = 10.6
3
This is a fantas c result. By considering n equally–spaced subintervals, we ob-
tained a formula for an approxima on of the definite integral that involved our
variable n. As n grows large – without bound – the error shrinks to zero and we
obtain the exact area.
This sec on started with a fundamental calculus technique: make an ap-

proxima on, refine the approxima on to make it be er, then use limits in the
refining process to get an exact answer. That is precisely what we just did.
Let’s prac ce this again.
Example 5.3.7 Approxima ng ∫definite integrals with a formula, using sums

5
Find a formula that approximates −1 x3 dx using the Right Hand Rule and n
equally spaced subintervals, then take the limit as n → ∞ to find the exact
area.
S Following Key Idea 5.3.1, we have ∆x = 5−(−1) n = 6/n.

We have xi = (−1) + (i − 1)∆x; as the Right Hand Rule uses xi+1 , we have
xi+1 = (−1) + i∆x.
The Riemann sum corresponding to the Right Hand Rule is (followed by sim-
Notes:
229
plifica ons):
∫ 5 ∑
n
x3 dx ≈ f(xi+1 )∆x
−1 i=1
∑
n
= f(−1 + i∆x)∆x
i=1
∑n
= (−1 + i∆x)3 ∆x
i=1
∑
n
( )
= (i∆x)3 − 3(i∆x)2 + 3i∆x − 1 ∆x (now distribute ∆x)
i=1
∑n
(3 4 )
= i ∆x − 3i2 ∆x3 + 3i∆x2 − ∆x (now split up summa on)
i=1
∑
n ∑
n ∑
n ∑
n
= ∆x4 i3 − 3∆x3 i2 + 3∆x2 i− ∆x
i=1 i=1 i=1 i=1
( )2
n(n + 1) n(n + 1)(2n + 1) n(n + 1)
= ∆x4 − 3∆x3 + 3∆x2 − n∆x
y 2 6 2
(use ∆x = 6/n)
100
1296 n2 (n + 1)2 216 n(n + 1)(2n + 1) 36 n(n + 1)
= · −3 3 · +3 2 −6
n4 4 n 6 n 2
(now do a sizable amount of algebra to simplify)
50
378 216
= 156 + + 2
n n
. x Once again, we have found a compact formula for approxima ng the definite
−1 1 2 3 4 5
integral with n equally spaced subintervals and the Right Hand Rule. Using 10
∫5 subintervals, we have an approxima on of 195.96 (these rectangles are shown
Figure 5.3.9: Approxima ng −1 x3 dx us-
ing the Right Hand Rule and 10 evenly in Figure 5.3.9). Using n = 100 gives an approxima on of 159.802.
spaced subintervals. Now find the exact answer using a limit:
∫ 5 ( )
378 216
x3 dx = lim 156 + + 2 = 156.
−1 n→∞ n n
Limits of Riemann Sums
We have used limits to evaluate given definite integrals. Will this always
work? We will show, given not–very–restric ve condi ons, that yes, it will al-
ways work.
Notes:
230
5.3 Riemann Sums
The previous two examples demonstrated how an expression such as
∑
n
f(xi+1 )∆x
i=1
can be rewri en as an expression explicitly involving n, such as 32/3(1 − 1/n2 ).

Viewed in this manner, we can think of the summa on as a func on of n.
An n value is given (where n is a posi ve integer), and the sum of areas of n
equally spaced rectangles is returned, using the Le Hand, Right Hand, or Mid-
point Rules.
∫b
Given a definite integral a f(x) dx, let:
∑
n
• SL (n) = f(xi )∆x, the sum of equally spaced rectangles formed using
i=1
the Le Hand Rule,
∑
n
• SR (n) = f(xi+1 )∆x, the sum of equally spaced rectangles formed us-
i=1
ing the Right Hand Rule, and
∑n ( )
xi + xi+1
• SM (n) = f ∆x, the sum of equally spaced rectangles
i=1
2
formed using the Midpoint Rule.
Recall the defini on of a limit as n → ∞: lim SL (n) = K if, given any ε > 0,
n→∞
there exists N > 0 such that
|SL (n) − K| < ε when n ≥ N.
The following theorem states that we can use any of our three rules to find
∫b
the exact value of a definite integral a f(x) dx. It also goes two steps further.
The theorem states that the height of each rectangle doesn’t have to be deter-
mined following a specific rule, but could be f(ci ), where ci is any point in the i th
subinterval, as discussed before Riemann Sums were defined in Defini on 5.3.2.
The theorem goes on to state that the rectangles do not need to be of the
same width. Using the nota on of Defini on 5.3.1, let ∆xi denote the length
of the i th subinterval in a par on of [a, b] and let ||∆x|| represent the length
of the largest subinterval in the par on: that is, ||∆x|| is the largest of all the
∆xi ’s. If ||∆x|| is small, then [a, b] must be par oned into many subintervals,
since all subintervals must have small lengths. “Taking the limit as ||∆x|| goes
to zero” implies that the number n of subintervals in the par on is growing to
Notes:
231
infinity, as the largest subinterval length is becoming arbitrarily small. We then

interpret the expression
∑n
lim f(ci )∆xi
||∆x||→0
i=1
as “the limit of the sum of the areas of rectangles, where the width of each
rectangle can be different but ge ng small, and the height of each rectangle is
not necessarily determined by a par cular rule.” The theorem states that this
Riemann Sum also gives the value of the definite integral of f over [a, b].
Theorem 5.3.2 Definite Integrals and the Limit of Riemann Sums

Let f be con nuous on the closed interval [a, b] and let SL (n), SR (n),
SM (n), ∆x, ∆xi and ci be defined as before. Then:
∑
n
1. lim SL (n) = lim SR (n) = lim SM (n) = lim f(ci )∆x,
n→∞ n→∞ n→∞ n→∞
i=1
∑
n ∫ b
2. lim f(ci )∆x = f(x) dx, and
n→∞ a
i=1
∑
n ∫ b
3. lim f(ci )∆xi = f(x) dx.
∥∆x∥→0 a
i=1
We summarize what we have learned over the past few sec ons here.
• Knowing the “area under the curve” can be useful. One common example:
the area under a velocity curve is displacement.
∫b
• We have defined the definite integral, a f(x) dx, to be the signed area
under f on the interval [a, b].
• While we can approximate a definite integral many ways, we have focused

on using rectangles whose heights can be determined using the Le Hand
Rule, the Right Hand Rule and the Midpoint Rule.
• Sums of rectangles of this type are called Riemann sums.
• The exact value of the definite integral can be computed using the limit of
a Riemann sum. We generally use one of the above methods as it makes
the algebra simpler.
Notes:
232
5.3 Riemann Sums
We first learned of deriva ves through limits then learned rules that made
the process simpler. We know of a way to evaluate a definite integral using limits;
in the next sec on we will see how the Fundamental Theorem of Calculus makes
the process simpler. The key feature of this theorem is its connec on between
the indefinite integral and the definite integral.
Notes:
233
Exercises 5.3
Terms and Concepts 16. 1 − e + e2 − e3 + e4
1. A fundamental calculus technique is to use to re-

In Exercises 17 – 24, evaluate the summa on using Theorem
fine approxima ons to get an exact answer.
5.3.1.
2. What is the upper bound in the summa on
∑
14 ∑
10
(48i − 201)? 17. 5

i=7 i=1
3. This sec on approximates definite integrals using what ge- ∑

25
ometric shape? 18. i

i=1
4. T/F: A sum using the Right Hand Rule is an example of a

Riemann Sum. ∑
10
19. (3i2 − 2i)
i=1
Problems
∑
15
In Exercises 5 – 12, write out each term of the summa on and 20. (2i3 − 10)
i=1
compute the sum.
∑
4 ∑
10
5. i2 21. (−4i3 + 10i2 − 7i + 11)

i=2 i=1
∑
3
∑
10
6. (4i − 2) 22. (i3 − 3i2 + 2i + 7)
i=−1 i=1
∑
2
23. 1 + 2 + 3 + . . . + 99 + 100
7. sin(πi/2)
i=−2
24. 1 + 4 + 9 + . . . + 361 + 400
∑
10
8. 5
Theorem 5.3.1 states
i=1
∑n ∑k ∑
n
ai = ai + ai , so
∑5
1 i=1 i=1 i=k+1
9.
i=1
i ∑
n ∑
n ∑
k
ai = ai − ai .
∑
6 i=k+1 i=1 i=1
10. (−1)i i Use this fact, along with other parts of Theorem 5.3.1, to eval-
i=1 uate the summa ons given in Exercises 25 – 28.
4 (
∑ )
1 1 ∑
20
11. −
i=1
i i+1 25. i
i=11
∑
5
12. (−1)i cos(πi) ∑
25
i=0 26. i3
i=16
In Exercises 13 – 16, write each sum in summa on nota on.
∑
12
13. 3 + 6 + 9 + 12 + 15 27. 4
i=7
14. −1 + 0 + 3 + 8 + 15 + 24 + 35 + 48 + 63
∑
10
1 2 3 4 28. 4i3
15. + + +
2 3 4 5 i=5
234
∫ 1
In Exercises
∫ 29 – 34, a definite integral 35. x3 dx, using the Right Hand Rule.
b
0
f(x) dx is given.
a ∫ 1
(a) Graph f(x) on [a, b]. 36. 3x2 dx, using the Le Hand Rule.
(b) Add to the sketch rectangles using the provided rule. −1
∫ b
∫ 3
(c) Approximate f(x) dx by summing the areas of the
a 37. (3x − 1) dx, using the Midpoint Rule.
rectangles. −1
∫ 3 ∫ 4
29. x2 dx, with 6 rectangles using the Le Hand Rule. 38. (2x2 − 3) dx, using the Le Hand Rule.
−3 1
∫ 2 ∫ 10
30. (5 − x2 ) dx, with 4 rectangles using the Midpoint Rule. 39. (5 − x) dx, using the Right Hand Rule.
0 −10
∫ π ∫ 1
31. sin x dx, with 6 rectangles using the Right Hand Rule.
0
40. (x3 − x2 ) dx, using the Right Hand Rule.
0
∫ 3
32. 2x dx, with 5 rectangles using the Le Hand Rule.
0
Review
∫ 2
33. ln x dx, with 3 rectangles using the Midpoint Rule. In Exercises 41 – 46, find an an deriva ve of the given func-
1
on.
∫ 9
1
34. dx, with 4 rectangles using the Right Hand Rule.
1 x 41. f(x) = 5 sec2 x
In Exercises
∫ 35 – 40, a definite integral 7
b 42. f(x) =
f(x) dx is given. As demonstrated in Examples 5.3.6 x
a
and 5.3.7, do the following. 43. g(t) = 4t5 − 5t3 + 8
∫ b
(a) Find a formula to approximate f(x) dx using n
a 44. g(t) = 5 · 8t
subintervals and the provided rule.
(b) Evaluate the formula using n = 10, 100 and 1, 000. 45. g(t) = cos t + sin t
∫ the formula, as n → ∞, to find the
(c) Find the limit of
b
1
exact value of f(x) dx. 46. f(x) = √
a x
235
5.4 The Fundamental Theorem of Calculus

∫b
Let f(t) be a con nuous func on defined on [a, b]. The definite integral a f(x) dx
y is the “area under f ” on [a, b]. We can turn this concept into a func on by le ng
the upper (or lower) bound vary.
∫x
Let F(x) = a f(t) dt. It computes the area under f on [a, x] as illustrated
in Figure 5.4.1. We can study this func∫ on using our knowledge of the definite
a
integral. For instance, F(a) = 0 since a f(t) dt = 0.
We can also apply calculus ideas to F(x); in par cular, we can compute its
deriva ve. While this may seem like an innocuous thing to do, it has far–reaching
t
a x b implica ons, as demonstrated by the fact that the result is given as an important
. theorem.
Figure 5.4.1: The
∫ x area of the shaded re-
gion is F(x) = a f(t) dt.
Theorem 5.4.1 The Fundamental Theorem of Calculus, Part 1
∫x
Let f be con nuous on [a, b] and let F(x) = a f(t) dt. Then F is a differ-
en able func on on (a, b), and
F ′ (x) = f(x).
Ini ally this seems simple, as demonstrated in the following example.
Example 5.4.1
∫ Using the Fundamental Theorem of Calculus, Part 1
x
Let F(x) = (t + sin t) dt. What is F ′ (x)?
2
−5
S Using the Fundamental Theorem of Calculus, we have F ′ (x) =

2
x + sin x.
This simple example reveals something incredible: F(x) is an an deriva ve

of x2 + sin x! Therefore, F(x) = 31 x3 − cos x + C for some value of C. (We can
find C, but generally we do not care. We know that F(−5) = 0, which allows us
to compute C. In this case, C = cos(−5) + 125 3 .)
We have done more than found a complicated way of compu ng an an-

deriva ve. Consider a func on f defined on an open interval containing a, b
∫b ∫x
and c. Suppose we want to compute a f(t) dt. First, let F(x) = c f(t) dt. Using
Notes:
236
the proper es of the definite integral found in Theorem 5.2.1, we know

∫ b ∫ c ∫ b
f(t) dt = f(t) dt + f(t) dt
a a c
∫ a ∫ b
=− f(t) dt + f(t) dt
c c
= −F(a) + F(b)
= F(b) − F(a).
We now see how indefinite integrals and definite integrals are related: we can
evaluate a definite integral using an deriva ves! This is the second part of the
Fundamental Theorem of Calculus.
Theorem 5.4.2 The Fundamental Theorem of Calculus, Part 2

Let f be con nuous on [a, b] and let F be any an deriva ve of f. Then
∫ b
f(x) dx = F(b) − F(a).
a
Example 5.4.2 Using the Fundamental Theorem of Calculus,∫ 4 Part 2

We spent a great deal of me in the previous sec on studying 0 (4x − x2 ) dx.
Using the Fundamental Theorem of Calculus, evaluate this definite integral.
S We need an an deriva ve of f(x) = 4x − x2 . All an deriva-

ves of f have the form F(x) = 2x2 − 13 x3 + C; for simplicity, choose C = 0.
The Fundamental Theorem of Calculus states
∫ 4 ( 1 ) ( ) 64
(4x − x2 ) dx = F(4) − F(0) = 2(4)2 − 43 − 0 − 0 = 32 − = 32/3.
0 3 3
This is the same answer we obtained using limits in the previous sec on, just
with much less work.
Nota on: A special nota on is o en used in the process of evalua ng definite

integrals using the Fundamental Theorem of Calculus. Instead of explicitly writ-
b

ing F(b) − F(a), the nota on F(x) is used. Thus the solu on to Example 5.4.2
a
would be wri en as:
∫ 4 ( )4
1 ( 1 ) ( )
(4x − x2 ) dx = 2x2 − x3 = 2(4)2 − 43 − 0 − 0 = 32/3.
0 3 0 3
Notes:
237
The Constant C: Any an deriva ve F(x) can be chosen when using the Funda-
mental Theorem of Calculus to evaluate a definite integral, meaning any value
of C can be picked. The constant always cancels out of the expression when
evalua ng F(b) − F(a), so it does not ma er what value is picked. This being
the case, we might as well let C = 0.
Example 5.4.3 Using the Fundamental Theorem of Calculus, Part 2

Evaluate the following definite integrals.
∫ 2 ∫ π ∫ 5 ∫ 9 ∫
3 t
√ 5
1. x dx 2. sin x dx 3. e dt 4. u du 5. 2 dx
−2 0 0 4 1
S
∫ 2 ( ) ( )
2
1 4 1 4 1
1. 3
x dx = x = 2 − 4
(−2) = 0.
−2 4 −2 4 4
∫ π π
( )
2. sin x dx = − cos x = − cos π − − cos 0 = 1 + 1 = 2.
0 0
(This is interes ng; it says that the area under one “hump” of a sine curve
is 2.)
∫ 5 5

3. et dt = et = e5 − e0 = e5 − 1 ≈ 147.41.
0 0
∫ ∫ 9
9 √ 9
1 2 3 2( 3 3
) 2( ) 38
4. u du = u du = u 2 =
2 9 2 − 4 2 = 27 − 8 = .
4 4 3 4 3 3 3
∫ 5 5

5. 2 dx = 2x = 2(5) − 2 = 2(5 − 1) = 8.
1 1
This integral is interes ng; the integrand is a constant func on, hence we
are finding the area of a rectangle with width (5 − 1) = 4 and height 2.
No ce how the evalua on of the definite integral led to 2(4) = 8.
∫b
In general, if c is a constant, then a c dx = c(b − a).
Understanding Mo on with the Fundamental Theorem of

Calculus
We established, star ng with Key Idea 2.2.1, that the deriva ve of a posi on
func on is a velocity func on, and the deriva ve of a velocity func on is an ac-
celera on func on. Now consider definite integrals of velocity and accelera on
∫ b
func ons. Specifically, if v(t) is a velocity func on, what does v(t) dt mean?
a
Notes:
238
The Fundamental Theorem of Calculus states that

∫ b
v(t) dt = V(b) − V(a),
a
where V(t) is any an deriva ve of v(t). Since v(t) is a velocity func on, V(t)
must be a posi on func on, and V(b) − V(a) measures a change in posi on, or
displacement.
Example 5.4.4 Finding displacement

A ball is thrown straight up with velocity given by v(t) = −32t + 20 /s, where
∫ 1
t is measured in seconds. Find, and interpret, v(t) dt.
0
S Using the Fundamental Theorem of Calculus, we have

∫ 1 ∫ 1
v(t) dt = (−32t + 20) dt
0 0
1

= −16t2 + 20t
0
= 4.
Thus if a ball is thrown straight up into the air with velocity v(t) = −32t + 20,
the height of the ball, 1 second later, will be 4 feet above the ini al height. (Note
that the ball has traveled much farther. It has gone up to its peak and is falling
down, but the difference between its height at t = 0 and t = 1 is 4 .)
Integra ng a rate of change func on gives total change. Velocity is the rate
of posi on change; integra ng velocity gives the total change of posi on, i.e.,
displacement.
Integra ng a speed func on gives a similar, though different, result. Speed
is also the rate of posi on change, but does not account for direc on. So inte-
gra ng a speed func on gives total change of posi on, without the possibility
of “nega ve posi on change.” Hence the integral of a speed func on gives dis-
tance traveled.
As accelera on is the rate of velocity change, integra ng an accelera on
func on gives total change in velocity. We do not have a simple term for this
analogous to displacement. If a(t) = 5miles/h2 and t is measured in hours,
then ∫ 3
a(t) dt = 15
0
means the velocity has increased by 15m/h from t = 0 to t = 3.
Notes:
239
The Fundamental Theorem of Calculus and the Chain Rule
∫ x Part 1 of the Fundamental Theorem of Calculus (FTC) states that given F(x) =
d( )
f(t) dt, F ′ (x) = f(x). Using other nota on, F(x) = f(x). While we have
a dx
just prac ced evalua ng definite integrals, some mes finding an deriva ves is
impossible and we need to rely on other techniques ∫ xto approximate the value
of a definite integral. Func ons wri en as F(x) = a f(t) dt are useful in such
situa ons.
It may be of further use to compose such a func on with another. As an
example, we may compose F(x) with g(x) to get
∫
( ) g(x)
F g(x) = f(t) dt.
a
What is the deriva ve of such a func on? The Chain Rule can be employed to
state
d( ( )) ( ) ( )
F g(x) = F ′ g(x) g ′ (x) = f g(x) g ′ (x).
dx
An example will help us understand this.
Example 5.4.5 The FTC, Part 1, and the Chain Rule

∫ x2
Find the deriva ve of F(x) = ln t dt.
2
∫ x
S We can view F(x) as being the func on G(x) = ln t dt
2
( ) 2
composed with g(x) = x ; that is, F(x) = G g(x) . The Fundamental Theorem
of Calculus states that G ′ (x) = ln x. The Chain Rule gives us
( )
F ′ (x) = G ′ g(x) g ′ (x)
= ln(g(x))g ′ (x)
= ln(x2 )2x
= 2x ln x2
Normally, the steps defining G(x) and g(x) are skipped.
Prac ce this once more.
Example 5.4.6 The FTC, Part

∫ 1, and the Chain Rule
5
Find the deriva ve of F(x) = t3 dt.
cos x
Notes:
240
∫ cos x
S Note that F(x) = − t3 dt. Viewed this way, the deriva-
5
ve of F is straigh orward:
y
′ 3
F (x) = sin x cos x.
f(x)
Area Between Curves

g(x)
Consider con nuous func ons f(x) and g(x) defined on [a, b], where f(x) ≥
g(x) for all x in [a, b], as demonstrated in Figure 5.4.2. What is the area of the x
shaded region bounded by the two curves over [a, b]? a b
The area can be found by recognizing that this area is “the area under f − .
y
the area under g.” Using mathema cal nota on, the area is
∫ b ∫ b f(x)
f(x) dx − g(x) dx.
a a
g(x)
Proper es of the definite integral allow us to simplify this expression to
∫ x
b ( ) a b
f(x) − g(x) dx. .
a
Figure 5.4.2: Finding the area bounded by
two func ons on an interval; it is found
by subtrac ng the area under g from the
Theorem 5.4.3 Area Between Curves area under f.
Let f(x) and g(x) be con nuous func ons defined on [a, b] where f(x) ≥
g(x) for all x in [a, b]. The area of the region bounded by the curves
y = f(x), y = g(x) and the lines x = a and x = b is
∫ b ( )
f(x) − g(x) dx.
a y
15 y = x2 + x − 5
10
Example 5.4.7 Finding area between curves
Find the area of the region enclosed by y = x2 + x − 5 and y = 3x − 2. 5
S It will help to sketch these two func ons, as done in Figure x

−2 −1 1 2 3 4
5.4.3. The region whose area we seek is completely bounded by these two
func ons; they seem to intersect at x = −1 and x = 3. To check, set x2 +x−5 =
.y = 3x − 2
Figure 5.4.3: Sketching the region en-
closed by y = x2 + x − 5 and y = 3x − 2
Notes: in Example 5.4.7.
241
3x − 2 and solve for x:

y
x2 + x − 5 = 3x − 2
(x2 + x − 5) − (3x − 2) = 0
x2 − 2x − 3 = 0
(x − 3)(x + 1) = 0
x x = −1, 3.
1 2 3 4
.
Following Theorem 5.4.3, the area is
Figure 5.4.4: A graph of a func on f to in-
troduce the Mean Value Theorem.
∫ ∫
3 ( ) 3
y
3x − 2 − (x2 + x − 5) dx = (−x2 + 2x + 3) dx
−1 −1
( )3
1 3
= − x + x + 3x
2
3 −1
( )
1 1
= − (27) + 9 + 9 − +1−3
3 3
2
= 10 = 10.6
x 3
1 2 3 4
.
(a)
y
The Mean Value Theorem and Average Value
∫ 4 Consider the graph of a func on f in Figure 5.4.4 and the area defined by
1
f(x) dx. Three rectangles are drawn in Figure 5.4.5; in (a), the height of the
rectangle is greater than f on [1, 4], hence the area of this rectangle is is greater
∫4
than 0 f(x) dx.
x In (b), the height of the rectangle is smaller than f on [1, 4], hence the area
1 2 3 4 ∫4
of this rectangle is less than 1 f(x) dx.
.
(b) Finally, in (c) the
∫ 4height of the rectangle is such that the area of the rectangle
y is exactly that of 0 f(x) dx. Since rectangles that are “too big”, as in (a), and
∫4
rectangles that are “too li le,” as in (b), give areas greater/lesser than 1 f(x) dx,
it makes sense that there is a rectangle, whose top intersects f(x) somewhere
on [1, 4], whose area is exactly that of the definite integral.
We state this idea formally in a theorem.
x
1 2 3 4
. Notes:
(c)
Figure 5.4.5: Differently sized rectan-

gles
∫ 4 give upper and lower bounds on
1
f(x) dx; the last rectangle matches the
area exactly.
242
y
Theorem 5.4.4 The Mean Value Theorem of Integra on
Let f be con nuous on [a, b]. There exists a value c in [a, b] such that 1
∫ b
f(x) dx = f(c)(b − a). sin 0.69
a
This is an existen al statement; c exists, but we do not provide a method of x

c 1 2 π
finding it. Theorem 5.4.4 is directly connected to the Mean Value Theorem of .
Differen a on, given as Theorem 3.2.1; we leave it to the reader to see how. . 5.4.6: A graph of y = sin x on
Figure
We demonstrate the principles involved in this version of the Mean Value
[0, π] and the rectangle guaranteed by
Theorem in the following example. the Mean Value Theorem.
Example 5.4.8
∫π Using the Mean Value Theorem
Consider 0 sin x dx. Find a value c guaranteed by the Mean Value Theorem.
∫π
S We first need to evaluate 0 sin x dx. (This was previously
done in Example 5.4.3.)
∫ π π y

sin x dx = − cos x = 2.
0 0 y = f(x)
Thus we seek a value c in [0, π] such that π sin c = 2.

f(c)
π sin c = 2 ⇒ sin c = 2/π ⇒ c = arcsin(2/π) ≈ 0.69.
In Figure 5.4.6 sin x is sketched along with a rectangle with height sin(0.69). x
a c b
The area of the rectangle is the same as the area under sin x on [0, π].
.
∫b (a)
Let f be a func on on [a, b] with c such that f(c)(b−a) = f(x) dx. Consider
∫b( ) a y
a
f(x) − f(c) dx:
∫ ∫ ∫
b ( ) b b
y = f(x) − f(c)
f(x) − f(c) dx = f(x) − f(c) dx
a a a f(c)
= f(c)(b − a) − f(c)(b − a)
= 0.
x
When f(x) is shi ed by −f(c), the amount of area under f above the x–axis on a c b
[a, b] is the same as the amount of area below the x–axis above f; see Figure .
5.4.7 for an illustra on of this. In this sense, we can say that f(c) is the average (b)
value of f on [a, b].
Figure 5.4.7: In (a), a graph of y =
f(x) and the rectangle guaranteed by the
Mean Value Theorem. In (b), y = f(x) is
shi ed down by f(c); the resul ng “area
Notes: under the curve” is 0.
243
The value f(c) is the average value in another sense. First, recognize that the
Mean Value Theorem can be rewri en as
∫ b
1
f(c) = f(x) dx,
b−a a
for some value of c in [a, b]. Next, par on the interval [a, b] into n equally
spaced subintervals, a = x1 < x2 < . . . < xn+1 = b and choose any ci in
[xi , xi+1 ]. The average of the numbers f(c1 ), f(c2 ), …, f(cn ) is:
1( ) 1∑ n
f(c1 ) + f(c2 ) + . . . + f(cn ) = f(ci ).
n n i=1
(b−a)
Mul ply this last expression by 1 in the form of (b−a) :
1∑ ∑
n n
1
f(ci ) = f(ci )
n i=1 i=1
n
∑
n
1 (b − a)
= f(ci )
i=1
n (b − a)
1 ∑
n
b−a
= f(ci )
b − a i=1 n
1 ∑
n
= f(ci )∆x (where ∆x = (b − a)/n)
b − a i=1
Now take the limit as n → ∞:

∫
1 ∑
n b
1
lim f(ci )∆x = f(x) dx = f(c).
n→∞ b − a b−a a
i=1
This tells us this: when we evaluate f at n (somewhat) equally spaced points in

[a, b], the average value of these samples is f(c) as n → ∞.
This leads us to a defini on.
Defini on 5.4.1 The Average Value of f on [a, b]

Let f be con nuous on [a, b]. The average value of f on [a, b] is f(c),
where c is a value in [a, b] guaranteed by the Mean Value Theorem. I.e.,
∫ b
1
Average Value of f on [a, b] = f(x) dx.
b−a a
Notes:
244
An applica on of this defini on is given in the following example.
Example 5.4.9 Finding the average value of a func on

An object moves back and forth along a straight line with a velocity given by
v(t) = (t − 1)2 on [0, 3], where t is measured in seconds and v(t) is measured
in /s.
What is the average velocity of the object?
S By our defini on, the average velocity is:

∫ ∫ ( )3
1 3
1 3 ( ) 1 1 3
(t − 1)2 dt = t2 − 2t + 1 dt = t − t2 + t = 1 /s.
3−0 0 3 0 3 3 0
We can understand the above example through a simpler situa on. Suppose
you drove 100 miles in 2 hours. What was your average speed? The answer is
simple: displacement/ me = 100 miles/2 hours = 50 mph.
What was the displacement of the object
∫3 in Example 5.4.9? We calculate
this by integra ng its velocity func on: 0 (t − 1)2 dt = 3 . Its final posi on
was 3 feet from its ini al posi on a er 3 seconds: its average velocity was 1 /s.
This sec on has laid the groundwork for a lot of great mathema cs to fol-
low. The most important lesson is this: definite integrals can be evaluated using
an deriva ves. Since the previous sec on established that definite integrals are
the limit of Riemann sums, we can later create Riemann sums to approximate
values other than “area under the curve,” convert the sums to definite integrals,
then evaluate these using the Fundamental Theorem of Calculus. This will allow
us to compute the work done by a variable force, the volume of certain solids,
the arc length of curves, and more.
The downside is this: generally speaking, compu ng an deriva ves is much
more difficult than compu ng deriva ves. The next chapter is devoted to tech-
niques of finding an deriva ves so that a wide variety of definite integrals can
be evaluated. Before that, the next sec on explores techniques of approximat-
ing the value of definite integrals beyond using the Le Hand, Right Hand and
Midpoint Rules. These techniques are invaluable when an deriva ves cannot
be computed, or when the actual func on f is unknown and all we know is the
value of f at certain x-values.
Notes:
245
Exercises 5.4 ∫
Terms and Concepts 18.
2
1
dx
1 x
1. How are definite and indefinite integrals related?
∫ 2
2. What constant of integra on is most commonly used when 1
19. dx
evalua ng definite integrals? 1 x2
∫ x
3. T/F: If f is a con nuous func on, then F(x) = f(t) dt is ∫ 2
1
a 20. dx
also a con nuous func on. 1 x3
4. The definite integral can be used to find “the area under a ∫ 1

curve.” Give two other uses for definite integrals. 21. x dx
0
Problems ∫ 1
22. x2 dx
In Exercises 5 – 28, evaluate the definite integral. 0
∫ 3
∫
5. (3x2 − 2x + 1) dx 1
1 23. x3 dx
0
∫ 4
6. (x − 1)2 dx ∫
0 1
24. x100 dx
∫ 1 0
7. (x3 − x5 ) dx
−1
∫ 4
∫ π 25. dx
8. cos x dx −4
π/2
∫ π/4
∫ −5
9. sec2 x dx 26. 3 dx
0 −10
∫ e
1 ∫
10. dx 2
1 x 27. 0 dx
−2
∫ 1
11. 5x dx
−1 ∫ π/3
∫ −1
28. csc x cot x dx
3 π/6
12. (4 − 2x ) dx
−2
∫ π 29. Explain why:

13. (2 cos x − 2 sin x) dx
0
∫ 3 ∫ 1
14. ex dx
1
(a) xn dx = 0, when n is a posi ve, odd integer, and
−1
∫ 4 √
15. t dt ∫ ∫
0 1 1
(b) xn dx = 2 xn dx when n is a posi ve, even
∫ 25 −1 0
1 integer.
16. √ dt
9 t
∫ ∫
8 √
3
a+2π
17. x dx 30. Explain why sin t dt = 0 for all values of a.
1 a
246
In Exercises 31 – 34, find a value c guaranteed by the Mean In Exercises 47 – 50, an accelera on func on of an object
Value Theorem. moving along a straight line is given. Find the change of the
∫ 2 object’s velocity over the given me interval.
31. x2 dx
0
47. a(t) = −32 /s2 on [0, 2]
∫ 2
32. x2 dx
−2 48. a(t) = 10 /s2 on [0, 5]
∫ 1
33. ex dx 49. a(t) = t /s2 on [0, 2]
0
∫ 16 √ 50. a(t) = cos t /s2 on [0, π]

34. x dx
0
In Exercises 35 – 40, find the average value of the func on on In Exercises 51 – 54, sketch the given func ons and find the
the given interval. area of the enclosed region.
35. f(x) = sin x on [0, π/2]

51. y = 2x, y = 5x, and x = 3.
36. y = sin x on [0, π]
52. y = −x + 1, y = 3x + 6, x = 2 and x = −1.
37. y = x on [0, 4]
38. y = x2 on [0, 4] 53. y = x2 − 2x + 5, y = 5x − 5.
39. y = x3 on [0, 4] 54. y = 2x2 + 2x − 5, y = x2 + 3x + 7.
40. g(t) = 1/t on [1, e]

In Exercises 55 – 58, find F ′ (x).
In Exercises 41 – 46, a velocity func on of an object moving
along a straight line is given. Find the displacement of the
object over the given me interval. ∫ x3 +x
1
55. F(x) = dt
2 t
41. v(t) = −32t + 20 /s on [0, 5]
∫ 0
42. v(t) = −32t + 200 /s on [0, 10] 56. F(x) = t3 dt
x3
43. v(t) = 10 /s on [0, 3].
∫ x2
44. v(t) = 2t mph on [−1, 1] 57. F(x) = (t + 2) dt
x
45. v(t) = cos t /s on [0, 3π/2]

∫ ex
√
4 58. F(x) = sin t dt
46. v(t) = t /s on [0, 16]
ln x
247
5.5 Numerical Integra on

The Fundamental Theorem of Calculus gives a concrete technique for finding
y the exact value of a definite integral. That technique is based on compu ng an-
deriva ves. Despite the power of this theorem, there are s ll situa ons where
2
1 y = e−x we must approximate the value of the definite integral instead of finding its ex-
act value. The first situa on we explore is where we cannot compute the an-
deriva ve of the integrand. The second case is when we actually do not know
0.5 the func on in the integrand, but only its value when evaluated at certain points.
An elementary func on is any func on that is a combina on of polynomial,

x nth root, ra onal, exponen al, logarithmic and trigonometric func ons. We can
0.5 1 compute the deriva ve of any elementary func on, but there are many elemen-
.
tary func ons of which we cannot compute an an deriva ve. For example, the
(a) following func ons do not have an deriva ves that we can express with ele-
mentary func ons:
y
2 sin x
1 e−x , sin(x3 ) and .
y = sin(x )3
x
2
∫ The simplest way to refer to the an deriva ves of e−x is to simply write
0.5 −x2
e dx.
This sec on outlines three common methods of approxima ng the value of
x definite integrals. We describe each as a systema c method of approxima ng
−1 1
area under a curve. By approxima ng this area accurately, we find an accurate
−0.5
approxima on of the corresponding definite integral.
.. We will apply the methods we learn in this sec on to the following definite
integrals:
(b)
∫ 1 ∫ π ∫ 4π
2 2 sin(x)
y
e−x dx, sin(x3 ) dx, and dx,
1 0 − π4 0.5 x
sin x as pictured in Figure 5.5.1.

y=
x
0.5 The Le and Right Hand Rule Methods
In Sec on 5.3 we addressed the problem of evalua ng definite integrals by

x approxima ng the area under the curve using rectangles. We revisit those ideas
5 10 15 here before introducing other methods of approxima ng definite integrals.
. We start with a review of nota on. Let f be a con nuous func on on the
∫ b
(c) interval [a, b]. We wish to approximate f(x) dx. We par on [a, b] into n
a
b−a
Figure 5.5.1: Graphically represen ng equally spaced subintervals, each of length ∆x = . The endpoints of these
three definite integrals that cannot be n
evaluated using an deriva ves.
Notes:
248
subintervals are labeled as
x1 = a, x2 = a + ∆x, x3 = a + 2∆x, . . . , xi = a + (i − 1)∆x, . . . , xn+1 = b.
Key Idea 5.3.1 states that to use the Le Hand Rule we use the summa on
∑
n ∑ n
f(xi )∆x and to use the Right Hand Rule we use f(xi+1 )∆x. We review
i=1 i=1
the use of these rules in the context of examples.
Example 5.5.1∫ Approxima ng definite integrals with rectangles

1
−x2
Approximate e dx using the Le and Right Hand Rules with 5 equally
0
spaced subintervals.
y
S We begin by par oning the interval [0, 1] into 5 equally
2
spaced intervals. We have ∆x = 1−0
5 = 1/5 = 0.2, so
1 y = e−x
x1 = 0, x2 = 0.2, x3 = 0.4, x4 = 0.6, x5 = 0.8, and x6 = 1.

0.5
Using the Le Hand Rule, we have:
∑
n
( ) x
f(xi )∆x = f(x1 ) + f(x2 ) + f(x3 ) + f(x4 ) + f(x5 ) ∆x 0.2 0.4 0.6 0.8 1
.
i=1
( )
= f(0) + f(0.2) + f(0.4) + f(0.6) + f(0.8) ∆x (a)
(
≈ 1 + 0.961 + 0.852 + 0.698 + 0.527)(0.2) y
≈ 0.808.
2
1 y = e−x
Using the Right Hand Rule, we have:
0.5
∑
n
( )
f(xi+1 )∆x = f(x2 ) + f(x3 ) + f(x4 ) + f(x5 ) + f(x6 ) ∆x
i=1
( )
= f(0.2) + f(0.4) + f(0.6) + f(0.8) + f(1) ∆x x
( . 0.2 0.4 0.6 0.8 1
≈ 0.961 + 0.852 + 0.698 + 0.527 + 0.368)(0.2)
≈ 0.681. (b)
Figure 5.5.2 shows the rectangles used in each method to approximate the ∫1 2
Figure 5.5.2: Approxima ng 0
e−x dx in
definite integral. These graphs show that in this par cular case, the Le Hand Example 5.5.1.
Rule is an over approxima on and the Right Hand Rule is an under approxima-
on. To get a be er approxima on, we could use more rectangles, as we did in
Notes:
249
xi Exact Approx. sin(x3i )

x1 −π/4 −0.785 −0.466 Sec on 5.3. We could also average the Le and Right Hand Rule results together,
x2 −7π/40 −0.550 −0.165 giving
x3 −π/10 −0.314 −0.031 0.808 + 0.681
= 0.7445.
x4 −π/40 −0.0785 0 2
x5 π/20 0.157 0.004 The actual answer, accurate to 4 places a er the decimal, is 0.7468, showing
x6 π/8 0.393 0.061 our average is a good approxima on.
x7 π/5 0.628 0.246
x8 11π/40 0.864 0.601
Example 5.5.2 Approxima ng definite integrals with rectangles
x9 7π/20 1.10 0.971 ∫ π
2
x10 17π/40 1.34 0.690 Approximate sin(x3 ) dx using the Le and Right Hand Rules with 10 equally
x11 π/2 1.57 −0.670 − π4
spaced subintervals.
Figure 5.5.3: Table of values used to
∫π
approximate −2 π sin(x3 ) dx in Example S We begin by finding ∆x:
4
5.5.2.
b−a π/2 − (−π/4) 3π
= = ≈ 0.236.
n 10 40
It is useful to write out the endpoints of the subintervals in a table; in Figure
5.5.3, we give the exact values of the endpoints, their decimal approxima ons,
y
and decimal approxima ons of sin(x3 ) evaluated at these points.
1 Once this table is created, it is straigh orward to approximate the definite
y = sin(x3 )
integral using the Le and Right Hand Rules. (Note: the table itself is easy to
0.5
create, especially with a standard spreadsheet program on a computer. The last
two columns are all that are needed.) The Le Hand Rule sums the first 10 values
of sin(x3i ) and mul plies the sum by ∆x; the Right Hand Rule sums the last 10
x
values of sin(x3i ) and mul plies by ∆x. Therefore we have:
−1 1 ∫ π2
−0.5 Le Hand Rule: sin(x3 ) dx ≈ (1.91)(0.236) = 0.451.
. − π4
∫ π
2
Right Hand Rule: sin(x3 ) dx ≈ (1.71)(0.236) = 0.404.
(a) − π4
Average of the Le and Right Hand Rules: 0.4275.
y
3
The actual answer, accurate to 3 places a er the decimal, is 0.460. Our ap-
y = sin(x )
1 proxima ons were once again fairly good. The rectangles used in each approx-
ima on are shown in Figure 5.5.4. It is clear from the graphs that using more
0.5
rectangles (and hence, narrower rectangles) should result in a more accurate
approxima on.
x
−1 1 The Trapezoidal Rule
∫ 1
−0.5 2
. In Example 5.5.1 we approximated the value of e−x dx with 5 rectangles
0
of equal width. Figure 5.5.2 shows the rectangles used in the Le and Right
(b)
Figure 5.5.4: Approxima ng

∫ π2 3
− π sin(x ) dx in Example 5.5.2.
4
Notes:
250
Hand Rules. These graphs clearly show that rectangles do not match the shape
of the graph all that well, and that accurate approxima ons will only come by
using lots of rectangles.
Instead of using rectangles to approximate the area, we can instead use
2
trapezoids. In Figure 5.5.5, we show the region under f(x) = e−x on [0, 1] y
approximated with 5 trapezoids of equal width; the top “corners” of each trape- 2
1 y = e−x
zoid lies on the graph of f(x). It is clear from this figure that these trapezoids
more accurately approximate
∫1 the area under f and hence should give a be er
2
approxima on of 0 e−x dx. (In fact, these trapezoids seem to give a great ap-
proxima on of the area!) 0.5
The∫ 1formula for the area of a trapezoid is given in Figure 5.5.6. We approxi-
2
mate 0 e−x dx with these trapezoids in the following example.
x
0.2 0.4 0.6 0.8 1
Example 5.5.3 Approxima ng definite integrals using trapezoids .
∫ 1
−x2 ∫1 2
Use 5 trapezoids of equal width to approximate e dx. Figure 5.5.5: Approxima ng 0 e−x dx
0 using 5 trapezoids of equal widths.
S To compute the areas of the 5 trapezoids in Figure 5.5.5, it
will again be useful to create a table of values as shown in Figure 5.5.7.
The le most trapezoid has legs of length 1 and 0.961 and a height of 0.2.
Thus, by our formula, the area of the le most trapezoid is:
1 + 0.961 a+b
b Area = 2 h
(0.2) = 0.1961. a
2
.
Moving right, the next trapezoid has legs of length 0.961 and 0.852 and a height h
of 0.2. Thus its area is:
Figure 5.5.6: The area of a trapezoid.
0.961 + 0.852
(0.2) = 0.1813.
2
The sum of the areas of all 5 trapezoids is: xi e−xi
2
1 + 0.961 0.961 + 0.852 0.852 + 0.698 0 1

(0.2) + (0.2) + (0.2)+ 0.2 0.961
2 2 2
0.4 0.852
0.698 + 0.527 0.527 + 0.368
(0.2) + (0.2) = 0.7445. 0.6 0.698
2 2 0.8 0.527
∫ 1 1 0.368
2
We approximate e−x dx ≈ 0.7445.
2
0 Figure 5.5.7: A table of values of e−x .
There are many things to observe in this example. Note how each term in
the final summa on was mul plied by both 1/2 and by ∆x = 0.2. We can factor
these coefficients out, leaving a more concise summa on as:
1 [ ]
(0.2) (1+0.961)+(0.961+0.852)+(0.852+0.698)+(0.698+0.527)+(0.527+0.368) .
2
Notes:
251
Now no ce that all numbers except for the first and the last are added twice.
Therefore we can write the summa on even more concisely as
0.2 [ ]
1 + 2(0.961 + 0.852 + 0.698 + 0.527) + 0.368 .
2
∫ b
This is the heart of the Trapezoidal Rule, wherein a definite integral f(x) dx
a
is approximated by using trapezoids of equal widths to approximate the corre-
sponding area under f. Using n equally spaced subintervals with endpoints x1 ,
b−a
x2 , . . ., xn+1 , we again have ∆x = . Thus:
n
∫ b ∑n
f(xi ) + f(xi+1 )
f(x) dx ≈ ∆x
a i=1
2
∆x ∑ (
n
)
= f(xi ) + f(xi+1 )
2 i=1
∆x [ ∑ ]
n
= f(x1 ) + 2 f(xi ) + f(xn+1 ) .
2 i=2
Example 5.5.4 Using the Trapezoidalπ Rule

∫ 2
Revisit Example 5.5.2 and approximate sin(x3 ) dx using the Trapezoidal Rule
− π4
and 10 equally spaced subintervals.
S We refer back to Figure 5.5.3 for the table of values of sin(x3 ).

Recall that ∆x = 3π/40 ≈ 0.236. Thus we have:
∫
0.236 [ ( ) ]
π
2
sin(x3 ) dx ≈ − 0.466 + 2 − 0.165 + (−0.031) + . . . + 0.69 + (−0.67)
− π4 2
= 0.4275.
No ce how “quickly” the Trapezoidal Rule can be implemented once the ta-
ble of values is created. This is true for all the methods explored in this sec on;
the real work is crea ng a table of xi and f(xi ) values. Once this is completed, ap-
proxima ng the definite integral is not difficult. Again, using technology is wise.
Spreadsheets can make quick work of these computa ons and make using lots
of subintervals easy.
Also no ce the approxima ons the Trapezoidal Rule gives. It is the average
of the approxima ons given by the Le and Right Hand Rules! This effec vely
Notes:
252
renders the Le and Right Hand Rules obsolete. They are useful when first learn-
ing about definite integrals, but if a real approxima on is needed, one is gener-
ally be er off using the Trapezoidal Rule instead of either the Le or Right Hand
Rule.
How can we improve on the Trapezoidal Rule, apart from using more and
more trapezoids? The answer is clear once we look back and consider what we
have really done so far. The Le Hand Rule is not really about using rectangles to
approximate area. Instead, it approximates a func on f with constant func ons
on small subintervals and then computes the definite integral of these constant
func ons. The Trapezoidal Rule is really approxima ng a func on f with a linear
func on on a small subinterval, then computes the definite integral of this linear
func on. In both of these cases the definite integrals are easy to compute in
geometric terms.
So we have a progression: we start by approxima ng f with a constant func-
on and then with a linear func on. What is next? A quadra c func on. By
approxima ng the curve of a func on with lots of parabolas, we generally get
an even be er approxima on of the definite integral. We call this process Simp-
son’s Rule, named a er Thomas Simpson (1710-1761), even though others had
used this rule as much as 100 years prior.
y
Simpson’s Rule
3
Given one point, we can create a constant func on that goes through that
point. Given two points, we can create a linear func on that goes through those 2
points. Given three points, we can create a quadra c func on that goes through
those three points (given that no two have the same x–value).
1
Consider three points (x1 , y1 ), (x2 , y2 ) and (x3 , y3 ) whose x–values are equally
spaced and x1 < x2 < x3 . Let f be the quadra c func on that goes through these
. x
three points. It is not hard to show that 1 2 3
∫ x3
x3 − x1 ( )
f(x) dx = y1 + 4y2 + y3 . (5.4) Figure 5.5.8: A graph of a func on f and
x1 6 a parabola that approximates it well on
Consider Figure 5.5.8. A func on f goes through the 3 points shown and the [1, 3].
parabola g that also goes through those points is graphed with a dashed line.
Using our equa on from above, we know exactly that
∫ 3
3 − 1( )
g(x) dx = 3 + 4(1) + 2 = 3.
1 6
Since g is a good approxima on for f on [1, 3], we can state that
∫ 3
f(x) dx ≈ 3.
1
Notes:
253
No ce how the interval [1, 3] was split into two subintervals as we needed 3
2
points. Because of this, whenever we use Simpson’s Rule, we need to break the
xi e−xi interval into an even number of∫ subintervals.
0 1 b
0.25 0.939 In general, to approximate f(x) dx using Simpson’s Rule, subdivide [a, b]
0.5 0.779 a
into n subintervals, where n is even and each subinterval has width ∆x = (b −
0.75 0.570
1 0.368
a)/n. We approximate f with n/2 parabolic curves, using Equa on (5.4) to com-
pute the area under these parabolas. Adding up these areas gives the formula:
(a) ∫ b
∆x [ ]
y f(x) dx ≈ f(x1 )+4f(x2 )+2f(x3 )+4f(x4 )+. . .+2f(xn−1 )+4f(xn )+f(xn+1 ) .
a 3
1 −x2
y=e Note how the coefficients of the terms in the summa on have the pa ern 1, 4,
2, 4, 2, 4, . . ., 2, 4, 1.
Let’s demonstrate Simpson’s Rule with a concrete example.
0.5
Example 5.5.5∫ Using Simpson’s Rule

1
−x2
x
Approximate e dx using Simpson’s Rule and 4 equally spaced subintervals.
0
0.25 0.5 0.75 1
.
S We begin by making a table of values as we have in the past,
(b) as shown in Figure 5.5.9(a). Simpson’s Rule states that
∫
0.25 [ ]
Figure 5.5.9: A table of values to approxi- 1
∫1 2
e−x dx ≈
2
mate 0 e−x dx, along with a graph of the 1 + 4(0.939) + 2(0.779) + 4(0.570) + 0.368 = 0.74683.
func on. 0 3
Recall in Example 5.5.1 we stated that the correct answer, accurate to 4

places a er the decimal, was 0.7468. Our approxima on with Simpson’s Rule,
with 4 subintervals, is be er than our approxima on with the Trapezoidal Rule
xi sin(x3i ) using 5!
2
−0.785 −0.466 Figure 5.5.9(b) shows f(x) = e−x along with its approxima ng parabolas,
−0.550 −0.165 demonstra ng how good our approxima on is. The approxima ng curves are
−0.314 −0.031 nearly indis nguishable from the actual func on.
−0.0785 0
0.157 0.004 Example 5.5.6 Using Simpson’s Rule
0.393 0.061 ∫ π
2
0.628 0.246 Approximate sin(x3 ) dx using Simpson’s Rule and 10 equally spaced inter-
0.864 0.601 − π4
1.10 0.971 vals.
1.34 0.690
1.57 −0.670 S Figure 5.5.10 shows the table of values that we used in the
past for this problem, shown here again for convenience. Again, ∆x = (π/2 +
Figure 5.5.10: Table of values used to π/4)/10 ≈ 0.236.
∫π
approximate −2 π sin(x3 ) dx in Example
4
5.5.6.
Notes:
254
y
Simpson’s Rule states that
∫ 1
0.236 [ y = sin(x3 )
π
2
3
sin(x ) dx ≈ (−0.466) + 4(−0.165) + 2(−0.031) + . . .
− π4 3
] 0.5
. . . + 2(0.971) + 4(0.69) + (−0.67)
= 0.4701 x
−1 1
Recall that the actual value, accurate to 3 decimal places, is 0.460. Our ap-
−0.5
proxima on is within one 1/100th of the correct value. The graph in Figure 5.5.11
.
shows how closely the parabolas match the shape of the graph.
Figure 5.5.11: Approxima ng
∫ π2
Summary and Error Analysis 3
− π4 sin(x ) dx in Example 5.5.6 with
Simpson’s Rule and 10 equally spaced
We summarize the key concepts of this sec on thus far in the following Key intervals.
Idea.
Key Idea 5.5.1 Numerical Integra on

b−a
Let f be a con nuous func on on [a, b], let n be a posi ve integer, and let ∆x = .
n
Set x1 = a, x2 = a + ∆x, . . ., xi = a + (i − 1)∆x, xn+1 = b.
∫ b
Consider f(x) dx.
a ∫ b [ ]
Le Hand Rule: f(x) dx ≈ ∆x f(x1 ) + f(x2 ) + . . . + f(xn ) .
a
∫ [
b ]
Right Hand Rule: f(x) dx ≈ ∆x f(x2 ) + f(x3 ) + . . . + f(xn+1 ) .
a
∫ b
∆x [ ]
Trapezoidal Rule: f(x) dx ≈ f(x1 ) + 2f(x2 ) + 2f(x3 ) + . . . + 2f(xn ) + f(xn+1 ) .
a 2
∫ b
∆x [ ]
Simpson’s Rule: f(x) dx ≈ f(x1 ) + 4f(x2 ) + 2f(x3 ) + . . . + 4f(xn ) + f(xn+1 ) (n even).
a 3
In our examples, we approximated the value of a definite integral using a

given method then compared it to the “right” answer. This should have raised
several ques ons in the reader’s mind, such as:
1. How was the “right” answer computed?
2. If the right answer can be found, what is the point of approxima ng?
3. If there is value to approxima ng, how are we supposed to know if the

approxima on is any good?
Notes:
255
These are good ques ons, and their answers are educa onal. In the exam-
ples, the right answer was never computed. Rather, an approxima on accurate
to a certain number of places a er the decimal was given. In Example 5.5.1, we
do not know the exact answer, but we know it starts with 0.7468. These more
accurate approxima ons were computed using numerical integra on but with
more precision (i.e., more subintervals and the help of a computer).
Since the exact answer cannot be found, approxima on s ll has its place.
How are we to tell if the approxima on is any good?
“Trial and error” provides one way. Using technology, make an approxima-
on with, say, 10, 100, and 200 subintervals. This likely will not take much me
at all, and a trend should emerge. If a trend does not emerge, try using yet more
subintervals. Keep in mind that trial and error is never foolproof; you might
stumble upon a problem in which a trend will not emerge.
A second method is to use Error Analysis. While the details are beyond the
scope of this text, there are some formulas that give bounds for how good your
approxima on will be. For instance, the formula might state that the approx-
ima on is within 0.1 of the correct answer. If the approxima on is 1.58, then
one knows that the correct answer is between 1.48 and 1.68. By using lots of
subintervals, one can get an approxima on as accurate as one likes. Theorem
5.5.1 states what these bounds are.
Theorem 5.5.1 Error Bounds in the Trapezoidal Rule and

Simpson’s Rule
∫ b
1. Let ET be the error in approxima ng f(x) dx using the Trape-
a
zoidal Rule with n subintervals.
If f has a con nuous
2nd deriva ve on [a, b] and M is any upper
bound of f (x) on [a, b], then
′′
(b − a)3
ET ≤ M.
12n2
∫ b
2. Let ES be the error in approxima ng f(x) dx using Simpson’s
a
Rule with n subintervals.
If f has a con nuous 4th deriva ve on [a, b] and M is any upper
bound of f (4) on [a, b], then
(b − a)5
ES ≤ M.
180n4
Notes:
256
There are some key things to note about this theorem.

1. The larger the interval, the larger the error. This should make sense intu-
i vely.
2. The error shrinks as more subintervals are used (i.e., as n gets larger).
3. The error in Simpson’s Rule has a term rela ng to the 4th deriva ve of f.
Consider a cubic polynomial: it’s 4th deriva ve is 0. Therefore, the error in
approxima ng the definite integral of a cubic polynomial with Simpson’s
Rule is 0 – Simpson’s Rule computes the exact answer!
We revisit Examples 5.5.3 and 5.5.5 and compute the error bounds using
Theorem 5.5.1 in the following example.
Example 5.5.7 Compu ng error bounds ∫

1
2
Find the error bounds when approxima ng e−x dx using the Trapezoidal
0
Rule and 5 subintervals, and using Simpson’s Rule with 4 subintervals.
S
Trapezoidal Rule with n = 5:
2
We start by compu ng the 2nd deriva ve of f(x) = e−x : y
2
f ′′ (x) = e−x (4x2 − 2).
x
Figure 5.5.12 shows a graph of f ′′ (x) on [0, 1]. It is clear that the largest value of 0.5 1
f ′′ , in absolute value, is 2. Thus we let M = 2 and apply the error formula from
Theorem 5.5.1. −1
(1 − 0)3 2
y = e−x (4x2 − 2)
ET = · 2 = 0.006.
12 · 52
−2
.
Our error es ma on formula states that our approxima on of 0.7445 found
in Example 5.5.3 is within 0.0067 of the correct answer, hence we know that Figure 5.5.12: Graphing f ′′ (x) in Example
∫ 1 5.5.7 to help establish error bounds.
2
0.7445 − 0.0067 = .7378 ≤ e−x dx ≤ 0.7512 = 0.7445 + 0.0067.
0
We had earlier computed the exact answer, correct to 4 decimal places, to be

0.7468, affirming the validity of Theorem 5.5.1.
Simpson’s Rule with n = 4:

2
We start by compu ng the 4th deriva ve of f(x) = e−x :
2
f (4) (x) = e−x (16x4 − 48x2 + 12).
Notes:
257
Figure 5.5.13 shows a graph of f (4) (x) on [0, 1]. It is clear that the largest value
y
of f (4) , in absolute value, is 12. Thus we let M = 12 and apply the error formula
2
y = e−x (16x4 − 48x2 + 12) from Theorem 5.5.1.
10
5 (1 − 0)5
Es = · 12 = 0.00026.
180 · 44
x
0.5 1
Our error es ma on formula states that our approxima on of 0.74683 found
−5 in Example 5.5.5 is within 0.00026 of the correct answer, hence we know that
.
∫ 1
Figure 5.5.13: Graphing f (4) (x) in Exam- 0.74683 − 0.00026 = .74657 ≤
2
e−x dx ≤ 0.74709 = 0.74683 + 0.00026.
ple 5.5.7 to help establish error bounds. 0
Once again we affirm the validity of Theorem 5.5.1.
At the beginning of this sec on we men oned two main situa ons where
numerical integra on was desirable. We have considered the case where an
an deriva ve of the integrand cannot be computed. We now inves gate the
Speed situa on where the integrand is not known. This is, in fact, the most widely
Time
(mph) used applica on of Numerical Integra on methods. “Most of the me” we ob-
0 0 serve behavior but do not know “the” func on that describes it. We instead
1 25 collect data about the behavior and make approxima ons based on this data.
2 22 We demonstrate this in an example.
3 19
4 39
5 0 Example 5.5.8 Approxima ng distance traveled
6 43 One of the authors drove his daughter home from school while she recorded
7 59 their speed every 30 seconds. The data is given in Figure 5.5.14. Approximate
8 54 the distance they traveled.
9 51
10 43 S Recall that by integra ng a speed func on we get distance
11 35 traveled. We have informa on about v(t); we will use Simpson’s Rule to approx-
12 40 ∫ b
13 43 imate v(t) dt.
14 30 a
15 0 The most difficult aspect of this problem is conver ng the given data into the
16 0 form we need it to be in. The speed is measured in miles per hour, whereas the
17 28 me is measured in 30 second increments.
18 40 We need to compute ∆x = (b − a)/n. Clearly, n = 24. What are a and b?
19 42
Since we start at me t = 0, we have that a = 0. The final recorded me came
20 40
21 39
a er 24 periods of 30 seconds, which is 12 minutes or 1/5 of an hour. Thus we
22 40 have
23 23 b−a 1/5 − 0 1 ∆x 1
24 0 ∆x = = = ; = .
n 24 120 3 360
Figure 5.5.14: Speed data collected at 30
second intervals for Example 5.5.8. Notes:
258
Thus the distance traveled is approximately:

∫ 0.2
1 [ ]
v(t) dt ≈ f(x1 ) + 4f(x2 ) + 2f(x3 ) + · · · + 4f(xn ) + f(xn+1 )
0 360
1 [ ]
= 0 + 4 · 25 + 2 · 22 + · · · + 2 · 40 + 4 · 23 + 0
360
≈ 6.2167 miles.
We approximate the author drove 6.2 miles. (Because we are sure the reader
wants to know, the author’s odometer recorded the distance as about 6.05
miles.)
We started this chapter learning about an deriva ves and indefinite inte-
grals. We then seemed to change focus by looking at areas between the graph
of a func on and the x-axis. We defined these areas as the definite integral of
the func on, using a nota on very similar to the nota on of the indefinite inte-
gral. The Fundamental Theorem of Calculus ed these two seemingly separate
concepts together: we can find areas under a curve, i.e., we can evaluate a def-
inite integral, using an deriva ves.
We ended the chapter by no ng that an deriva ves are some mes more
than difficult to find: they are impossible. Therefore we developed numerical
techniques that gave us good approxima ons of definite integrals.
We used the definite integral to compute areas, and also to compute dis-
placements and distances traveled. There is far more we can do than that. In
Chapter 7 we’ll see more applica ons of the definite integral. Before that, in
Chapter 6 we’ll learn advanced techniques of integra on, analogous to learning
rules like the Product, Quo ent and Chain Rules of differen a on.
Notes:
259
Exercises 5.5 ∫
Terms and Concepts 5 √
15. x2 + 1 dx
0
1. T/F: Simpson’s Rule is a method of approxima ng an- ∫ π
deriva ves. 16. x sin x dx
0
2. What are the two basic situa ons where approxima ng the
∫
value of a definite integral is necessary? π/2 √
17. cos x dx
0
3. Why are the Le and Right Hand Rules rarely used?
∫ 4
4. Simpson’s Rule is based on approxima ng por ons of a 18. ln x dx

1
func on with what type of func on?
∫ 1
1
19. dx
−1 sin x + 2
Problems
∫ 6
1
In Exercises 5 – 12, a definite integral is given. 20. dx
0 sin x + 2
(a) Approximate the definite integral with the Trapezoidal
Rule and n = 4.
In Exercises 21 – 24, find n such that the error in approximat-
(b) Approximate the definite integral with Simpson’s Rule ing the given definite integral is less than 0.0001 when using:
and n = 4.
(a) the Trapezoidal Rule
(c) Find the exact value of the integral.
(b) Simpson’s Rule
∫ 1
2
5. x dx ∫ π
−1
21. sin x dx
∫ 10
0
6. 5x dx ∫ 4
1
0
22. √ dx
∫ 1 x
π
7. sin x dx ∫
0
π ( )
23. cos x2 dx
∫ 0
4 √
8. x dx ∫ 5
0
24. x4 dx
∫ 3 0
9. (x3 + 2x2 − 5x + 7) dx
0 In Exercises 25 – 26, a region is given. Find the area of the
∫ 1
region using Simpson’s Rule:
10. x4 dx
0 (a) where the measurements are in cen meters, taken in
1 cm increments, and
∫ 2π
11. cos x dx (b) where the measurements are in hundreds of yards,
0 taken in 100 yd increments.
∫ 3 √
12. 9 − x2 dx
−3
In Exercises 13 – 20, approximate the definite integral with

the Trapezoidal Rule and Simpson’s Rule, with n = 6.
4.7
6.3
∫
6.9
25.
6.6
1 ( ) .
5.1
13. cos x2 dx
0
∫ 1 2
14. ex dx
−1
260
26.
.
3.6
3.6
4.5
6.6
5.6
261
6: T
A
The previous chapter introduced the an deriva ve and connected it to signed
areas under a curve through the Fundamental Theorem of Calculus. The next
chapter explores more applica ons of definite integrals than just area. As eval-
ua ng definite integrals will become important, we will want to find an deriva-
ves of a variety of func ons.
This chapter is devoted to exploring techniques of an differen a on. While
not every func on has an an deriva ve in terms of elementary func ons (a
concept introduced in the sec on on Numerical Integra on), we can s ll find
an deriva ves of a wide variety of func ons.
6.1 Subs tu on
We mo vate this sec on with an example. Let f(x) = (x2 + 3x − 5)10 . We can
compute f ′ (x) using the Chain Rule. It is:
f ′ (x) = 10(x2 + 3x − 5)9 · (2x + 3) = (20x + 30)(x2 + 3x − 5)9 .

∫
Now consider this: What is (20x + 30)(x2 + 3x − 5)9 dx? We have the answer
in front of us;
∫
(20x + 30)(x2 + 3x − 5)9 dx = (x2 + 3x − 5)10 + C.
How would we have evaluated this indefinite integral without star ng with f(x)
as we did?
This sec on explores integra on by subs tu on. It allows us to “undo the
Chain Rule.” Subs tu on allows us to evaluate the above integral without know-
ing the original func on first.
∫ The underlying principle is to rewrite a ∫“complicated” integral of the form
f(x) dx as a not–so–complicated integral h(u) du. We’ll formally establish
later
∫ how this is done. First, consider again our introductory indefinite integral,
(20x + 30)(x2 + 3x − 5)9 dx. Arguably the most “complicated” part of the
integrand is (x2 + 3x − 5)9 . We wish to make this simpler; we do so through a
subs tu on. Let u = x2 + 3x − 5. Thus
(x2 + 3x − 5)9 = u9 .
Chapter 6 Techniques of An differen a on
We have established u as a func on of x, so now consider the differen al of u:
du = (2x + 3)dx.
Keep in mind that (2x+3) and dx are mul plied; the dx is not “just si ng there.”
Return to the original integral and do some subs tu ons through algebra:
∫ ∫
(20x + 30)(x2 + 3x − 5)9 dx = 10(2x + 3)(x2 + 3x − 5)9 dx
∫
= 10(x2 + 3x − 5)9 (2x + 3) dx
| {z } | {z }
u du
∫
= 10u9 du
= u10 + C 2
(replace u with x + 3x − 5)
= (x + 3x − 5) + C
2 10
One might well look at this and think “I (sort of) followed how that worked,
but I could never come up with that on my own,” but the process is learnable.
This sec on contains numerous examples through which the reader will gain
understanding and mathema cal maturity enabling them to regard subs tu on
as a natural tool when evalua ng integrals.
We stated before that integra on by subs tu on “undoes” the Chain Rule.
Specifically, let F(x) and g(x) be differen able func ons and consider the deriva-
ve of their composi on:
d( ( ))
F g(x) = F ′ (g(x))g ′ (x).
dx
Thus ∫
F ′ (g(x))g ′ (x) dx = F(g(x)) + C.
Integra on by subs tu on works by recognizing the “inside” func on g(x) and

replacing it with a variable. By se ng u = g(x), we can rewrite the deriva ve
as
d ( ( ))
F u = F ′ (u)u ′ .
dx
Since du = g ′ (x)dx, we can rewrite the above integral as
∫ ∫
F (g(x))g (x) dx = F ′ (u)du = F(u) + C = F(g(x)) + C.
′ ′
This concept is important so we restate it in the context of a theorem.
Notes:
264
6.1 Subs tu on
Theorem 6.1.1 Integra on by Subs tu on

Let F and g be differen able func ons, where the range of g is an interval
I contained in the domain of F. Then
∫
F ′ (g(x))g ′ (x) dx = F(g(x)) + C.
If u = g(x), then du = g ′ (x)dx and

∫ ∫
F ′ (g(x))g ′ (x) dx = F ′ (u) du = F(u) + C = F(g(x)) + C.
The
∫ point of subs tu on is to make the integra on step easy. Indeed, the
step F ′ (u) du = F(u) + C looks easy, as the an deriva ve of the deriva ve of F
is just F, plus a constant. The “work” involved is making the proper subs tu on.
There is not a step–by–step process that one can memorize; rather, experience
will be one’s guide. To gain experience, we now embark on many examples.
Example ∫6.1.1 Integra ng by subs tu on

2
Evaluate x sin(x + 5) dx.
S Knowing that subs tu on is related to the Chain Rule, we

choose to let u be the “inside” func on of sin(x2 + 5). (This is not always a good
choice, but it is o en the best place to start.)
Let u = x2 + 5, hence du = 2x dx. The integrand has an x dx term, but
not a 2x dx term. (Recall that mul plica on is commuta ve, so the x does not
physically have to be next to dx for there to be an x dx term.) We can divide both
sides of the du expression by 2:
1
du = 2x dx ⇒ du = x dx.
2
We can now subs tute.

∫ ∫
x sin(x2 + 5) dx = sin(x2 + 5) |{z}
x dx
| {z }
u 1
2 du
∫
1
= sin u du
2
Notes:
265
1
= − cos u + C (now replace u with x2 + 5)
2
1
= − cos(x2 + 5) + C.
2
∫
Thus x sin(x2 + 5) dx = − 12 cos(x2 + 5) + C. We can check our work by eval-
ua ng the deriva ve of the right hand side.

Evaluate cos(5x) dx.
S Again let u replace the “inside” func on. Le ng u = 5x, we

have du = 5dx. Since our integrand does not have a 5dx term, we can divide
the previous equa on by 5 to obtain 15 du = dx. We can now subs tute.
∫ ∫
cos(5x) dx = cos(|{z} 5x ) |{z}
dx
u 1
5 du
∫
1
= cos u du
5
1
= sin u + C
5
1
= sin(5x) + C.
5
We can again check our work through differen a on.
The previous example exhibited a common, and simple, type of subs tu on.
The “inside” func on was a linear func on (in this case, y = 5x). When the
inside func on is linear, the resul ng integra on is very predictable, outlined
here.
Key Idea 6.1.1 Subs tu on With A Linear Func on

∫ ′
Consider F (ax + b) dx, where a ̸= 0 and b are constants. Le ng
u = ax + b gives du = a · dx, leading to the result
∫
1
F ′ (ax + b) dx = F(ax + b) + C.
a
∫
Thus sin(7x − 4) dx = − 71 cos(7x − 4) + C. Our next example can use Key
Idea 6.1.1, but we will only employ it a er going through all of the steps.
Notes:
266
6.1 Subs tu on
Example ∫6.1.3 Integra ng by subs tu ng a linear func on

7
Evaluate dx.
−3x + 1
S View the integrand as the composi on of func ons f(g(x)),

where f(x) = 7/x and g(x) = −3x + 1. Employing our understanding of subs -
tu on, we let u = −3x + 1, the inside func on. Thus du = −3dx. The integrand
lacks a −3; hence divide the previous equa on by −3 to obtain −du/3 = dx.
We can now evaluate the integral through subs tu on.
∫ ∫
7 7 du
dx =
−3x + 1 u −3
∫
−7 du
=
3 u
−7
= ln |u| + C
3
7
= − ln | − 3x + 1| + C.
3
Using Key Idea 6.1.1 is faster, recognizing that u is linear and a = −3. One may
want to con nue wri ng out all the steps un l they are comfortable with this
par cular shortcut.
Not all integrals that benefit from subs tu on have a clear “inside” func on.
Several of the following examples will demonstrate ways in which this occurs.

Evaluate sin x cos x dx.
S There is not a composi on of func on here to exploit; rather,

just a product of func ons. Do not be afraid to experiment; when given an inte-
gral to evaluate, it is o en beneficial to think “If I let u be this, then du must be
that …” and see if this helps simplify the integral at all.
In this example, let’s set u = sin x. Then du = cos x dx, which we have as
part of the integrand! The subs tu on becomes very straigh orward:
∫ ∫
sin x cos x dx = u du
1 2
= u +C
2
1
= sin2 x + C.
2
Notes:
267
One would do well to ask “What would happen if we let u = cos x?” The result
is just as easy to find, yet looks very different. The challenge to the reader is to
evaluate the integral le ng u = cos x and discover why the answer is the same,
yet looks different.
Our examples so far have required “basic subs tu on.” The next example
demonstrates how subs tu ons can be made that o en strike the new learner
as being “nonstandard.”

√
Evaluate x x + 3 dx.
S Recognizing the composi on of func ons, set u = x + 3.

Then du = dx, giving what seems ini ally to be a simple subs tu on. But at this
stage, we have: ∫ ∫
√ √
x x + 3 dx = x u du.
We cannot evaluate an integral that has both an x and an u in it. We need to

convert the x to an expression involving just u.
Since we set u = x+3, we can also state that u−3 = x. Thus √ we can replace
1
x in the integrand with u − 3. It will also be helpful to rewrite u as u 2 .
∫ ∫
√ 1
x x + 3 dx = (u − 3)u 2 du
∫
( 3 1)
= u 2 − 3u 2 du
2 5 3
= u 2 − 2u 2 + C
5
2 5 3
= (x + 3) 2 − 2(x + 3) 2 + C.
5
Checking your work is always a good idea. In this par cular case, some algebra
will be needed to make one’s answer match the integrand in the original prob-
lem.

1
Evaluate dx.
x ln x
S This is another example where there does not seem to be

an obvious composi on of func ons. The line of thinking used in Example 6.1.5
is useful here: choose something for u and consider what this implies du must
Notes:
268
6.1 Subs tu on
be. If u can be chosen such that du also appears in the integrand, then we have
chosen well.
Choosing u = 1/x makes du = −1/x2 dx; that does not seem helpful. How-
ever, se ng u = ln x makes du = 1/x dx, which is part of the integrand. Thus:
∫ ∫
1 1 1
dx = dx
x ln x ln x |{z}
|{z} x
u du
∫
1
= du
u
= ln |u| + C
= ln | ln x| + C.
The final answer is interes ng; the natural log of the natural log. Take the deriva-
ve to confirm this answer is indeed correct.
Integrals Involving Trigonometric Func ons
Sec on 6.3 delves deeper into integrals of a variety of trigonometric func-

ons; here we use subs tu on to establish a founda on that we will build upon.
The next three examples will help fill in some missing pieces of our an deriva-
ve knowledge. We know the an deriva ves of the sine and cosine func ons;
what about the other standard func ons tangent, cotangent, secant and cose-
cant? We discover these next.
Example ∫6.1.7 Integra on by subs tu on: an deriva ves of tan x

Evaluate tan x dx.
S The previous paragraph established that we did not know

the an deriva ves of tangent, hence we must assume that we have learned
something in this sec on that can help us evaluate this indefinite integral.
Rewrite tan x as sin x/ cos x. While the presence of a composi on of func-
ons may not be immediately obvious, recognize that cos x is “inside” the 1/x
func on. Therefore, we see if se ng u = cos x returns usable results. We have
Notes:
269
that du = − sin x dx, hence −du = sin x dx. We can integrate:

∫ ∫
sin x
tan x dx = dx
cos x
∫
1
= sin
| {zx dx}
cos
|{z}x −du
u
∫
−1
= du
u
= − ln |u| + C
= − ln | cos x| + C.
Some texts prefer to bring the −1 inside the logarithm as a power of cos x, as in:
− ln | cos x| + C = ln |(cos x)−1 | + C

1

= ln +C
cos x
= ln | sec x| + C.
∫
Thus the result they give is tan x dx = ln | sec x| + C. These two answers are
equivalent.
Example ∫6.1.8 Integra ng by subs tu on: an deriva ves of sec x

Evaluate sec x dx.
S This example employs a wonderful trick: mul ply the inte-

grand by “1” so that we see how to integrate more clearly. In this case, we write
“1” as
sec x + tan x
1= .
sec x + tan x
This may seem like it came out of le field, but it works beau fully. Consider:
∫ ∫
sec x + tan x
sec x dx = sec x · dx
sec x + tan x
∫
sec2 x + sec x tan x
= dx.
sec x + tan x
Notes:
270
6.1 Subs tu on
Now let u = sec x + tan x; this means du = (sec x tan x + sec2 x) dx, which is
our numerator. Thus:
∫
du
=
u
= ln |u| + C
= ln | sec x + tan x| + C.
We can use similar techniques to those used in Examples 6.1.7 and 6.1.8
to find an deriva ves of cot x and csc x (which the reader can explore in the
exercises.) We summarize our results here.
Theorem 6.1.2 An deriva ves of Trigonometric Func ons

∫ ∫
1. sin x dx = − cos x + C 4. csc x dx = − ln | csc x + cot x| + C
∫ ∫
2. cos x dx = sin x + C 5. sec x dx = ln | sec x + tan x| + C
∫ ∫
3. tan x dx = − ln | cos x|+C 6. cot x dx = ln | sin x| + C
We explore one more common trigonometric integral.
Example ∫6.1.9 Integra on by subs tu on: powers of cos x and sin x

2
Evaluate cos x dx.
( )2
S We have a composi on of func ons as cos2 x = cos x .
However, se ng u = cos x means du = − sin x dx, which we do not have in the
integral. Another technique is needed.
The process we’ll employ is to use a Power Reducing formula for cos2 x (per-
haps consult the back of this text for this formula), which states
1 + cos(2x)
cos2 x = .
2
The right hand side of this equa on is not difficult to integrate. We have:
∫ ∫
1 + cos(2x)
cos2 x dx = dx
2
∫ ( )
1 1
= + cos(2x) dx.
2 2
Notes:
271
Now use Key Idea 6.1.1:
1 1 sin(2x)
= x+ +C
2 2 2
1 sin(2x)
= x+ + C.
2 4
We’ll make significant use of this power–reducing technique in future sec ons.
Simplifying the Integrand
It is common to be reluctant to manipulate the integrand of an integral; at

first, our grasp of integra on is tenuous and one may think that working with
the integrand will improperly change the results. Integra on by subs tu on
works using a different logic: as long as equality is maintained, the integrand can
be manipulated so that its form is easier to deal with. The next two examples
demonstrate common ways in which using algebra first makes the integra on
easier to perform.
Example ∫6.1.10 Integra on by subs tu on: simplifying first

x3 + 4x2 + 8x + 5
Evaluate dx.
x2 + 2x + 1
S One may try to start by se ng u equal to either the numer-

ator or denominator; in each instance, the result is not workable.
When dealing with ra onal func ons (i.e., quo ents made up of polynomial
func ons), it is an almost universal rule that everything works be er when the
degree of the numerator is less than the degree of the denominator. Hence we
use polynomial division.
We skip the specifics of the steps, but note that when x2 + 2x + 1 is divided
into x3 + 4x2 + 8x + 5, it goes in x + 2 mes with a remainder of 3x + 3. Thus
x3 + 4x2 + 8x + 5 3x + 3
=x+2+ 2 .
x2 + 2x + 1 x + 2x + 1
Integra ng x + 2 is simple. The frac on can be integrated by se ng u = x2 +

2x + 1, giving du = (2x + 2) dx. This is very similar to the numerator. Note that
Notes:
272
6.1 Subs tu on
du/2 = (x + 1) dx and then consider the following:

∫ 3 ∫ ( )
x + 4x2 + 8x + 5 3x + 3
dx = x + 2 + dx
x2 + 2x + 1 x2 + 2x + 1
∫ ∫
3(x + 1)
= (x + 2) dx + 2
dx
x + 2x + 1
∫
1 2 3 du
= x + 2x + C1 +
2 u 2
1 2 3
= x + 2x + C1 + ln |u| + C2
2 2
1 2 3
= x + 2x + ln |x2 + 2x + 1| + C.
2 2
In some ways, we “lucked out” in that a er dividing, subs tu on was able to be
done. In later sec ons we’ll develop techniques for handling ra onal func ons
where subs tu on is not directly feasible.
Example ∫6.1.11 Integra on by alternate methods

x2 + 2x + 3
Evaluate √ dx with, and without, subs tu on.
x
S √ We already know how to integrate this par cular example.

1
Rewrite x as x 2 and simplify the frac on:
x2 + 2x + 3
= x 2 + 2x 2 + 3x− 2 .
3 1 1
x 1/2
We can now integrate using the Power Rule:

∫ 2 ∫ ( )
x + 2x + 3
2 + 2x 2 + 3x− 2
3 1 1
dx = x dx
x1/2
2 5 4 3 1
= x 2 + x 2 + 6x 2 + C
5 3
This is a perfectly fine approach. We demonstrate how this can also be solved
using subs tu√ on as1 its implementa on is rather clever.
Let u = x = x 2 ; therefore
1 −1 1 1
du = x 2 dx = √ dx ⇒ 2du = √ dx.
2 2 x x
∫ 2 ∫
x + 2x + 3
This gives us √ dx = (x2 + 2x + 3) · 2 du. What are we to do
x
1
with the other x terms? Since u = x 2 , u2 = x, etc. We can then replace x2 and
Notes:
273
x with appropriate powers of u. We thus have

∫ 2 ∫
x + 2x + 3
√ dx = (x2 + 2x + 3) · 2 du
x
∫
= 2(u4 + 2u2 + 3) du
2 5 4 3
= u + u + 6u + C
5 3
2 5 4 3 1
= x 2 + x 2 + 6x 2 + C,
5 3
which is obviously the same answer we obtained before. In this situa on, sub-
s tu on is arguably more work than our other method. The fantas c thing is
that it works. It demonstrates how flexible integra on is.
Subs tu on and Inverse Trigonometric Func ons

When studying deriva ves of inverse func ons, we learned that
d( ) 1
tan−1 x = .
dx 1 + x2
Applying the Chain Rule to this is not difficult; for instance,
d( ) 5
tan−1 5x = .
dx 1 + 25x2
We now explore how Subs tu on can be used to “undo” certain deriva ves that
are the result of the Chain Rule applied to Inverse Trigonometric func ons. We
begin with an example.
Example ∫6.1.12 Integra ng by subs tu on: inverse trigonometric func ons

1
Evaluate dx.
25 + x2
S The integrand looks similar to the deriva ve of the arctan-

gent func on. Note:
1 1
= x2
25 + x2 25(1 + 25 )
1
= ( )2
25(1 + 5x )
1 1
= ( ) .
25 1 + x 2
5
Notes:
274
6.1 Subs tu on
Thus ∫ ∫
1 1 1
2
dx = ( x )2 dx.
25 + x 25 1+ 5
This can be integrated using Subs tu on. Set u = x/5, hence du = dx/5 or
dx = 5du. Thus
∫ ∫
1 1 1
dx = ( )2 dx
25 + x2 25 1 + 5x
∫
1 1
= du
5 1 + u2
1
= tan−1 u + C
5
1 (x)
= tan−1 +C
5 5
Example 6.1.12 demonstrates a general technique that can be applied to

other integrands that result in inverse trigonometric func ons. The results are
summarized here.
Theorem 6.1.3 Integrals Involving Inverse Trigonometric Func ons

Let a > 0.
∫ (x)
1 1
1. 2 2
dx = tan−1 +C
a +x a a
∫ (x)
1
2. √ dx = sin−1 +C
a2 − x2 a
∫ ( )
1 1 |x|
3. √ dx = sec−1 +C
x x2 − a2 a a
Let’s prac ce using Theorem 6.1.3.
Example 6.1.13 Integra ng by subs tu on: inverse trigonometric func ons

Evaluate the given indefinite integrals.
∫ ∫ ∫
1 1 1
1. dx, 2. √ dx 3. √ dx.
9 + x2 x x − 100
2 1 5 − x2
Notes:
275
S Each can be answered using a straigh orward applica on of

Theorem 6.1.3.
∫
1 1 x
1. 2
dx = tan−1 + C, as a = 3.
9+x 3 3
∫
1
2. √ dx = 10 sec−1 10x + C, as a = 1
10 .
x x2 − 1
100
∫
1 x √
3. √ = sin−1 √ + C, as a = 5.
5 − x2 5
Most applica ons of Theorem 6.1.3 are not as straigh orward. The next
examples show some common integrals that can s ll be approached with this
theorem.
Example ∫6.1.14 Integra ng by subs tu on: comple ng the square

1
Evaluate dx.
x2 − 4x + 13
S Ini ally, this integral seems to have nothing in common with

the integrals in Theorem 6.1.3. As it lacks a square root, it almost certainly is not
related to arcsine or arcsecant. It is, however, related to the arctangent func on.
We see this by comple ng the square in the denominator. We give a brief
reminder of the process here.
Start with a quadra c with a leading coefficient of 1. It will have the form of
x2 +bx+c. Take 1/2 of b, square it, and add/subtract it back into the expression.
I.e.,
b2 b 2
x2 + bx + c = x2 + bx + − +c
| {z 4} 4
(x+b/2)2
( )2
b b2
= x+ +c−
2 4
In our example, we take half of −4 and square it, ge ng 4. We add/subtract it
into the denominator as follows:
1 1
= 2
x2 − 4x + 13 x − 4x + 4 −4 + 13
| {z }
(x−2)2
1
=
(x − 2)2 + 9
Notes:
276
6.1 Subs tu on
We can now integrate this using the arctangent rule. Technically, we need to
subs tute first with u = x − 2, but we can employ Key Idea 6.1.1 instead. Thus
we have
∫ ∫
1 1 1 x−2
dx = dx = tan−1 + C.
x2 − 4x + 13 (x − 2)2 + 9 3 3
Example ∫6.1.15 Integrals requiring mul ple methods

4−x
Evaluate √ dx.
16 − x2
S This integral requires two different methods to evaluate it.

We get to those methods by spli ng up the integral:
∫ ∫ ∫
4−x 4 x
√ dx = √ dx − √ dx.
16 − x2 16 − x2 16 − x2
The first integral is handled using a straigh orward applica on of Theorem 6.1.3;
the second integral is handled by subs tu on, with u = 16−x2 . We handle each
separately.
∫
4 x
√ dx = 4 sin−1 + C.
16 − x2 4
∫
x
√ dx: Set u = 16 − x2 , so du = −2xdx and xdx = −du/2. We
16 − x2
have
∫ ∫
x −du/2
√ dx = √
16 − x 2 u
∫
1 1
=− √ du
2 u
√
=− u+C
√
= − 16 − x2 + C.
Combining these together, we have
∫
4−x x √
√ dx = 4 sin−1 + 16 − x2 + C.
16 − x2 4
Subs tu on and Definite Integra on

This sec on has focused on evalua ng indefinite integrals as we are learning
a new technique for finding an deriva ves. However, much of the me integra-
on is used in the context of a definite integral. Definite integrals that require
subs tu on can be calculated using the following workflow:
Notes:
277
∫ b
1. Start with a definite integral f(x) dx that requires subs tu on.
a
∫
2. Ignore the bounds; use subs tu on to evaluate f(x) dx and find an an-
deriva ve F(x).
b

3. Evaluate F(x) at the bounds; that is, evaluate F(x) = F(b) − F(a).
a
This workflow works fine, but subs tu on offers an alterna ve that is powerful
and amazing (and a li le me saving).
At its heart, (using
∫ the nota on of Theorem 6.1.1) subs tu on ∫ converts inte-
grals of the form F ′ (g(x))g ′ (x) dx into an integral of the form F ′ (u) du with
the subs tu on of u = g(x). The following theorem states how the bounds of
a definite integral can be changed as the subs tu on is performed.
Theorem 6.1.4 Subs tu on with Definite Integrals

Let F and g be differen able func ons, where the range of g is an interval
I that is contained in the domain of F. Then
∫ b ∫ g(b)
( )
F ′ g(x) g ′ (x) dx = F ′ (u) du.
a g(a)
In effect, Theorem 6.1.4 states that once you convert to integra ng with re-
spect to u, you do not need to switch back to evalua ng with respect to x. A few
examples will help one understand.
Example ∫6.1.16 Definite integrals and subs tu on: changing the bounds
2
Evaluate cos(3x − 1) dx using Theorem 6.1.4.
0
S Observing the composi on of func ons, let u = 3x − 1,

hence du = 3dx. As 3dx does not appear in the integrand, divide the la er
equa on by 3 to get du/3 = dx.
By se ng u = 3x − 1, we are implicitly sta ng that g(x) = 3x − 1. Theorem
6.1.4 states that the new lower bound is g(0) = −1; the new upper bound is
Notes:
278
6.1 Subs tu on
y
g(2) = 5. We now evaluate the definite integral: 1 y = cos(3x − 1)
∫ 2 ∫ 5
du
cos(3x − 1) dx = cos u 0.5
0 −1 3

1 5
x
= sin u −1 1 2 3 4 5
3 −1
1( ) −0.5
= sin 5 − sin(−1) ≈ −0.039.
3
. −1
No ce how once we converted the integral to be in terms of u, we never went
back to using x. (a)
y
The graphs in Figure 6.1.1 tell more of the story. In (a) the area defined by
1
the original integrand is shaded, whereas in (b) the area defined by the new in-
tegrand is shaded. In this par cular situa on, the areas look very similar; the 1
0.5 y= cos(u)
new region is “shorter” but “wider,” giving the same area. 3
u
Example 6.1.17 Definite integrals and subs tu on: changing the bounds
∫ π/2 −1 1 2 3 4 5
Evaluate sin x cos x dx using Theorem 6.1.4. −0.5

0
. −1
S We saw the corresponding indefinite integral in Example 6.1.4.
In that example we set u = sin x but stated that we could have let u = cos x. (b)
For variety, we do the la er here.
Let u = g(x) = cos x, giving du = − sin x dx and hence sin x dx = −du. The Figure 6.1.1: Graphing the areas de-
new upper bound is g(π/2) = 0; the new lower bound is g(0) = 1. Note how fined by the definite integrals of Example
6.1.16.
the lower bound is actually larger than the upper bound now. We have
∫ π/2 ∫ 0 y
sin x cos x dx = −u du (switch bounds & change sign) 1
0 1
∫ 1
y = sin x cos x
= u du 0.5
0
1 1
= u2 = 1/2.
2 0 x
1 π
2
In Figure 6.1.2 we have again graphed the two regions defined by our definite
integrals. Unlike the previous example, they bear no resemblance to each other.
However, Theorem 6.1.4 guarantees that they have the same area. . −0.5
(a)
y
Integra on by subs tu on is a powerful and useful integra on technique. y=u
1
The next sec on introduces another technique, called Integra on by Parts. As
subs tu on “undoes” the Chain Rule, integra on by parts “undoes” the Product
Rule. Together, these two techniques provide a strong founda on on which most 0.5
other integra on techniques are based.
u
1 π
2
Notes: . −0.5
(b)
Figure 6.1.2: Graphing the areas de-

fined by the definite integrals of Example
6.1.17.
279
Exercises 6.1 ∫
Terms and Concepts 17. cos(3 − 6x)dx
1. Subs tu on “undoes” what deriva ve rule? ∫

18. sec2 (4 − x)dx
2. T/F: One can use algebra to rewrite the integrand of an in-
tegral to make it easier to evaluate. ∫
19. sec(2x)dx
∫
Problems 20. tan2 (x) sec2 (x)dx
In Exercises 3 – 14, evaluate the indefinite integral to develop ∫

( )
an understanding of Subs tu on. 21. x cos x2 dx
∫
( )7 ∫
3. 3x2 x3 − 5 dx
22. tan2 (x)dx
∫
( )3 ∫
4. (2x − 5) x2 − 5x + 7 dx
23. cot x dx. Do not just refer to Theorem 6.1.2 for the an-
∫ swer; jus fy it through Subs tu on.
( )8
5. x x2 + 1 dx ∫
24. csc x dx. Do not just refer to Theorem 6.1.2 for the an-
∫
( )5 swer; jus fy it through Subs tu on.
6. (12x + 14) 3x2 + 7x − 1 dx
In Exercises 25 – 32, use Subs tu on to evaluate the indefi-
∫
1 nite integral involving exponen al func ons.
7. dx
2x + 7 ∫
∫ 25. e3x−1 dx
1
8. √ dx
2x + 3 ∫
3
∫ 26. ex x2 dx
x
9. √ dx ∫
x+3 2
−2x+1
27. ex (x − 1)dx
∫
x −x
3
10. √ dx ∫
x ex + 1
28. dx
ex
∫ √
e x ∫
11. √ dx ex
x 29. dx
ex +1
∫
x4 ∫
12. √ dx ex − e−x
x5 + 1 30. dx
e2x
∫ 1 ∫
x
+1
13. dx 31. 33x dx
x2
∫ ∫
ln(x)
14. dx 32. 42x dx
x
In Exercises 15 – 24, use Subs tu on to evaluate the indefi- In Exercises 33 – 36, use Subs tu on to evaluate the indefi-
nite integral involving trigonometric func ons. nite integral involving logarithmic func ons.
∫ ∫
ln x
15. 2
sin (x) cos(x)dx 33. dx
x
∫ ∫ ( )2
3 ln x
16. cos (x) sin(x)dx 34. dx
x
280
∫ ( )
ln x3 In Exercises 53 – 78, evaluate the indefinite integral.
35. dx
x
∫
∫ x2
53. dx
1 (x3 + 3)2
36. dx
x ln (x2 )
∫
( 2 )( )8
54. 3x + 2x 5x3 + 5x2 + 2 dx
In Exercises 37 – 42, use Subs tu on to evaluate the indefi-
nite integral involving ra onal func ons.
∫
x
∫ 55. √ dx
x2 + 3x + 1 1 − x2
37. dx
x ∫
( )
∫ 56. x2 csc2 x3 + 1 dx
x3 + x2 + x + 1
38. dx
x ∫ √
∫ 57. sin(x) cos(x)dx
x3 − 1
39. dx ∫
x+1 ( )
58. sin 5x + 1 dx
∫
x2 + 2x − 5
40.
x−3
dx ∫
1
59. dx
∫ x−5
3x2 − 5x + 7
41. dx ∫
x+1 7
60. dx
3x + 2
∫
x2 + 2x + 1 ∫
42. dx
x3 + 3x2 + 3x 3x3 + 4x2 + 2x − 22
61. dx
x2 + 3x + 5
In Exercises 43 – 52, use Subs tu on to evaluate the indefi- ∫
nite integral involving inverse trigonometric func ons. 2x + 7
62. dx
x2 + 7x + 3
∫
7 ∫
43. dx 9(2x + 3)
x2 + 7 63. dx
3x2 + 9x + 7
∫
3 ∫
44. √ dx −x3 + 14x2 − 46x − 7
9 − x2 64. dx
x2 − 7x + 1
∫
14 ∫
45. √ dx x
5 − x2 65. dx
x4 + 81
∫
2 ∫
46. √ dx 2
x x2 − 9 66. dx
4x2 + 1
∫ ∫
5
47. √ dx 67. √
1
dx
x4 − 16x2 x 4x2 − 1
∫ ∫
x 1
48. √ dx 68. √ dx
1 − x4 16 − 9x2
∫ ∫
1 3x − 2
49. dx 69. dx
x2 − 2x + 8 x2 − 2x + 10
∫ ∫
2 7 − 2x
50. √ dx 70. dx
−x + 6x + 7
2 x2 + 12x + 61
∫ ∫
3 x2 + 5x − 2
51. √ dx 71. dx
−x2 + 8x + 9 x2− 10x + 32
∫ ∫
5 x3
52. dx 72. dx
x2 + 6x + 34 x2 + 9
281
∫ ∫
x3 − x 6 √
73. 2
dx 80. x x − 2dx
x + 4x + 9 2
∫ ∫
sin(x) π/2
74. dx 81. sin2 x cos x dx
cos2 (x) + 1
−π/2
∫
75.
cos(x)
dx ∫ 1
sin2 (x) + 1 82. 2x(1 − x2 )4 dx
0
∫
cos(x) ∫
76. dx −1
1 − sin2 (x) 83. (x + 1)ex
2
+2x+1
dx
∫ −2
3x − 3
77. √ dx ∫
x2 − 2x − 6 1
1
84. dx
∫ −1 1 + x2
x−3
78. √ dx
x2 − 6x + 8 ∫ 4
1
85. dx
In Exercises 79 – 86, evaluate the definite integral. 2 x2 − 6x + 10
∫ 3 ∫ √
3
1 1
79. dx 86. √ dx
1 x−5 1 4 − x2
282
A: S T S P
Chapter 1 7. Let ε > 0 be given. We wish to find δ > 0 such that when
|x − 3| < δ, |f(x) − 6| < ε.
Sec on 1.1 Consider |f(x) − 6| < ε, keeping in mind we want to make a
statement about |x − 3|:
1. Answers will vary. |f(x) − 6| < ε
3. F |x2 − 3 − 6| < ε
5. Answers will vary. |x2 − 9| < ε
7. −1 |x − 3| · |x + 3| < ε
9. Limit does not exist |x − 3| < ε/|x + 3|
11. 1.5
13. Limit does not exist. Since x is near 3, we can safely assume that, for instance,
2 < x < 4. Thus
15. 1
f(a+h)−f(a) 2+3<x+3<4+3
h h
−0.1 −7 5<x+3<7
17. −0.01 −7 The limit seems to be exactly 7. 1 1 1
< <
0.01 −7 7 x+3 5
0.1 −7 ε ε ε
< <
f(a+h)−f(a) 7 x+3 5
h h
−0.1 4.9
19. −0.01 4.99 The limit is approx. 5. Let δ = 7ε . Then:
0.01 5.01
|x − 3| < δ
0.1 5.1
ε
f(a+h)−f(a) |x − 3| <
h h 7
−0.1 29.4 ε
|x − 3| <
21. −0.01 29.04 The limit is approx. 29. x+3
0.01 28.96 ε
|x − 3| · |x + 3| < · |x + 3|
0.1 28.6 x+3
f(a+h)−f(a)
h h
−0.1 −0.998334 Assuming x is near 3, x + 3 is posi ve and we can drop the
23. −0.01 −0.999983 The limit is approx. −1. absolute value signs on the right.
0.01 −0.999983 ε
|x − 3| · |x + 3| < · (x + 3)
0.1 −0.998334 x+3
Sec on 1.2 |x2 − 9| < ε
2
|(x − 3) − 6| < ε,
1. ε should be given first, and the restric on |x − a| < δ implies
|f(x) − K| < ε, not the other way around. which is what we wanted to prove.
3. T 9. Let ε > 0 be given. We wish to find δ > 0 such that when

|x − 1| < δ, |f(x) − 6| < ε.
5. Let ε > 0 be given. We wish to find δ > 0 such that when
Consider |f(x) − 6| < ε, keeping in mind we want to make a
|x − 4| < δ, |f(x) − 13| < ε.
Consider |f(x) − 13| < ε:
|f(x) − 6| < ε
|f(x) − 13| < ε
|(2x2 + 3x + 1) − 6| < ε
|(2x + 5) − 13| < ε
|2x − 8| < ε |2x2 + 3x − 5| < ε
2|x − 4| < ε |2x + 5| · |x − 1| < ε
−ε/2 < x − 4 < ε/2. |x − 1| < ε/|2x + 5|
Since x is near 1, we can safely assume that, for instance,

This implies we can let δ = ε/2. Then:
0 < x < 2. Thus
|x − 4| < δ
0 + 5 < 2x + 5 < 4 + 5
−δ < x − 4 < δ
5 < 2x + 5 < 9
−ε/2 < x − 4 < ε/2
1 1 1
−ε < 2x − 8 < ε < <
9 2x + 5 5
−ε < (2x + 5) − 13 < ε ε ε ε
< <
|(2x + 5) − 13| < ε, 9 2x + 5 5
which is what we wanted to prove.
Let δ = 9ε . Then: 13. 3
|x − 1| < δ 15. 1
ε 17. 0
|x − 1| <
9 19. 7
ε
|x − 1| < 21. 1/2
2x + 5
ε 23. Limit does not exist
|x − 1| · |2x + 5| < · |2x + 5|
2x + 5
25. 2
π 2 +3π+5
27. 5π 2 −2π−3
≈ 0.6064
Assuming x is near 1, 2x + 5 is posi ve and we can drop the
absolute value signs on the right. 29. −8
ε 31. 10
|x − 1| · |2x + 5| < · (2x + 5)
2x + 5 33. −3/2
|2x2 + 3x − 5| < ε 35. 0
|(2x2 + 3x + 1) − 6| < ε, 37. 1
which is what we wanted to prove. 39. 3
11. Let ε > 0 be given. We wish to find δ > 0 such that when 41. 1
|x − 2| < δ, |f(x) − 5| < ε. However, since f(x) = 5, a constant
func on, the la er inequality is simply |5 − 5| < ε, which is 43. (a) Apply Part 1 of Theorem 1.3.1.
always true. Thus we can choose any δ we like; we arbitrarily (b) Apply Theorem 1.3.6; g(x) = xx is the same as g(x) = 1
choose δ = ε. everywhere except at x = 0. Thus lim g(x) = lim 1 = 1.
x→0 x→0
13. Let ε > 0 be given. We wish to find δ > 0 such that when ( )
(c) The func on f(x) is always 0, so g f(x) is never defined as
|x − 1| < δ, |f(x) − 1| < ε. g(x) is not defined at x = 0. Therefore the limit does not
Consider |f(x) − 1| < ε, keeping in mind we want to make a exist.
(d) The Composi on Rule requires that lim g(x) be equal to
x→0
|f(x) − 1| < ε g(0). They are not equal, so the condi ons of the
|1/x − 1| < ε Composi on Rule are not sa sfied, and hence the rule is
|(1 − x)/x| < ε not violated.
|x − 1|/|x| < ε Sec on 1.4
|x − 1| < ε · |x|
1. The func on approaches different values from the le and right;
the func on grows without bound; the func on oscillates.
Since x is near 1, we can safely assume that, for instance,
3. F
1/2 < x < 3/2. Thus ε/2 < ε · x.
Let δ = 2ε . Then: 5. (a) 2
(b) 2
|x − 1| < δ
ε (c) 2
|x − 1| < (d) 1
2
|x − 1| < ε · x (e) As f is not defined for x < 0, this limit is not defined.
|x − 1|/x < ε (f) 1
7. (a) Does not exist.
Assuming x is near 1, x is posi ve and we can bring it into the (b) Does not exist.
absolute value signs on the le . (c) Does not exist.
|(x − 1)/x| < ε (d) Not defined.
|1 − 1/x| < ε (e) 0
|(1/x) − 1| < ε, (f) 0
which is what we wanted to prove. 9. (a) 2
Sec on 1.3 (b) 2
(c) 2
1. Answers will vary. (d) 2
3. Answers will vary. 11. (a) 2
5. As x is near 1, both f and g are near 0, but f is approximately twice (b) 2
the size of g. (I.e., f(x) ≈ 2g(x).)
(c) 2
7. 9 (d) 0
9. 0 (e) 2
11. 3 (f) 2
A.2
(g) 2 31. (0, ∞)
(h) Not defined 33. (−∞, 0]
13. (a) 2 35. Yes, by the Intermediate Value Theorem.
(b) −4 37. We cannot say; the Intermediate Value Theorem only applies to
(c) Does not exist. func on values between −10 and 10; as 11 is outside this range,
we do not know.
(d) 2
39. Approximate root is x = 1.23. The intervals used are:
15. (a) 0 [1, 1.5] [1, 1.25] [1.125, 1.25]
(b) 0 [1.1875, 1.25] [1.21875, 1.25] [1.234375, 1.25]
(c) 0 [1.234375, 1.2421875] [1.234375, 1.2382813]
(d) 0 41. Approximate root is x = 0.69. The intervals used are:

[0.65, 0.7] [0.675, 0.7] [0.6875, 0.7]
(e) 2 [0.6875, 0.69375] [0.690625, 0.69375]
(f) 2
43. (a) 20
(g) 2 (b) 25
(h) 2 (c) Limit does not exist
17. (a) 1 − cos2 a = sin2 a (d) 25
(b) sin2 a 45. Answers will vary.
(c) sin2 a
Sec on 1.6
(d) sin2 a
19. (a) 4 1. F
(b) 4 3. F
(c) 4 5. T
(d) 3 7. Answers will vary.
21. (a) −1 9. (a) ∞
(b) 1 (b) ∞
(c) Does not exist 11. (a) 1
(d) 0 (b) 0
23. 2/3 (c) 1/2
25. −9 (d) 1/2
Sec on 1.5 13. (a) Limit does not exist
(b) Limit does not exist
1. Answers will vary.
15. Tables will vary.
3. A root of a func on f is a value c such that f(c) = 0.
x f(x)
5. F 2.9 −15.1224
(a) It seems limx→3− f(x) = −∞.
7. T 2.99 −159.12
2.999 −1599.12
9. F
x f(x)
11. No; lim f(x) = 2, while f(1) = 1. 3.1 16.8824
x→1 (b) It seems limx→3+ f(x) = ∞.
3.01 160.88
13. No; f(1) does not exist. 3.001 1600.88
15. Yes (c) It seems limx→3 f(x) does not exist.
17. (a) No; lim f(x) ̸= f(−2) 17. Tables will vary.
x→−2
(b) Yes x f(x)

(c) No; f(2) is not defined. (a) 2.9 132.857 It seems limx→3− f(x) = ∞.
2.99 12124.4
19. (a) Yes
x f(x)
(b) Yes (b) 3.1 108.039 It seems limx→3+ f(x) = ∞.
21. (a) Yes 3.01 11876.4
(b) Yes (c) It seems limx→3 f(x) = ∞.
23. (−∞, ∞) 19. Horizontal asymptote at y = 2; ver cal asymptotes at x = −5, 4.
25. [−2, 2] 21. Horizontal asymptote at y = 0; ver cal asymptotes at x = −1, 0.
√ √
27. (−∞, − 6] and [ 6, ∞) 23. No horizontal or ver cal asymptotes.
29. (−∞, ∞) 25. ∞
A.3
f(0+h)−f(0)
27. −∞ 35. limh→0+ h
= 0; note also that limx→0+ f ′ (x) = 0. So f
is differen able at x = 0.
29. Solu on omi ed. f(1+h)−f(1)
limh→0− h
= −∞; note also that
31. Yes. The only “ques onable” place is at x = 3, but the le and limx→1− f ′ (x) = −∞. So f is not differen able at x = 1.
right limits agree. f is differen able on [0, 1), not its en re domain.
Chapter 2 37. Approximately 24.
39. (a) (−∞, ∞)
Sec on 2.1 (b) (−∞, −1) ∪ (−1, 1) ∪ (1, ∞)
1. T (c) (−∞, 5]
√ √
(d) [− 5, 5]
Sec on 2.2
7. f ′ (x) = 0 1. Velocity
9. f ′ (t) = −3 3. Linear func ons.
11. h′ (x) = 3x2 5. −17
−1
13. r ′ (x) = x2
7. f(10.1) is likely most accurate, as accuracy is lost the farther from
x = 10 we go.
15. (a) y = 6
9. 6
(b) x = −2
11. /s2
17. (a) y = −3x + 4
13. (a) thousands of dollars per car
(b) y = 1/3(x − 7) − 17
(b) It is likely that P(0) < 0. That is, nega ve profit for not
19. (a) y = 48(x − 4) + 64 producing any cars.
1
(b) y = − 48 (x − 4) + 64 15. f(x) = g′ (x)
21. (a) y = −1/4(x + 2) − 1/2 17. Either g(x) = f ′ (x) or f(x) = g′ (x) is acceptable. The actual
(b) y = 4(x + 2) − 1/2 answer is g(x) = f ′ (x), but is very hard to show that f(x) ̸= g′ (x)
given the level of detail given in the graph.
23. y = 8.1(x − 3) + 16
19. f ′ (x) = 10x
25. y = 7.77(x − 2) + e2 , or y = 7.77(x − 2) + 7.39.
21. f ′ (π) ≈ 0.
27. (a) Approxima ons will vary; they should match (c) closely.
Sec on 2.3
(b) f ′ (x) = 2x
(c) At (−1, 0), slope is −2. At (0, −1), slope is 0. At (2, 3), 1. Power Rule.
slope is 4.
3. One answer is f(x) = 10ex .
y
5. g(x) and h(x)
3
7. One possible answer is f(x) = 17x − 205.
2 9. f ′ (x) is a velocity func on, and f ′′ (x) is accelera on.
1 11. f ′ (x) = 14x − 5
x
13. m′ (t) = 45t4 − 38 t2 + 3
−2 −1 1 2 3 4
15. f ′ (r) = 6er
−1
29. . 17. f ′ (x) = 2
x
−1
y 19. h′ (t) = et − cos t + sin t
5
21. f ′ (t) = 0
23. g′ (x) = 24x2 − 120x + 150
x
25. f ′ (x) = 18x − 12
−2 −1 1 2
27. f ′ (x) = 6x5 f ′′ (x) = 30x4 f ′′′ (x) = 120x3 f(4) (x) = 360x2
29. h′ (t) = 2t − et h′′ (t) = 2 − et h′′′ (t) = −et h(4) (t) = −et
−5 31. f ′ (θ) = cos θ + sin θ f ′′ (θ) = − sin θ + cos θ
31. .
f ′′′ (θ) = − cos θ − sin θ f(4) (θ) = sin θ − cos θ
33. (a) Approximately on (−2, 0) and (2, ∞).
33. Tangent line: y = 2(x − 1)
(b) Approximately on (−∞, −2) and (0, 2). Normal line: y = −1/2(x − 1)
(c) Approximately at x = 0, ±2. 35. Tangent line: y = x − 1
(d) Approximately on (−∞, −1) and (1, ∞). Normal line: y = −x + 1
√ √
(e) Approximately on (−1, 1). 37. Tangent line: y = 2(x − π4 ) − 2
−1
√
(f) Approximately at x = ±1. Normal line: y = √ (x − π4 ) − 2
2
A.4
y
39. The tangent line to f(x) = ex at x = 0 is y = x + 1; thus
6
e0.1 ≈ y(0.1) = 1.1.
4
Sec on 2.4 2
x
1. F −2 −1 1 2 3 4 5
−2
3. T
−4
5. F
. −6
51. .
7. (a) f ′ (x) = (x2 + 3x) + x(2x + 3)
Sec on 2.5
(b) f ′ (x) = 3x2 + 6x
(c) They are equal. 1. T
9. (a) h′ (s) = 2(s + 4) + (2s − 1)(1) 3. F
(b) h′ (s) = 4s + 7 5. T
(c) They are equal. 7. f ′ (x) = 10(4x3 − x)9 · (12x2 − 1) = (120x2 − 10)(4x3 − x)9
x(2x)−(x2 +3)1
9. g′ (θ) = 3(sin θ + cos θ)2 (cos θ − sin θ)
11. (a) f ′ (x) = ( )
x2 11. f ′ (x) = 3 ln x + x2 2( 1x + 2x)
(b) f ′ (x) = 1 − 3 ( )3 ( )
x2 13. f ′ (x) = 4 x + 1x 1 − x12
(c) They are equal.
15. g′ (x) = 5 sec2 (5x)
13. (a) h′ (s) =
4s3 (0)−3(12s2 ) ( )( 4 )
16s6 17. g ′ (t) = cos t5 + 1t 5t − 1
t3
(b) h′ (s) = −9/4s−4
19. p′ (t) = −3 cos2 (t2 + 3t + 1) sin(t2 + 3t + 1)(2t + 3)
(c) They are equal.
21. f ′ (x) = 2/x
15. f ′ (x) = sin x + x cos x 23. g′ (r) = ln 4 · 4r
17. f ′ (x) = ex ln x + ex 1x 25. g′ (t) = 0
( ) ( )
−12 (3t +2) (ln 2)2t −(2t +3) (ln 3)3t
19. g′ (x) = (x−5)2 27. f ′ (x) = (3t +2)2
21. h′ (x) = − csc2 x − ex 2 2 2
2x (ln 3·3x 2x+1)−(3x +x)(ln 2·2x 2x)
2
29. f ′ (x) = 2
22x
23. h′ (t) = 14t + 6
( ) ( ) 31. f ′ (x) =
25. f ′ (x) = 6x + 8 ex + 3x2 + 8x + 7 ex 5(x2 + x)4 (2x + 1)(3x4 + 2x)3 + 3(x2 + x)5 (3x4 + 2x)2 (12x3 + 2)
27. f ′ (x) = 7 33. f ′ (x) = 3 cos(3x + 4) cos(5 − 2x) + 2 sin(3x + 4) sin(5 − 2x)
4(5x−9)3 cos(4x+1)−15 sin(4x+1)(5x−9)2
35. f ′ (x) =
2 2
sin (x)+cos (x)+3 cos(x)
29. f ′ (x) = (cos(x)+3)2 (5x−9)6
37. Tangent line: y = 0
−x sin x−cos x tan x−x sec2 x
31. f ′ (x) = x2
+ tan2 x Normal line: x = 0
33. g′ (t) = 12t2 et + 4t3 et − cos2 t + sin2 t 39. Tangent line: y = −3(θ − π/2) + 1
Normal line: y = 1/3(θ − π/2) + 1
35. f ′ (x) = 2xex tan x = x2 ex tan x + x2 ex sec2 x
41. In both cases the deriva ve is the same: 1/x.
37. Tangent line: y = 2x + 2 ◦
43. (a) F/mph
Normal line: y = −1/2x + 2
(b) The sign would be nega ve; when the wind is blowing at
39. Tangent line: y = 4 10 mph, any increase in wind speed will make it feel colder,
Normal line: x = 2 i.e., a lower number on the Fahrenheit scale.
41. x = 3/2 Sec on 2.6
43. f ′ (x) is never 0.
45. f ′′ (x) = 2 cos x − x sin x 3. T
47. f ′′ (x) = cot2 x csc x + csc3 x 5. f ′ (x) = 12 x−1/2 − 21 x−3/2 = 1
√ − 1
√
2 x 2 x3
y
7. f ′ (t) = √−t
6 1−t2
√
4 9. h′ (x) = 1.5x0.5 = 1.5 x
√
x(1)−(x+7)(1/2x−1/2 )
11. g′ (x) =
2 1 7
x
= √
2 x
− √
2 x3
x
−3 −2 −1 1 2 3 dy −4x3
−2
13. dx
= 2y+1
−4
dy
15. dx
= sin(x) sec(y)
. 49. . −6 dy y
17. dx
= x
A.5
dy 2 sin(y) cos(y)
19. dx
=− x
31. 3x2 + 1
21. dy
dx
= 1
2y+2
Chapter 3
23. If one takes the deriva ve of the equa on, as shown, using the
− cos(x)(x+cos(y))+sin(x)+y Sec on 3.1
Quo ent Rule, one finds dy dx
= sin(y)(sin(x)+y)+x+cos(y) .
If one first clears the denominator and writes 1. Answers will vary.
sin(x) + y = cos(y) + x then takes the deriva ve of both sides,
1−cos(x) 3. Answers will vary.
one finds dydx
= 1+sin(y) .
These expressions, by themselves, are not equal. However, for 5. F
values of x and y that sa sfy the original equa on (i.e, for x and y 7. A: none; the func on isn’t defined here. B: abs. max & rel. max C:
sin(x)+y)
such that cos(y)+x) = 1), these expressions are equal. rel. min D: none; the func on isn’t defined here. E: none F: rel.
min G: rel. max
dy
25. = − 2x+y
dx 2y+x 9. f ′ (0) = 0
27. (a) y = 0 11. f ′ (π/2) = 0 f ′ (3π/2) = 0
(b) y = −1.859(x − 0.1) + 0.281 13. f ′ (2) is not defined f ′ (6) = 0
29. (a) y = 4 15. f ′ (0) = 0
√
(b) y = 0.93(x − 2) + 4
108
√
17. min: (−0.5, 3.75)
31. (a) y= − 72 ) + 6+32 3
− √1 (x max: (2, 10)
3 √
√ √ 19. min: (π/4, 3 2/2)
(b) y = 3(x − 4+32 3 ) + 32
( )
max: (π/2, 3)
3 √ √
(2y+1)(−12x2 )+4x3 2 −4x 21. min: ( 3, 2 3)
d2 y 2y+1
33. dx2
= (2y+1)2 max: (5, 28/5)
2
d y cos x cos y+sin2 x tan y 23. min: (π, −eπ )
35. dx2
= cos2 y √ π/4
( ln(1+x) ) max: (π/4, 2e
)
37. y′ = (1 + x)1/x 1
x(x+1)
− x2
2
25. min: (1, 0)
Tangent line: y = (1 − 2 ln 2)(x − 1) + 2
( ) max: (e, 1/e)
xx
39. y′ = x+1 1
ln x + 1 − x+1 dy y(y−2x)
27. =
Tangent line: y = (1/4)(x − 1) + 1/2 dx x(x−2y)
( 1 ) 29. 3x2 + 1
41. y′ = x+1
x+2 x+1
1
− x+2
Tangent line: y = 1/9(x − 1) + 2/3 Sec on 3.2
Sec on 2.7 1. Answers will vary.
1. F 3. Any c in [−1, 1] is valid.
3. The point (10, 1) lies on the graph of y = f−1 (x) (assuming f is 5. c = −1/2
inver ble). 7. Rolle’s Thm. does not apply.
5. Compose f(g(x)) and g(f(x)) to confirm that each equals x. 9. Rolle’s Thm. does not apply.
7. Compose f(g(x)) and g(f(x)) to confirm that each equals x. 11. c = 0
( )′ √
9. f−1 (20) = f ′ 1(2) = 1/5 13. c = 3/ 2
( )′ √ 15. The Mean Value Theorem does not apply.
11. f−1 ( 3/2) = f ′ (π/6) 1
=1
√
( −1 )′ 17. c = ± sec−1 (2/ π)
1
13. f (1/2) = f ′ (1) = −2 √
5−7 7
19. c = 6
15. h′ (t) = √ 2
1−4t2 21. Max value of 19 at x = −2 and x = 5; min value of 6.75 at
x = 1.5.
17. g′ (x) = 2
1+4x2
23. They are the odd, integer valued mul ples of π/2 (such as
sin(t)
19. g′ (t) = cos−1 (t) cos(t) − √ 2 0, ±π/2, ±3π/2, ±5π/2, etc.)
1−t
sin −1
(x)+cos−1 (x)
Sec on 3.3
21. h′ (x) = √
1−x2 cos−1 (x)2
23. f ′ (x) = − √ 1
1−x2 3. Answers will vary; graphs should be steeper near x = 0 than near
25. (a) f(x) = x, so f ′ (x) = 1 x = 2.
(b) f ′ (x) = cos(sin−1 x) √ 1 = 1. 5. False; for instance, y = x3 is always increasing though it has a
1−x2 cri cal point at x = 0.
√ √
27. y = 2(x − 2/2) + π/4 7. Graph and verify.
dy y(y−2x)
29. dx
= x(x−2y)
9. Graph and verify.
A.6
11. Graph and verify. 37. max: x = 0
13. Graph and verify. 39. max: x = π/4; min: x = −3π/4
√
15. domain: (−∞, ∞) 41. min: x = 1/ e
c.p. at c = −1; 43. f ′ has no maximal or minimal value.
decreasing on (−∞, −1);
45. f ′ has a minimal value at x = 0
increasing on (−1, ∞);
rel. min at x = −1. 47. Possible points of inflec on: x = −2/3, 0; f ′ has a rela ve min
at: x = 0 ; rela ve max at: x = −2/3
17. domain=(−∞, ∞)
√ 49. f ′ has no rela ve extrema
c.p. at c = 16 (−1 ± 7); √ √
√ √ 51. f ′ has a rela ve max at x = −1/ 3; rela ve min at x = 1/ 3
decreasing on ( 61 (−1 − 7), 16 (−1 + 7)));
√ √ 53. f ′ has a rela ve min at x = 3π/4; rela ve max at x = −π/4
increasing on (−∞, 61 (−1 − 7)) and ( 16 (−1 + 7), ∞);
√ √
rel. min at x = 16 (−1 + 7); 55. f ′ has a rela ve min at x = 1/ e3 = e−3/2
√
rel. max at x = 16 (−1 − 7). Sec on 3.5
19. domain=(−∞, ∞)
c.p. at c = 1;
decreasing on (1, ∞) 3. T
increasing on (−∞, 1); 5. T
rel. max at x = 1. 7. A good sketch will include the x and y intercepts and draw the
21. domain=(−∞, −2) ∪ (−2, 4) ∪ (4, ∞) appropriate line.
no c.p.; 9. Use technology to verify sketch.
decreasing on en re domain, (−∞, −2), (−2, 4) and (4, ∞)
11. Use technology to verify sketch.
23. domain=(−∞, ∞)
c.p. at c = −3π/4, −π/4, π/4, 3π/4;
decreasing on (−3π/4, −π/4) and (π/4, 3π/4);
increasing on (−π, −3π/4), (−π/4, π/4) and (3π/4, π); 17. Use technology to verify sketch.
rel. min at x = −π/4, 3π/4; 19. Use technology to verify sketch.
rel. max at x = −3π/4, π/4. 21. Use technology to verify sketch.
25. c = 1/2 23. Use technology to verify sketch.
Sec on 3.4 25. Use technology to verify sketch.
√
27. Cri cal point: x = 0 Points of inflec on: ±b/ 3
nπ/2−b
3. Yes; Answers will vary. 29. Cri cal points: x = a
, where n is an odd integer Points of
inflec on: (nπ − b)/a, where n is an integer.
5. Graph and verify.
dy
31. dx
= −x/y, so the func on is increasing in second and fourth
7. Graph and verify. quadrants, decreasing in the first and third quadrants.
9. Graph and verify. d2 y
dx2
= −1/y − x2 /y3 , which is posi ve when y < 0 and is
11. Graph and verify. nega ve when y > 0. Hence the func on is concave down in the
first and second quadrants and concave up in the third and fourth
13. Graph and verify. quadrants.
15. Possible points of inflec on: none; concave up on (−∞, ∞)
Chapter 4
17. Possible points of inflec on: x = 0; concave down on (−∞, 0);
concave up on (0, ∞)
Sec on 4.1
19. Possible points of inflec on: x = −2/3, 0; concave down on
(−2/3, 0); concave up on (−∞, −2/3) and (0, ∞) 1. F
21. Possible points of inflec on: x = 1; concave up on (−∞, ∞) 3. x0 = 1.5, x1 = 1.5709148, x2 = 1.5707963, x3 = 1.5707963,
√ x4 = 1.5707963, x5 = 1.5707963
23. Possible
√ points√ of inflec on: x = ±1/ 3; concave
√ down on
√
(−1/ 3, 1/ 3); concave up on (−∞, −1/ 3) and (1/ 3, ∞) 5. x0 = 0, x1 = 2, x2 = 1.2, x3 = 1.0117647, x4 = 1.0000458,
x5 = 1
25. Possible points of inflec on: x = −π/4, 3π/4; concave down on
(−π/4, 3π/4) concave up on (−π, −π/4) and (3π/4, π) 7. x0 = 2, x1 = 0.6137056389, x2 = 0.9133412072,
x3 = 0.9961317034, x4 = 0.9999925085, x5 = 1
27. Possible points of inflec on: x = 1/e3/2 ; concave down on
(0, 1/e3/2 ) concave up on (1/e3/2 , ∞) 9. roots are: x = −5.156, x = −0.369 and x = 0.525
29. min: x = 1 11. roots are: x = −1.013, x = 0.988, and x = 1.393
√ √
31. max: x = −1/ 3 min: x = 1/ 3 13. x = ±0.824,
33. min: x = 1 15. x = ±0.743
35. min: x = 1 17. The approxima ons alternate between x = 1 and x = 2.
A.7
√ √
3
Sec on 4.2 13. Use y = 3 x; dy = 1/(3 x2 ) · dx with x = 64 and dx = −1.
Thus dy = −1/48 ≈ 0.0208; we could√use
1. T −1/48
√ ≈ −1/50 = −0.02; knowing 3 64 = 4, we have
3
63 ≈ 3.98.
3. (a) 5/(2π) ≈ 0.796cm/s
15. Use y = sin x; dy = cos x · dx with x = π and dx ≈ −0.14. Thus
(b) 1/(4π) ≈ 0.0796 cm/s
dy = 0.14; knowing sin π = 0, we have sin 3 ≈ 0.14.
(c) 1/(40π) ≈ 0.00796 cm/s
17. dy = (2x + 3)dx
5. 63.14mph −2
19. dy = 4x3
dx
7. Due to the height of the plane, the gun does not have to rotate ( )
very fast. 21. dy = 2xe3x + 3x2 e3x dx
2(tan x+1)−2x sec2 x
(a) 0.0573 rad/s 23. dy = (tan x+1)2
dx
(b) 0.0725 rad/s 25. dy = (ex sin x + ex cos x)dx
(c) In the limit, rate goes to 0.0733 rad/s 27. dy = 1
dx
(x+2)2
9. (a) 0.04 /s
29. dy = (ln x)dx
(b) 0.458 /s
31. dV = ±0.157
(c) 3.35 /s
33. ±15π/8 ≈ ±5.89in2
(d) Not defined; as the distance approaches 24, the rates
approaches ∞. 35. (a) 297.8 feet
11. (a) 50.92 /min (b) ±62.3

(c) ±20.9%
(b) 0.509 /min
(c) 0.141 /min 37. (a) 298.9 feet
As the tank holds about 523.6 3, it will take about 52.36 minutes. (b) ±8.67
13. (a) The rope is 80 long. (c) ±2.9%
(b) 1.71 /sec 39. 1%

(c) 1.84 /sec Chapter 5
(d) About 34 feet.
Sec on 5.1
15. The cone is rising at a rate of 0.003 /s.
Sec on 4.3 1. Answers will vary.
1. T
3. 2500; the two numbers are each 50.
7. velocity
5. There is no maximum sum; the fundamental equa on has only 1
cri cal value that corresponds to a minimum. 9. 3/4x4 + C
√ 11. 10/3x3 − 2x + C
7. Area = 1/4, with sides of length 1/ 2.
9. The radius should be about 3.84cm and the height should be 13. s + C
2r = 7.67cm. No, this is not the size of the standard can. 15. −3/(t) + C
11. The height and width should be 18 and the length should be 36, 17. tan θ + C
giving a volume of 11, 664in3 .
√ 19. sec x − csc x + C
13. 5 − 10/ 39 ≈ 3.4 miles should be run underground, giving a
21. 3t / ln 3 + C
minimum cost of $374,899.96.
23. 4/3t3 + 6t2 + 9t + C
15. The dog should run about 19 feet along the shore before star ng
to swim. 25. x6 /6 + C
√
17. The largest area is 2 formed by a square with sides of length 2. 27. ax + C
Sec on 4.4 29. − cos x + 3

31. x4 − x3 + 7
1. T
33. 7x / ln 7 + 1 − 49/ ln 7
3. F
7x3 9x 40
35. 6
− 2
+ 3
37. θ − sin(θ) − π + 4
7. Use y = x2 ; dy = 2x · dx with x = 2 and dx = 0.05. Thus
dy = .2; knowing 22 = 4, we have 2.052 ≈ 4.2. 39. 3x − 2
9. Use y = x3 ; dy = 3x2 · dx with x = 5 and dx = 0.1. Thus 41. dy = (2xex cos x + x2 ex cos x − x2 ex sin x)dx
dy = 7.5; knowing 53 = 125, we have 5.13 ≈ 132.5. Sec on 5.2
√ √
11. Use y = x; dy = 1/(2 √ x) · dx with x =√16 and dx = 0.5. Thus
dy = .0625; knowing 16 = 4, we have 16.5 ≈ 4.0625. 1. Answers will vary.
A.8
3. 0 19. 1045
5. (a) 3 21. −8525
(b) 4 23. 5050
(c) 3 25. 155
(d) 0 27. 24
(e) −4 29. 19
(f) 9 √
31. π/3 + π/(2 3) ≈ 1.954
7. (a) 4 33. 0.388584
(b) 2 (1+n)2
35. (a) Exact expressions will vary; 4n2
.
(c) 4 (b) 121/400, 10201/40000, 1002001/4000000
(d) 2 (c) 1/4
(e) 1
37. (a) 8.
(f) 2 (b) 8, 8, 8
9. (a) π (c) 8
(b) π 39. (a) Exact expressions will vary; 100 − 200/n.
(c) 2π (b) 80, 98, 499/5
(d) 10π (c) 100
11. (a) −59 41. F(x) = 5 tan x + 4
(b) −48 43. G(t) = 4/6t6 − 5/4t4 + 8t + 9
(c) −27 45. G(t) = sin t − cos t − 78
(d) −33
Sec on 5.4
13. (a) 4
(b) 4 1. Answers will vary.
(c) −4 3. T
(d) −2 5. 20
15. (a) 2 /s 7. 0
(b) 2 9. 1
(c) 1.5 11. (5 − 1/5)/ ln 5
17. (a) 64 /s 13. −4
(b) 64 15. 16/3
(c) t = 2 17. 45/4
√
(d) t = 2 + 7 ≈ 4.65 seconds 19. 1/2
19. 2 21. 1/2
21. 16 23. 1/4
23. 24 25. 8
25. −7 27. 0
27. 1/4x4 − 2/3x3 + 7/2x2 − 9x + C 29. Explana ons will vary. A sketch will help.
√
29. 3/4t4/3 − 1/t + 2t / ln 2 +C 31. c = 2/ 3
Sec on 5.3 33. c = ln(e − 1) ≈ 0.54

35. 2/π
1. limits
37. 2
3. Rectangles.
39. 16
5. 22 + 32 + 42 = 29 41. −300
7. 0 − 1 + 0 + 1 + 0 = 0 43. 30
9. 1 + 1/2 + 1/3 + 1/4 + 1/5 = 137/60 45. −1
11. 1/2 + 1/6 + 1/12 + 1/20 = 4/5 47. −64 /s
∑
13. Answers may vary; 5i=1 3i 49. 2 /s
∑
15. Answers may vary; 4i=1 i+1
i
51. 27/2
17. 5 · 10 = 50 53. 9/2
A.9
55. F′ (x) = (3x2 + 1) x3 1+x 17. − 61 sin(3 − 6x) + C
57. F′ (x) = 2x(x2 + 2) − (x + 2) 19. 1

2
ln | sec(2x) + tan(2x)| + C
Sec on 5.5 sin(x2 )
21. 2
+C
1. F 23. The key is to rewrite cot x as cos x/ sin x, and let u = sin x.
1 3x−1
3. They are superseded by the Trapezoidal Rule; it takes an equal 25. 3
e +C
amount of work and is generally more accurate.
1 (x−1)2
27. 2
e +C
5. (a) 3/4
( )
(b) 2/3 29. ln ex +1 +C
(c) 2/3 27x
31. +C
1
√ ln 27
7. (a) (1 + 2)π ≈ 1.896 1
4 33. ln2 (x) + C
1
√ 2
(b) 6
(1 + 2 2)π ≈ 2.005 3
35. 2
(ln x)2 +C
(c) 2
x2
9. (a) 38.5781 37. 2
+ 3x + ln |x| + C
(b) 147/4 ≈ 36.75 39. x3
− x2
+ x − 2 ln |x + 1| + C
3 2
(c) 147/4 ≈ 36.75 3 2
41. 2
− 8x + 15 ln |x + 1| + C
x
11. (a) 0 ( )
√
(b) 0 43. 7 tan−1 √x + C
7
(c) 0 ( )
45. 14 sin−1 √x +C
5
13. Trapezoidal Rule: 0.9006
Simpson’s Rule: 0.90452 47. 5
4
sec−1 (|x|/4) + C
( )
15. Trapezoidal Rule: 13.9604 tan−1 x−1
√
49. √ 7 +C
Simpson’s Rule: 13.9066 7
(
−1 x−4
)
17. Trapezoidal Rule: 1.1703 51. 3 sin 5
+C
Simpson’s Rule: 1.1873
53. − 31 +C
3(x +3)
19. Trapezoidal Rule: 1.0803
√
Simpson’s Rule: 1.077 55. − 1 − x2 + C
( )
21. (a) n = 161 (using max f ′′ (x) = 1) 57. − 32 cos 2 (x) + C
3
( (4) )
(b) n = 12 (using max f (x) = 1) 59. ln |x − 5| + C
( )
23. (a) n = 1004 (using max f ′′ (x) = 39)
61. 3x2 + ln x2 + 3x + 5 − 5x + C
2
( )
(b) n = 62 (using max f (4) (x) = 800)
63. 3 ln 3x2 + 9x + 7 + C
25. (a) Area is 30.8667 cm2 . ( 2)
(b) Area is 308, 667 yd2 .
1
65. 18 tan−1 x9 + C
Chapter 6 67. sec−1 (|2x|) + C

( x−1 )
69. 3 ln x2 − 2x + 10 +
2
1
3
tan−1 3
+C
Sec on 6.1 ( )
41 tan−1 x−5√
71. 15
ln x2 − 10x + 32 + x + √ 7
+ C
1. Chain Rule. 2 7
( )
3. 1 3
(x − 5)8 + C 24 tan−1 x+2 √
8
73. x2
+ 3 ln x2 + 4x + 9 − 4x + √ 5
+C
1
( 2 )9 2 5
5. x +1 +C
18
75. tan−1 (sin(x)) + C
7. 1
ln |2x + 7| + C √
2
√ 77. 3 x2 − 2x − 6 + C
2
9. (x + 3)3/2 − 6(x + 3)1/2 + C = 32 (x − 6) x + 3 + C
3 79. − ln 2
√
11. 2e x + C 81. 2/3
13. − 2x12 − 1
x
+C 83. (1 − e)/2
sin3 (x)
15. 3
+C 85. π/2
A.10
Index
!, 405 Constant Mul ple Rule

Absolute Convergence Theorem, 456 of deriva ves, 84
absolute maximum, 129 of integra on, 201
absolute minimum, 129 of series, 427
Absolute Value Theorem, 410 constrained op miza on, 754
accelera on, 77, 651 con nuous func on, 37, 696
Alterna ng Harmonic Series, 427, 454, 467 proper es, 40, 697
Alterna ng Series Test, 450 vector–valued, 638
aN , 669, 679 contour lines, 684
analy c func on, 488 convergence
angle of eleva on, 656 absolute, 454, 456
an deriva ve, 197 Alterna ng Series Test, 450
of vector–valued func on, 646 condi onal, 454
arc length, 379, 527, 553, 648, 673 Direct Comparison Test, 437
arc length parameter, 673, 675 for integra on, 347
asymptote Integral Test, 434
horizontal, 50 interval of, 462
ver cal, 48 Limit Comparison Test, 438
aT , 669, 679 for integra on, 349
average rate of change, 635 nth –term test, 429
average value of a func on, 777 of geometric series, 422
average value of func on, 244 of improper int., 342, 347, 349
of monotonic sequences, 416
Binomial Series, 489 of p-series, 423
Bisec on Method, 42 of power series, 461
boundary point, 690 of sequence, 408, 413
bounded sequence, 412 of series, 419
convergence, 413 radius of, 462
bounded set, 690 Ra o Comparison Test, 443
Root Comparison Test, 446
center of mass, 791–793, 795, 822 coordinates
Chain Rule, 101 cylindrical, 828
mul variable, 721, 724 polar, 533
nota on, 107 spherical, 831
circle of curvature, 678 cri cal number, 131
circula on, 870 cri cal point, 131, 749–751
closed, 690 cross product
closed disk, 690 and deriva ves, 643
concave down, 151 applica ons, 605
concave up, 151 area of parallelogram, 606
concavity, 151, 524 torque, 608
inflec on point, 152 volume of parallelepiped, 607
test for, 152 defini on, 601
conic sec ons, 498 proper es, 603, 604
degenerate, 498 curl, 853
ellipse, 501 of conserva ve fields, 868
hyperbola, 504 curvature, 675
parabola, 498 and mo on, 679
connected, 865 equa ons for, 677
simply, 865 of circle, 677, 678
conserva ve field, 865, 866, 868 radius of, 678
curve differen al, 189
parametrically defined, 511 nota on, 189
rectangular equa on, 511 Direct Comparison Test
smooth, 517 for integra on, 347
curve sketching, 159 for series, 437
cusp, 517 direc onal deriva ve, 729, 731, 732, 735, 736
cycloid, 633 directrix, 498, 563
cylinder, 563 Disk Method, 364
cylindrical coordinates, 828 displacement, 238, 634, 648
distance
decreasing func on, 142 between lines, 619
finding intervals, 143 between point and line, 619
definite integral, 209 between point and plane, 628
and subs tu on, 278 between points in space, 560
of vector–valued func on, 646 traveled, 659
proper es, 211 divergence, 852, 853
del operator, 851 Alterna ng Series Test, 450
deriva ve Direct Comparison Test, 437
accelera on, 78 for integra on, 347
as a func on, 66 Integral Test, 434
at a point, 62 Limit Comparison Test, 438
basic rules, 82 for integra on, 349
Chain Rule, 101, 107, 721, 724 nth –term test, 429
Constant Mul ple Rule, 84 of geometric series, 422
Constant Rule, 82 of improper int., 342, 347, 349
differen al, 189 of p-series, 423
direc onal, 729, 731, 732, 735, 736 of sequence, 408
exponen al func ons, 107 of series, 419
First Deriv. Test, 145 Ra o Comparison Test, 443
Generalized Power Rule, 102 Root Comparison Test, 446
higher order, 85 Divergence Theorem
interpreta on, 86 in space, 900
hyperbolic funct., 324 in the plane, 877
implicit, 111, 726 dot product
interpreta on, 75 and deriva ves, 643
inverse func on, 122 defini on, 588
inverse hyper., 327 proper es, 589, 590
inverse trig., 125 double integral, 770, 771
logarithmic differen a on, 118 in polar, 781
Mean Value Theorem, 138 proper es, 774
mixed par al, 704
mo on, 78 eccentricity, 503, 507
mul variable differen ability, 713, 718 elementary func on, 248
normal line, 63 ellipse
nota on, 66, 85 defini on, 501
parametric equa ons, 521 eccentricity, 503
par al, 700, 708 parametric equa ons, 517
Power Rule, 82, 95, 116 reflec ve property, 504
power series, 465 standard equa on, 502
Product Rule, 89 extrema
Quo ent Rule, 92 absolute, 129, 749
second, 85 and First Deriv. Test, 145
Second Deriv. Test, 155 and Second Deriv. Test, 155
Sum/Difference Rule, 84 finding, 132
tangent line, 62 rela ve, 130, 749, 750
third, 85 Extreme Value Theorem, 130, 754
trigonometric func ons, 94 extreme values, 129
vector–valued func ons, 639, 640, 643
velocity, 78 factorial, 405
differen able, 62, 713, 718 First Deriva ve Test, 145
first octant, 560 by subs tu on, 265
floor func on, 38 definite, 209
flow, 870, 872 and subs tu on, 278
fluid pressure/force, 397, 399 proper es, 211
flux, 870, 872, 893, 894 Riemann Sums, 232
focus, 498, 501, 504 displacement, 238
Fubini’s Theorem, 771 distance traveled, 659
func on double, 770
of three variables, 687 fluid force, 397, 399
of two variables, 683 Fun. Thm. of Calc., 236, 237
vector–valued, 631 general applica on technique, 353
Fundamental Theorem of Calculus, 236, 237 hyperbolic funct., 324
and Chain Rule, 240 improper, 342, 345, 347, 349
Fundamental Theorem of Line Integrals, 864, 866 indefinite, 197
inverse hyper., 327
Gabriel’s Horn, 384 iterated, 761
Gauss’s Law, 904 Mean Value Theorem, 243
Generalized Power Rule, 102 mul ple, 761
geometric series, 421, 422 nota on, 198, 209, 237, 761
gradient, 731, 732, 735, 736, 746 numerical, 248
and level curves, 732 Le /Right Hand Rule, 248, 255
and level surfaces, 746 Simpson’s Rule, 253, 255, 256
Green’s Theorem, 874 Trapezoidal Rule, 251, 255, 256
of mul variable func ons, 759
Harmonic Series, 427
Head To Tail Rule, 578 of power series, 465
Hooke’s Law, 390 of trig. func ons, 271
hyperbola of trig. powers, 294, 299
defini on, 504 of vector–valued func on, 646
eccentricity, 507 of vector–valued func ons, 646
parametric equa ons, 517 par al frac on decomp., 314
reflec ve property, 507 Power Rule, 202
standard equa on, 505 Sum/Difference Rule, 202
hyperbolic func on surface area, 383, 529, 554
defini on, 321 trig. subst., 305
deriva ves, 324 triple, 808, 819–821
iden es, 324 volume
integrals, 324 cross-sec onal area, 362
inverse, 325 Disk Method, 364
deriva ve, 327 Shell Method, 371, 375
integra on, 327 Washer Method, 366, 375
logarithmic def., 326 with cylindrical coordinates, 829
with spherical coordinates, 833
implicit differen a on, 111, 726 work, 387
improper integra on, 342, 345 interior point, 690
incompressible vector field, 852 Intermediate Value Theorem, 42
increasing func on, 142 interval of convergence, 462
finding intervals, 143 iterated integra on, 761, 770, 771, 808, 819–821
indefinite integral, 197 changing order, 765
of vector–valued func on, 646 proper es, 774, 814
indeterminate form, 2, 49, 335, 336
inflec on point, 152 L’Hôpital’s Rule, 332, 334
ini al point, 574 lamina, 787
ini al value problem, 202 Le Hand Rule, 218, 223, 248
Integral Test, 434 Le /Right Hand Rule, 255
integra on level curves, 684, 732
arc length, 379 level surface, 688, 746
area, 209, 762, 763 limit
area between curves, 241, 354 Absolute Value Theorem, 410
average value, 244 at infinity, 50
by parts, 283 defini on, 10
difference quo ent, 6 level curves, 684
does not exist, 4, 32 level surface, 688
indeterminate form, 2, 49, 335, 336 limit, 691, 692, 698
L’Hôpital’s Rule, 332, 334 range, 683, 687
le handed, 30
of infinity, 46 Newton’s Method, 168
of mul variable func on, 691, 692, 698 norm, 574
of sequence, 408 normal line, 63, 521, 742
of vector–valued func ons, 637 normal vector, 623
one sided, 30 nth –term test, 429
proper es, 18, 692 numerical integra on, 248
pseudo-defini on, 2 Le /Right Hand Rule, 248, 255
right handed, 30 Simpson’s Rule, 253, 255
Squeeze Theorem, 22 error bounds, 256
Limit Comparison Test Trapezoidal Rule, 251, 255
for integra on, 349 error bounds, 256
for series, 438
line integral octant
Fundamental Theorem, 864, 866 first, 560
over scalar field, 841, 843, 859 one to one, 880
over vector field, 860 open, 690
path independent, 865, 866 open ball, 698
proper es over a scalar field, 846 open disk, 690
proper es over a vector field, 863 op miza on, 181
lines, 612 constrained, 754
distances between, 619 orientable, 880
equa ons for, 614 orthogonal, 592, 742
intersec ng, 615 decomposi on, 596
parallel, 615 orthogonal decomposi on of vectors, 596
skew, 615 orthogonal projec on, 594
logarithmic differen a on, 118 oscula ng circle, 678
outer unit normal vector, 900
Möbius band, 881
Maclaurin Polynomial, see Taylor Polynomial p-series, 423
defini on, 474 parabola
Maclaurin Series, see Taylor Series defini on, 498
defini on, 485 general equa on, 499
magnitude of vector, 574 reflec ve property, 501
mass, 787, 788, 822, 847 parallel vectors, 582
center of, 791, 847 Parallelogram Law, 578
maximum parametric equa ons
absolute, 129, 749 arc length, 527
and First Deriv. Test, 145 concavity, 524
and Second Deriv. Test, 155 defini on, 2
511
rela ve/local, 130, 749, 752 finding ddxy2 , 525
Mean Value Theorem finding dydx , 521
of differen a on, 138 normal line, 521
of integra on, 243 of a surface, 880
Midpoint Rule, 218, 223 surface area, 529
minimum tangent line, 521
absolute, 129, 749 parametrized surface, 880
and First Deriv. Test, 145, 155 par al deriva ve, 700, 708
rela ve/local, 130, 749, 752 high order, 708
moment, 793, 795, 822 meaning, 702
monotonic sequence, 414 mixed, 704
mul ple integra on, see iterated integra on second deriva ve, 704
mul variable func on, 683, 687 total differen al, 712, 718
con nuity, 696–698, 714, 719 par on, 225
differen ability, 713, 714, 718, 719 size of, 225
domain, 683, 687 path independent, 865, 866
perpendicular, see orthogonal Root Comparison Test
piecewise smooth curve, 846 for series, 446
planes
coordinate plane, 562 saddle point, 751, 752
distance between point and plane, 628 Second Deriva ve Test, 155, 752
equa ons of, 624 sensi vity analysis, 717
introduc on, 562 sequence
normal vector, 623 Absolute Value Theorem, 410
tangent, 745 posi ve, 437
point of inflec on, 152 sequences
polar boundedness, 412
coordinates, 533 convergent, 408, 413, 416
func on defini on, 405
arc length, 553 divergent, 408
gallery of graphs, 540 limit, 408
surface area, 554 limit proper es, 411
func ons, 536 monotonic, 414
area, 549 series
area between curves, 551 absolute convergence, 454
finding dy
dx , 546
Absolute Convergence Theorem, 456
graphing, 536 alterna ng, 449
polar coordinates, 533 Approxima on Theorem, 452
plo ng points, 533 Alterna ng Series Test, 450
poten al func on, 857, 866 Binomial, 489
Power Rule condi onal convergence, 454
differen a on, 82, 89, 95, 116 convergent, 419
integra on, 202 defini on, 419
power series, 460 Direct Comparison Test, 437
algebra of, 491 divergent, 419
convergence, 461 geometric, 421, 422
deriva ves and integrals, 465 Integral Test, 434
projec le mo on, 656, 657, 670 interval of convergence, 462
Limit Comparison Test, 438
quadric surface Maclaurin, 485
defini on, 566 nth –term test, 429
ellipsoid, 568 p-series, 423
ellip c cone, 567 par al sums, 419
ellip c paraboloid, 567 power, 460, 461
gallery, 567–569 deriva ves and integrals, 465
hyperbolic paraboloid, 569 proper es, 427
hyperboloid of one sheet, 568 radius of convergence, 462
hyperboloid of two sheets, 569 Ra o Comparison Test, 443
sphere, 568 rearrangements, 455, 456
trace, 566 Root Comparison Test, 446
Quo ent Rule, 92 Taylor, 485
telescoping, 424, 425
R, 574 Shell Method, 371, 375
radius of convergence, 462 signed area, 209
radius of curvature, 678 signed volume, 770, 771
Ra o Comparison Test simple curve, 865
for series, 443 simply connected, 865
rearrangements of series, 455, 456 Simpson’s Rule, 253, 255
related rates, 174 error bounds, 256
Riemann Sum, 218, 222, 225 smooth, 642
and definite integral, 232 curve, 517
Right Hand Rule, 218, 223, 248 surface, 880
right hand rule smooth curve
of Cartesian coordinates, 560 piecewise, 846
of the cross product, 605 speed, 651
Rolle’s Theorem, 138 sphere, 561
spherical coordinates, 831 vector field, 850
Squeeze Theorem, 22 conserva ve, 865, 866
Stokes’ Theorem, 906 curl of, 853
Sum/Difference Rule divergence of, 852, 853
of deriva ves, 84 over vector field, 860
of integra on, 202 poten al func on of, 857, 866
of series, 427 vector–valued func on
summa on algebra of, 632
nota on, 219 arc length, 648
proper es, 221 average rate of change, 635
surface, 880 con nuity, 638
smooth, 880 defini on, 631
surface area, 800 deriva ves, 639, 640, 643
of parametrized surface, 886, 887 describing mo on, 651
solid of revolu on, 383, 529, 554 displacement, 634
surface integral, 891 distance traveled, 659
surface of revolu on, 564, 565 graphing, 631
integra on, 646
tangent line, 62, 521, 546, 641 limits, 637
direc onal, 739 of constant length, 645, 655, 656, 665
tangent plane, 745 projec le mo on, 656, 657
Taylor Polynomial smooth, 642
defini on, 474 tangent line, 641
Taylor’s Theorem, 477 vectors, 574
Taylor Series algebra of, 577
common series, 491 algebraic proper es, 580
defini on, 485 component form, 575
equality with genera ng func on, 487 cross product, 601, 603, 604
Taylor’s Theorem, 477 defini on, 574
telescoping series, 424, 425 dot product, 588–590
terminal point, 574 Head To Tail Rule, 578
torque, 608 magnitude, 574
total differen al, 712, 718 norm, 574
sensi vity analysis, 717 normal vector, 623
total signed area, 209 orthogonal, 592
trace, 566 orthogonal decomposi on, 596
Trapezoidal Rule, 251, 255 orthogonal projec on, 594
error bounds, 256 parallel, 582
triple integral, 808, 819–821 Parallelogram Law, 578
proper es, 814 resultant, 578
standard unit vector, 584
unbounded sequence, 412 unit vector, 580, 582
unbounded set, 690 zero vector, 578
unit normal vector velocity, 77, 651
aN , 669 volume, 770, 771, 806
and accelera on, 668, 669
Washer Method, 366, 375
and curvature, 679
work, 387, 599
defini on, 666
in R2 , 668
unit tangent vector
and accelera on, 668, 669
and curvature, 675, 679
aT , 669
defini on, 664
in R2 , 668
unit vector, 580
proper es, 582
standard unit vector, 584
unit normal vector, 666
unit tangent vector, 664
Differen a on Rules
d d x d ( −1 ) 1 d
1. (cx) = c 10. (a ) = ln a · ax 19. sin x = √ 28. (sech x) = − sech x tanh x
dx dx dx 1 − x2 dx
d d 1 d ( −1 ) −1 d
2. (u ± v) = u′ ± v′ 11. (ln x) = 20. cos x = √ 29. (csch x) = − csch x coth x
dx dx x dx 1 − x2 dx
d d 1 1 d ( −1 ) −1 d
3. (u · v) = uv′ + u′ v 12. (loga x) = · 21. csc x = √ 30. (coth x) = − csch2 x
dx dx ln a x dx |x| x2 − 1 dx
( )
d u vu′ − uv′ d d ( −1 ) 1 d ( ) 1
4. = 13. (sin x) = cos x 22. sec x = √ 31. cosh−1 x = √
dx v v2 dx dx |x| x2 − 1 dx x2 − 1
d d d ( −1 ) 1 d ( ) 1
5. (u(v)) = u′ (v)v′ 14. (cos x) = − sin x 23. tan x = 32. sinh−1 x = √
dx dx dx 1 + x2 dx x2 + 1
d d d ( −1 ) −1 d ( ) −1
6. (c) = 0 15. (csc x) = − csc x cot x 24. cot x = 33. sech−1 x = √
dx dx dx 1 + x2 dx x 1 − x2
d d d d ( ) −1
7. (x) = 1 16. (sec x) = sec x tan x 25. (cosh x) = sinh x 34. csch−1 x = √
dx dx dx dx |x| 1 + x2
d n d d d ( ) 1
8. (x ) = nxn−1 17. (tan x) = sec2 x 26. (sinh x) = cosh x 35. tanh−1 x =
dx dx dx dx 1 − x2
d x d d d ( ) 1
9. (e ) = ex 18. (cot x) = − csc2 x 27. (tanh x) = sech2 x 36. coth−1 x =
dx dx dx dx 1 − x2
Integra on Rules
∫ ∫ ∫ ∫ ( )
1 x
1. c · f(x) dx = c f(x) dx 12. tan x dx = − ln | cos x| + C 23. √ dx = sin−1
+C
− a2 x2 a
∫ ∫ ∫ ( )
1 1 |x|
2. f(x) ± g(x) dx = 13. sec x dx = ln | sec x + tan x| + C 24. √ dx = sec−1 +C
x x2 − a 2 a a
∫ ∫ ∫ ∫
f(x) dx ± g(x) dx 14. csc x dx = − ln | csc x + cot x| + C 25. cosh x dx = sinh x + C
∫ ∫ ∫
3. 0 dx = C 15. cot x dx = ln | sin x| + C 26. sinh x dx = cosh x + C
∫ ∫ ∫
4. 1 dx = x + C 16. sec2 x dx = tan x + C 27. tanh x dx = ln(cosh x) + C
∫ ∫ ∫
1 n+1
5. xn dx = x + C, n ̸= −1 17. csc2 x dx = − cot x + C 28. coth x dx = ln | sinh x| + C
n+1
∫ ∫ ∫
1 √
6. ex dx = ex + C 18. sec x tan x dx = sec x + C 29. √ dx = ln x + x2 − a2 + C
x2 − a2
∫ ∫ ∫
1 √
7. ln x dx = x ln x − x + C 19. csc x cot x dx = − csc x + C 30. √ dx = ln x + x2 + a2 + C
x2 + a 2
∫ ∫
1 ( ) ∫
8. ax dx = · ax + C 1 1 1 1 a + x
20. cos2 x dx = x + sin 2x + C 31. dx = ln +C
2a a − x
ln a 2 4 a 2 − x2
∫ ∫
1 1 1 ( ) ∫ ( )
9. dx = ln |x| + C 21. sin2 x dx =
x − sin 2x + C 1 1 x
x 2 4 32. √ dx = ln √ +C
∫ x a 2 − x2 a a + a 2 − x2
∫ ( ) ∫
10. cos x dx = sin x + C 1 1 −1 x 1 1 x
22. dx = tan +C 33. √ dx = ln √ +C
∫
x2 + a 2 a a x x2 + a 2 a a + x2 + a 2
11. sin x dx = − cos x + C
The Unit Circle Defini ons of the Trigonometric Func ons
y
Unit Circle Defini on

( √ ) (0, 1) ( √ ) y
−1, 3 1
2,
3
( √ √ 2) 2 2
(√ √ )
− 22 , 22 π/2 2
, 22 (x, y)
2π/3 π/3
2 sin θ = y cos θ = x
( √ ) (√ )
− 23 , 12 3π/4 90◦ π/4
3
, 1
120◦ 60◦ 2 2 y θ 1 1
5π/6 135
◦
45◦ csc θ = sec θ =
π/6
x y x
150◦ 30◦ x
y x
tan θ = cot θ =
x y
(−1, 0) π 180◦ 0◦ 0 (1, 0) x
210◦ 330◦ Right Triangle Defini on

7π/6 11π/6 O H
225◦ 315◦
( √ ) ◦
300◦
(√ )
e sin θ = csc θ =
− 23 , − 2 5π/4 240 us H O
2 ,−2
1 7π/4 1 3
Opposite
270◦ n
( √ √ ) 4π/3 5π/3 (√ √ ) te
− 22 , − 22 ypo A H
2 ,− 2
2 2
( √ ) 3π/2 ( √ )
H cos θ = sec θ =
− 12 , − 23
H A
2,− 2
1 3
θ
(0, −1)
O A
Adjacent tan θ = cot θ =
A O
Common Trigonometric Iden es

Pythagorean Iden es Cofunc on Iden es Double Angle Formulas
(π ) (π )
2
sin x + cos x = 12
sin − x = cos x csc − x = sec x sin 2x = 2 sin x cos x
2 2
2
tan x + 1 = sec x 2 (π ) (π) cos 2x = cos2 x − sin2 x
cos − x = sin x sec − x = csc x
1 + cot2 x = csc2 x 2 2 = 2 cos2 x − 1
(π ) (π )
tan − x = cot x cot − x = tan x = 1 − 2 sin2 x
2 2
2 tan x
tan 2x =
1 − tan2 x
Sum to Product Formulas Power–Reducing Formulas Even/Odd Iden es

( ) ( )
x+y x−y 1 − cos 2x sin(−x) = − sin x
sin x + sin y = 2 sin cos sin2 x =
2 2 2
( ) ( ) cos(−x) = cos x
x−y x+y 2 1 + cos 2x
sin x − sin y = 2 sin cos cos x = tan(−x) = − tan x
2 2 2
( ) ( ) 1 − cos 2x csc(−x) = − csc x
x+y x−y tan2 x =
cos x + cos y = 2 cos cos 1 + cos 2x
2 2 sec(−x) = sec x
( ) ( )
x+y x−y cot(−x) = − cot x
cos x − cos y = −2 sin sin
2 2
Product to Sum Formulas Angle Sum/Difference Formulas

1( )
sin(x ± y) = sin x cos y ± cos x sin y
sin x sin y = cos(x − y) − cos(x + y)
2
1( ) cos(x ± y) = cos x cos y ∓ sin x sin y
cos x cos y = cos(x − y) + cos(x + y) tan x ± tan y
2 tan(x ± y) =
1( ) 1 ∓ tan x tan y
sin x cos y = sin(x + y) + sin(x − y)
2
Areas and Volumes
Triangles Right Circular Cone
h = a sin θ c Volume = 13 πr2 h
h a h
Area = 12 bh Surface Area =
θ √ r
Law of Cosines: πr r2 + h2 + πr2
b
c2 = a2 + b2 − 2ab cos θ
Parallelograms Right Circular Cylinder

r
Area = bh Volume = πr2 h
h
Surface Area = h
b 2πrh + 2πr2
Trapezoids Sphere
1
a
Area = 2 (a + b)h Volume = 43 πr3
r
h Surface Area =4πr2
Circles General Cone

Area = πr2 Area of Base = A
r h
Circumference = 2πr Volume = 13 Ah
A
Sectors of Circles General Right Cylinder

θ in radians s Area of Base = A
1 2
Area = 2 θr Volume = Ah
θ h
s = rθ r
A
Algebra
Factors and Zeros of Polynomials

Let p(x) = an xn + an−1 xn−1 + · · · + a1 x + a0 be a polynomial. If p(a) = 0, then a is a zero of the polynomial and a solu on of
the equa on p(x) = 0. Furthermore, (x − a) is a factor of the polynomial.
Fundamental Theorem of Algebra

An nth degree polynomial has n (not necessarily dis nct) zeros. Although all of these zeros may be imaginary, a real
polynomial of odd degree must have at least one real zero.
Quadra c Formula √
If p(x) = ax2 + bx + c, and 0 ≤ b2 − 4ac, then the real zeros of p are x = (−b ± b2 − 4ac)/2a
Special Factors
x2 − a2 = (x − a)(x + a) x3 − a3 = (x − a)(x2 + ax + a2 )
x + a = (x + a)(x − ax + a )
3 3 2 2
x4 − a4 = (x2 − a2 )(x2 + a2 )
n(n−1)
(x + y)n = xn + nxn−1 y + 2! xn−2 y2 + · · · + nxyn−1 + yn
(x − y)n = xn − nxn−1 y + n(n−1)
2! x y − · · · ± nxyn−1 ∓ yn
n−2 2
Binomial Theorem
(x + y) = x + 2xy + y2
2 2
(x − y)2 = x2 − 2xy + y2
(x + y)3 = x3 + 3x2 y + 3xy2 + y3 (x − y)3 = x3 − 3x2 y + 3xy2 − y3
(x + y)4 = x4 + 4x3 y + 6x2 y2 + 4xy3 + y4 (x − y)4 = x4 − 4x3 y + 6x2 y2 − 4xy3 + y4
Ra onal Zero Theorem

If p(x) = an xn + an−1 xn−1 + · · · + a1 x + a0 has integer coefficients, then every rational zero of p is of the form x = r/s,
where r is a factor of a0 and s is a factor of an .
Factoring by Grouping
acx3 + adx2 + bcx + bd = ax2 (cs + d) + b(cx + d) = (ax2 + b)(cx + d)
Arithme c Opera ons

a c ad + bc a+b a b
ab + ac = a(b + c) + = = +
b d bd c c c
(a) (a)
(a) (d) ad a a ac
( bc ) = = b = ( )=
b c bc c bc b b
d c
( )
b ab a−b b−a ab + ac
a = = =b+c
c c c−d d−c a
Exponents and Radicals

√ ax √
a0 = 1, a ̸= 0 (ab)x = ax bx ax ay = ax+y a = a1/2 = ax−y n
a = a1/n
ay
( a )x √ √
ax √ −x 1 √ √ √ a n
a
= x n
am =a m/n
a = x n
ab = n a n b x y
(a ) = a xy n
=√
b b a b n
b
Addi onal Formulas
Summa on Formulas:
∑n ∑
n
n(n + 1)
c = cn i=
2
∑
n
i=1 i=1
∑n ( )2
n(n + 1)(2n + 1) n(n + 1)
i2 = 3
i =
i=1
6 i=1
2
Trapezoidal
∫
Rule:
b
∆x [ ]
f(x) dx ≈ f(x1 ) + 2f(x2 ) + 2f(x3 ) + ... + 2f(xn ) + f(xn+1 )
a 2
(b − a)3 [ ]
with Error ≤ 2
max f ′′ (x)
12n
Simpson’s
∫
Rule:
b
∆x [ ]
f(x) dx ≈ f(x1 ) + 4f(x2 ) + 2f(x3 ) + 4f(x4 ) + ... + 2f(xn−1 ) + 4f(xn ) + f(xn+1 )
a 3
(b − a)5 [ ]
with Error ≤ 4
max f (4) (x)
180n
Arc Length: Surface of Revolu on:

∫ ∫ √
b √ b
L= 1 + f ′ (x)2 dx S = 2π f(x) 1 + f ′ (x)2 dx
a a
(where f(x) ≥ 0)
∫ b √
S = 2π x 1 + f ′ (x)2 dx
a
(where a, b ≥ 0)
Work Done by a Variable Force: Force Exerted by a Fluid:

∫ b ∫ b
W= F(x) dx F= w d(y) ℓ(y) dy
a a
Taylor Series Expansion for f(x):

f ′′ (c) f ′′′ (c) f (n) (c)
pn (x) = f(c) + f ′ (c)(x − c) + (x − c)2 + (x − c)3 + ... + (x − c)n
2! 3! n!
Maclaurin Series Expansion for f(x), where c = 0:

f ′′ (0) 2 f ′′′ (0) 3 f (n) (0) n
pn (x) = f(0) + f ′ (0)x + x + x + ... + x
2! 3! n!
Summary of Tests for Series:
Condi on(s) of Condi on(s) of
Test Series Comment
Convergence Divergence
∞
∑ This test cannot be used to
nth-Term an lim an ̸= 0
n→∞ show convergence.
n=1
∑∞
1
Geometric Series rn |r| < 1 |r| ≥ 1 Sum =
n=0
1−r
∞
( a )
∑ ∑
Telescoping Series (bn − bn+a ) lim bn = L Sum = bn − L
n→∞
n=1 n=1
∑∞
1
p-Series p>1 p≤1
n=1
(an + b)p
∫ ∞ ∫ ∞
∞
∑ a(n) dn a(n) dn an = a(n) must be
Integral Test an 1 1
con nuous
n=0 is convergent is divergent
∞
∑ ∞
∑
∞
∑ bn bn
Direct Comparison an n=0 n=0
n=0 converges and diverges and

0 ≤ a n ≤ bn 0 ≤ bn ≤ a n
∞
∑ ∞
∑
∞
bn bn
∑ n=0 n=0
Also diverges if
Limit Comparison an lim an /bn = ∞
converges and diverges and n→∞
n=0
lim an /bn ≥ 0 lim an /bn > 0
n→∞ n→∞
∞
{an } must be posi ve
∑ an+1 an+1
Ra o Test an lim <1 lim >1 Also diverges if
n→∞ an n→∞ an
n=0 lim an+1 /an = ∞
n→∞
{an } must be posi ve

∞
∑ ( )1/n ( )1/n
Root Test an lim an <1 lim an >1 Also diverges if
n→∞ n→∞ ( )1/n
n=0 lim an =∞
n→∞

APEX Calculus1 Version 4

Uploaded by

Copyright:

Available Formats

APEX Calculus1 Version 4

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

APEX Calculus1 Version 4

Uploaded by

Copyright:

Available Formats

APEX C I

Gregory Hartman, Ph.D.

Virginia Military Ins tute

Virginia Military Ins tute

Brian Heinold, Ph.D.

Mount Saint Mary’s University

Dimplekumar Chalishajar, Ph.D.

The College of Wooster

Table of Contents iii

3 The Graphical Behavior of Func ons 129

4 Applica ons of the Deriva ve 167

6 Techniques of An diﬀeren a on 263

A Solu ons To Selected Problems A.1

For Students: How to Read this Text

Mathema cs textbooks have a reputa on for being hard to read. High–level

Interac ve, 3D Graphics

• More useful numbering of Examples, Theorems, etc. “Deﬁni on 11.4.2”

1.1 An Introduc on To Limits

lim y = lim f(x) = L.

Figure 1.1.3: Values of sin(x)/x with x

Later we will be able to prove that the limit is exactly 1.

S Again we graph f(x) and create a table of its values near x =

1. The func on f(x) may approach diﬀerent values on either side of c.

2. The func on may grow without upper or lower bound as x approaches c.

3. The func on may oscillate as x approaches c without approaching a spe-

We’ll explore each of these in turn. y

Example 1.1.3 Diﬀerent Values Approached From Le and Right

S A graph of f(x) around x = 1 and a table are given in Figures . x

does not exist.

Example 1.1.5 The Func on Oscillates . x

Figure 1.1.12: Values of f(x) near x = 1 in

0, f(x) approaches 0. However, Figure 1.1.13(b) zooms in on sin(1/x), on the

(a) (b) (c)

Limits of Diﬀerence Quo ents

through those two points. See Figure 1.1.14.

f(1 + h) − f(1) f(1 + h) − f(1)

is as h approaches 0. That is,

Proper understanding of limits is key to understanding calculus. With limits,

Figure 1.1.16: The diﬀerence quo ent

15. lim f(x), where

sin x 16. lim f(x), where

Problems In Exercises 17 – 24, a func on f and a value a are

7. lim x2 + 3x − 5 17. f(x) = −7x + 2, a = 3

18. f(x) = 9x + 0.06, a = −1

1.2 Epsilon-Delta Deﬁni on of a Limit

1. if “y tends to L” as “x tends to c.”

2. if “y approaches L” as “x approaches c.”

3. if “y is near L” whenever “x is near c.”

The deﬁni on we describe in this sec on comes from formalizing 3. A quick

3′ . If x is within a certain tolerance level of c, then the corresponding value y =

3′′ . If x is within δ units of c, then the corresponding value of y is within ε units

We can write “x is within δ units of c” mathema cally as

|x − c| < δ, which is equivalent to c − δ < x < c + δ.

|x − c| < δ −→ |y − L| < ε or c − δ < x < c + δ −→ L − ε < y < L + ε.

Deﬁni on 1.2.1 The Limit of a Func on f

With ε = 0.5, we pick any δ < 1.75.

Figure 1.2.1: Illustra ng the ε−δ process.

Let’s try another value of ε.

3.9601 < x < 4.0401.

Example 1.2.2 Evalua ng a limit using the deﬁni on