Reform Calculus: Part I

Marcel B. Finan
Arkansas Tech University
All
c Rights Reserved

1
 PREFACE
This supplement consists of my lectures of a freshmen-level mathematics class
offered at Arkansas Tech University. The lectures are designed to accompany
the textbook ”Calculus: Single and Multivariable” written by the Harvard
Consortium.
This book has been written in a way that can be read by students. That is,
the text represents a serious effort to produce exposition that is accessible to
a student at the freshmen or high school levels.
The lectures cover sections from Chapters 1, 2, 3, 4, 5, and 6. These chap-
ters are basically well suited for a 4-hour one semester course in a first course
Calculus.

Marcel B. Finan
January 2003

2
Contents
1 Functions and Change 4

2 Exponential Functions 17

3 Building New Functions from Old Ones 25

4 Logarithmic Functions 36

5 Trigonometric Functions 42

6 Powers, Polynomial, and Rational Functions 58

7 Velocity as a Rate of Change 74

8 The Formal Definition of Limit 79

9 The Concept of Continuity 87

10 The Derivative at a Point 92

11 The Derivative Function 95

12 Leibniz Notation for The Derivative 98

13 The Second Derivative 100

14 Continuity and Differentiability 103

15 Derivatives of Power and Polynomial Functions 107

16 Derivatives of Exponential Functions 110

17 The Product and Quotient Rules 112

18 The Chain Rule 114

19 Derivatives of Trigonometric Functions 117

20 Applications of the Chain Rule 120

3
21 Implicit Differentiation 123

22 Parametric Equations 126

23 Linear Approximations and Differentials 131

24 Indeterminate Forms and L’Hôpital’s Rule 135

25 Using First and Second Derivatives 138

26 Special Families of Curves 145

27 Global Maxima and Minima 148

28 Applications to Economics 151

29 Mathematical Modeling and Optimization 159

30 Hyperbolic Functions 162

31 The Mean Value Theorem 165

32 Measuring The Distance Traveled 168

33 The Definite Integral 174

34 Interpretations of the Definite Integral 179

35 Theorems About Definite Integrals 182

36 Finding Antiderivatives Graphically and Numerically 185

37 Analytical Construction of Antiderivatives 189

38 Applications of Antiderivatives 193

39 The Second Fundamental Theorem of Calculus 198

4
1 Functions and Change
Functions play a crucial role in mathematics. A function describes how one
quantity depends on others. More precisely, when we say that a quantity y
is a function of a quantity x we mean a rule that assigns to every possible
value of x exactly one value of y. We call x the input and y the output. In
function notation we write

y = f (x)

Since y depends on x it makes sense to call x the independent variable

and y the dependent variable.
In applications of mathematics, functions are often representations of real
world phenomena. Thus, the functions in this case are referred to as math-
ematical models. If the set of input values is a finite set then the models
are known as discrete models. Otherwise, the models are known as con-
tinuous models. For example, if H represents the temperature after t hours
for a specific day, then H is a discrete model. If A is the area of a circle of
radius r then A is a continuous model.
There are four common ways in which functions are presented and used: By
words, by tables, by graphs, and by formulas.

Example 1.1
The sales tax on an item is 6%. So if p denotes the price of the item and C
the total cost of buying the item then if the item is sold at $ 1 then the cost
is 1 + (0.06)(1) = $1.06 or C(1) = $1.06. If the item is sold at $2 then the
cost of buying the item is 2 + (0.06)(2) = $2.12, or C(2) = $2.12, and so on.
Thus we have a relationship between the quantities C and p such that each
value of p determines exactly one value of C. In this case, we say that C is
a function of p. Describes this function using words, a table, a graph, and a
formula.

Solution.
•Words: To find the total cost, multiply the price of the item by 0.06 and
add the result to the price.
•Table: The chart below gives the total cost of buying an item at price p as
a function of p for 1 ≤ p ≤ 6.
p 1 2 3 4 5 6
C 1.06 2.12 3.18 4.24 5.30 6.36

5
•Graph: The graph of the function C is obtained by plotting the data in
the above table. See Figure 1.
•Formula: The formula that describes the relationship between C and p is
given by
C(p) = 1.06p.

Figure 1

Emphasis of the Four Representations

A formula has the advantage of being both compact and precise. However,
not much insight can be gained from a formula as from a table or a graph.
A graph provides an overall view of a function and thus makes it easy to
deduce important properties. Tables often clearly show trends that are not
easily discerned from formulas, and in many cases tables of values are much
easier to obtain than a formula.
Now, most of the functions that we will encounter in this course have for-
mulas. For example, the area A of a circle is a function of its radius r. In
function notation, we write A(r) = πr2 . However, there are functions that
can not be represented by a formula. For example, the value of Dow Jones
Industrial Average at the close of each business day. In this case the value
depends on the date, but there is no known formula. Functions of this nature,
are mostly represented by either a graph or a table of numerical data.

6
Example 1.2
The table below shows the daily low temperature for a one-week period in
New York City during July.
(a) What was the low temperature on July 19?
(b) When was the low temperature 73◦ F ?
(c) Is the daily low temperature a function of the date?Explain.
(d) Can you express T as a formula?

D 17 18 19 20 21 22 23
T 73 77 69 73 75 75 70

Solution.
(a) The low temperature on July 19 was 69◦ F.
(b) On July 17 and July 20 the low temperature was 73◦ F.
(c) T is a function of D since each value of D determines exactly one value
of T.
(d) T can not be expressed by an exact formula.

So far, we have introduced rules between two quantities that define func-
tions. Unfortunately, it is possible for two quantities to be related and yet
for neither quantity to be a function of the other.

Example 1.3
Let x and y be two quantities related by the equation

x2 + y 2 = 4.

(a) Is x a function of y? Explain.

(b) Is y a function of x? Explain.

Solution.
(a) For y = 0 we have two values of x, namely, x = −2 and x = 2. So x is
not a function of y.
(b) For x = 0 we have two values of y, namely, y = −2 and y = 2. So y is
not a function of x.

Next, suppose that the graph of a relationship between two quantities x

and y is given. To say that y is a function of x means that for each value of
x there is exactly one value of y. Graphically, this means that each vertical

7
line must intersect the graph at most once. Hence, to determine if a graph
represents a function one uses the following test:

Vertical Line Test: A graph is a function if and only if every vertical

line crosses the graph at most once.

According to the vertical line test and the definition of a function, if a ver-
tical line cuts the graph more than once, the graph could not be the graph
of a function since we have multiple y values for the same x-value and this
violates the definition of a function.

Example 1.4
Which of the graphs (a), (b), (c) in Figure 2 represent y as a function of x?

Figure 2

Solution.
By the vertical line test, (b) represents a function whereas (a) and (c) fail to
represent functions since one can find a vertical line that intersects the graph
more than once.

8
Domain and Range of a Function
If we√try to find the possible input values that can be used in the function
y = x − 2 we see that we must restrict x to the interval [2, ∞), that is
x ≥ 2. Similarly, the function y = x12 takes only certain values for the out-
put, namely, y > 0. Thus, a function is often defined for certain values of x
and the dependent variable often takes certain values.
The above discussion leads to the following definitions: By the domain of a
function we mean all possible input values that yield an input value. Graph-
ically, the domain is part of the horizontal axis. The range of a function is
the collection of all possible output values. The range is part of the vertical
axis.
The domain and range of a function can be found either algebraically or
graphically.

• Finding the Domain and the Range Algebraically

When finding the domain of a function, ask yourself what values can’t be
used. Your domain is everything else. There are simple basic rules to con-
sider:

• The domain of all polynomial functions, i.e. functions of the form f (x) =
an xn + an−1 xn−1 + · · · + a1 x + a0 , where n is nonnegative integer, is the Real
numbers IR.
• Square root functions can not contain a negative underneath the radical.
Set the expression under the radical greater than or equal to zero and solve
for the variable. This will be your domain.
• Fractional functions, i.e. ratios of two functions, determine for which input
values the numerator and denominator are not defined and the domain is
everything else. For example, make sure not to divide by zero!

Example 1.5
Find, algebraically, the domain and the range of each of the following func-
tions. Write your answers in interval notation:

(a) y = πx2 (b) y = √1 (c) y = 2 + x1 .

x−4

Solution.
(a) Since the function is a polynomial then its domain is the interval (−∞, ∞).
To find the range, solve the given equation for x in terms of y obtaining

9
 q
y
x=± π
. Thus, x exists for y ≥ 0. So the range is the interval [0, ∞).

1
(b) The domain of y = √x−4 consists of all numbers x such that x − 4 > 0 or
x > 4. That is, the interval (4, ∞). To find the range, we solve for x in terms
of y > 0 obtaining x = 4 + y12 . x exists for all y > 0. Thus, the range is the
interval (0, ∞).

(c) The domain of y = 2 + x1 is the interval (−∞, 0) ∪ (0, ∞). To find

1
the range, write x in terms of y to obtain x = y−2 . The values of y for
which this later formula is defined is the range of the given function, that is,
(−∞, 2) ∪ (2, ∞).

Remark 1.1
Note that the domain of the function y = πx2 of the previous problem consists
of all real numbers. If this function is used to model a real-world situation,
that is, if the x stands for the radius of a circle and y is the corresponding
area then the domain of y in this case consists of all numbers x ≥ 0. In
general, for a word problem the domain is the set of all x values such that
the problem makes sense.

Finding the Domain and the Range Graphically

We often use a graphing calculator to find the domain and range of functions.
In general, the domain will be the set of all x values that has corresponding
points on the graph. We note that if there is an asymptote (shown as a
vertical line on the TI series) we do not include that x value in the domain.
To find the range, we seek the top and bottom of the graph. The range will
be all points from the top to the bottom (minus the breaks in the graph).

Example 1.6
Use a graphing calculator to find the domain and the range of each of the
following functions. Write your answers in interval notation:

(a) y = πx2 (b) y = √1 (c) y = 2 + x1 .

x−4

Solution.
(a) The graph of y = πx2 is given in Figure 3.

10
 Figure 3
The domain is the set (−∞, ∞) and the range is [0, ∞).
1
(b) The graph of y = √x−4 is given in Figure 4.

Figure 4
The domain is the set (4, ∞) and the range is (0, ∞).
(c) The graph of y = 2 + x1 is given in Figure 5.

Figure 5

11
The domain is the set (−∞, 0) ∪ (0, ∞) and the range is (−∞, 2) ∪ (2, ∞).

Average Rate of Change

Functions given by tables of values have their limitations in that nearly al-
ways leave gaps. One way to fill these gaps is by using the average rate
of change. For example, Table 1 below gives the population of the United
States between the years 1950 - 1990.
d(year) 1950 1960 1970 1980 1990
N(in millions) 151.87 179.98 203.98 227.23 249.40
Table 1

This table does not give the population in 1972. One way to estimate
N (1972), is to find the average yearly rate of change of N from 1970 to
1980 given by
227.23 − 203.98
= 2.325 million people per year.
10
Then,
N (1972) = N (1970) + 2(2.325) = 208.63 million.

Average rates of change can be calculated not only for functions given by
tables but also for functions given by formulas. The average rate of change
of a function y = f (x) from x = a to x = b is given by the difference
quotient

∆y Change in function value f (b) − f (a)

= = .
∆x Change in x value b−a
Geometrically, this quantity represents the slope of the secant line going
through the points (a, f (a)) and (b, f (b)) on the graph of f (x). See Figure 6.
The average rate of change of a function on an interval tells us how much
the function changes, on average, per unit change of x within that interval.
On some part of the interval, f may be changing rapidly, while on other
parts f may be changing slowly. The average rate of change evens out these
variations.

12
 Figure 6

Example 1.7
Find the average value of the function f (x) = x2 from x = 3 to x = 5.

Solution.
The average rate of change is

∆y f (5) − f (3) 25 − 9
= = = 8.
∆x 5−3 2
Average Rate of Change and Increasing/Decreasing Functions
Now, we would like to use the concept of the average rate of change to test
whether a function is increasing or decreasing on a specific interval. First,
we introduce the following definition: We say that a function is increasing
if its graph climbs as x moves from left to right. That is, the function values
increase as x increases. It is said to be decreasing if its graph falls as x
moves from left to right. This means that the function values decrease as x
increases.

As an application of the average rate of change, we can use such quantity

to decide whether a function is increasing or decreasing. If a function f is
increasing on an interval I then by taking any two points in the interval I,
say a < b, we see that f (a) < f (b) and in this case

f (b) − f (a)
> 0.
b−a

13
Going backward with this argument we see that if the average rate of change
is positive in an interval then the function is increasing in that interval.
Similarly, if the average rate of change is negative in an interval I then the
function is decreasing there.

Some functions can be increasing on some intervals and decreasing on other

intervals. These intervals can often be identified from the graph.

Example 1.8
Determine the intervals where the function is increasing and decreasing. See
Figure 7.

Figure 7

Solution.
The function is increasing on (−∞, −1) ∪ (1, ∞) and decreasing on the in-
terval (−1, 1).

Linear Functions
In general, the average rate of change of a function is different on different
intervals. For example, consider the function f (x) = x2 . The average rate of
change of f (x) on the interval [0, 1] is
f (1) − f (0)
= 1.
1−0
The average rate of change of f (x) on [1, 2] is
f (2) − f (1)
= 3.
2−1

14
A linear function is a function with the property that the average rate of
change on any interval is the same. We say that y is changing at a constant
rate with respect to x. Thus, y changes by the some amount for every unit
change in x. Geometrically, the graph is a straight line ( and thus the term
linear).

Now, suppose that f (x) is a linear function of x. Then f changes at a constant

rate m. That is, if we pick two points (0, f (0)) and (x, f (x)) then
f (x) − f (0)
m= .
x−0
That is, f (x) = mx + f (0). This is the function notation of the linear
function f (x). Another notation is the equation notation, y = mx + f (0).
We will denote the number f (0) by b. In this case, the linear function will be
written as f (x) = mx+b or y = mx+b. Since b = f (0) then the point (0, b) is
the point where the line crosses the vertical line. We call it the y-intercept.
So the y-intercept is the output corresponding to the input x = 0, sometimes
known as the initial value of y.
If we pick any two points (x1 , y1 ) and (x2 , y2 ) on the graph of f (x) = mx + b
then we must have
y2 − y1
m= .
x2 − x1
We call m the slope of the line.
Example 1.9
The value of a new computer equipment is $20,000 and the value drops at a
constant rate so that it is worth $ 0 after five years. Let V (t) be the value
of the computer equipment t years after the equipment is purchased.

(a) Find the slope m and the y-intercept b.

(b) Find a formula for V (t).
Solution.
(a) Since V (0) = 20, 000 and V (5) = 0 then the slope of V (t) is
0 − 20, 000
m= = −4, 000
5−0
and the vertical intercept is V (0) = 20, 000.
(b) A formula of V (t) is V (t) = −4, 000t + 20, 000. In financial terms, the

15
function V (t) is known as the straight-line depreciation function.

The Significance of the Parameters m and b

Next, we consider the significance of the parameters m and b in the slope-
intercept form of a linear funtion y = mx + b.
We have seen that the graph of a linear function f (x) = mx + b is a straight
line. But a line can be horizontal, vertical, rising to the right or falling to
the right. The slope is the parameter that provides information about the
steepness of a straight line.

• If m = 0 then f (x) = b is a constant function whose graph is a hori-

zontal line at (0, b).
• For a vertical line, the slope is undefined since any two points on the line
have the same x-value and this leads to a division by zero in the formula for
the slope. The equation of a vertical line has the form x = a.
• Suppose that the line is neither horizontal nor vertical. If m > 0 then by
Section 3, f (x) is increasing. That is, the line is rising to the right.
• If m < 0 then f (x) is decreasing. That is, the line is falling to the right.
• The slope, m, tells us how fast the line is climbing or falling. The larger
the value of m the more the line rises and the smaller the value of m the
more the line falls.
The parameter b tells us where the line crosses the vertical axis.
Example 1.10
Arrange the slopes of the lines in Figure 8 from largest to smallest.

Figure 8

16
Solution.
According to Figure 8 we have

mG > m F > m D > m A > m E > m B > m C .

Proportionality
Consider the linear function y = mx + b. If b = 0 then we say that the quan-
tity y is directly proportional to x with constant of proportionality
m. If y = xk then in this case we say that y is inversely proportional to x.

Recommended Problems (pp. 7 - 9): 3, 5, 8, 10, 12, 14, 17, 19, 20,
21, 23, 25, 27, 30, 32, 33, 35, 37, 38.

17
2 Exponential Functions
Exponential functions appear in many applications such as population growth,
radioactive decay, and interest on bank loans.
Recall that linear functions are functions that change at a constant rate. For
example, if f (x) = mx + b then f (x + 1) = m(x + 1) + b = f (x) + m. So
when x increases by 1, the y value increases by m. In contrast, an exponential
function with base a is one that changes by constant multiples of a. That is,
f (x + 1) = af (x). Writing a = 1 + r we obtain f (x + 1) = f (x) + rf (x).
Thus, an exponential function is a function that changes at a constant per-
cent rate.
Exponential functions are used to model increasing quantities such as pop-
ulation growth problems.

Example 2.1
Suppose that you are observing the behavior of cell duplication in a lab. In
one experiment, you started with one cell and the cells doubled every minute.
That is, the population cell is increasing at the constant rate of 100%. Write
an equation to determine the number (population) of cells after one hour.

Solution.
Table 2 below shows the number of cells for the first 5 minutes. Let P (t) be
the number of cells after t minutes.
t 0 1 2 3 4 5
P(t) 1 2 4 8 16 32
Table 2

At time 0, i.e t=0, the number of cells is 1 or 20 = 1. After 1 minute, when

t = 1, there are two cells or 21 = 2. After 2 minutes, when t = 2, there are 4
cells or 22 = 4.
Therefore, one formula to estimate the number of cells (size of population)
after t minutes is the equation (model)

f (t) = 2t .

18
It follows that f (t) is an increasing function. Computing the rates of change
to obtain
f (1)−f (0)
1−0
= 1
f (2)−f (1)
2−1
= 2
f (3)−f (2)
3−2
= 4
f (4)−f (3)
4−3
= 8
f (5)−f (4)
5−4
= 16.
Thus, the rate of change is increasing. Geometrically, this means that the
graph of f (t) is concave up. See Figure 9.

Figure 9

Now, to determine the number of cells after one hour we convert to minutes
to obtain t = 60 minutes so that f (60) = 260 = 1.15 × 1018 cells.

Exponential functions can also model decreasing quantities known as de-

cay models.

Example 2.2
If you start a biology experiment with 5,000,000 cells and 45% of the cells
are dying every minute, how long will it take to have less than 50,000 cells?

Solution.
Let P (t) be the number of cells after t minutes. Then P (t + 1) = P (t) −
45%P (t) or P (t + 1) = 0.55P (t). By constructing a table of data we find

19
 t P(t)
0 5,000,000
1 2,750,000
2 1,512,500
3 831,875
4 457,531.25
5 251,642.19
6 138,403.20
7 76,121.76
8 41,866.97

So it takes 8 minutes for the population to reduce to less than 50,000 cells.
A formula of P (t) is P (t) = 5, 000, 000(0.55)t . The graph of P (t) is given in
Figure 10.

Figure 10

From the previous two examples, we see that an exponential function has the
general form
P (t) = b · at , a > 0 a 6= 1.
Since b = P (0) then we call b the initial value. We call a the base of P (t).
If a > 1, then P (t) shows exponential growth with growth factor a. The
graph of P will be similar in shape to that in Figure 27.
If 0 < a < 1, then P shows exponential decay with decay factor a. The
graph of P will be similar in shape to that in Figure 28.
Since P (t + 1) = aP (t) then P (t + 1) = P (t) + rP (t) where r = a − 1. We
call r the percent growth rate.

20
Remark 2.1
Why a is restricted to a > 0 and a 6= 1? Since t is allowed to have
√ any value
then a negative a will create meaningless expressions such as a (if t = 12 ).
Also, for a = 1 the function P (t) = b is called a constant function and its
graph is a horizontal line.

Doubling Time
In some exponential models one is interested in finding the time for an expo-
nential growing quantity to double. We call this time the doubling time.
To find it, we start with the equation b · at = 2b or at = 2. Solving for t we
log 2
find t = log a
.

Example 2.3
Find the doubling time of a population growing according to P = P0 e0.2t .

Solution.
Setting the equation P0 e0.2t = 2P0 and dividing both sides by P0 to obtain
e0.2t = 2. Take ln of both sides to obtain 0.2t = ln 2. Thus, t = ln
0.2
2
≈ 3.47.

Half-Life
On the other hand, if a quantity is decaying exponentially then the time
required for the initial quantity to reduce into half is called the half-life. To
find it, we start with the equation bat = 2b and we divide both sides by b
to obtain at = 0.5. Take the log of both sides to obtain t log a = log (0.5).
Solving for t we find t = loglog(0.5)
a
.

Example 2.4
The half-life of Iodine-123 is about 13 hours. You begin with 50 grams of
this substance. What is a formula for the amount of Iodine-123 remaining
after t hours?

Solution.
Since the problem involves exponential decay then if Q(t) is the quantity
remaining after t hours then Q(t) = 50at with 0 < a < 1. But Q(13) = 25.
1
That is, 50a13 = 25 or a13 = 0.5. Thus a = (0.5) 13 ≈ 0.95 and Q(t) =
50(0.95)t .

21
The Effect of the Parameters a and b
Recall that an exponential function with base a and initial value b is a func-
tion of the form f (x) = b · ax . In this section, we assume that b > 0. Since
b = f (0) then (0, b) is the vertical intercept of f (x). In this section we con-
sider graphs of exponential functions.

Let’s see the effect of the parameter b on the graph of f (x) = bax .
Example 2.5
Graph, on the same axes, the exponential functions f1 (x) = 2·(1.1)x , f2 (x) =
(1.1)x , and f3 (x) = 0.75(1.1)x .
Solution.
The three functions as shown in Figure 11.

Figure 11
Note that these functions have the same growth factor but different b and
therefore different vertical intercepts.

We know that the slope of a linear function measures the steepness of the
graph. Similarly, the parameter a measures the steepness of the graph of an
exponential function. First, we consider the effect of the growth factor on
the graph.
Example 2.6
Graph, on the same axes, the exponential functions f1 (x) = 4x , f2 (x) = 3x ,
and f3 (x) = 2x .

22
Solution.
Using a graphing calculator we find

Figure 12

It follows that the greater the value of a, the more rapidly the graph rises.
That is, the growth factor a affects the steepness of an exponential function.
Also note that as x decreases, the function values approach the x-axis. Sym-
bolically, as x → −∞, y → 0.

Next, we study the effect of the decay factor on the graph.

Example 2.7  x
Graph, on the same axes, the exponential functions f1 (x) = 2−x = 21 , f2 (x) =
3−x , and f3 (x) = 4−x .

Solution.
Using a graphing calculator we find

23
 Figure 13
It follows that the smaller the value of a, the more rapidly the graph falls.
Also as x increases, the function values approach the x-axis. Symbolically,
as x → ∞, y → 0.

• General Observations
(i) For a > 1, as x decreases, the function values get closer and closer to 0.
Symbolically, as x → −∞, y → 0. For 0 < a < 1, as x increases, the function
values gets closer and closer to the x-axis. That is, as x → ∞, y → 0. We
call the x-axis, a horizontal asymptote.
(ii) The domain of an exponential function consists of the set of all real num-
bers whereas the range consists of the set of all positive real numbers.
(iii) The graph of f (x) = bax with b > 0 is always concave up.

Exponential Functions with Base e

When writing y = bet then we say that y is an exponential function with
base e. Look at your calculator and locate the key ln. (This is called the
natural logarithm function which will be discussed in the next section) Pick
any positive number of your choice, say c, and compute eln c . What do you
notice? For any positive number c, you notice that eln c = c. Thus, any posi-
tive number a can be written in the form a = ek where k = ln a.

Now, suppose that Q(t) = bat . Then by the above paragraph we can write
a = ek . Thus,
Q(t) = b(ek )t = bekt .

24
Note that if k > 0 then ek > 1 so that Q(t) represents an exponential growth
and if k < 0 then ek < 1 so that Q(t) is an exponential decay.
We call the constant k the continuous growth rate.

Example 2.8
If f (t) = 3(1.072)t is rewritten as f (t) = 3ekt , find k.

Solution.
By comparison of the two functions we find ek = 1.072. Solving this equation
graphically (e.g. using a calculator) we find k ≈ 0.695.

Example 2.9
A population increases from its initial level of 7.3 million at the continuous
rate of 2.2% per year. Find a formula for the population P (t) as a function
of the year t. When does the population reach 10 million?

Solution.
We are given the initial value 7.3 million and the continuous growth rate
k = 0.022. Therefore, P (t) = 7.3e0.022t . Next,we want to find the time when
P (t) = 10. That is , 7.3e0.022t = 10. Divide both sides by 7.3 to obtain
e0.022t ≈ 1.37. Solving this equation graphically to obtain t ≈ 14.3.

Next, in order to convert from Q(t) = bekt to Q(t) = bat we let a = ek .

For example, to convert the formula Q(t) = 7e0.3t to the form Q(t) = bat we
let b = 7 and a = e0.3 ≈ 1.35. Thus, Q(t) = 7(1.35)t .

Example 2.10
Find the annual percent rate and the continuous percent growth rate of
Q(t) = 200(0.886)t .

Solution.
The annual percent of decrease is r = a − 1 = 0.886 − 1 = −0.114 = −11.4%.
To find the continuous percent growth rate we let ek = 0.886 and solve for k
graphically to obtain k ≈ −0.121 = −12.1%.

Recommended Problems:
pp. 14 - 17: 1, 2, 3, 7, 11, 14, 15, 19, 21, 23, 25, 27, 30, 33, 38.

25
3 Building New Functions from Old Ones
In this section we will discuss some procedures for building new functions
from old ones. The first one is known as the composition of functions.
We start with an example of a real-life situation where composite functions
are applied.

Example 3.1
You have two money machines, both of which increase any money inserted
into them. The first machine doubles your money. The second adds five
dollars. The money that comes out is described by f (x) = 2x in the first
case, and g(x) = x + 5 in the second case, where x is the number of dollars
inserted. The machines can be hooked up so that the money coming out of
one machine goes into the other. Find formulas for each of the two possible
composition machines.

Solution.
Suppose first that x dollars enters the first machine. Then the amount of
money that comes out is f (x) = 2x. This amount enters the second machine.
The final amount coming out is given by g(f (x)) = f (x) + 5 = 2x + 5.
Now, if x dollars enters the second machine first, then the amount that comes
out is g(x) = x + 5. If this amount enters the second machine then the final
amount coming out is f (g(x)) = 2(x + 5) = 2x + 10. The function f (g(x)) is
called the composition of f with g.

In general, suppose we are given two functions f and g such that the range
of g is contained in the domain of f so that the output of g can be used as
input for f. We define a new function, called the composition of f with g,
by the formula
(f ◦ g)(x) = f (g(x)).

26
Using a Venn diagram (See Figure 14) we have

Figure 14

Example 3.2
Suppose that f (x) = 2x + 1 and g(x) = x2 − 3.

(a) Find f ◦ g and g ◦ f.

(b) Calculate (f ◦ g)(5) and (g ◦ f )(−3).
(c) Are f ◦ g and g ◦ f equal?

Solution.
(a) (f ◦ g)(x) = f (g(x)) = f (x2 − 3) = 2(x2 − 3) + 1 = 2x2 − 5. Similarly,
(g ◦ f )(x) = g(f (x)) = g(2x + 1) = (2x + 1)2 − 3 = 4x2 + 4x − 2.
(b) (f ◦ g)(5) = 2(5)2 − 5 = 45 and (g ◦ f )(−3) = 4(−3)2 + 4(−3) − 2 = 22.
(c) f ◦ g 6= g ◦ f.

With only one function you can build new functions using composition of
the function with itself. Also, there is no limit on the number of functions
that can be composed.

Example 3.3
Suppose that f (x) = 2x + 1 and g(x) = x2 − 3.

(a) Find (f ◦ f )(x).

(b) Find [f ◦ (f ◦ g)](x).

27
Solution.
(a) (f ◦ f )(x) = f (f (x)) = f (2x + 1) = 2(2x + 1) + 1 = 4x + 3.
(b) [f ◦(f ◦g)](x) = f (f (g(x))) = f (f (x2 −3)) = f (2x2 −5) = 2(2x2 −5)+1 =
4x2 − 9.

Decomposition of Functions
If a formula for (f ◦ g)(x) is given then the process of finding the formulas
for f and g is called decomposition.

Example 3.4 √
Decompose (f ◦ g)(x) = 1 − 4x2 .

Solution. √ 2
√ is f (x) = x and g(x) = 1 − 4x . Another possible
One possible answer
answer is f (x) = 1 − x2 and g(x) = 2x. Thus, decomposition of functions
in not unique.

Inverse Functions
We have seen that when every vertical line crosses a curve at most once then
the curve is the graph of a function f. We called this procedure the vertical
line test. Now, if every horizontal line crosses the graph at most once then
the function can be used to build a new function, called the inverse func-
tion and is denoted by f −1 , such that if f takes an input x to an output y
then f −1 takes y as its input and x as its output. That is

f (x) = y if and only if f −1 (y) = x.

When a function has an inverse then we say that the function is invertible.

Remark 3.1
The test used to identify invertible functions which we discussed above is
referred to as the horizontal line test.

Example 3.5
Use a graphing calculator to decide whether or not the function is invertible,
that is, has an inverse function:
(a) f (x) = x3 + 7 (b) g(x) = |x|.

28
Solution.
(a) Using a graphing calculator, the graph of f (x) is given in Figure 15.

Figure 15

We see that every horizontal line crosses the graph once so the function is
invertible.
(b) The graph of g(x) = |x| (See Figure 16) shows that there are horizontal
lines that cross the graph twice so that g is not invertible.

Figure 16

Remark 3.2
It is important not to confuse between f −1 (x) and (f (x))−1 . The later is just
the reciprocal of f (x), that is, (f (x))−1 = f (x)
1
whereas the former is how the
inverse function is represented.

Domain and Range of an Inverse Function

Figure 17 shows the relationship between f and f −1 .

29
 Figure 17
This figure shows that we get the inverse of a function by simply reversing
the direction of the arrows. That is, the outputs of f are the inputs of f −1
and the outputs of f −1 are the inputs of f. It follows that
Domain of f −1 = Range of f and Range of f −1 = Domain of f.
Example 3.6 √
Consider the function f (x) = x − 4.

(a) Find the domain and the range of f (x).

(b) Use the horizontal line test to show that f (x) has an inverse.
(c) What are the domain and range of f −1 ?
Solution.
(a) The function f (x) is defined for all x ≥ 4. The range is the interval [0, ∞).
(b) Graphing f (x) we see that f (x) satisfies the horizontal line test and so
f has an inverse. See Figure 18.
(c) The domain of f −1 is the range of f, i.e. the interval [0, ∞). The range
of f −1 is the domain of f , that is, the interval [4, ∞).

30
Finding a Formula for the Inverse Function
How do you find the formula for f −1 from the formula of f ? The procedure
consists of the following steps:

1. Replace f (x) with y.

2. Interchange the letters x and y.
3. Solve for y in terms of x.
4. Replace y with f −1 (x).

Example 3.7
Find the formula for the inverse function of f (x) = x3 + 7.

Solution.
As seen in Example 3.5, f (x) is invertible. We find its inverse as follows:

1. Replace f (x) with y to obtain y = x3 + 7.

2. Interchange x and y to obtain x = y 3 + √ 7.
3
3. Solve for y to obtain y = x − 7 or y = 3 x√ − 7.
4. Replace y with f −1 (x) to obtain f −1 (x) = 3 x − 7.

Compositions of f and its Inverse

Suppose that f is an invertible function. Then the expressions y = f (x) and
x = f −1 (y) are equivalent. So if x is in the domain of f then

f −1 (f (x)) = f −1 (y) = x

and for y in the domain of f −1 we have

f (f −1 (y)) = f (x) = y

It follows that for two functions f and g to be inverses of each other we must
have f (g(x)) = x for all x in the domain of g and g(f (x)) = x for x in the
domain of f.

Example 3.8
Check that the pair of functions f (x) = x4 − 32 and g(x) = 4(x+ 32 ) are inverses
of each other.

31
Solution.
The domain and range of both functions consist of the set of all real numbers.
Thus, for any real number x we have
3 4x + 6 3
f (g(x)) = f (4(x + )) = f (4x + 6) = − = x.
2 4 2
and
x 3 x 3 3
g(f (x)) = g( − ) = 4( − + ) = x.
4 2 4 2 2
So f and g are inverses of each other.

Combinations of Functions
Next, we are going to construct new functions from old ones using the oper-
ations of addition, subtraction, multiplication, and division.
Let f (x) and g(x) be two given functions. Then for all x in the common
domain of these two functions we define new functions as follows:

• Sum: (f + g)(x) = f (x) + g(x).

• Difference: (f − g)(x) = f (x) − g(x).
• Product: (f
 · g)(x) = f (x) · g(x).
• Division: g (x) = fg(x)
f (x)
provided that g(x) 6= 0.

In the following example we see how to construct the four functions discussed
above when the individual functions are defined by formulas.

Example 3.9 √
Let f (x) = x + 1 and g(x) = x + 3. Find the common domain and then
find a formula for each of the functions f + g, f − g, f · g, fg .

Solution.
The domain of f (x) consists of all real numbers whereas the domain of g(x)
consists of all numbers x ≥ 3. Thus, the common domain is the interval
[−3, ∞). For any x in this domain we have
√
(f + g)(x) = x + 1 + √x + 3
(f − g)(x) = √x+1− √ x+3
(f · g)(x) = x x+3+ x+3
f x+1
g
(x) = √x+3 provided x > −3.

32
Transformations of Functions
We close this section by giving a summary of the various transformations
obtained when either the input or the output of a function is altered.

Vertical Shifts: The graph of f (x) + k with k > 0 is a vertical transla-

tion of the graph of f (x), k units upward, whereas for k < 0 it is a shift by
k units downward.
Horizontal Shifts: The graph of f (x + k) with k > 0 is a horizontal trans-
lation of the graph of f (x), k units to the left, whereas for k < 0 it is a shift
by k units to the right.
Reflections about the x-axis: For a given function f (x), the graph of
−f (x) is a reflection of the graph of f (x) about the x-axis.
Reflections about the y-axis: For a given function f (x), the graph of
f (−x) is a reflection of the graph of f (x) about the y-axis.
Vertical Stretches and Compressions: If a function f (x) is given, then
the graph of kf (x) is a vertical stretch of the graph of f (x) by a factor of k
for k > 1, and a vertical compression for 0 < k < 1.
What about k < 0? If |k| > 1 then the graph of kf (x) is a vertical stretch
of the graph of f (x) followed by a reflection about the x-axis. If 0 < |k| < 1
then the graph of kf (x) is a vertical compression of the graph of f (x) by a
factor of k followed by a reflection about the x-axis.
Horizontal Stretches and Compressions: If a function f (x) is given,
then the graph of f (kx) is a horizontal stretch of the graph of f (x) by a
factor of k1 for 0 < k < 1, and a horizontal compression for k > 1.
What about k < 0? If |k| > 1 then the graph of f (kx) is a horizontal com-
pression of the graph of f (x) followed by a reflection about the y-axis. If
0 < |k| < 1 then the graph of f (kx) is a horizontal stretch of the graph of
f (x) by a factor of k1 followed by a reflection about the y-axis.

Odd and Even Functions

For a given function f (x), the graph of the new function −f (x) is the reflec-
tion of the graph of f (x) about the x-axis.
Example 3.10
Graph the functions f (x) = 2x and −f (x) = −2x on the same axes.
Solution.
The graph of both f (x) = 2x and −f (x) are shown in Figure 19.

33
 Figure 19

We know that the points x and −x are on opposite sides of the x-axis. So
the graph of the new function f (−x) is the reflection of the graph of f (x)
about the y-axis.
If the reflection of the graph of f (x) about the y-axis is the same as the graph
of f (x),i.e, f (−x) = f (x), then we say that the graph of f (x) is symmetric
about the y-axis. We call such a function an even function.

Example 3.11
(a) Using a graphing calculator show that the function f (x) = (x − x3 )2 is
even.
(b) Now show that f (x) is even algebraically.

Solution.
(a) The graph of f (x) is symmetric about the y-axis so that f (x) is even.
See Figure 20.

34
 Figure 20

(b) Since f (−x) = (−x − (−x)3 )2 = (−x + x3 )2 = [−(x − x3 )]2 = (x − x3 )2 =

f (x) then f (x) is even.

Now, if the images f (x) and f (−x) are of opposite signs i.e, f (−x) = −f (x),
then the graph of f (x) is symmetric about the origin. In this case, we say
that f (x) is odd. Alternatively, since f (x) = −f (−x), if the graph of a
function is reflected first across the y-axis and then across the x-axis and you
get the graph of f (x) again then the function is odd.

Example 3.12
1+x2
(a) Using a graphing calculator show that the function f (x) = x−x3
is odd.
(b) Now show that f (x) is odd algebraically.

Solution.
(a) The graph of f (x) is symmetric about the origin so that f (x) is odd. See
Figure 21.

35
 Figure 21
1+(−x)2 1+x2 1+x2
(b) Since f (−x) = (−x)−(−x)3
= −x+x3
= −(x−x3 )
= −f (x) then f (x) is odd.

A function can be either even, odd, or neither.

Example 3.13
(a) Show that the function f (x) = x2 is even but not odd.
(b) Show that the function f (x) = x3 is odd but not even.
(c) Show that the function f (x) = x + x2 is neither odd nor even.
(d) Is there a function that is both even and odd? Explain.

Solution.
(a) Since f (−x) = f (x) and f (−x) 6= −f (x) then f (x) is even but not odd.
(b) Since f (−x) = −f (x) and f (−x) 6= f (x) then f (x) is odd but not even.
(c) Since f (−x) = −x + x2 6= ±f (x) then f (x) is neither even nor odd.
(d) We are looking for a function such that f (−x) = f (x) and f (−x) =
−f (x). This implies that f (x) = −f (x) or 2f (x) = 0. Dividing by 2 to ob-
tain f (x) = 0. This function is both even and odd.

Recommended Problems (pp. 21 - 23): 2, 3, 5, 7, 11, 12, 13,

14, 17, 18, 25, 29, 31, 36.

36
4 Logarithmic Functions
An equation of the form ax = b can be solved graphically. That is, using a
calculator we graph the horizontal line y = b and the exponential function
y = ax and then find the point of intersection.
In this section we discuss an algebraic way to solve equations of the form
ax = b where a and b are positive constants. For this, we introduce two
functions that are found in today’s calculators, namely, the functions log x
and ln x.

If x > 0 then we define log x to be a number y that satisfies the equality

10y = x. For example, log 100 = 2 since 102 = 100. Similarly, log 0.01 = −2
since 10−2 = 0.01. We call log x the common logarithm of x. Thus,

y = log x if and only if 10y = x.

Similarly, we have

y = ln x if and only if ey = x.

We call ln x the natural logarithm of x.

Example 4.1
(a) Rewrite log 30 = 1.477 using exponents instead of logarithms.
(b) Rewrite 100.8 = 6.3096 using logarithms instead of exponents.

Solution.
(a) log 30 = 1.477 is equivalent to 101.477 = 30.
(b) 100.8 = 6.3096 is equivalent to log 6.3096 = 0.8.

Example 4.2
Without a calculator evaluate the following expressions:

(a) log 1 (b) log 100 (c) log ( √110 ) (d) 10log 100 (e) 10log (0.01)

Solution.
(a) log 1 = 0 since 100 = 1.
(b) log 100 = log 1 = 0 by (a).
1
(c) log ( √110 ) = log 10− 2 = − 12 .

37
(d) 10log 100 = 102 = 100.
(e) 10log (0.01) = 10−2 = 0.01.

Properties of Logarithms

(i) Since 10x = 10x we can write

log 10x = x

(ii) Since log x = log x then

10log x = x
(iii) log 1 = 0 since 100 = 1.
(iv) log 10 = 1 since 101 = 10.
(v) Suppose that m = log a and n = log b. Then a = 10m and b = 10n . Thus,
a · b = 10m · 10n = 10m+n . Rewriting this using logs instead of exponents, we
see that
log (a · b) = m + n = log a + log b.
a 10m
(vi) If, in (v), instead of multiplying we divide, that is b
= 10n
= 10m−n
then using logs again we find
a
 
log = log a − log b.
b
(vii) It follows from (vi) that if a = b then log a − log b = log 1 = 0 that is
log a = log b.
(viii) Now, if n = log b then b = 10n . Taking both sides to the power k
we find bk = (10n )k = 10nk . Using logs instead of exponents we see that
log bk = nk = k log b that is

log bk = k log b.

Example 4.3
Solve the equation: 4(1.171)x = 7(1.088)x .

Solution.  x
Rewriting the equation into the form 1.171
1.088
= 7
4
and then using properties
(vii) and (viii) to obtain
1.171 7
 
x log = log .
1.088 4

38
Thus,
log 74
x= 
1.171
.
log 1.088

Example 4.4
Solve the equation log (2x + 1) + 3 = 0.

Solution.
Subtract 3 from both sides to obtain log (2x + 1) = −3. Switch to exponential
form to get 2x + 1 = 10−3 = 0.001. Subtract 1 and then divide by 2 to obtain
x = −0.4995.

Remark 4.1
• All of the above arguments are valid for the function ln x for which we
replace the number 10 by the number e = 2.718 · · · . That is, ln (a · b) =
ln a + ln b, ln ab = ln a − ln b etc.
• Keep in mind the following:
log (a + b) 6= log a + log b. For example, log 2 6= log 1 + log 1 = 0.
log (a − b) 6= log a − log b. For example, log (2 − 1) = log 1 = 0 whereas
log 2 − log 1 = log 2 6= 0.
log (ab) 6= log a · log b. For example, log 1 = log (2 · 12 ) = 0 whereas log 2 ·
log 12 =

− log2 2 6= 0.
log ab 6= log a
log b
. For example, letting a = b = 2 we find that log ab = log 1 = 0
whereas log a
log b
= 1.
 
1 1 1 1
log a
6= log a
. For example, log 1 = log 2 whereas log 12
= − log1 2 .
2

Logarithmic Functions and Their Graphs

Next, we will graph logarithmic functions and determine a number of their
general features.
We have seen that the notation y = log x is equivalent to 10y = x. Since 10
raised to any power is always positive then the domain of the function log x
consists of all positive numbers. That is, log x cannot be used with negative
numbers.
Now, let us sketch the graph of this function by first constructing the follow-
ing chart:

39
 x log x Average Rate of Change
0 undefined -
0.001 -3 -
0.01 -2 111.11
0.1 -1 11.11
1 0 1.11
10 1 0.11
100 2 0.011
1000 3 0.0011

From the chart we see that the graph is always increasing. Since the average
rate of change is decreasing then the graph is always concave down. Now
plotting these points and connecting them with a smooth curve to obtain

Figure 22

From the graph we observe the following properties:

(a) The range of log x consists of all real numbers.
(b) The graph never crosses the y-axis since a positive number raised to any
power is always positive.
(c) The graph crosses the x-axis at x = 1.
(d) As x gets closer and closer to 0 from the right the function log x decreases
without bound. That is, as x → 0+ , x → −∞. We call the y-axis a vertical
asymptote. In general, if a function increases or decreases without bound
as x gets closer to a number a then we say that the line x = a is a vertical
asymptote.

40
Next, let’s graph the function y = 10x by using the above process:

x 10x Average Rate of Change

-3 0.001 -
-2 0.01 0.009
-1 0.1 0.09
0 1 0.9
1 10 9
2 100 90
3 1000 900
Note that this chart can be obtained from the chart of log x discussed above
by interchanging the variables x and y. This means, that the graph of y = 10x
is a reflection of the graph of y = log x about the line y = x as seen in Figure
23.

Figure 23

Example 4.5
Sketch the graphs of the functions y = ln x and y = ex on the same axes.

Solution.
The functions y = ln x and y = ex are inverses of each like the functions
y = log x and y = 10x . So their graphs are reflections of one another across
the line y = x as shown in Figure 24.

41
 Figure 24

Recommended Problems (pp. 27 - 28): 2, 8, 14, 15, 19, 25, 30,

33, 35, 36, 37, 39, 40, 42, 44.

42
5 Trigonometric Functions
In this section, we review some of the results of trigonometry that we need
in the sequel.

An angle is determined by rotating a ray ( or a half-line) from one posi-

tion, called the initial side, to a terminal position, called the terminal
side, as shown in Figure 25. The point V is called the vertex of the angle.

Figure 25

If the initial side is the positive x-axis then we say that the angle is in
standard position. See Figure 26.

Figure 26
Angles that are obtained by a counterclockwise rotation are considered pos-
itive and those that are obtained clockwise are negative angles. See Figure
27.

43
 Figure 27

Angle Measure
To be able to compare angles, one needs to assign a unit of measurement for
angles. The measure of an angle is determined by the amount of rotation
from the initial side to the terminal side, this is how much the angle ”opens”.
There are two commonly used measures of angles: degrees and radians

• Degree Measure:
If we rotate counterclockwise a ray about a fixed vertex and then return back
to its initial position then we say that we have a one complete revolution.
The angle in this case is said to have measure of 360 degrees, in symbol 360◦ .
Thus, 1◦ is 3601
th of a revolution. See Figure 28.

Figure 28

• Radian Measure:
A more natural method of measuring angles used in calculus and other
branches of mathematics is the radian measure. The amount an angle opens
is measured along the arc of the unit circle with its center at the vertex of
the angle.( An angle whose vertex is the center of a circle is called a central
angle.) One radian, abbreviated rad, is defined to be the measure of a
central angle that intercepts an arc s of length one unit. See Figure 29.

44
 Figure 29

Since one complete revolution measured in radians is 2π radians and mea-

sured in degrees is 360◦ then we have the following conversion formulas:
π 180 ◦
1◦ = rad ≈ 0.01745 rad and 1 rad = ( ) ≈ 57.296◦ .
180 π
Example 5.1
Complete the following chart.
degree 30◦ 45◦ 60◦ 90◦ 180◦ 270◦
radian

Solution.
degree 30◦ 45◦ 60◦ 90◦ 180◦ 270◦
radian π6 π
4
π
3
π
2
π 3π
2
 
By the conversion formulas, we have, for example 30◦ = 30(1◦ ) = 30 π
180
=
π
6
. In a similar way we convert the remaining angles.

Example 5.2
Convert each angle in degrees to radians: (a) 150◦ (b) −45◦ .

Solution.
(a) 150◦ = 150(1◦ ) = 150( 180
π
) = 5π
6
rad.
◦ ◦
(b) −45 = −45(1 ) = −45( 180 ) = − π4 rad.
π

Example 5.3
Convert each angle in radians to degrees: (a) − 3π
4
(b) 7π
3
.

45
Solution.
(a) − 3π
4
= − 3π 4
(1 rad) = − 3π
4
( 180
π
)◦ = −135◦ .
(b) 7π
3
= 7π
3
( 180
π
)◦ = 420◦

Remark 5.1
When we write a trigonometric function, such as sin t, then it is assumed
that t is in radians. If we want to evaluate the trigonometric function of an
angle measured in degrees we will use the degree notation such as cos 30◦ .

We next discuss a relationship between a central angle θ, measured in radians,

and the length of the arc s that it intercepts.

Theorem 5.1
For a circle of radius r, a central angle of θ radians subtends an arc whose
length s is given by the formula:
s = rθ

Proof.
Suppose that r > 1. (A similar argument holds for 0 < r < 1.) Draw the unit
circle with the same center C (See Figure 30).

Figure 30
By definition of radian measure, the length of the arc determined by θ on the
unit circle is also θ. From elementary geometry, we know that the ratio of the
measures of the arc lengths are the same as the ratio of the corresponding
radii. That is,
r s
= .
1 θ
46
Now the formula follows by cross-multiplying.

The above formula allows us to define the radian measure using a circle
of any radius r. (See Figure 31).

Figure 31

Example 5.4
Find the length of the arc of a circle of radius 2 meters subtended by a central
angle of measure 0.25 radians.

Solution.
We are given that r = 2 m and θ = 0.25 rad. By the previous theorem we
have:
s = rθ = 2(0.25) = 0.5 m

Example 5.5
Suppose that a central angle of measure 30◦ is subtended by an arc of length
π
2
feet. Find the radius r of the circle.

Solution.
π
Substituting in the formula s = rθ we find 2
= r π6 . Solving for r to obtain
r = 3f eet.

The Trigonometric Functions

By the unit circle we mean the circle with center at the point O(0, 0) and
radius 1. Such a circle has the equation x2 + y 2 = 1.
Now, let t be any real number. From the point A(1, 0), walk on the unit
_
circle a distance t arriving at some point P (a, b). Then the arc AP subtends

47
a central angle θ. See Figure 32.

Figure 32

We define the following trigonometric functions:

b
sin t = b cos t = a tan t = a
csc t = 1b sec t = a1 cot t = a
b

where a 6= 0 and b 6= 0. If a = 0 then the functions sec t and tan t are

undefined. If b = 0 then the functions csct and cot t are undefined.
The above trigonometric functions are referred to as circular functions.
It follows from the above definition that

sin2 x + cos2 x = 1.

The table below lists the domain, range, and the periodicity of each of the
triginometric functions discussed above.
f (x) Domain Range Period
sin x (−∞, ∞) [-1,1] 2π
cos x (−∞, ∞) [-1,1] 2π
tan x x 6= (2n + 1) π2 (−∞, ∞) π
cot x x 6= nπ (−∞, ∞) π
sec x x 6= (2n + 1) π2 (−∞, −1] ∪ [1, ∞) 2π
csc x x 6= nπ (−∞, −1] ∪ [1, ∞) 2π

48
Example 5.6
Referring to Figure 33, answer the following questions.
(a) Express the area of ∆OBC in terms of sin θ and cos θ.
(b) Express the area of ∆OBD in terms of tan θ.
(c) Use parts (a) and (b) to show

θ 1
1< < .
sin θ cos θ

Figure 33

Solution.
|EC|
(a) Area ∆OBC = 12 |EC||OB| = 12 |EC| = 21 sin θ since sin θ = |OC|
= |EC|.
|DB|
(b) Area ∆OBD = 21 |BD||OB| = 21 |BD| = 1
2
tan θ since tan θ = |OB|
=
|DB|.
(c) Using Figure 33 we see that

Area ∆OBC < Area circular sector OBC < Area ∆OBD

But the area of the circular sector OBC is 12 θ. Hence,

1 1 1
sin θ < θ < tan θ
2 2 2
2
Multiplying through by sin θ
to obtain

θ 1
1< < .
sin θ cos θ

49
Graphs of Sine, Cosine, and Tangent Functions
Figure 34 shows the graph of y = sinx.

Figure 34

Figure 35 shows the graph of y = cosx.

Figure 35

50
Figure 36 shows the graph of y = tanx.

Figure 36

Amplitude and period of y = a sin (bx), y = a cos (bx), b > 0

We now consider graphs of functions that are transformations of the sine and
cosine functions.

• The parameter a: This is outside the function and so deals with the
output (i.e. the y values). Since −1 ≤ sin (bx) ≤ 1 and −1 ≤ cos (bx) ≤ 1
then −a ≤ a sin (bx) ≤ a and −a ≤ a cos (bx) ≤ a. So, the range of the
function y = a sin (bx) or the function y = a cos (bx) is the closed interval
[−a, a]. The number |a| is called the amplitude. Graphically, this number
describes how tall the graph is. The amplitude is half the distance from the
top of the curve to the bottom of the curve. If b = 1, the amplitude |a|
indicates a vertical stretch of the basic sine or cosine curve if a > 1, and a
vertical compression if 0 < a < 1. If a < 0 then a reflection about the x-axis
is required.
Figure 37 shows the graph of y = 2 sin x and the graph of y = 3 sin x.

51
 Figure 37

• The parameter b: This is inside the function and so effects the input (i.e.
x values). Now, the graph of either y = a sin (bx) or y = a cos (bx) completes
one period from bx = 0 to bx = 2π. By solving for x we find the interval of
one period to be [0, 2π
b
]. Thus, the above mentioned functions have a period
2π
of b . The number b tells you the number of cylces in the interval [0, 2π].
Graphically,b either stretches (if b < 1) or compresses (if b > 1) the graph
horizontally.
Figure 38 shows the function y = sin x with period 2π and the function
y = sin (2x) with period π.

Figure 38

52
The Inverse Sine Function
The function f (x) = sin x is increasing on the interval [− π2 , π2 ]. See Figure
39. Thus, f (x) is one-to-one and consequently it has an inverse denoted by

f −1 (x) = sin−1 x.

We call this new function the inverse sine function.

Figure 39

From the definition of inverse functions discussed in Section 4, we have the

following properties of f −1 (x) :

(i) Dom(sin−1 x) = Range(sin x) = [−1, 1].

(ii) Range(sin−1 x) = Dom(sin x) = [− π2 , π2 ].
(iii) sin (sin−1 x) = x for all −1 ≤ x ≤ 1.
(iv) sin−1 (sin x) = x for all − π2 ≤ x ≤ π2 .
(v) y = sin−1 x if and only if sin y = x. Using words, the notation y = sin−1 x
gives the angle y whose sine value is x.

Remark 5.2
If x is outside the interval [− π2 , π2 ] then we look for the angle y in the interval
[− π2 , π2 ] such that sin x = sin y. In this case, sin−1 (sin x) = y. For example,
sin−1 (sin 5π 6
) = sin−1 (sin π6 ) = π6 .

The graph of y = sin−1 x is the reflection of the graph of y = sin x about the
line y = x as shown in Figure 40.

53
 Figure 40

Example 5.7
Find the exact value √of:
(a) sin−1 1 (b) sin−1 23 (c) sin−1 (− 12 )

Solution.
(a) Since sin π2 = 1√ then sin−1 1 =
√ 2
π
.
π 3 −1 3
(b) Since sin 3 = 2 then sin 2
= π3 .
−1
π
(c) Since sin (− 6 ) = − 2 then sin (− 21 ) = − π6 .
1

Example 5.8
Find the exact value of:
(a) sin (sin−1 2) (b) sin−1 (sin π3 ).

Solution.
(a) sin (sin−1 2) is undefined since 2 is not in the domain of sin−1 x.
(b) sin (sin−1 π3 ) = Π3 .

The Inverse Cosine function

In order to define the inverse cosine function, we will restrict the function
f (x) = cos x over the interval [0, π]. There the function is always decreasing.
See Figure 41. Therefore f (x) is one-to-one function. Hence, its inverse will
be denoted by
f −1 (x) = cos−1 x.

54
We call cos−1 x the inverse cosine function.

Figure 41

The following are consequences of the definition of inverse functions:

(i) Dom(cos−1 x) = Range(cos x) = [−1, 1].
(ii) Range(cos−1 x) = Dom(cos x) = [0, π].
(iii) cos (cos−1 x) = x for all −1 ≤ x ≤ 1.
(iv) cos−1 (cos x) = x for all 0 ≤ x ≤ π.
(v) y = cos−1 x if and only if cos y = x. Using words, the notation y = cos−1 x
gives the angle y whose cosine value is x.

Remark 5.3
If x is outside the interval [0, π] then we look for the angle y in the interval
[0, π] such that cos x = cos y. In this case, cos−1 (cos x) = y. For example,
cos−1 (cos 7π
6
) = cos−1 (cos 5π
6
) = 5π
6
.

The graph of y = cos−1 x is the reflection of the graph of y = cos x about the
line y = x as shown in Figure 42.

55
 Figure 42

Example 5.9
Let θ = cos−1 x. Find the six trigonometric functions of θ.

Solution.
−1 2 2
Let
√ u = cos x. Then cos u = x. Since sin u + cos u = √ 1 then sin u =
± 1 − x2 . Since 0 ≤ u ≤ π then sin u ≥ 0 so that sin u = 1 − x2 . Thus,
√
sin (cos−1 x) = 1 − x2
−1
cos (cos x) = x
−1 1 1
csc (cos x) = sin (cos−1 x) = √1−x 2

sec (cos−1 x) = 1
−1
cos (cos x)
= 1
x √
sin (cos−1 x) 1−x2
tan (cos−1 x) = −1
cos (cos x)
= x
cot (cos−1 x) = 1
−1
tan (cos x)
= √ x
1−x2
.

Example 5.10
Find the exact
√ value of:
(a) cos−1 22 (b) cos−1 (− 12 ).

Solution.√ √
(a) cos−1 22 = π4 since cos π4 = 2
2
.
(b) cos−1 (− 12 ) = 2π
3
.

The Inverse Tangent Function

Although not one-to-one on its full domain, the tangent function is one-to-
one when restricted to the interval (− π2 , π2 ) since it is increasing there (See

56
Figure 43).

Figure 43

Thus, the inverse function exists and is denoted by

f −1 (x) = tan−1 x.

We call this function the inverse tangent function.

As before, we have the following properties:
(i) Dom(tan−1 x) = Range(tan x) = (−∞, ∞).
(ii) Range(tan−1 x) = Dom(tan x) = (− π2 , π2 ).
(iii) tan (tan−1 x) = x for all x.
(iv) tan−1 (tan x) = x for all − π2 < x < π2 .
(v) y = tan−1 x if and only if tan y = x. In words, the notation y = tan−1 x
means that y is the angle whose tangent value is x.

Remark 5.4
If x is outside the interval (− π2 , π2 ) and x 6= (2n + 1) π2 , where n is an
integer, then we look for the angle y in the interval (− π2 , π2 ) such that
tan x = tan y. In this case, tan−1 (tan x) = y. For example, tan−1 (tan 5π
6
)=
tan−1 (tan (− π6 )) = − π6 .

The graph of y = tan−1 x is the reflection of y = tan x about the line y = x

as shown in Figure 44.

57
 Figure 44

Example 5.11
Find the exact value of:
(a) tan−1 (tan π4 ) (b) tan−1 (tan 7π
5
).

Solution.
(a) tan−1 (tan π4 ) = π4 .
(b) tan−1 (tan 7π5
) = tan−1 (tan ( 2π
5
)) = 2π
5
.

Example 5.12
Let u = tan−1 x. Find the six trigonometric functions of u.

Solution. √
Since 1 + tan2 u = sec2 u then sec u = ± 1 + x2 . But − π2 < u < π2 then
√
sec u > 0 so that sec u = 1 + x2 . Also, cot u = tan1 u = x1 . In summary,

sin (tan−1 x) = 1
csc (tan−1 x)
x
= √1+x 2

cos (tan−1 x) = 1
sec (tan−1
1
= 1+x2
√
√x)
1+x2
csc (tan−1 x) = √ x
sec (tan−1 x) = 1 + x2
tan (tan−1 x) = x
cot (tan−1 x) = 1
x

Recommended Problems (pp. 35 - 6): 1, 5, 7, 8, 9, 10, 11, 13,

15, 16, 17, 18, 19, 20, 23, 25, 27, 28, 33, 35.

58
6 Powers, Polynomial, and Rational Functions
In this section we discuss three important families of functions: powers, poly-
nomials, and rational functions.

Power Functions
A function f (x) is a power function of x if there is a nonzero constant k
such that
f (x) = kxn
The number n is called the power of x.

Remark 6.1
Recall that an exponential function has the form f (x) = bax , where the base
a is fixed ans the exponent x varies. For a power function these properties
are reversed- the base varies and the exponent remains constant.

Domains of Power Functions

If n is a non-negative integer then the domain of f (x) = kxn consists of all
real numbers. If n is a negative integer then the domain of f consists of all
nonzero real numbers.
If n = rs , where r and s have no common factors, then the domain of f (x) is
all real numbers for s odd and n > 0 (all non zero real numbers for s odd and
n < 0.) If s is even and n > 0 then the domain consists of all non-negative
real numbers( all positive real numbers if n < 0.)

The Effect of n on the Graph of xn

We assume that k = 1 and we will compare the graphs of f (x) = xn for
various values of n. We will use graphing calculator to illustrate how power
functions work and the role of n.
When n = 0 then the graph is a horizontal line at (0, 1). When n = 1 then the
graph is a straight line through the origin with slope equals to 1. See Figure .

59
 Figure 45
The graphs of all power functions with n = 2, 4, 6, · · · have the same charac-
teristic ∪ − shape and they satisfy the following properties:

1. Pass through the points (0, 0), (1, 1), and (−1, 1).
2. Decrease for negative values of x and increase for positive values of x.
3. Are symmetric about the y-axis because the functions are even.
4. Are concave up.
5. The graph of y = x4 is flatter near the origin and steeper away from the
origin than the graph of y = x2 . See Figure 46.

Figure 46

The graphs of power functions with n = 1, 3, 5, · · · resemble the side view of

60
a chair and satisfy the following properties:

1. Pass through (0, 0) and (1, 1) and (−1, −1).

2. Increase on every interval.
3. Are symmetric about the origin because the functions are odd.
4. Are concave down for negative values of x and concave up for positive
values of x.
5. The graph of y = x5 is flatter near the origin and steeper far from the
origin than the graph of y = x3 . See Figure 47.

Figure 47

Graphs of y = xn with n = −1, −3, · · · .

1. Passes through (1, 1) and (−1, −1) and does not have a y-intercept.
2. Is decreasing everywhere that it is defined.
3. Is symmetric about the origin because the function is odd.
4. Is concave down for negative values of x and concave up for positive values
of x.
5. Has the x-axis as a horizontal asymptote and they y-axis as a vertical
asymptote.
6. For −1 < x < 1, the graph of y = x1 approaches the y-axis more rapidly
then the graph of y = x13 . For x < −1 or x > 1 the graph of y = x13 approach
the x-axis more rapidly then the graph of y = x1 . See Figure 48.

61
 Figure 48

Graphs of y = xn with n = −2, −4, · · ·

1. Passes through (1, 1) and (−1, 1) and does not have a y- or x-intercept.
2. Is increasing for negative values of x and decreasing for positive values of
x.
3. Is symmetric about the y-axis because the function is even.
4. Is concave up everywhere that it is defined.
5. Has the x-axis as a horizontal asymptote and the y-axis as vertical asymp-
tote.
6. For −1 < x < 1, the graph of y = x12 approaches the y-axis more rapidly
then the graph of y = x14 . For x < −1 or x > 1 the graph of y = x14 approach
the x-axis more rapidly then the graph of y = x12 . See Figure 49.

62
 Figure 49

1
Graphs of y = x r where r = 2, 4, · · · has the following properties:
1. Domain consists of all non-negative real numbers.
2. Pass through (0, 0) and (1, 1).
3. Are increasing for x > 0.
4. Are concave down for x > 0.
1
5. The graph of y = x 4 is steeper near the origin and flatter away from the
1
origin then the graph of y = x 2 See Figure 50.

Figure 50

63
 1
Graphs of y = x r where r = 3, 5, · · · has the following properties:
1. Domain consists for all real numbers.
2. Pass through (0, 0), (1, 1) and (−1, −1).
3. Are increasing.
4. Are concave down for x > 0 and concave up for x < 0.
1
5. The graph of y = x 5 is steeper near the origin and flatter away from the
1
origin then the graph of y = x 3 . See Figure 51.

Figure 51

Polynomial Functions
Polynomial functions are among the simplest, most important, and most
commonly used mathematical functions. These functions consist of one or
more terms of variables with whole number exponents. (Whole numbers are
positive integers and zero.) All such functions in one variable (usually x) can
be written in the form

f (x) = an xn + an−1 xn−1 + · · · + a1 x + a0 , an 6= 0

where an , an−1 , · · · , a1 , a0 are all real numbers, called the coefficients of f (x).
The number n is a non-negative integer. It is called the degree of the
polynomial. A polynomial of degree zero is just a constant function. A
polynomial of degree one is a linear function, of degree two a quadratic
function, etc. The number an is called the leading coefficient and a0 is

64
called the constant term.
Note that the terms in a polynomial are written in descending order of the
exponents. Polynomials are defined for all values of x.
Example 6.1
Find the leading coefficient, the constant term and the degreee of the poly-
nomial f (x) = 4x5 − x3 + 3x2 + x + 1.
Solution.
The given polynomial is of degree 5, leading coefficient 4, and constant term
1.

A polynomial function will never involve terms where the variable occurs
in a denominator, underneath a radical, as an input of either an exponential,
logarithmic or trigonometric function.
Example 6.2
Determine whether the function is a polynomial function or not:

(a) f (x) = 3x4 − 4x2 + 5x − 10

(b) g(x) = x3 − ex + 3
(c) h(x) = x2 − √3x + x1 + 4
(d) i(x) = x2 − x − 5
(e) j(x) = x3 − 3x2 + 2x − 5 ln x − 3.
Solution.
(a) f (x) is a polynomial function of degree 4.
(b) g(x) is not a ploynomial because one of the terms is an exponential func-
tion.
(c) h(x) is not a polynomial because x is in the denominator of a fraction.
(d) i(x) is not a polynomial because it contains a radical sign.
(e) j(x) is not a polynomial because one of the terms is a logarithm of x.

Long-Run Behavior of a Polynomial Function

If f (x) and g(x) are two functions such that f (x) − g(x) ≈ 0 as x increases
without bound then we say that f (x) resembles g(x) in the long run. For
example, if n is any positive integer then x1n ≈ 0 in the long run.
Now, if f (x) = an xn + an−1 xn + · · · + a1 x + a0 then
an−1 an−2 a1 a0
 
n
f (x) = x an + + 2 + · · · + n−1 + n
x x x x

65
 1
Since xk
≈ 0 in the long run, for each 1 ≤ k ≤ n then

f (x) ≈ an xn

in the long run.

Example 6.3
The polynomial function f (x) = 1 − 2x4 + x3 resembles the function g(x) =
−2x4 in the long run.

Zeros of a Polynomial Function

If f is a polynomial function in one variable, then the following statements
are equivalent:

• x = a is a zero or root of the function f.

• x = a is a solution of the equation f (x) = 0.
• (a, 0) is an x-intercept of the graph of f. That is, the point where the graph
crosses the x-axis.

Example 6.4
Find the x-intercepts of the polynomial f (x) = x3 − x2 − 6x.

Solution.
Factoring the given function to obtain

f (x) = x(x2 − x − 6)
= x(x − 3)(x + 2)

Thus, the x-intercepts are the zeros of the equation

x(x − 3)(x + 2) = 0

That is, x = 0, x = 3, or x = −2.

Graphs of a Polynomial Function

Polynomials are continuous and smooth everywhere:

• A continuous function means that it can be drawn without picking up

your pencil. There are no jumps or holes in the graph of a polynomial func-
tion.

66
• A smooth curve means that there are no sharp turns (like an absolute
value) in the graph of the function.
• The y-intercept of the polynomial is the constant term a0 .

The shape of a polynomial depends on the degree and leading coefficient:

• If the leading coefficient, an , of a polynomial is positive, then the right

hand side of the graph will rise towards +∞.
• If the leading coefficient, an , of a polynomial is negative, then the right
hand side of the graph will fall towards −∞.
• If the degree, n, of a polynomial is even, the left hand side will do the same
as the right hand side.
• If the degree, n, of a polynomial is odd, the left hand side will do the
opposite of the right hand side.

Example 6.5
According to the graphs given below, indicate the sign of an and the parity
of n for each curve.

Figure 52
Solution.
(a) an < 0 and n is odd.

67
(b) an > 0 and n is odd.
(c) an > 0 and n is even.
(d) an < 0 and n is even.

Short-Run Behavior of Polynomial Functions

We have seen in the discussion above that the two functions f (x) = x3 +
3x2 + 3x + 1 and g(x) = x3 − x2 − 6x resemble the function h(x) = x3 at the
long-term behavior. See Figure 53.

Figure 53

Although f and g have similar long-run behavior, they are not identical
functions. This can be seen by studying the short-run (or local) behavior
of these functions.
The short-run behavior of the graph of a function concerns graphical feature
of the graph such as its intercepts or the number of bumps on the graph. For
example, the function f (x) has an x-intercept at x = −1 and y-intercept at
y = 1 and no bumps. On the other hand, the function g(x) has x-intercepts
at x = −2, 0, 3, y-intercept at y = 0, and two bumps.
As you have noticed, the zeros (or roots) of a polynomial function is one of the
important part of the short-run behavior. To find the zeros of a polynomial
function, we can write it in factored form and then use the zero product rule
which states that if a · b = 0 then either a = 0 or b = 0. To illustrate, let us
find the zeros of the function g(x) = x3 − x2 − 6x.

68
Factoring, we find
g(x) = x(x2 − x − 6) = x(x − 3)(x + 2).
Setting g(x) = 0 and solving we find x = 0, x = −2, and x = 3. The number
of zeros determines the number of bumps that a graph has since between any
two consecutive zeros, there is a bump because the graph changes direction.
Thus, as we see from the graph of g(x) that g(x) has two bumps.
Now, the function f (x) has only one zero at x = −1. We call x = −1 a zero
of multiplicity three.
It is easy to see that when a polynomial function has a zero of even multiplic-
ity than the graph does not cross the x-axis at that point; on the contrary,
if the zero is of odd multiplicity than the graph crosses the x-axis.
Example 6.6
Sketch the graph of f (x) = (x + 1)3 and g(x) = (x + 1)2 .
Solution.
The graphs of f (x) and g(x) are shown in Figure 54.

Figure 54

Rational Functions
A rational function is a function that is the ratio of two polynomial func-
tions fg(x)
(x)
. The domain consists of all numbers such that g(x) 6= 0.
Example 6.7
x−2
Find the domain of the function f (x) = x2 −x−6
.

69
Solution.
The domain consists of all numbers x such that x2 − x − 6 6= 0. But this last
quadratic expression is 0 when x = −2 or x = 3. Thus, the domain is the set
(−∞, −2) ∪ (−2, 3) ∪ (3, ∞).

The Long-Run Behavior of Rational Functions

Given a rational function
am xm + am−1 xm−1 + · · · + a1 x + a0
f (x) = .
bn xn + bn−1 xn−1 + · · · + b1 x + b0
We know that the top polynomial resembles am xm and the bottom poly-
nomial resembles bn xn in the long run. It follows that, in the long run,
m
f (x) ≈ abmn xxn .

Example 6.8
Discuss the long run behavior of each of the following functions:
2
(a) f (x) = 3x +2x−4
2x2 −x+1
.
2x+3
(b) f (x) = x3 −2x2 +4 .
2 −3x−1
(c) f (x) = 2x x−2 .

Solution.
2 +2x−4 2
(a) f (x) = 3x
2x2 −x+1
≈ 3x
2x2
= 32 .
2x+3 2x 2
(b) f (x) = x3 −2x 2 +4 ≈ x3 = x2 .
2 −3x−1 2
(c) f (x) = 2x x−2 ≈ 2xx = 2x.

Horizontal Asymptote
Contrary to polynomial functions, it is possible for a rational function to
level at x → ∞ or x → −∞. That is, f (x) approaches a value b as x → ∞ or
x → −∞. We call y = b a horizontal asymptote. The graph of a rational
function may cross its horizontal asymptote.

Example 6.9
Find the horizontal asymptote, if it exists, for each of the following functions:
2
(a) f (x) = 3x +2x−4
2x2 −x+1
.
2x+3
(b) f (x) = x3 −2x2 +4 .
2 −3x−1
(c) f (x) = 2x x−2 .

70
Solution.
(a) As x → ±∞, we have
3x2 +2x−4
f (x) = 2x2 −x+1
x2 3+ x2 − 42 3
= x2
· 2− x1 +
x
1 → 2
x2

so the line y = 32 is the horizontal asymptote.

(b) As x → ±∞, we have
2x+3
f (x) = x3 −2x2 +4
x 2+ x3
= x3
· 1− x2 + 43
→ 0
x

so the x-axis is the horizontal asymptote.

(c) As x → ±∞, we have
2x2 −3x−1
f (x) = x−2
x2 2− x3 − 12
= x
· 1− x2
x
→∞

so the function has no horizontal asymptote.

The Short-Run Behavior of Rational Functions

We close this section by studying the local behavior of rational functions
which includes the zeros and the vertical asymptotes.

The Zeros of a Rational Function

The zeros of a rational function or its x-intercepts. They are those numbers
that make the numerator zero and the denominator nonzero.

Example 6.10
Find the zeros of each of the following functions:

x2 +x−2 x2 +x−2
(a) f (x) = x−3
(b) g(x) = x−1
.

Solution.
(a) Factoring the numerator we find x2 + x − 2 = (x − 1)(x + 2). Thus, the
zeros of the numerator are 1 and −2. Since the denominator is different from
zeros at these values then the zeros of f (x) are 1 and -2.

71
(b) The zeros of the numerator are 1 and -2. Since 1 is also a zero of the
denominator then g(x) has -2 as the only zero.

Vertical Asymptote
When the graph of a function either grows without bounds or decay without
bounds as x → a then we say that x = a is a vertical asymptote. For
rational functions, the vertical asymptotes are the zeros of the denominator.
Thus, if x = a is a vertical asymptote then as x approaches a from either
sides the function becomes either positively large or negatively large. The
graph of a function never crosses its vertical asymptotes.

Example 6.11
2x−11
Find the vertical asymptotes of the function f (x) = x2 +2x−8

Solution.
Factoring x2 + 2x − 8 = 0 we find (x − 2)(x + 4) = 0. Thus, the vertical
asymptotes are the lines x = 2 and x = −4.

Graphing Rational Functions

f (x)
To graph a rational function h(x) = g(x)
:

1. Find the domain of h(x) and therefore sketch the vertical asymptotes
of h(x).
2. Sketch the horizontal or the oblique asymptote if they exist.
3. Find the x − intercepts of h(x) by solving the equation f (x) = 0.
4. Find the y-intercept: h(0)
5. Draw the graph

Example 6.12
x(4−x)
Sketch the graph of the function f (x) = x2 −6x+5

Solution.

1. Domain = (−∞, 1) ∪ (1, 5) ∪ (5, ∞). The vertical asymptotes are x = 1

and x = 5.
2. As x → ±∞, f (x) ≈ −1 so the line y = 1 is the horizontal asymptote.
3. The x-intercepts are at x = 0 and x = 4.
4. The y-intercept is y = 0.

72
5. The graph is given in Figure 55.

Figure 55

When Numerator and Denominator Have Common Zeros

2 +x−2
We have seen in Example ??, that the function g(x) = x x−1 has a common
zero a4 x = 1. You might wonder What the graph looks like. For x 6= 1, the
function reduces to g(x) = x + 2. Thus, the graph of g(x) is a straight line
with a hole at x = 1 as shown in Figure 56.

73
 Figure 56

Recommended Problems (pp. 42 - 5): 3, 4, 6, 7, 9, 11, 13, 14,

23, 25, 26, 27, 32, 34.

74
7 Velocity as a Rate of Change
In this section, we discuss the concepts of the average rate of change and the
instantaneous rate of change of a given function. As an application, we use
the velocity of a moving object.

The motion of an object along a line at a particular instant is very difficult

to define precisely. The modern approach consists of computing the average
velocity over smaller and smaller time intervals containing the instant. To
be more precise, let s(t) be the position function of a moving object at time
t. We would like to compute the velocity of the object at the instant t = t0 :

Average Velocity
We start by finding the average velocity of the object over the time interval
t0 ≤ t ≤ t0 + ∆t given by the expression
Distance T raveled s(t0 + ∆t) − s(t0 )
v= = .
Elapsed T ime ∆t
Geometrically, the average velocity over the time interval [t0 , t0 + ∆t] is just
the slope of the line joining the points (t0 , s(t0 )) and (t0 + ∆t, s(t0 + ∆t)) on
the graph of s(t).(See Figure 57)

Figure 57
Example 7.1
An object moves a distance s(t) = 16t2 feet from its starting point in t sec-
onds. Complete the following table

time interval [1.8,2] [1.9,2] [1.99,2] [1.999,2] [2,2.0001] [2,2.001] [2,2.01]

Average velocity

75
Solution.

time interval [1.8,2] [1.9,2] [1.99,2] [1.999,2] [2,2.0001] [2,2.001] [2,2.01]

Average velocity 60.8 62.4 63.84 63.98 64.0016 64.016 64.16

In general, the average rate of change of a function y = f (x) over the

interval [a, b] is the difference quotient

f (b) − f (a)
.
b−a
Instantaneous Velocity and Speed
The next step is to calculate the average velocity on smaller and smaller time
intervals containing t0 ( that is, make ∆t close to zero). The average velocity
in this case approaches what we would intuitively call the instantaneous
velocity at time t = t0 which is defined using the limit notation by

s(t0 + ∆t) − s(t0 )

v(t0 ) = lim
∆t→0 ∆t
Geometrically, the instantaneous velocity at t0 is the slope of the tangent
line to the graph of s(t) at the point (t0 , s(t0 )).(See Figure 58)

Figure 58

Example 7.2
For the distance function in Example 7.1, find the instantaneous velocity at
t = 2.

76
Solution.
Examining the bottom row of the table in Example 7.1, we see that the aver-
age velocity seems to be approaching the value 64 as we shrink the time in-
tervals. Thus, it is reasonable to expect the velocity to be v(2) = 64 f t/sec.

Speed of a Moving Object

We define the speed of a moving object to be the absolute value of the veloc-
ity function. Sometimes there is confusion between the words ”speed” and
”velocity”. Speed is a nonnegative number that indicates how fast an object
is moving, whereas velocity indicates both speed and direction(relative to a
coordinate system). For example, if the object is moving along a vertical line
we define a positive velocity when the object is going upward and a negative
velocity when the object is going downward.

The Informal Definition of Limit

The notation of limit was introduced in defining the instantaneous velocity
v(t). Here is a general, informal, definition of the limit of a function.
The notation
lim f (x) = L
x→a

is read ”the limit of f(x) as x approaches a is L” and means that the values of
f (x) approach a unique number L when the variable x is chosen sufficiently
close to a, from either side, but not equal to a.
Sometimes we will find it convenient to represent the notation

lim f (x) = L
x→a

by writing f (x) → L as x → a.

Example 7.3 (Graphical evaluation of a limit)

Evaluate graphically, limθ→0 sinθ θ .

Solution.
sin θ
The graph of f (θ) = θ
is given by Figure 59. From this, we see that

sin θ
lim = 1.
θ→0 θ

77
 Figure 59

Example 7.4 (Evaluating a limit numerically)

2 2
Evaluate numerically limt→0 16(2+t) t−16(2) .

Solution.

t -.2 -.1 -.01 -0.001 0 .0001 .001 .01

16(2+t)2 −16(2)2
t
60.8 62.4 63.84 63.98 undefined 64.0016 64.016 64.16
From the table we conclude that
16(2 + t)2 − 16(2)2
lim = 64.
t→0 t
Example 7.5 Evaluating a limit algebraically
2 2
Use algebra to find limt→0 16(2+t) t−16(2) .

Solution.
Expanding the numerator gives

16(2 + t)2 − 16(2)2 16t2 + 64t

= .
t t
Thus,
16(2 + t)2 − 16(2)2
lim = lim(16t + 64) = 64.
t→0 t t→0

78
Generalizing the idea of instantaneous velocity, we define the instantaneous
rate of change of a function y = f (x) at x = a to be

f (x) − f (a)
lim
x→a x−a
or alternatively, by letting h = x − a

f (a + h) − f (a)
lim
h→0 h
Example 7.6
Evaluate
f (a + h) − f (a)
lim
h→0 h
when f (x) = x2 .

Solution.
Since f (x + h) = (x + h)2 = x2 + 2xh + h2 and f (x) = x2 then

f (a + h) − f (a) 2xh + h2
lim = lim = lim (2x + h) = 2x.
h→0 h h→0 h h→0

Recommended Problems (pp. 61 - 62): 1, 2, 3, 4, 5, 6, 8, 9, 10, 11,

13, 15, 16, 18, 20.

79
8 The Formal Definition of Limit
In this section we will explore the concept of the limit of a function and then
develop algebraic techniques for computing limits in a systematic way.

Definition and Properties of Limit

The limit of a function f is a tool for investigating the behavior of f (x) as
x gets closer and closer to a particular number a. For example, the function
s(t) = 16t2 of Example 7.1 approaches the value 64 as t approaches 2.
The notation

lim f (x) = L (1)

x→a

is read ”the limit of f(x) as x approaches a is L” and means that the values of
f (x) approach a unique number L when the variable x is chosen sufficiently
close to a, from either side, but not equal to a.
Sometimes we will find it convenient to represent the notation

lim f (x) = L
x→a

by writing f (x) → L as x → a.

Geometrically, the notation (1) means that for any tiny interval around L
that you choose, you can find a tiny interval around a such that the func-
tion f takes any point in the tiny interval around a inside the tiny interval
around L. This means that for any  > 0 (as small as we want), there is a
δ > 0 (sufficiently small) such that if the distance between a point x and a
is less than δ (i.e. the point x is inside the tiny interval around a) then the
distance between f (x) and L is less than  (i.e, the point f (x) is inside the
tiny interval around L.). See Figure 60.

80
 Figure 60

Symbolically, (1) is equivalent to

For every  > 0, there exists a δ > 0 such that if 0 < |x − a| < δ then |f(x) − L| < . (2)

(2) is known as the formal definition of the limit of a function.

Remark 8.1
Realize that in the above definition the  is given and the δ is to be found.
In practice, one finds an explicit expression for δ in terms of .

Example 8.1
Show, by using the formal definition of limit, that limx→2 x2 = 4.

Solution.
Let  > 0 be a small given positive number. We must find a δ > 0 such that
if 0 < |x − 2| < δ then |x2 − 4| < . Note that |x2 − 4| = |x − 2||x + 2|. Since
0 < |x − 2| < δ then |x + 2| = |(x − 2) + 4| ≤ |x − 2| + 4 < δ + 4. Since δ is
supposed to be small so we can assume that δ < 1 so that |x + 2| < 4 + δ < 5.
Thus, |x2 − 4| < 5δ. So choose 0 < δ < 1 such that  = 5δ, i.e., δ = 5 .

Using the  − δ definition in finding the limit of a function is sometimes

cumbersome. Instead, one can use the following basic properties listed in the
next theorem.

81
Theorem 8.1
Suppose that limx→a f (x) = L and limx→a g(x) = L0 . Then

(a) limx→a (f (x) + g(x)) = L + L0 .

(b) limx→a (f (x)g(x))
 = L · L0 .
(c) limx→a fg(x)(x)
= LL0 , provided that L0 6= 0.
(d) limx→a k = k, f or any constant k.
(e) limx→a x = a.
(f) If f (x) is a polynomial then limx→a f (x) = f (a).

Proof.
(a) Let  > 0 be given. Then there exist δ1 > 0 and δ2 > 0 such that

0 < |x − a| < δ1 implies |f (x) − L| <
2
and

0 < |x − a| < δ2 implies |g(x) − L0 | < .
2
Let δ be the smallest of the two numbers δ1 and δ2 . Assume that 0 < |x−a| <
δ. Then 0 < |x − a| < δ1 and 0 < |x − a| < δ2 . Moreover,
 
|(f (x)+g(x))−(L+L0 )| = |(f (x)−L)+(g(x)−L0 )| ≤ |f (x)−L|+|g(x)−L0 | < + = .
2 2
This establishes the first property.
(b) First, note the following

|f (x)g(x) − LL0 | = |(f (x) − L)g(x) + (g(x) − L0 )L|

≤ |f (x) − L||g(x)| + |g(x) − L0 ||L|
= |f (x) − L||g(x) − L0 + L0 | + |g(x) − L0 ||L|
≤ |f (x) − L|(|g(x) − L0 | + |L0 |) + |g(x) − L0 ||L|.

Let  > 0 be given. Then there is δ1 > 0 such that if 0 < |x − a| < δ1 then
|g(x) − L0 | < 2L

. Also, there is a δ2 > 0 such that if 0 < |x − a| < δ2 then

( +|L0 |)
|f (x) − L| < 2|L|2 . Now, choose δ to be the smallest of δ1 and δ2 . Then
for 0 < |x − a| < δ we have
  1 
|f (x)g(x) − LL0 | < ( + |L0 |)  0
+ |L| = .
2 2|L| 2|L|
+ |L | 2|L|

82
This proves (b).
1
(c) We will show that limx→a g(x)
= L10 . Let δ1 > 0 be such that whenever
0 0
0 < |x − a| < δ then |g(x) − L0 | < |L2 | . But |L0 | − |g(x)| ≤ |g(x) − L0 | < |L2 | .
0
This implies that |g(x)| > |L2 | for 0 < |x − a| < δ1 . On the other hand, there
0 2
is a δ2 > 0 such that whenever 0 < |x − a| < δ2 then |g(x) − L0 | < |L2| . Let
δ be the smallest of the δ1 and δ2 . Then
0
   
=  g(x)−L − L0 | < .
 1 1  2
 g(x) −  < |g(x)
 
L0 g(x)L0 |L0 |2


This result and (b) yield a proof for (c).

(d) Let  > 0 be given. Choose δ = . Then, if 0 < |x − a| < δ then
|k − k| = 0 < .
(e) Let  > 0 be given. Choose δ = . Then 0 < |x−a| < δ implies |x−a| < .
(f) This follows from (a) - (e).

Example 8.2
Evaluate the following limits:
(i) limx→2 (2x5 − 9x3 + 3x2 − 11).
x3 −3x+7
(ii) limx→−1 5x 2 +9x+6 .

Solution.
(i) limx→2 (2x5 − 9x3 + 3x2 − 11) = 2(2)5 − 9(2)3 + 3(2)2 − 11 = −7
x3 −3x+7 (−1)3 −3(−1)+7 9
(ii) limx→−1 5x 2 +9x+6 = 5(−1)2 +9(−1)+6 = 2 .

Sided Limits
By graphing the function f (x) = |x| x
, x 6= 0, (See Figure 61) one notices that
limx→0 f (x) does not exist. However, there is a limit if we approach zero
from either the right ( in this case the limit is 1) or the left of zero( the limit
is -1).

83
 Figure 61
This leads to the following definitions:

Right-Hand Limit
We write limx→a+ f (x) = L if we can make the values of f (x) as close to L
as we please by choosing x sufficiently close to a from the right.
Left-Hand Limit
We write limx→a− f (x) = L if we can make the values of f (x) as close to L
as we please by choosing x sufficiently close to a from the left.
Remark 8.2
When we write limx→a f (x) = L we mean that f (x) approaches L from
both sides of a. That is, limx→a f (x) = L if and only if limx→a− f (x) =
limx→a+ f (x) = L.

Limits That Do Not Exist

Whenever there is no number L such that limx→a f (x) = L, we say that
limx→a f (x) does not exist. Below we exhibit three examples in which limits
fail to exist.
Example 8.3
|x|
Explain why limx→0 x
does not exist.
Solution.
According to Figure 61, we have that limx→0+ |x|
x
= 1 and limx→0− |x|
x
= −1.
|x|
Thus, by the above remark, limx→0 x does not exist.

84
Example 8.4
Use a calculator to graph the function f (x) = sin ( x1 ) and then conclude that
limx→0 f (x) does not exist.

Solution.
The graph of f (x) = sin ( x1 ) is given by Figure 62.

Figure 62

It follows that the graph oscillates between -1 and 1 as x approaches 0. Hence,

limx→0 sin ( x1 ) does not exist.

Example 8.5
(i) Use a calculator to graph f (x) = x12 .
(ii) Use part (i) to show that limx→0 f (x) = ∞.

Solution.
According to Figure 63, we see that the function increases without bound as
x approaches 0. We write this using the notation

lim f (x) = ∞.
x→0

85
 Figure 63

Infinite Limits and Vertical Asymptotes

A function f that decreases or increases without bound as x approaches a is
said to tend to infinity at a. Symbolically, we write

lim f (x) = ∞
x→a

if f increases without bound and

lim f (x) = −∞
x→a

if f decreases without bound.

In this case, we say that the graph of f has a vertical asymptote at x = a.

Warning. It is important to keep in mind that ∞ is not a number, but

is merely a symbol denoting unrestricted growth in the magnitude of the
function.

Example 8.6
1
Investigate limx→0+ x
and limx→0− x1 .

Solution.
1 1
According to Figure 64, we see that limx→0+ x
= ∞ and limx→0− x
= −∞.
Thus, the y-axis is a vertical asymptote.

86
 Figure 64

Limits at Infinity and Horizontal Asymptotes

Sometimes we want to know what happens to f (x) as x increases without
bound (i.e. x → ∞) or x decreases without bound (i.e., x → −∞). If f (x)
gets as close to a number L as we please when x gets sufficiently large then
we write
lim f (x) = L.
x→∞

Similarly, if f (x) gets as close to a number L as we please when x decreases

without bound then we write

lim f (x) = L.
x→−∞

In this case, the graph of f (x) has a horizontal asymptote at y = L.

Example 8.7
1
Investigate limx→∞ x
and limx→−∞ x1 .

Solution.
1 1
According to Figure 64, we see that limx→∞ x
= 0 and limx→−∞ x
= 0. Thus,
the x-axis is a horizontal asymptote.

Recommended Problems (pp. 68 - 9): 1, 2, 7, 9, 12, 13, 15, 18, 20,

22, 24, 31, 33, 35, 37, 39.

87
9 The Concept of Continuity
The goal of this section is to introduce the concept of continuity. Graphically,
a function is said to be continuous if its graph has no holes or jumps. Stated
differently, a continuous function has a graph which can be drawn without
lifting the pencil from the paper.

Continuity at a Point
We say that a function f (x) is continuous at x = a if and only if the func-
tional values f (x) get closer and closer to the value f (a) as x is sufficiently
close to a. We write
lim f (x) = f (a).
x→a

This means, that for any given  > 0 we can find a δ > 0 such that

|x − a| < δ implies |f (x) − f (a)| < .

In words, we say ”f is continuous at a” if, for each open interval J containing

f (a), we can find an open interval I containing a so that for each point x in
I, f (x) lies in the interval J. That is, given  > 0 we can find δ > 0 such that

|x − a| < δ implies |f (x) − f (a)| < .

Example 9.1
(i) Theorem 8.1(f) shows that polynomials P (x) are continuous everywhere.
(ii) Similarly, if h(x) = fg(x)
(x)
is a rational function and a is a number such
that g(a) 6= 0 then
f (x) f (a)
lim =
x→a g(x) g(a)
That is, h(x) is continuous at numbers that are not zeroes of the function
g(x). (iii) Trigonometric functions are continuous at values where they are
defined.

Discontinuity
A function f (x) that is not continuous at x = a is said to be discontinuous
there. We exhibit three examples of discontinuous functions.

Example 9.2 (Removable Discontinuity)

2 +x−2
Show that the function f (x) = x x−1 is discontinuous at x = 1.

88
Solution.
Graphing the given function (see Figure 65) er find

Figure 65
The dot indicates an excluded point on the graph. Thus, we see that f (1) is
undefined and therefore f (x) is discontinuous at x = 1. Note that

lim f (x) = 3.
x→1

Thus, if we redefine f (x) in such a way that f (1) = 3 then we create a

continuous function at x = 1. That is, the discontinuity is removable.

Example 9.3 (Infinite Discontinuity)

Show that f (x) = x12 is discontinuous at x = 0.

Solution.
According to Figure 63, we have that limx→0 x12 does not exist. Thus, f (x) is
discontinuous at x = 0. Since limx→0 f (x) = ∞ then we call x = 0 an infinite
discontinuity.

Example 9.4 (Jump Discontinuity)

Show that f (x) = |x|
x
is discontinuous at x = 0.

Solution.
The fact that f (x) is discontinuous at x = 0 follows from Figure 61.

The limit properties of Theorem 8.1 can be used to prove the following prop-
erties of continuous functions.

89
Theorem 9.1
If f and g are two continuous functions at x = a and k is a constant then all
of the following functions are continuous at x = a.

◦ Scalar Multiple kf
◦ Sum and Difference f ± g
◦ Product f · g
◦ Quotient fg , provided that g(a) 6= 0.

Theorem 9.2 (Continuity of Composite Functions)

The composition of two continuous functions is continuous.

Proof.
Let g be a function continuous at a and f a function continuous at g(a).
we will show that f ◦ g is continuous at a. Let  > 0 be given. Since f is
continuous at g(a) then there is a δ 0 > 0 such that

|x − g(a)| < δ 0 implies |f (x) − f (g(a))| < .

Now, since g is continuous at a then there is a δ 0 > 0 such that

|x − a| < δ implies |g(x) − g(a)| < δ 0 .

Thus,
|x − a| < δ implies |f (g(x)) − f (g(a))| < .
This completes a proof of the theorem.

Continuity on an Interval
We say that a function f is continuous on the open interval (a, b) if it
is continuous at each number in this interval. If in addition, the function is
continuous from the right of a, i.e. limx→a+ f (x) = f (a), then we say that
f is continuous on the interval [a, b). If f is continuous from the left of b,i.e.
limx→b− f (x) = f (b) then we say that f is continuous on the interval (a, b].
Finally, if f is continuous on the open interval (a, b), from the right at a and
from the left at b then we say that f is continuous in the interval [a, b].

Example 9.5
Find the interval(s) on which each of the given functions is continuous.

90
 2
(i) f1 (x) = xx2 −1
−4
.
(ii) f2 (x) = csc x
sin ( x1 ).
(iii) f3 (x) = (
3 − x, if −5 ≤ x < 2
(iv) f5 (x) =
x − 2, if 2 ≤ x < 5

Solution.
(i) (−∞, −2) ∪ (−2, 2) ∪ (2, ∞).
(ii) x 6= nπ where n is an integer.
(iii) (−∞, 0) ∪ (0, ∞) (See Example 8.4).
(iv) Since limx→2− f5 (x) = limx→2− (3 − x) = 1 and limx→2+ (x − 2) = 0 then
f5 is continuous on the interval [−5, 2) ∪ (2, 5).

The Intermediate Value Theorem

Continuity can be a very useful tool in solving equations. So if a function is
continuous on an interval and changes sign then definitely it has to cross the
x-axis. This shows that the function possesses a zero in that interval. This
is a special case of the following theorem.

Theorem 9.3 (Bolzano Theorem)

Suppose that f is continuous on [a, b] and f (a) and f (b) have opposite signs.
Then there is a number a ≤ c ≤ b such that f (c) = 0.

Proof.
Without loss of generality we assume that f (a) < 0 < f (b). The proof uses a
result from advanced calculus known as the ”Least Upper Bound Axiom”:

”If S is a nonempty set of real numbers such that |x| ≤ M for all x ∈ S
then there is a number c such that x ≤ c for all x ∈ S. Moreover, if d is any
number such that x ≤ d for all x ∈ S then c ≤ d. We call c a least upper
bound of S.”

Now, let s = {a ≤ x ≤ b : f (x) < 0}. Since f (a) < 0 then a belongs to
S so that S is nonempty. Also, |x| ≤ b for all x ∈ S. By the Least upper
Bound Axiom there is a least upper bound number a ≤ c ≤ b. We will show
that f (c) = 0.
Suppose that f (c) < 0. Since f is continuous at c and (2f (c), 0) is an open
interval around f (c) then there is an open interval I = (s, t) around c such

91
that 2f (c) < f (x) < 0 for all x in I. This shows that I is contained in S.
Thus, t ∈ S and c < t. This contradicts the definition of c. Hence, f (c) ≥ 0.
A similar argument shows that f (c) cannot be positive. Hence, f (c) = 0.
This concludes a proof of the theorem.

Theorem 9.4 (Intermediate Value Theorem)

Let f be a continuous function on [a, b] with f (a) < f (b). If f (a) < d < f (b)
then there is a ≤ c ≤ b such that f (c) = d.

Proof.
Define F (x) = f (x) − d. Then, clearly, F (x) is continuous on [a, b]. Moreover,
F (a) = f (a) − d < 0 and F (b) = f (b) − d > 0. By the Bolzano Theorem
there is a ≤ c ≤ b such that F (c) = 0. That is, f (c) = d.

Example 9.6
Show that cos x = x3 − x has at least one zero on the interval [ π4 , π2 ].

Solution.
Let f (x) = cos x − x3 + x. Since f ( π4 ) ≈ 1.008 > 0 and f ( π2 ) ≈ −2.305 < 0
then by the IVT there is at least one number c in the interval ( π4 , π2 ) such
that f (c) = 0.

Recommended Problems (pp. 47 - 8): 1, 2, 5, 7, 9, 11, 14, 15,

16, 18, 19.

92
10 The Derivative at a Point
In this section we introduce the definition of the derivative and its geometrical
significance.
In Section 7, we defined the instantaneous rate of change of a function f (x)
at a point x = a to be the value that the difference quotient or the average
rate of change
f (a + h) − f (a)
h
approaches over smaller and smaller intervals(i.e. when h → 0). This instan-
taneous rate of change is called the derivative of f (x) with respect to x
at x = a and will be denoted by f 0 (a). Thus,

f (a + h) − f (a)
f 0 (a) = lim .
h→0 h
If this limit exists then we say that f is differentiable at a. To differ-
entiate a function f (x) at x = a means to find its derivative at the point
(a, f (a)). The process of finding the derivative of a function is known as
differentiation.

Example 10.1 √
Use the definition of the derivative to find f 0 (x) where f (x) = x, x > 0.

Solution.

f 0 (x) = limh→0 f √ (x+h)−f (x)

h √
x+h− x
= lim
√ h→0 √ h√ √ 
= limh→0 x+h− h
x √x+h+ x
x+h+ x
√
x+h−x
= limh→0 h(√x+h+√x)
1 √
= limh→0 √x+h+ x
1
= √
2 x

Going back to Figure 58 of Section 7, we conclude that the number f 0 (a) is

the slope of the tangent line to the graph of f (x) at the point (a, f (a)). The
equation of the tangent line to the graph of f (x) at x = a is then given by
the formula
y − f (a) = f 0 (a)(x − a).

93
The equation of the normal line to the graph of f (x) at x = a is given by
1
y − f (a) = − (x − a),
f 0 (a)
assuming that f 0 (a) 6= 0.

Example 10.2
(i) Find the derivative of the function f (x) = x2 at x = 1.
(ii) Write the equation of the tangent line to the graph of f at the point
(1, f (1)).

Solution.
(i)
f 0 (1) = limh→0 f (1+h)−f
h
(1)
2 −1
= limh→0 (1+h)h
2 −1
= limh→0 1+2h+h h
= limh→0 h(2+h)
h
= limh→0 (2 + h) = 2.
(ii) The equation of the tangent line is given by

y − f (1) = f 0 (1)(x − 1)

or in slope-intercept form
y = 2x − 1.

Example 10.3
Find the equation of the line that is perpendicular to the tangent line to
f (x) = x2 at x = 1.

Solution.
The equation of the line is given by

y = mx + b.

Since m × f 0 (1) = −1 and f 0 (1) = 2 then m = − 21 . Thus, y = − 12 x + b.

Since the line crosses the point (1, 1) then 1 = − 12 + b or b = 32 . Hence, the
equation of the normal line is
1 3
y =− x+ .
2 2

94
Remark 10.1
By letting x = a + h in the definition of f 0 (a) we obtain an alternative form
of f 0 (a) which is useful in computations and is given by

f (x) − f (a)
f 0 (a) = lim .
x→a x−a
Example 10.4
Show that the derivative of f (x) = sin x at x = 0 is 1.

Solution.
Using the above remark and Example 7.3 of Section 7 we find

f 0 (0) = limx→0 f (x)−f

x−0
(0)

= limx→0 sin x−sin

x
0
sin x
= limx→0 x = 1

Recommended Problems (pp. 76 - 8): 1, 4, 7, 8, 11, 13, 15, 17, 19,

20, 22, 24, 25, 29, 30, 33.

95
11 The Derivative Function
Recall that a function f is differentiable at x if the following limit exists

f (x + h) − f (x)
f 0 (x) = lim . (3)
h→0 h
Thus, we associate with the function f , a new function f 0 whose domain is
the set of points x at which the limit (31.1) exists. We call the function f 0
the derivative function of f.

The Derivative Function Graphically

Since the derivative at a point represents the slope of the tangent line then
one can obtain the graph of the derivative function from the graph of the
original function. It is important to keep in mind the relationship between
the graphs of f and f 0 . If f 0 (x) > 0 then the tangent line must be tilted
upward and the graph of f is rising or increasing. Similarly, if f 0 (x) < 0 then
the tangent line is tilted downward and the graph of f is falling or decreasing.
If f 0 (a) = 0 then the tangent line is horizontal at x = a.

Example 11.1
Sketch the graph of the derivative of the function shown in Figure 66.

Figure 66

Solution.
Note that for x < −1.12 the derivative is positive and getting less and less
positive. At x ≈ −1.12 we have f 0 (−1.12) = 0. For −1.12 < x < 0 the

96
derivative is negative and getting more and more negative till reaching x = 0.
For 0 < x < 1.79 the derivative is less and less negative and at x = 1.79 we
have f 0 (1.79) = 0. Finally, for x > 1.79 the derivative is getting more and
more positive. Thus, a possible graph of f 0 is given in Figure 67.

Figure 67

The Derivative Function Numerically

Here, we want to estimate the derivative of a function defined by a table.
The derivative can be estimated by using the average rate of change or the
difference quotient
f (a + h) − f (a)
f 0 (a) ≈ .
h
If a is a left-endpoint then f 0 (a) is estimated by
f (b) − f (a)
f 0 (a) ≈
b−a
where b > a. If a is a right-endpoint then f 0 (a) is estimated by
f (a) − f (b)
f 0 (a) ≈
a−b
where b < a. If a is an interior point then f 0 (a) is estimated by
!
0 1 f (a) − f (b) f (c) − f (a)
f (a) ≈ +
2 a−b c−a
where b < a < c.

97
Example 11.2
Find approximate values for f 0 (x) at each of the x-values given in the follow-
ing table
x 0 5 10 15 20
f(x) 100 70 55 46 40
Solution.
f (5)−f (0)
f 0 (0) ≈  5  = −6
1 f (10)−f (5) f (5)−f (0)
f 0 (5) ≈ 2 5
+ 5
= −4.5

1 f (15)−f (10) f (10)−f (5)
f 0 (10) ≈ 2 5
+ 5
= −2.4

0 1 f (20)−f (15) f (15)−f (10)
f (15) ≈ 2 5
+ 5
= −1.5
f (20)−f (15)
f 0 (20) ≈ 5
= −1.2
The Derivative Function From a Formula
Now, if a formula for f is given then by applying the definition of f 0 (x) as
the limit of the difference quotient we can find a formula of f 0 as shown in
the following two problems.
Example 11.3 (Derivative of a Constant Function)
Suppose that f (x) = k for all x. Find a formula for f 0 (x).
Solution.

f 0 (x) = limh→0 f (x+h)−f

h
(x)

= limh→0 k−kh
= 0.
Thus, f 0 (x) = 0.
Example 11.4 (Derivative of a Linear Function)
Find the derivative of the linear function f (x) = mx + b.
Solution.

f 0 (x) = limh→0 f (x+h)−f

h
(x)

= limh→0 m(x+h)+b−(mx+b)
h
= limh→0 mhh
= m.
Thus, f 0 (x) = m.

Recommended Problems (pp. 83 - 5): 1, 3, 5, 8, 11, 15, 16, 19, 21,

27, 28, 29, 33, 35, 40.

98
12 Leibniz Notation for The Derivative
When dealing with mathematical models that involve derivatives it is conve-
nient to denote the prime (or Newton) notation of the derivative of a function
dy
y = f (x) by dx . That is,
dy
= f 0 (x).
dx
This notation is called Leibniz notation (due to W.G. Leibniz). For exam-
dy
ple, we can write dx = 2x for y 0 = 2x.
When using Leibniz notation to denote the value of the derivative at a point
a we will write 
dy 

dx x=a
dy
Thus, to evaluate dx
= 2x at x = 2 we would write

dy 
 = 2x|x=2 = 2(2) = 4.
dx x=2

Remark 12.1
When you think about it, the Leibniz notation better indicates what is going
on when you take a derivative than does the Newton notation. For one
thing, it clearly shows that a derivative of a function is taken with respect
to a particular independent variable. This will prove to be handy when we
deal with the applications of the derivative in Section 29.

One of the advantages of Leibniz notation is the recognition of the units

of the derivative. For example, if the position function s(t) is expressed in
meters and the time t in seconds then the units of the velocity function ds dt
are meters/sec.
In general, the units of the derivative are the units of the dependent variable
divided by the units of the independent variable.

Example 12.1
The cost, C ( in dollars) to produce x gallons of ice cream can be expressed
as C = f (x). What are the units of measurements and the meaning of the
statement dC = 1.4?

dx

x=200

99
Solution.
dC
dx
is measured in dollars per gallon. The notation

dC 
 = 1.4
dx x=200

means that if 200 gallons of ice cream have already been produced then the
cost of producing the next gallon will be roughly 1.4 dollars.

Recommended Problems (pp. 88 - 9): 1, 3, 5, 7, 9, 11, 13, 15,

16, 19.

100
13 The Second Derivative
Let f (x) be a differentiable function. If the limit

f 0 (x + h) − f 0 (x)
lim
h→0 h
exists then we say that the function f 0 (x) is differentiable and we denote its
derivative by f 00 (x) or using Leibniz notation

d2 y
!
d dy
= .
dx2 dx dx

We call f 00 (x) the second derivative of f (x).

Now, recall that if f 0 (x) > 0 ( resp. f 0 (x) < 0) over an interval I then
the function f (x) is increasing (resp. decreasing on I). So if f 00 (x) > 0 on I
then f 0 (x) is increasing on I. This means that the slope of the tangent line
to the graph of f (x) is either getting more and more positive or less and less
negative. This occurs only when the graph of f is bending up on the interval
I. In this case, we say that f is concave up on I. Similarly, if f 00 (x) < 0 on I
then f 0 (x) is decreasing. This means that the slope of the tangent line to the
graph of f is either getting less and less positive or more and more negative.
This occurs when the graph of f is bending down. In this case, we say that
f is concave down.

Remark 13.1
Note that when a curve is concave up then the tangent lines lie below the
curve whereas when it is concave down then the tangent lines lie above the
curve.

Example 13.1
Give the signs of f 0 and f 00 for the function whose graph is given in Figure
68.

101
 Figure 68

Solution.
Since f is always increasing then f 0 is always positive. Since the graph is
concave up then f 00 is always positive.

Example 13.2
Find where the graph of f (x) = x3 + 3x + 1 is concave up and where it is
concave down.

Solution.
Finding the first and second derivatives of f we obtain f 0 (x) = 3x2 + 3 and
f 00 (x) = 6x. Thus, the graph of f is concave up for x > 0 and concave down
for x < 0.

As an application to the second derivative, we consider the motion of an

object determined by the position function s(t). Recall that the velocity of
the object is defined to be the first derivative of s(t), i.e.
ds
v(t) = s0 (t) =
dt
and the absolute value of v(t) is the speed. When the object speeds up we say
that he/she accelerates and when the object slows down we say that he/she
decelerates. We define the acceleration of an object as the derivative of the
velocity function and consequently as the second derivative of the position
function
d2 s ds
a(t) = 2 = .
dt dt

102
Example 13.3
A particle is moving along a straight line. If its distance, s, to the right of a
fixed point is given by Figure 69, estimate:

(a) When the particle is moving to the right and when it is moving to the
left.
(b) When the particle has positive acceleration and when it has negative
acceleration.

Figure 69

Solution.
(a) When s is increasing (i.e. v > 0) then the particle moves to the right.
This occurs when 0 < t < 23 and for t > 2. On the other hand, the particle
moves to the left when s is decreasing (i.e. v < 0). This happens when
2
3
< t < 2.
(b) Positive acceleration occurs when the graph is concave up. This occurs
when t > 34 . The particle has negative acceleration when the curve is concave
down, i.e for t < 43 .

Recommended Problems (pp. 93 - 4): 1, 2, 3, 4, 5, 9, 10, 11,

13, 15, 20, 22, 23.

103
14 Continuity and Differentiability
So far we have been discussing functions that are differentiable. What about
nondifferentiable functions? How do we know when a function is not differ-
entiable?
A function fails to be differentiable at a point if:

• the point is a sharp corner point. In this case, the left-hand derivative
and the right-hand derivative are different and therefore the limit of the dif-
ference quotient does not exist.
• The tangent line is vertical at the point since vertical lines have no slopes.
• The function is discontinuous at a point. (See Theorem below.)

Example 14.1
Show that the function f (x) = |x| is not differentiable at x = 0. This is an
example of a nondifferentiable function at a corner point.

Solution.
f 0 (0) would exist if the following limit exists and is equal to f 0 (0)

f (0 + h) − f (0) |h|
lim = lim .
h→0 h h→0 h

According to Figure 61 of Section 8 of these notes, limh→0− |h|

h
= −1 whereas
|h| |h|
limh→0+ h = 1. Thus, limh→0 h does not exist. This shows that f (x) is not
differentiable at x = 0.

Example 14.2
1
Show that f (x) = x 3 is not differentiable at x = 0. This is an example of a
nondifferentiable function at a point where the tangent line is vertical.

Solution.
Figure 70 shows the graph of f (x).

104
 Figure 70
Notice that at x = 0 the tangent line is vertical. Looking at the difference
quotient at x = 0 we find
1
f (0 + h) − f (0) h3 1
lim = lim = lim 2 = ∞.
h→0 h h→0 h h→0 h 3

Thus, f (x) has a vertical tangent at x = 0 and f 0 (0) does not exist.
Example 14.3
Show that the piecewise defined function
(
x + 1 if x ≤ 1
f (x) =
3x − 1 if x > 1
is not differentiable at x = 1. This shows an example of a nondifferentiable
function at a sharp corner point.
Solution.
Finding the limit of the difference quotient from both the left of 1 and the
right of 1 we obtain
1+h+1−2 h
lim− = lim− = 1
h→0 h h→0 h

and
3 + 3h − 1 − 2 3h
lim+ = lim+ = 3.
h→0 h h→0 h
Thus,
f (1 + h) − f (1)
lim
h→0 h
0
does not exist and consequently, f (1) does not exist.

105
Example 14.4
(
−x2 + 5 if x ≤ 2
f (x) =
x − 2 if x > 2
is not differentiable at x = 2. This shows an example of a nondifferentiable
function at a point of discontinuity.

Solution.
Finding the left hand derivative we obtain
2
limh→0− f (2+h)−f (2)
h
= limh→0− 5−(2+h)h −(5−4)
= limh→0− (−h − 4) = −4.

Similarly, the right hand derivative is

f (2+h)−f (2)
limh→0+ h
= limh→0+ (2+h)−2−1
h
= limh→0+ h−1
h
= −∞.

It follows that f 0 (2) does not exist.

Now, if the graph of a function has a tangent line at a point, then we would
expect to draw the graph continuously (without the pencil leaving the pa-
per) near the point. This suggests that a differentiable function at a point
is continuous at that point. We state this result as a theorem and then we
provide a proof.

Theorem 14.1
If a function f (x) is differentiable at x = a then it is continuous there.

Proof.
Since f 0 (a) exists then

f (x) − f (a)
lim = f 0 (a).
x→a x−a
Thus,
 
f (x)−f (a)
limx→a [f (x) − f (a)] = limx→a x−a
(x − a) = f 0 (a) · 0 = 0

That is, limx→a f (x) = f (a) and this shows that f is continuous at x = a.

106
Remark 14.1
According to Problem 14.3, a continuous function need not be differentiable.
That is, the converse of the above theorem is not true in general. So be
careful not to consider all continuous functions to be differentiable.

Recommended Problems (pp. 98 - 9): 1, 3, 5, 7, 9, 10.

107
15 Derivatives of Power and Polynomial Func-
tions
Finding the derivative function by using the limit of the difference quotient is
sometimes difficult for functions with complicated expressions. Fortunately,
there is an indirect way for computing derivatives that does not compute
limits but instead uses formulas which we will derive in this section and in
the coming sections.
We first derive a couple of formulas of differentiation.
Theorem 15.1
If f is differentiable and k is a constant then the new function kf (x) is
differentiable with derivative given by
[kf (x)]0 = kf 0 (x).
Proof.

[kf 0 (x) = limh→0 kf (x+h)−kf

h
(x)
= limh→0 k(f (x+h)−f
h
(x))

= k limh→0 f (x+h)−f
h
(x)
= kf 0 (x)
where we used the fact that a constant can be taking across the limit sign
by the properties of limits.
Theorem 15.2
If f (x) and g(x) are two differentiable functions then the functions f + g and
f − g are also differentiable with derivatives
[f (x) ± g(x)]0 = f 0 (x) ± g 0 (x)
Proof.
Again by using the definition of the derivative and the fact that the limit of
a sum/difference is the sum/difference of limits we find
[f (x) + g(x)]0 = limh→0 (f (x+h)+g(x+h))−(f
h
(x)+g(x))

= limh→0 (f (x+h)−f (x))+(g(x+h)−g(x))

h
f (x+h)−f (x)
= limh→0 h
+ limh→0 g(x+h)−g(x)
h
= f 0 (x) + g 0 (x).
The same proof is valid for the difference formula.

Next, we state and give a partial proof of a rule for finding the derivative of
a power function of the form f (x) = xn .

108
Theorem 15.3 (Power Rule)
For any real number n, the derivative of the function y = xn is given by the
formula
dy
= nxn−1
dx
Proof.
We prove the result when n is a positive integer. We start by writing the
definition of the derivative of any function f (x)

f (x + h) − f (x)
f 0 (x) = lim .
h→0 h
Letting h = ax − x we can rewrite the previous definition in the form

f (ax) − f (x)
f 0 (x) = lim .
a→1 ax − x
Thus,
(ax)n − xn an − 1
f 0 (x) = lim = xn−1 lim .
a→1 ax − x a→1 a − 1

Dividing an − 1 by a − 1 by the method of synthetic division we find

an − 1 = (a − 1)(1 + a + a2 + a3 + · · · + an−1 ).

Thus,
f 0 (x) = xn−1 lim (1 + a + a2 + · · · + an−1 ) = nxn−1 .
a→1

Example 15.1
Use the power rule to differentiate the following:
4 1
(a) y = x 3 (b) y = √
3x (c) y = xπ .

Solution.
1
(a) Using the power rule with n = 43 to obtain y 0 = 43 x 3 .
1
(b) Since y = x− 3 then using the power rule with n = − 13 to obtain y 0 =
4
− 31 x− 3 .
(c) Using the power rule with n = π to obtain y 0 = πxπ−1 .

109
Remark 15.1
The derivative of a function of the form y = 2x is not y 0 = x2x−1 because
y = 2x is an exponential function and not a power function. A formula for
finding the derivative of an exponential function will be discussed in the next
section.

Now, combining the results discussed above, we can find the derivative of
functions that are combinations of power functions of the form axn . In par-
ticular, the derivative of a polynomial function f (x) = an xn + an−1 xn−1 +
· · · + a1 x + a0 is given by the formula

f 0 (x) = nan xn−1 + (n − 1)an−1 xn−2 + · · · + a1 .

Example 15.2 √ x5
Find the derivative of the function y = 3x7 − 5
+ π.

Solution. √
The derivative is f 0 (x) = 7 3x6 − x4 .

Example 15.3 √ 10 1
Find the second derivative of y = 5 3 x − x4
+ √
2 x
.

Solution.
1
Note that the given function can be written in the form y = 5x 3 − 10x−4 +
1 −21
2
x . Thus, the first derivative is

5 2 1 3
y 0 = x− 3 + 40x−5 − x− 2 .
3 4
The second derivative is
10 − 5 3 5
y 00 = − x 3 − 200x−6 + x− 2 .
9 8
Recommended Problems (pp. 111 - 3): 1, 2, 5, 9, 10, 12, 13, 14,
22, 24, 26, 29, 35, 43, 45, 47, 49, 53, 57.

110
16 Derivatives of Exponential Functions
We start this section by looking at the limit
eh − 1
lim .
h→0 h
The chart below suggests that the limit is 1.

h -0.01 -0.001 -0.0001 0 0.0001 0.001 0.01

eh −1
h
0.995 0.9995 0.99995 undefined 1.0000 1.0005 1.005
Now, let’s try and find the derivative of the function f (x) = ex at any
number x. By the definition of the derivative and the limit above we see that

f 0 (x) = limh→0 f (x+h)−f

h
(x)
x+h x
= limh→0 e h−e
x h
= limh→0 e (eh −1)
h
= ex limh→0 e h−1 = ex .
This means that ex is its own derivative:
d x
(e ) = ex .
dx
Now, suppose that the x in ex is replaced by a differentiable function of x,
say u(x). We would like to find the derivative of eu with respect to x, i.e.,
d
what is dx (eu )?

Theorem 16.1
d u du
(e ) = eu .
dx dx
Proof.
By the definition of the derivative we have
d u eu(x+h) − eu(x)
(e ) = lim .
dx h→0 h
Since u is differentiable at x then by letting
u(x + h) − u(x)
v= − u0 (x)
h
111
we find
u(x + h) = u(x) + (v + u0 (x))h
with limh→0 v = 0. Similarly, we can write

ey+k = ey + (w + ey )k

with limk→0 w = 0. In particular, letting y = u(x) and k = (v + u0 (x))h we

find
0
eu(x)+(v+u (x))h = eu(x) + (w + eu(x) )(v + u0 (x))h.
Hence,
0
eu(x+h) − eu(x) = eu(x)+(v+u (x))h − eu(x)
= eu(x) + (w + eu(x) )(v + u0 (x))h − eu(x)
= (w + eu )(v + u0 (x))h

Thus,
d u(x+h)
−e u(x)
dx
(eu ) = limh→0 e h
= limh→0 (w + eu )(v + u0 (x))
= eu u0

We end this section, by finding the derivative of the function f (x) = ax ,

where a > 0 and a 6= 1. First, note that f (x) = ax = (eln a )x = ex ln a . Thus,
by Theorem 16.1 we see that
d x d x ln a d
(a ) = (e ) = ex ln a (x ln a) = ax ln a.
dx dx dx
Example 16.1
Find the derivative of each of the following functions:

(a) f (x) = 3x (b) y = 2 · 3x + 5 · e3x−4 .

Solution.

(a) f 0 (x) = 3x ln 3.
(b) y 0 = 2(3x )0 + 5(e3x−4 )0 = 2 · 3x ln 3 + 5(3)e3x−4 = 2 · 3x ln 3 + 15 · e3x−4 .

Recommended Problems (pp. 116 - 7): 1, 4, 8, 10, 12, 16, 21,

24, 26, 27, 29, 30, 32, 36, 40, 43, 45.

112
17 The Product and Quotient Rules
At this point we don’t have the tools to find the derivative of either the
2 2
function f (x) = x3 ex or the function g(x) = xex . Looking closely at the
function f (x) we notice that this function is the product of two functions,
2
namely, x3 and ex . On the other hand, the function g(x) is the ratio of two
functions. Thus, we hope to have a rule for differentiating a product of two
functions and one for differentiating the ratio of two functions.
We start by finding the derivative of the product u(x)v(x), where u and v
are differentiable functions:
(u(x)v(x))0 = limh→0 u(x+h)v(x+h)−u(x)v(x)
h
= limh→0 u(x+h)(v(x+h)−v(x))+v(x)(u(x+h)−u(x))
h
= limh→0 u(x + h) limh→0 v(x+h)−v(x)
h
+ v(x) limh→0 u(x+h)−u(x)
h
= u(x)v 0 (x) + u0 (x)v(x).
Note that since u is differentiable so it is continuous and therefore
lim u(x + h) = u(x).
h→0

The formula
d d d
(u(x)v(x)) = u(x) (v(x)) + (u(x))v(x). (4)
dx dx dx
is called the product rule.
Example 17.1
2
Find the derivative of f (x) = x3 ex .
Solution.
2 2
Let u(x) = x3 and v(x) = ex . Then u0 (x) = 3x2 and v 0 (x) = 2xex . Thus,
by the product rule we have
2 2 2 2
f 0 (x) = x3 (2x)ex + 3x2 ex = 2x4 ex + 3x2 ex .

The quotient rule is obtained from the product rule as follows: Let f (x) =
u(x)
v(x)
. Then u(x) = f (x)v(x). Using the product rule, we find u0 (x) = f (x)v 0 (x)+
f 0 (x)v(x). Solving for f 0 (x) to obtain
u0 (x) − f (x)v 0 (x)
f 0 (x) = .
v(x)

113
 u(x)
Now replace f (x) by v(x)
to obtain
u(x)

u(x) 0
 u0 (x)− v(x) v 0 (x)
v(x)
= v(x)
u0 (x)v(x)−u(x)v 0 (x)
v(x)
= v(x)
u0 (x)v(x)−u(x)v 0 (x)
= (v(x))2
.

Example 17.2
x2
Find the derivative of g(x) = ex
.

Solution.
Let u(x) = x2 and v(x) = ex . Then by the quotient rule we have
(x2 )0 ex −x2 (ex )0
f 0 (x) = (ex )2
2xex −x2 ex
= e2x

Example 17.3
Prove the power rule for integer exponents.

Solution.
In Section 15, we proved the result for positive integers. The result is trivially
true when the exponent is zero. So suppose that y = xn with n a negative
1
integer. Then y = x−n where −n is a positive integer. Applying both the
quotient rule and the power rule we find

(0)(x−n ) − (−nx−n−1 )
y0 = = nxn−1 .
x−2n

Recommended Problems (pp. 121 - 2): 1, 5, 9, 11, 22, 25, 29, 31,
33, 38, 41, 43, 44.

114
18 The Chain Rule
In this section we want to find the derivative of a composite function f (g(x))
where f (x) and g(x) are two differentiable functions.

Theorem 18.1
If f and g are differentiable then f (g(x)) is differentiable with derivative
given by the formula
d
f (g(x)) = f 0 (g(x)) · g 0 (x).
dx
This result is known as the chain rule. Thus, the derivative of f (g(x)) is
the derivative of f (x) evaluated at g(x) times the derivative of g(x).

Proof.
By the definition of the derivative we have
d f (g(x + h)) − f (g(x))
f (g(x)) = lim .
dx h→0 h
Since g is differentiable at x then by letting
g(x + h) − g(x)
v= − g 0 (x)
h
we find
g(x + h) = g(x) + (v + g 0 (x))h
with limh→0 v = 0. Similarly, we can write

f (y + k) = f (y) + (w + f 0 (y))k

with limk→0 w = 0. In particular, letting y = g(x) and k = (v + g 0 (x))h we

find

f (g(x) + (v + g 0 (x))h) = f (g(x)) + (w + f 0 (g(x)))(v + g 0 (x))h.

Hence,

f (g(x + h)) − f (g(x)) = f (g(x) + (v + g 0 (x))h) − f (g(x))

= f (g(x)) + (w + f 0 (g(x)))(v + g 0 (x))h − f (g(x))
= (w + f 0 (g(x)))(v + g 0 (x))h

115
Thus,
d
dx
f (g(x)) = limh→0 f (g(x+h))−fh
(g(x))

= limh→0 (w + f 0 (g(x)))(v + g 0 (x))

= f 0 (g(x))g 0 (x).
This completes a proof of the theorem.

Example 18.1
Find the derivative of y = (4x2 + 1)7 .

Solution.
First note that y = f (g(x)) where f (x) = x7 and g(x) = 4x2 + 1. Thus,
f 0 (x) = 7x6 , f 0 (g(x)) = 7(4x2 + 1)6 and g 0 (x) = 8x. So according to the chain
rule, y 0 = 7(4x2 + 1)6 (8x) = 56x(4x2 + 1)6 .

Example 18.2
x
Find the derivative of f (x) = x2 +1
.

Solution.
We already know one way to find the derivative of this function which is the
use of the quotient rule. Another way, is to use the product rule combined
with the chain rule since f (x) = x(x2 + 1)−1 .

f 0 (x) = (x)0 (x2 + 1)−1 + x[(x2 + 1)−1 ]0

= (x2 + 1)−1 − x(x2 + 1)−2 (2x)
1 2
= x2 +1
− (x22x+1)2

Example 18.3
Prove the power rule for rational exponents.

Solution. p
Suppose that y = x q , where p and q are integers with q > 0. Take the qth
power of both sides to obtain y q = xp . Differentiate both sides with respect
to x to obtain qy q−1 y 0 = pxp−1 . Thus,

p xp−1 p pq −1
y0 = p(q−1) = x .
qx q q
p
Note that we are assuming that x is chosen in such a way that x q is defined.

116
Example 18.4
d n
Show that dx x = nxn−1 for x > 0 and n is any real number.

Solution.
Since xn = en ln x then
d n d n ln x n
x = e = en ln x · = nxn−1 .
dx dx x

Recommended Problems (pp. 126 - 7): 1, 8, 11, 13, 18, 27, 31, 32,
38, 42, 49, 51, 53, 54, 55, 57, 64, 70.

117
19 Derivatives of Trigonometric Functions
The goal of this section is to find the derivatives of the six trigonometric
functions.
We start by finding the derivative of y = sin x. For this purpose, we remind
the reader of the following trigonometric identity:

sin (x + h) = sin x cos h + cos x sin h.

sin h
Using the definition of derivative, the above identity and the fact that limh→0 h
=
1, limh→0 cos hh−1 = 0 we find

y0 = limh→0 sin (x+h)−sin

h
x

= limh→0 sin x cos h+cosh x sin h−sin x

= limh→0 sin x(cos h−1)+cos
h
x sin h
cos h−1
= sin x limh→0 h + cos x limh→0 sinh h = cos x.

Now, if y = sin u where u is a function of x then by the chain rule we have

d du
(sin u) = cos u .
dx dx
   
π π
As a result of this rule and the fact that cos x = sin x + 2
and cos x + 2
=
− sin x we can obtain the derivative of cos x :
d d π π
   
(cos x) = sin x + = cos x + = − sin x.
dx dx 2 2
If u is a function of x then by the chain rule
d du
(cos u) = − sin u .
dx dx
Example 19.1
Differentiate: (a) 2 sin (3x) (b) cos (x2 ) (c) esin x .

Solution.
d
(a) dx (2 sin (3x)) = 2 cos (3x)(3x)0 = 6 cos (3x).
d
(b) dx (cos (x2 )) = − sin (x2 )(x2 )0 = −2x sin (x2 ).
d
(c) dx (esin x ) = esin x (sin x)0 = cos xesin x .

118
Example 19.2
(a) Use the quotient rule to find the derivative of the function sec x = cos1 x .
d
Then use the chain rule to find dx (sec u).
(b) Use the quotient rule to find the derivative of the function csc x = sin1 x .
d
Then use the chain rule to find dx (csc u).
sin x
(c) Use the quotient rule to find the derivative of the function tan x = cos x
.
d
Then use the chain rule to find dx (tan u).
(d) Use the quotient rule to find the derivative of the function cot x = tan1 x .
d
Then use the chain rule to find dx (cot u).
Solution.
1
(a) Since sec x = cos x
then by the quotient rule we have

0 (1)0 cos x − (1)(cos x)0 sin x

(sec x) = 2
= = sec x tan x.
cos x cos2 x
By the chain rule
d du
(sec u) = sec u tan u .
dx dx
1
(b) Since csc x = sin x then by the quotient rule we have
(1)0 sin x − (1)(sin x)0 − cos x
(csc x)0 = 2 = = − csc x cot x.
sin x sin2 x
By the chain rule
d du
(csc u) = − csc u cot u .
dx dx
sin x
(c) Since tan x = cos x
then by the quotient rule we have
(sin x)0 cos x − (sin x)(cos x)0 cos2 x + sin2 x 1
(tan x)0 = 2
= 2
= .
cos x cos x cos2 x
By the chain rule
d 1 du
(tan u) = .
dx cos2 u dx
(d) Since cot x = tan1 x then by the quotient rule we have
(1)0 tan x − (1)(tan x)0 − cos12 x 1
(cot x)0 = 2
= 2
=− 2 .
tan x tan x sin x
By the chain rule
d 1 du
(cot u) = − 2 .
dx sin u dx

119
Remark 19.1
When differentiating trigonometric functions, the domain must be given in
radians. Otherwise, conversion and possibly an application of the chain rule
will be required. For example, if f (x) = sin x, and x is in degrees, then
f 0 (x) = 180
π
cos x.

Recommended Problems (pp. 131 - 3): 6, 9, 15, 17, 27, 28, 33, 37,
39, 41, 44, 45, 51.

120
20 Applications of the Chain Rule
In this section we will use the chain rule to find the derivatives of logarithmic
and inverse trigonometric functions.
To find the derivative of ln x we recall the result
d u d
(e ) = eu (u).
dx dx
As a consequence of this rule, we can find the derivative of the function ln x.
Indeed, since eln x = x then differentiating both sides to obtain
d
eln x · (ln x) = 1.
dx
d
Solving for dx
(ln x) we find

d 1 1
(ln x) = ln x = .
dx e x
Example 20.1
(a) Use the chain rule to find the derivative of ln u where u is a function of
x.
(b) What is the derivative of loga x? loga u? where u is a function of x.

Solution.
(a) By the chain rule
d 1 du
(ln u) = .
dx u dx
(b) Since loga x = ln x
ln a
then (loga x)0 = x ln1 a . By the chain rule

d 1 du
(loga u) = .
dx u ln a dx
Example 20.2 q
Differentiate: (a) ln (x2 + 1) (b) 1 + ln (1 − y).

Solution.
(a) By the chain rule

(x2 + 1)0 2x
(ln (x2 + 1))0 = 2
= 2 .
x +1 x +1

121
 q 1
(b) Since 1 + ln (1 − y) = [1 + ln (1 − y)] 2 then by the chain rule
q 1 1 1 1 −y 0
1 + ln (1 − y) = [1 + ln (1 − y)]− 2 (1+ln (1 − y))0 = [1 + ln (1 − y)]− 2 .
2 2 1−y
Next, we find the derivatives of the inverse trigonometric functions.
Recall that y = arcsin x, − π2 ≤ x ≤ π2 , if and only if sin y = x. Differentiating
both sides of this equality with respect to x to obtain
cos y · y 0 = 1
or
1
y0 = .
cos y
√
Using the trigonometric identity cos (arcsin x) = 1 − x2 we see that
d 1
(arcsin x) = √ .
dx 1 − x2
Example 20.3
(a) Find the derivative of arccos x, 0 ≤ x ≤ π.
(b) Find the derivative of arccos u.
(c) Find the derivative of arctan x, − π2 < x < π2 .
(d) Find the derivative of arctan u, where u is a function of x.
Solution.
(a) y = arccos x, 0 ≤ x ≤ π, if and only if cos y = x. Differentiating both
sides of this equality with respect to x to obtain
− sin y · y 0 = 1
or
1
y0 = − .
sin y
√
Using the trigonometric identity sin (arccos x) = 1 − x2 we see that
d 1
(arccos x) = − √ .
dx 1 − x2
(b) By (a) and the chain rule we have
d 1 du
(arccos u) = − √ .
dx 1 − u2 dx

122
(c) y = arctan x, − π2 < x < π2 , if and only if tan y = x. Differentiating both
sides of this equality with respect to x to obtain
1
· y0 = 1
cos2 y
or
y 0 = cos2 y.
Using the trigonometric identity cos (arctan x) = √ 1 we see that
1+x2

d 1
(arctan x) = .
dx 1 + x2
(d) By (c) and the chain rule we have

d 1 du
(arctan u) = .
dx 1 + u2 dx
Example 20.4
Differentiate: (a) arctan (x2 ) (b) arcsin (tan x).

Solution.
(a) By (d) of the previous exercise we have

(x2 )0 2x
(arctan (x2 ))0 = 2 2
= .
1 + (x ) 1 + x4

(b)
1 1 1
(arcsin (tan x))0 = √ (tan x)0
= √ .
1 − tan2 x 1 − tan2 x cos2 x

Recommended Problems (pp. 136 - 8): 1, 4, 7, 8, 10, 15, 19, 23,

26, 29, 36, 39, 43, 44, 46, 49.

123
21 Implicit Differentiation
So far functions have been defined√explicitly, that is, they can be written
in the form y = f (x) such as y = 1 − x2 , −1 ≤ x ≤ 1. This same function
can be defined implicitely by the equation x2 + y 2 = 1 where 0 ≤ y ≤ 1. To
find the derivative of the explicit form we use the chain rule to obtain
d 1 1 1 −x
(1 − x2 ) 2 = (1 − x2 )− 2 (−2x) = √ .
dx 2 1 − x2
To find the derivative of the implicit form we start by differentiating both
sides of the equation with respect to x to obtain
dy
2x + 2y = 0.
dx
dy
Solving this equation for dx
we find
dy x x
= − = −√ .
dx y 1 − x2
The purpose of this section is to find the derivatives of implicit functions.
The process is known as implicit differentiation and consists of the fol-
lowing two steps:

Step 1. Differentiate both sides of the equation with respect to x. Remember

that y is a function of x for part of the curve and use the chain rule when
differentiating terms containing y.
dy
Step 2. Solve the differentiated equation in Step 1 algebraically for dx
.

Example 21.1
Suppose that y is a differentiable function of x such that

x2 y + 2y 3 = 3x + 2y.

Find the equation of the tangent line to the graph at the point (3, 1).

Solution.
Differentiating both sides to obtain

2xy + x2 y 0 + 6y 2 y 0 = 3 + 2y 0

124
Replacing x = 3 and y = 1 we find

6 + 9y 0 + 6y 0 = 3 + 2y 0

Solving for y 0 to obtain y 0 = − 13

3
. Thus, the equation of the tangent line is
given by
3
y − 1 = − (x − 3)
13
or in standard form
3x + 13y − 22 = 0.

Example 21.2
Consider the equation y 3 − xy = −6.
dy
(a) Find dx by implicit differentiation.
(b) Give a table of approximate y-values near (7, 2) for x = 6.8, 6.9, 7.0, 7.1, 7.2.
(c) Find the y-value for x = 6.8 by substituting 6.8 in the equation and solv-
ing for y using a calculator. Compare your answer in part (b).

Solution.
(a) Using implicit differentiation we find
d d d
dx
(y 3 ) − dx (xy) = dx
(−6)
2 dy dy
3y dx − y − x dx = 0
dy
(3y 2 − x) dx = y
dy y
dx
= 3y 2 −x

When x = 6 and y = 2 we have

dy 2 2
= = .
dx 12 − 7 5
(b)Using the tangent line approximation we find
2
y = 2 + (x − 7) = 0.4x − 0.8.
5
We use this equation to calculate the following approximate y-values
x 6.8 6.9 7.0 7.1 7.2
Approximate y 1.92 1.96 2.00 2.04 2.08

125
(c) Letting x = 6.8 in the original equation we find

y 3 − 6.8y + 6 = 0.

Using a graphing calculator we find that y ≈ 1.915.

Remark 21.1
It is important to make sure that y is a differentiable function of x before
applying the implicit differentiation process. For example, there is no real
valued function satisfying the equation x2 + y 2 = −1 so y 0 does not exist. A
careless application of the above technique might lead you to thinking that
y 0 exists and is equal to
dy x
=− .
dx y

Recommended Problems (pp. 140 - 1): 2, 9, 10, 15, 17, 21, 24, 26,
27.

126
22 Parametric Equations
In this section we define parametric equations in the plane and we illustrate
how to draw their graphs.
Parametric equations in the plane is a pair of functions

x = f (t) and y = g(t)

which describe the coordinates of the points of some curve in the plane. The
variable t is referred to as a parameter.

Example 22.1
Write the equation of a circle of radius r in both the Cartesian coordinates
form and the parametric form.

Solution.
The equation of the circle in the Cartesian coordinates system is given by

x2 + y 2 = r 2 .

The parametric equations of this circle are given by

x = cos t and y = sin t 0 ≤ t ≤ 2π.

Example 22.2
Figure 71 shows the graphs of two functions, f (t) and g(t). Describe the
motion of the particle whose coordinates at time t are x = f (t) and y = g(t).

Figure 71

Solution.
For 0 ≤ t ≤ 1, x moves at a constant rate from 0 to 1 and y moves at a

127
constant rate from 1 to 0. So the particle moves in a straight line from (0, 1)
to (1, 0). For 1 ≤ t ≤ 2, the particle moves on a straight line from (1, 0) to
(0, −1). For 2 ≤ t ≤ 3 the particle moves on a straight line from (0, −1) to
(−1, 0). Finally, for 3 ≤ t ≤ 4, the particle moves on the straight line from
(−1, 0) to (0, 1). Thus, the particle traces the diamond shown in Figure 72.

Figure 72

Remark 22.1
Parametric representations, in general, are not unique. That is, the same
curve can be expressed by a number of different parametrizations. For exam-
ple, the representations (x = cos t, y = sin t) and (x = cos (3t), y = sin (3t))
both represent the unit circle. These various representations describe various
motions of a moving particle on the unit circle. Looking at the tables below
we see that the particle on the second curve is moving three times faster than
the one moving on the curve of the previous example.

t cos t sin t t cos (3t) sin (3t)

0 1 0 0 1 0
π π
2
0 1 6
0 1
2π
π -1 0 6
-1 0
3π 3π
2
0 -1 6
0 -1
4π
2π 1 0 6
1 0

Parametric Equations of a Straight Line

Consider a straight line passing through two points (x0 , y0 ) and (x1 , y1 ). Then

128
the equation of the line is given by
y1 − y0
y − y0 = (x − x0 ).
x1 − x0
This is equivalent to
y − y0 x − x0
= .
y1 − y0 x1 − x0
Denote the common value by the parameter t to obtain
y − y0 x − x0
= = t.
y1 − y0 x1 − x0
Equating each fraction seperately to t and solving for x and y we find
x = x0 + (x1 − x0 )t and y = y0 + (y1 − y0 )t.
These equations are referred to as the parametric equations of the straight
line.
Example 22.3
Find the parametric equations of the line passing through the points (2, −1)
and (−1, 5).
Solution.
The parametric equations are given by
x = 2 − 3t and y = −1 + 6t.
Remark 22.2
Note that by restricting the parameter t in the previous problem to the
interval [0, 1] we obtain the parametric equations of the line segment from
(2, −1) to (−1, 5).
Speed and Velocity
Suppose that the motion of a moving particle is given by a curve with para-
metric equations
x = f (t) and y = g(t).
We define the instantaneous speed by the formula
v
u dx 2
u ! !2
dy
v= t
+ .
dt dt
The quantity vx = dxdt
is the instantaneous velocity in the x-direction and
dy
vy = dt is the instantaneous velocity in the y-direction.

129
Example 22.4
The motion of a particle is given by the parametric equations
x = t2 − 3t, y = t2 − 2t.
(a) Does the particle ever come to a stop? If so, when and where?
(b) Is the particle ever moving straight up or down? If so, when and where?
(c) Is the particle ever moving straight horizontally right or left? If so, when
and where?
Solution.
(a) Saying that the particle comes to a stop is equivalent to saying that v = 0,
i.e. vx = vy = 0. Solving the equations
dx
vx = dt
= 3t2 − 3 = 3(t − 1)(t + 1) = 0
dy
vy = dt
= 2t − 2 = 2(t − 1) = 0
give t = 1. Thus, the particle stops when t = 1 at the point (t3 − 3t, t2 −
2t) |t=1 = (−2, −1).
(b) Saying that the particle is traveling straight up or down is equivalent to
saying that vx = 0 and vy 6= 0. Solving the equation vx = dx dt
= 3(t2 − 1) = 0
we find t = ±1. Since vy = 0 at t = 1 then the only value of t is t = −1. The
position of the particle at that time is (t3 − 3t, t2 − 2t) |t=−1 = (2, 3).
(c) Saying that the particle is traveling straight left or right is equivalent to
saying that vx 6= 0 and vy = 0. Solving the equation vy = dy dt
= 2(t − 1) = 0
we find t = 1. Since vx = 0 at t = 1 then the particle is not moving at all.

Slopes of Parametric Curves

Suppose that a curve is defined parametrically by the equations x = f (t)
and y = g(t). The curve can be defined in Cartesian coordinates form either
dy
explicitely or implicitely. In the first case, we can write y = h(x) and find dx
using the chain rule:
dy dy dx
= .
dt dx dt
Thus, we obtain the slope of the curve as a function of t :
dy dy/dt
= .
dx dx/dt
Similar procedure applies if the curve is defined by an implicit function. In
this case we use implicit differentiation.

130
Example 22.5
Find the slope of the tangent line to the curve defined parametrically by
x = t3 − t and y = t2 at t = 2. Find the equation of the tangent line.

Solution.
The point corresponding to t = 2 is (6, 4). We have
  
dy  dy/dt  2t 4
= = t=2 =

dx (6,4) dx/dt (6,4) 3t2 −1 11
 

The equation of the tangent line is given by

4
y−4= (x − 6)
11
or n parametric form by

x = 6 + 11t, “y = 4 + 4t.

Parametrizing y = f (x)
The graph of any function defined in Cartesian coordinates form y = f (x)
can be parametrized by the equations:

x = t and y = f (t).

Example 22.6
Give parametric equations for the vertical line passing through the point
(−2, −3).

Solution.
A possible answer is: x = −2, y = t.

Recommended Problems (pp. 148 - 9): 1, 2, 5, 7, 9, 11, 15, 18, 19,

21, 22, 26, 30.

131
23 Linear Approximations and Differentials
Consider a function f and a point a. We want to find the best linear function
passing through (a, f (a)) that approximate the values of f (x) near a. If we
let g(x) be such a function and E(x) = f (x) − g(x) be the error in the
approximation then by best approximation we mean

E(x)
lim = 0.
x→a x−a
Since g(x) is linear and goes through (a, f (a)) then g(x) has the form

g(x) = g(a) + L(x − a).

Suppose that f is differentiable at a then we choose L = f 0 (a). In this case,

E(x) 0
limx→a x−a
= limx→a f (x)−f (a)−f
x−a
(a)(x−a)

= limx→a f (x)−f
x−a
(a)
− f 0 (a)
= f 0 (a) − f 0 (a) = 0

Thus, g(x) is the best linear approximation of f near a. We call g(x) the
tangent line approximation of f (x) at x = a and we write

f (x) ≈ f 0 (a)(x − a) + f (a)

It follows that the graph of f (x) near a can be thought of as a straight line.

Remark 23.1
The error of the approximation is defined to be E(x)= Exact - Approximate
and is given by
E(x) = f (x) − f 0 (a)(x − a) − f (a).
If E(x) < 0 then the approximation is an overestimate whereas if E(x) > 0
then the approximation is an underestimate. Figure 73 shows the tangent
line approximation and its error.

132
 Figure 73

Example 23.1
Find the tangent line approximation of f (x) = ekx near x = 0.

Solution.
According to the formula above, we have

f (h) ≈ f (0) + f 0 (0)h.

But f (0) = 1 and f 0 (0) = k since f 0 (x) = kekx . Hence, for small h we have

f (h) ≈ 1 + kh

That is, ekx ≈ 1 + kx for x close to 0.

Differentials
Now, suppose f is a differentiable function at x. Then we define the exact
change of x along the curve from x to x+h to be the quantity ∆x = h and the
exact change in y along the curve to be the quantity ∆y = f (x) − f (x + h).
We define the differential of x along the tangent line to the graph of f (x)
to be the quantity dx. Clearly, dx = ∆x. Also, we define the differential of
y along the tangent line to be the product

dy = f 0 (x)dx.

133
That is, dy is the estimated change in f or dy ≈ ∆y.
The Figure 74 exhibits all four quantities ∆x, ∆y, dx, and dy.

Figure 74

Example 23.2
Complete the following:

(a) d(sin x) =
(b) d(ln x) =
(c) d(ex ) =
(d) d(xn ) =

Solution.
(a) d(sin x) = cos xdx
(b) d(ln x) = dxx
(c) d(ex ) = ex dx
(d) d(xn ) = nxn−1 dx

Note that the linear approximation can be written now in terms of differen-
tials
f (a + dx) ≈ f (a) + dy
where dy = f 0 (a)dx.

134
Example 23.3
1
Approximate 3.98 using differentials. Use a calculator to compute the error.
Is the approxmation an overestimate or an underestimate?

Solution.
Let f (x) = 1
x
then f 0 (x) = − x12 . Let a = 4 and dx = −0.02. Then
1
3.98
= f (a + dx) ≈ f (4) + f 0 (4)dx
1 1
= 4
− 16 (−0.02) = 0.25125

Using a calculator, we find

1
Error = − 0.25125 ≈ 6.28 × 10−6 .
3.98
Hence the approximation is an underestimate.

Example 23.4
Estimate the change in the area of a square if its edge length is decreased
from 10 inches to 9.8 inches.

Solution.
Let x be the edge length, so the area is A = x2 = f (x). Since x changes from
10 to 9.8, the change in x is dx = −0.2. We have

dA = f 0 (x)dx = 2xdx,

so the estimated change in A is

dA = 2 · 10 · (−0.2) = −4.

The area will decrease by about 4 square inches.

Recommended Problems (pp. 153 - 4): 1, 2, 3, 4, 5, 6, 7, 9,

15.

135
24 Indeterminate Forms and L’Hôpital’s Rule
In this section we want to study limits of ratios of two functions
f (x)
lim .
x→a g(x)
We have seen that if limx→a f (x) = L and limx→a g(x) = L0 6= 0 then
f (x) L
lim = 0.
x→a g(x) L
Now, if L 6= 0 but L0 = 0 then we say that the limit does not exist. What if
both L and L0 are zero? Sometimes simple algebraic manipulations can be
used to find the limit. For example, if f (x) = x + 3 and g(x) = x2 + x − 6
then
x+3 x+3 1 1
lim = lim = lim =− .
x→−3 x2 + x − 6 x→−3 (x + 3)(x − 2) x→−3 x − 2 5
0
Another way of evaluating limits of the form 0
is by applying the following
result.

Theorem 24.1 (L’Hôpital’s Rule)

Suppose that f and g are differentiable functions on (a,b) and x0 is a point
inside the interval such that f (x0 ) = g(x0 ) = 0. If g 0 (x) 6= 0 in (a,b) except
possibly at x0 then
f (x) f 0 (x)
lim
x→x0 g(x)
= lim
x→x0 g 0 (x)
.

Proof.
We first show that
f (x) f 0 (x)
lim+ = lim+ 0 .
x→x0 g(x) x→x0 g (x)
The left limit case is similar, and the combination of these two cases estab-
lishes the result.
So fix x > x0 . Then the function g(t) is continuous on [x0 , x] and differen-
tiable on (x0 , x) so that we can apply the Mean Value Theorem (See Section
31) and obtain a number cx in (x0 , x) such that
g(x) − g(x0 )
= g 0 (cx ) 6= 0.
x − x0

136
This shows that g(x) 6= 0. Next, consider the function

f (x)
F (t) = f (t) − g(t).
g(x)

This function is continuous in [x0 , x] and differentiable on (x0 , x) with F (x) =

F (x0 ) = 0. So by the Mean Value Theorem there is a c in (x0 , x) such that
F 0 (c) = 0. That is,
f (x) 0
0 = f 0 (c) − g (c)
g(x)
or
f (x) f 0 (c)
= 0 .
g(x) g (c)
Thus,
f (x) f 0 (c) f 0 (x)
lim+ = lim+ 0 = lim+ 0 .
x→x0 g(x) c→x0 g (c) x→x0 g (x)

Remark 24.1 1
f (x) ∞ g(x)
1. In the case that limx→x0 g(x)
= ∞
then rewrite the limit as limx→x0 1
f (x)
and use the above rule.
2. The x0 in the theorem can be replaced by one of the symbols −∞ or ∞.

Example 24.1
e2x −1
Compute limx→0 x
.

Solution.
Using L’Hôpital’s rule we find

e2x − 1 2e2x
lim = lim = 2.
x→0 x x→0 1

Example 24.2
sin x
Use L’Hôpital’s rule to find limx→0 x
.

Solution.
d
sin x (sinx)
limx→0 x
= limx→0 dx
d
(x)
dx
= limx→0 cos x = cos 0 = 1

137
Example 24.3
ex −1−x
Calculate limx→0 x2
.

Solution. Using L’Hôpital’s rule twice we find

ex −1−x x
limx→0 x2
= limx→0 e 2x−1
x
= limx→0 e2 = 12

Example 24.4
Calculate limx→∞ xe−x .

Solution.
Since limx→∞ xe−x = limx→∞ exx is of the form ∞
∞
then we can apply L’Hôpital’s
rule and obtain
x 1
lim xe−x = lim x = lim x = 0.
x→∞ x→∞ e x→∞ e

Example 24.5
5x+e−x
Calculate limx→∞ 7x
.

Solution.
∞
The given limit is of the form ∞
so we can apply L’Hôpital’s rule and obtain

5x + e−x 5 − e−x 5
lim = x→∞
lim = .
x→∞ 7x 7 7
Example 24.6
f (x)
We say that a function g(x) dominates a function f (x) if and only limx→∞ g(x)
=
0, or equivalently, limx→∞ fg(x)
= ∞. Use L’Hôpital’s rule to show that the
(x)
√
function x dominates the function ln x.

Solution.
√ x as x → ∞ is of the form ∞ . So we can apply L’Hôpital’s rule
The limit ln x ∞
to find
1
ln x x 2
lim √ = lim 1 = lim √ = 0
x→∞ x x→∞ 2√x x→∞ x
√
Thus, x dominates ln x.

Recommended Problems (p. 158): 1, 2, 3, 4, 5, 6, 7, 8, 9, 11,

13, 14, 15, 18.

138
25 Using First and Second Derivatives
We start this section by reviewing what the first and second derivatives of a
function tell us about its graph:

• If f 0 (x) > 0 on an open interval I then f (x) is increasing on I.

• If f 0 (x) < 0 on an open interval I then f (x) is decreasing on I.
• If f 00 (x) > 0 on an open interval I then f (x) is concave up on I.
• If f 00 (x) < 0 on an open interval I then f (x) is concave down on I.

Example 25.1
Consider the function f (x) = x3 − 9x2 − 48x + 52.

(a) Find the intervals where the function is increasing/decreasing.

(b) Find the intervals where the function is concave up/down.

Solution.
(a) Finding the first derivative we obtain f 0 (x) = 3x2 − 18x − 48 = 3(x −
8)(x + 2). Constructing the chart of signs below

we see that f (x) is increasing on (−∞, −2)∪(8, ∞) and decreasing on (−2, 8).
(b) Finding the second derivative, we obtain f 00 (x) = 6x − 18 = 6(x − 3). So,
f (x) is concave up on (3, ∞) and concave down on (−∞, 3).

Local Maxima and Minima

Points of interest on the graph of a function are those points that are the
highest on the curve, or the lowest, in a specific interval. Such points are
called local extrema. The highest point, say f (a), is called a local maxi-
mum and satisfies f (x) ≤ f (a) for all x in an interval I. A local minimum
is a point f (a) such that f (a) ≤ f (x) for all x in an interval I containing a.

139
Example 25.2
Find the local maxima and the local minima of the function given in Figure
75.

Figure 75
Solution.
The local maxima occur at x = −2 and x = 14 whereas the local minimum
occurs at x = 8.

Next we will discuss two procedures for finding local extrema. We notice
from the previous example that local extrema occur at points p where the
derivative is either zero or undefined. We call p a critical number, f (p) a
critical value, and (p, f (p)) a critical point. Most of the critical numbers
that we will encounter in this book are of the form f 0 (p) = 0 type. The fol-
lowing theorem asserts that local extrema occur at the critical points. This
theorem is proved in Section 31.
Theorem 25.1
Suppose that f is defined on an interval I and has a local maximum or
minimum at an interior point a. If f is differentiable at a then f 0 (a) = 0.
Remark 25.1
By Theorem 25.1, local extrema are always critical points. The converse of
this statement is not true in general. That is, there are critical points that
are not local extrema of a function. An example of this situation is given
next.

140
Example 25.3
Show that f (x) = x3 has a critical point at x = 0 but 0 is neither a local
maximum nor a local minimum.

Solution.
Finding the derivative to obtain f 0 (x) = 3x2 . Setting this to 0 we find the
critical point x = 0. Since f 0 (x) does not change sign at 0 then 0 is neither
a local maximum nor a local minimum.

The graph in Example 25.2 suggests two tests for finding local extrema. The
first is known as the first derivative test and the second as the second
derivative test.

First-Derivative Test
Suppose that a continuous function f has a critical point at p.
• If f 0 changes sign from negative to positive at p, then f has a local mini-
mum at p.
• If f 0 changes sign from positive to negative at p, then f has a local maxi-
mum at p. See Figure 76.

Figure 76

Example 25.4
(a) Find the local extrema of the function f (x) = x3 − 9x2 − 48x + 52.
(b) Find the local extrema of the function f (x) = sin x + ex , x ≥ 0.

Solution.
(a) Using the chart of signs of f 0 discussed in Example 25.1, we find that
f (x) has a local maximum at x = −2 and a local minimum at x = 8.
(b) Finding the derivative to obtain f 0 (x) = cos x + ex . But for x ≥ 0,

141
1 ≤ ex . Since −1 ≤ cos x ≤ 1 then adding the two inequalities we see that
0 ≤ cos x+ex . This implies that f 0 does not change sign for x ≥ 0. Therefore,
there are no local maxima. The only local minimum occurs at (0, 1).

Second-Derivative Test
Let f be a continuous function such that f 0 (p) = 0.
• if f 00 (p) > 0 then f has a local minimum at p.
• if f 00 (p) < 0 then f has a local maximum at p.
• if f 00 (p) = 0 then the test fails. In this case, it is recommended that you
use the first derivative test.

Figure 77

Example 25.5
Use the second derivative test to find the local extrema of the function f (x) =
x3 − 9x2 − 48x + 52.

Solution.
The second derivative of f (x) is given by f 00 (x) = 6(x − 3). The critical
numbers are −2 and 8. Since f 00 (−2) = −30 < 0 then x = −2 is a local
maximum. Since f 00 (8) = 30 > 0 then x = 8 is a local minimum.

Example 25.6
Find the local extrema of the function f (x) = x4 .

Solution.
Let’s try and find the local extrema by using the second derivative test. Since
f 0 (x) = 4x3 then x = 0 is the only critical number. Since f 00 (x) = 12x2 then
f 00 (0) = 0. So the second derivative test is inconclusive. Now, using the first
derivative test, we see that f 0 (x) changes sign from negative to positive at

142
x = 0. Thus, x = 0 is a local minimum.

Concavity and Points of Inflection

We have seen that a local extremum is a point where the first derivative
changes sign. In this section we will discuss points where the second deriva-
tive changes sign. That is, the points where the graph of the function changes
concavity. We call such points points of inflection.
How do you find the points of inflection? Well, since f 00 changes sign on
the two sides of an inflection point then it makes sense to say that points of
inflection occur at points where either the second derivative is 0 or undefined.

Example 25.7
Find the point(s) of inflection of the function f (x) = xe−x .

Solution.
Using the product rule to obtain f 0 (x) = e−x − xe−x . Using the product rule
for the second time we find f 00 (x) = e−x (x − 2). Thus, a candidate for a point
of inflection is x = 2. Since f 00 (x) > 0 for x > 2 and f 00 (x) < 0 for x < 2 then
x = 2 is a point of inflection.

Remark 25.2
We have seen that not every value of x where the derivative is zero or un-
defined is a local maximum or minimum. The same thing applies for points
of inflection. That is, it is not always true that if the second derivative is 0
or undefined then automatically you have a point of inflection. It is critical
that f 00 changes sign at such a point in order to have a point of inflection.

Example 25.8
Consider the function f (x) = x4 . Show that f 00 (0) = 0 but 0 is not a point
of inflection.

Solution.
The second derivative is given by the formula f 00 (x) = 12x2 . Clearly, f 00 (0) =
0. Since f 00 (x) ≥ 0, that is, f 00 (x) does not change sign then 0 is not a point
of inflection.

Example 25.9
Graph the derivative of the function f (x) = x + sin x. Determine where f is
increasing most rapidly, and least rapidly.

143
Solution.
The derivative of f (x) is given by the expression f 0 (x) = cos x+1 ≥ 0 so that
f (x) is always increasing. Now, f (x) increases most rapidly at the maximum
values of f 0 (x) and increases least rapidly at the minimum values of f 0 (x).
Graphing the function f 0 (x) we find

Thus, f increases most rapidly at x = 2nπ and least rapidly at x = (2n + 1)π
where n is an integer.

Example 25.10
Graph a function with the following properties:

• f has a critical point at x = 4 and an inflection point at x = 8.

• f 0 < 0 for x < 4 and f 0 > 0 for x > 4.
• f 00 > 0 for x < 8 and f 00 < 0 for x > 8.

Solution.
The graph is given in Figure 78.

144
 Figure 78

Recommended Problems (pp. 172 - 5): 1, 2, 3, 4, 6, 7, 8, 9,

10, 11, 12, 13, 18, 20, 22, 24, 26, 32, 40, 41, 42, 43.

145
26 Special Families of Curves
Consider the line y = x. The collection of lines y = x + c is an example of a
one-parameter family of curves. We call c the parameter. In this section,
we will discuss certain family of functions that are of importance to us in
later discussions.

Motion Under Gravity

The motion of a particle along a horizontal or a vertical line is given by

s(t) = −4.9t2 + v0 t + s0

where the parameters v0 and s0 represent the initial velocity and the initial
distance respectively.

Harmonic Motion
A simple harmonic motion is modeled by either the function f (t) = a sin (bt)
or the function f (t) = a cos (bt). But the period of either function is known
to be p = 2π
|b|
. The number |a| is called the amplitude.

Remark 26.1
1. If maximum displacement occurs at t = 0 then the motion is modeled by
the cosine function.
2. If zero displacement occurs at t = 0 then the motion is modeled by the
sine function.

Example 26.1
Find the amplitude and the period of the simple harmoninc motion described
by the equation
2
y = 3 cos t.
3
Solution.
2π 2π
The amplitude is |a| = |3| = 3. The period is = b
= 2 = 3π.
3

Example 26.2
2
Find an equation of a simple harmoninc motion with period p = 3
and
amplitude 4. Assume that maximum displacement occurs at t = 0.

146
Solution.
Since maximum displacement occurs at t = 0 then y = a cos bt. But p = 2/3
and a = 4 so that y = 4 cos 3πt.

Bell Shaped-Curves
Bell-shaped curves occur in many common situations and have been studied
extensively by statisticians. More specifically, we look at the two-parameters
family of curves
(x−a)2
y = e− b

where b > 0. This family is related to the normal density function, studied
in probability and statistics.
2
Let’s first consider the case b = 1. In this case, y = e−(x−a) , and the graph
2
of this function is a horizontal shift of the graph of y = e−x , by a units. The
Figure 79 shows the cases, a = −2, 0, 2.

Figure 79
Next, consider the case where a = 0 and b being arbitrary. Using the chain
rule one finds
2x x2
y 0 = − e− b .
b
Thus, x = 0 is a critical number. Using the product rule together with the
chain rule one finds
2 2x2
!
x2
00
y = − 1 e− b
b b
Thus, y 00 (0) < 0 so that by the second derivative test, (0, 1) is a local max-
x2
imum. Also, notice that at 0 the graph is concave down. But e− b → 0 as

147
x → ±∞. Thus, the function must have points of inflection. The x-values of
these points are the zeros of y 00 = 0. That is,

2x2
!
2 x2
− 1 e− b = 0.
b b

Solving this equation for x we find

s
b
x=± .
2
(x−a)2
Summarizing, we see that the function
q y = e− b has a maximum at x = a
and inflection points at x = a ± 2b . Also, the parameter a gives the center
of the bell whereas the parameter b determines how narrow or wide the bell
is. The smaller b is, the closer the points of inflection are to the center a and
so the bell is sharply peaked near a; the larger b is, the farther the points of
inflection are from a,i.e. the bell is spread out.

Recommended Problems (pp. 179 - 180): 1, 4, 5, 7, 8, 13, 15,

16, 21, 23, 28.

148
27 Global Maxima and Minima
In this section we will look for the largest or the smallest values of a function
on its domain. Such points are called global extrema. If f (a) is the largest
value then it satisfies the inequality f (x) ≤ f (a) for all x in the domain of
f. We call f (a) the global or absolute maximum value of f and the
point (a, f (a)) the global maximum point. Similarly, if f (a) is the smallest
value of f (x) then f (a) ≤ f (x) for all x in the domain of f. We call f (a)
the absolute or global minimum value of f and the point (a, f (a)) the
global minimum.
The process of finding the global extrema is called optimization. Problems
that involve finding the global extrema are called optimization problems.

How do we find the global extrema?

• If the function is continuous on a closed interval then the global extrema
occur at either the critical points or the endpoints of the interval.

Problem 27.1
Find the global extrema of the function f (x) = x3 − 9x2 − 48x + 52 on the
closed interval [−5, 12].

Solution.
Finding the derivative of f (x) we get f 0 (x) = 3x2 − 18x − 48. Solving the
equation f 0 (x) = 0 that is, x2 − 6x − 16 = 0 we find the critical points at
x = 8 and x = −2. Now, evaluating the function at these points and at the
endpoints we find
f (−5) = −58
f (−2) = 104
f (8) = −396
f (12) = −92
It follows that (−2, 104) is the global maximum point and (8, −396) is the
global minimum point.
• If a function is continuous on an open interval or on all real numbers then
it is recommended to find the global extrema by graphing the function.

Problem 27.2
Find the global extrema of the function f (x) = 100(e−0.02x −e−0.1x ) for x ≥ 0.

149
Solution.
Let’s sketch the graph of this function. The standard process of graphing
consists of the following steps:

Step 1. Find the critical numbers. Setting f 0 (x) = 0 to obtain

100(−0.02e−0.02x + 0.1e−0.1x ) = 0
−0.02x
0.02e = 0.1e−0.1x
e−0.02x 0.1
e−0.1x
= 0.02
0.08x
e = 5
0.08x = ln 5
ln 5
x = 0.08
= 20.12

Step 2. We construct the following chart:

x 20.12
f’(x) + 0 -
f(x) % 53.50 &

Step 3. Find the second derivative to obtain f 00 (x) = 100(0.0004e−0.02x −

0.01e−0.1x ). Setting this to zero and solving for x as in Step 1 we find
x ≈ 40.25. Now we construct the table

x 40.25
f”(x) - 0 +
f(x) ∩ f(40.25) ∪
Step 4.Graph

150
 Figure 80

Thus, from the graph we see that (20.12, 53.50) is a global maximum. The
function has a global minimum at x = 0.

Recommended Problems (pp. 185 - 8): 2, 4, 7, 8, 9, 13, 19, 33, 34.

151
28 Applications to Economics
Management of most businesses always aim to maximizing profit. In this
section we will use the derivative to optimizie profit and revenue functions.
We begin by introducing the cost, revenue, and profit functions.

The cost function C gives the cost C(q) of manufacturing a quantity q

of some good. A linear cost function has the form
C(q) = mq + b,
where the y-intercept b is called the fixed cost, i.e. the costs incurred even
if nothing is produced, and the slope m is called the variable costs per
unit.
The function C(x) = C(x)
x
is called the average cost function.
Example 28.1
What is the cost function of manufacturing a product with fixed cost of $400
and variable costs of $40 per item, assuming the function is linear?
Solution.
The cost function is
C(q) = 40q + 400.
Example 28.2
Values of a linear cost function are shown below. What are the fixed costs
and the variable costs per units? Find a formula for the cost function.
q 0 5 10 15 20
C(q) 5000 5020 5040 5060 5080
Solution.
The fixed costs are b = C(0) = $5, 000, the variable costs are
5020 − 5000
m= =4
5−0
The cost function is
C(q) = 4q + 5, 000

A revenue function R gives the total revenue R(q) from the sale of a
quantity q at a unit price p dollars. Thus, R(q) = pq.

152
Example 28.3
A company that makes a certain brand of chairs has fixed costs of $5, 000
and marginal cost of $30 per chair. The company sells the chairs for $50
each. Find formulas for the cost and revenue functions.

Solution.
The cost function is C(q) = 30q + 5000. The revenue function is R(q) = pq =
30q.

Profit is defined to be revenue minus cost. That is

P (q) = R(q) − C(q).

The break-even point is the point where the profit is zero, i.e. R(q) = C(q).

Example 28.4
A company has cost and revenue functions, in dollars, given by C(q) =
6, 000 + 10q and R(q) = 12q.

(a) Graph the functions C(q) and R(q) on the same coordinate axes.
(b) Find the break-even point and illustrate it graphically.
(c) When does the company make a profit? Loses money?

Solution.
(a) The graph is given in Figure 81.

Figure 81

153
(b) The break-even point is the point of intersection of the two lines. To find
the point, set 12q = 10q + 6000 and solve for q to find q = 3000. Thus, the
break-even point is the point (3000, 36000).
(c) The company makes profit for q > 3000 and loses money for q < 3000.

Marginal Analysis
Marginal analysis is an area of economics concerned with estimating the effect
on quantities such as cost, revenue, and profit when the level of production
is changed by a unit amount. For example, if C(q) is the cost of produc-
ing q units of a certain commodity, then the marginal cost, MC(q), is the
additional cost of producing one more unit and is given by the difference
M C(q) = C(q + 1) − C(q). Using the estimation

C(q + 1) − C(q)
C 0 (q) ≈ = C(q + 1) − C(q)
(q + 1) − q

we find that
M C(q) ≈ C 0 (q)
and for this reason, we will compute the marginal cost by the derivative
C 0 (q).
Similarly, if R(q) is the revenue obtained from producing q units of a com-
modity, then the marginal revenue, MR(q), is the additional revenue ob-
tained from producing one more unit, and we compute M R(q) by the deriva-
tive R0 (q).

Example 28.5
Let C(q) represent the cost, R(q) the revenue, and P (q) the total profit, in
dollars, of producing q units.

(a) If C 0 (50) = 75 and R0 (50) = 84, approximately how much profit is earned
by the 51st item?
(b) If C 0 (90) = 71 and R0 (90) = 68, approximately how much profit is earned
by the 91st item?

Solution.
(a) P 0 (50) = R0 (50) − C 0 (50) = 84 − 75 = 9.
(b) P 0 (90) = R0 (90) − C 0 (90) = 68 − 71 = −3. A loss by 3 dollars.

154
Remark 28.1
Marginal cost and average cost can differ greatly. For example, suppose
it costs $1000 to produce 100 units and $1020 to produce 101 units. The
average cost per unit is $10, but the marginal cost of the 101st unit is $20.
Similar remarks apply for the marginal revenue and the marginal profit.

Supply and Demand Curves

The quantity q manufactured and sold depends on the unit price p. In gen-
eral, when the price goes up then manufacturers are willing to supply more
of the product whereas consumers are going to reduce their buyings. Since
consumers and manufacturers react differently to changes in price, there are
two curves relating p and q.

The supply curve is the quantity that producers are willing to make at
a given price. Thus, increasing price will increase quantity.
The demand curve is the amount that will be bought by consumers at a
given price. Thus, decreasing price will increase quantity.
Even though quantity is a function of price, it is the tradition to use the
vertical y-axis for the variable p and the horizontal x-axis for the variable q.
The supply and demand curves intersect at point (q ∗ , p∗ ) called the point of
equilibrium. We call p∗ the equilibrium price and q ∗ the equilibrium
quantity.

Example 28.6
Find the equilibrium point for the supply function S(p) = 3p − 50 and the
demand function D(p) = 100 − 2p.

Solution.
Setting the equation S(p∗ ) = D(p∗ ) to obtain 3p∗ −50 = 100−2p∗ . By adding
2p∗ + 50 to both sides we obtain 5p∗ = 150. Solving for p∗ we find p∗ = 30.
Substituting this value in S(p) we find q ∗ = 3(30) − 50 = 40.

Optimizing Profit
Recall that the profit resulting from producing and selling q items is defined
by
P (q) = R(q) − C(q)
where C(q) is the total cost of producing a quantity q and R(q) is the total
revenue from selling a quantity q of some good.

155
To maximize or minimize profit over a closed interval, we optimize the profit
function P. We know that global extrema occur at the critical numbers of
P or at the endpoints of the interval. Thus, the process of optimization
requires finding the critical numbers which are the zeros of the marginal
profit function
P 0 (q) = R0 (q) − C 0 (q) = 0
where R0 (q) is the marginal revenue function and C 0 (q) is the marginal cost
function. Thus, the global maximum or the global minimum of P occurs
when
M R(q) = M C(q)
or at the endpoints of the interval.

Example 28.7
Find the quantity q which maximizes profit given the total revenue and cost
functions
R(q) = 5q − 0.003q 2
C(q) = 300 + 1.1q.
where 0 ≤ q ≤ 800 units. What production level gives the minimum profit?

Solution.
The profit function is given by

P (q) = R(q) − C(q) = −0.003q 2 + 3.9q − 300.

The critical numbers of P are the solutions to the equation P 0 (q) = 0. That
is,
3.9 − 0.006q = 0
or q = 650 units. Since P (0) = −$300, P (800) = $900 and P (650) = $967.50
then the maximum profit occurs when q = 800 units and the minimum
profit(or loss) occurs when q = 0, i.e. when there is no production.

Example 28.8
The total revenue and total cost curves for a product are given in Figure 82.

156
 Figure 82

(a) Sketch the curves for the marginal revenue and marginal cost on the
same axes. Show on this graph the quantities where marginal revenue equals
marginal cost. What is the significance of these two quantities? At which
quantity is profit maximum?
(b) Graph the profit function P (q).

Solution.
(a) Since R is a straight line with positive slope then its derivative is a
positive constant. That is, the graph of the marginal revenue is a horizontal
line at some value a > 0. Since C is always increasing then its derivative M C
is always positive. For 0 < q < q3 the curve is concave down so that M C is
decreasing. For q > q3 the graph of C is concave up and so M C is increasing.
Thus, the graphs of C and R are shown in Figure 83.

157
 Figure 83
According to the graph, marginal revenue equals marginal cost at the values
q = q1 and q = q2 . So maximum profit occurs either at q1 , q2 or at the
endpoints. Notice that the production levels q1 and q2 correspond to the two
points where the tangent line to C is parallel to the tangent line to R. Now,
for 0 < q < q1 we have M R < M C so that P 0 = M R − M C < 0 and this
shows that P is decreasing. For q1 < q < q2 , M R > M C so that P 0 > 0
and hence P is increasing. So P changes from decreasing to increasing at
q1 which means that P has a minimum at q1 . Now, for q > q2 we have that
M R < M C so that P 0 < 0 and P is decreasing. Thus, P changes from
increasing to decreasing at q2 so q2 is a local maximum for P. So maximum
profit occurs either at the endpoints or at q2 . Since profit is negative for
q < q3 and q > q4 then the profit is maximum for q = q2 .
(b)

Figure 84

158
Recommended Problems (pp. 194 - 6): 1, 3, 5, 6, 7, 9, 10, 15.

159
29 Mathematical Modeling and Optimization
The branch of mathematics that consists of developing a mathematical frame-
work that is used to solve a real world problem is referred to as mathemat-
ical modeling.
With the calculus skills that we have developed so far, we will use mathemati-
cal modeling and calculus to solve optimization problems. By optimization
we mean the process of finding either the maximum or the minimum of a
function.
We outline a general optimization procedure:

Step 1. Draw a figure (if appropriate) and label all quantities relevant to
the problem.
Step 2. Name the quantity to be optimized. Find a formula for the quantity
to be optimized.
Step 3. Use conditions in the problem to eliminate variables in order to ex-
press the quantity to be optimized in terms of a single variable.
Step 4. Find the practical domain of the variable in Step 3.
Step 5. If possible, use the methods you have learned so far to obtain the
optimum value.

Example 29.1
A carpenter wants to make an open-topped box out of a rectangular sheet
of tin 24 inches wide and 45 inches long. The carpenter plans to cut con-
gruent squares out of each corner of the sheet and then bend and solder the
edges of the sheet upward to form the sides of the box, as shown in Figure 85.

Figure 85

For what dimensions does the box have the greatest possible volume?

160
Solution.
Let x be the side of each square. When constructing the box, the length of
the base is 45 − 2x, the width is 24 − 2x and the height of the box is x. Thus,
the volume is given by the formula
V = x(24 − 2x)(45 − 2x) = 4x3 − 138x2 + 1080x.
This quantity is to be maximized. The domain of this function consists of
those numbers that satisfy x ≥ 0, 24 − 2x ≥ 0, and 45 − 2x ≥ 0. Solving this
compound inequality we find that the domain is the interval [0, 12].
We next find the critical numbers of the function V (x) which are the zeros
of V 0 (x). That is, we solve the equation
12x2 − 276x + 1080 = 0
or after dividing by 12 and factoring
(x − 18)(x − 5) = 0.
Thus, the critical numbers are x = 18 and x = 5. The value x = 18 is
to be discarded since it is not in the interval [0, 12]. Evaluating V (x) at
x = 0, x = 5, and x = 12 we find V (0) = 1080, V (5) = 2450, and V (12) = 0.
Thus, V (x) is maximized at x = 5. So the box with the largest volume has
dimensions 5 in × 14 in × 35 in.
Example 29.2
You need to fence a rectangular play zone for children. What is the maximum
area for this play zone if it is to fit into a right-triangular plot with sides
measuring 4 m and 12 m?
Solution.
We start by drawing a picture of the play zone as shown in Figure 86.

161
 Figure 86

Let x and y be the dimensions of the inscribed rectangle. Then the quantity
to be maximized is the area of the rectangle given by A = xy. First, we
must express A as a function of a single variable. To do this, note that, the
triangles ABC and ADF are similar triangles so that the corresponding sides
are proportional and therefore we can write
4−y x
=
4 12
or y = 4 − 31 x. Substituting back into A to obtain A = x(4 − 31 x) = 4x − 13 x2
with 0 ≤ x ≤ 12. Next, we find the critical numbers that are the zeros of
the derivative function A0 (x) = 4 − 23 x. Solving for x we find x = 6. Evalu-
ating A at the endpoints and at the critical point we find A(0) = A(12) = 0
and A(6) = 12. Thus, the maximum area occurs when x = 6. In this case,
y = 4 − 13 (6) = 2. Thus, the largest rectangular play zone that can be built
in the triangular plot is a rectangle 6 m long and 2 m wide.

Recommended Problems (pp. 201 - 3): 1, 3, 4, 6, 7, 11, 14, 16,

24.

162
30 Hyperbolic Functions
The hyperbolic functions possess many parallels with the trigonometric func-
tions, both in naming and in their properties and inter-relationships.
The function ex and its reciprocal e−x can be combined to create three new
functions:

Hyperbolic Cosine
ex + e−x
cosh x = .
2
Hyperbolic Sine
ex − e−x
sinh x = .
2
Hyperbolic Tangent
ex − e−x
tanh x = .
ex + e−x
Example 30.1
Find cosh 0 and sinh 0.
Solution.
Substituting 0 for x in the formulas for cosh x and sinh x to obtain
e0 + e−0 1+1
cosh 0 = = =1
2 2
and
e0 − e−0 1−1
sinh 0 = = = 0.
2 2

Hyperbolic Functions Identities

Hyperbolic functions have similar identities then that of trigonometric func-
tions.
Substituting −x for x in the formula of cosh x to obtain
e−x + ex
cosh (−x) = = cosh x.
2
This says that cosh x is an even function.
Similarly, substituting −x for x in the formula of cosh x to obtain
e−x − ex
sinh (−x) = = − sinh x.
2
163
That is, sinh x is an odd function.
Next, we derive an identity for the hyperbolic functions similar to the Pythagorean
identity for the trigonometric functions.
2 2
ex +e−x ex −e−x
 
cosh2 x − sinh2 x = 2
− 2
e2x +2+e−2x e2x −2+e−2x
= 4
− 4
= 1.
Thus, (cosh x, sinh x) is a point on the hyperbola x2 − y 2 = 1, thus, the name
hyperbolic. (Recall that the functions sin x and cos x are called circular func-
tions since (cos x, sin x) is a point on the unit circle!)

Derivatives of Hyperbolic Functions

Sine (ex )0 = ex and (e−x )0 = −e−x then we have
ex + e−x ex − e−x
!
d d
(cosh x) = = = sinh x.
dx dx 2 2
Similarly,
ex − e−x ex + e−x
!
d d
(sinh x) = = = cosh x.
dx dx 2 2
To find the derivative of tanh x, we use the quotient rule
cosh2 x − sinh2 x
!
d d sinh x 1
(tanh x) = = 2 = .
dx dx cosh x cosh x cosh2 x

Graphs of Hyperbolic Functions

The graphs of cosh x and sinh x are given in Figure 87.

164
 Figure 87

Recommended Problems (p. 206): 1, 3, 5, 7, 12, 13, 14, 15, 20.

165
31 The Mean Value Theorem
We start this section by showing that the derivative of a function evaluated
at a local extremum is always zero.

Theorem 31.1
If a differentiable function f has a local extremum a inside an interval I then
f 0 (a) = 0.

Proof.
We will prove the theorem for a local maximum. The proof for a local
minimum is similar. So suppose that f has a local maximum at a on an
open interval I. Then f (x) ≤ f (a) for all x in I. In particular, for h close
to zero the point a + h is in the interval I and therefore f (a + h) ≤ f (a)
or f (a + h) − f (a) ≤ 0. If h is close to zero from the left (i.e. h < 0) then
Dividing both sides by h to obtain f (a+h)−f
h
(a)
≥ 0.(Note the inequality sign
is reversed since we are dividing by a negative h.) Now, take the limit of
both sides as h → 0− to obtain
f (a + h) − f (a)
lim− ≥0
h→0 h
That is, f 0 (a) ≥ 0. If h is close to zero from the right, i.e. h > 0 then by
dividing both sides of the inquality f (a + h) − f (a) ≤ 0 by h and then letting
h → 0+ we obtain
f (a + h) − f (a)
lim+ ≤0
h→0 h
That is, f 0 (a) ≤ 0. We have shown that 0 ≤ f 0 (a) ≤ 0. This is only true
when f 0 (a) = 0.

We have seen in Section 27 that a continuous function on a closed inter-

val has a maximum and a minimum at either the critical numbers of the
function or at the endpoints of the interval. This is a consequence of the
following theorem

Theorem 31.2 (The Extreme Value Theorem)

A continuous function on a closed interval [a, b] has both a global maximum
and a global minimum on that interval.

166
The proof of this theorem is beyond the scope of this course and is therefore
omitted.
Now consider the following problem: Let f (x) be a continuous function on
[a, b] and differentiable in (a, b). Is there a point on the graph of f where the
tangent line at that point is parallel to the line passing through the points
(a, f (a)) and (b, f (b))? The answer is yes according to the following theorem

Theorem 31.3 (Mean Value Theorem)

If f is continuous on [a, b] and differentiable in (a, b) then there is a number
a < c < b such that the tangent line to the graph at (c, f (c)) is parallel to
the line passing through the points (a, f (a)) and (b, f (b)). That is,
f (b) − f (a)
f 0 (c) = .
b−a
Proof.
We first consider the case when f (a) = f (b). If f is a constant function then
f 0 (x) = 0 in (a, b) and therefore letting c = a+b
2
we have f 0 (c) = 0 = f (b)−f
b−a
(a)
.
So suppose that f is not a constant function. By Theorem 31.2, f has a global
maximum and a global minimum and they are both different since f is not
constant. So one of them, called it c, must be in (a, b). By Theorem 31.1,
f 0 (c) = 0 = f (b)−f
b−a
(a)
.
Now, suppose that f (a) 6= f (b). Consider the function
f (b) − f (a)
g(x) = f (x) − f (a) − (x − a).
b−a
Note that g is differentiable in [a, b] and continuous in (a,b). Moreover,
g(a) = g(b) = 0. So by the first case, there is a a < c < b such that g 0 (c) = 0.
That is,
f (b) − f (a)
0 = g 0 (c) = f 0 (c) −
b−a
or
f (b) − f (a)
f 0 (c) = .
b−a
Example 31.1
Use the Mean Value Theorem to show the following: Suppose that f is
continuous on [a, b] and differentiable on (a, b). If f 0 (x) = 0 on (a, b) then
f is a constant function. That is, f (x) = c for all x in [a, b] where c is a
constant.

167
Solution.
We just need to show that any two values x1 and x2 in [a, b] have the same
y-value, that is, f (x1 ) = f (x2 ). By the MVT, there is x1 < c < x2 such that
f 0 (c) = f (xx22)−f
−x1
(x1 )
. But f 0 (c) = 0 so that f (xx22)−f
−x1
(x1 )
= 0. This implies that
f (x2 ) − f (x1 ) = 0 or f (x1 ) = f (x2 ).
Theorem 31.4 (The Increasing Function Theorem)
Suppose that f (x) is continuous on [a, b] and differentiable in (a, b).

(a) If f 0 (x) > 0 on [a, b] then f is increasing on [a,b](that is if x1 < x2

then f (x1 ) < f (x2 ), where a ≤ x1 , x2 ≤ b).
(b) If f 0 (x) ≥ 0 on (a,b) then f is nondecreasing on [a,b](that is if x1 < x2
then f (x1 ) ≤ f (x2 ), where a ≤ x1 , x2 ≤ b).
Proof.
Let x1 and x2 be two numbers in [a, b] such that x1 < x2 . By the Mean Value
Theorem, there is a number c inside the interval (x1 , x2 ) such that
f (x2 ) − f (x1 ) = f 0 (c)(x2 − x1 )
(a) Since f 0 (x) > 0 in (a,b) then the right hand side is positive. That is,
f (x2 ) − f (x1 ) > 0 or f (x1 ) < f (x2 ).
(b) Since f 0 (x) ≥ 0 then f (x2 )−f (x1 ) = f 0 (c)(x2 −x1 ) ≥ 0 or f (x1 ) ≤ f (x2 ).
Theorem 31.5 (The Racetrack Principle)
Suppose that f and g are continuous on [a, b] and differentiable in (a, b).
Furthemore, suppose that f 0 (x) ≤ g 0 (x) in (a, b).
(a) If f (a) = g(a) then f (x) ≤ g(x) for all a ≤ x ≤ b.
(b) If f (b) = g(b) then f (x) ≥ g(x) for all a ≤ x ≤ b.

Proof.
(a) Let h(x) = g(x) − f (x). Since f 0 (x) ≤ g 0 (x) for all a < x < b then
h0 (x) ≥ 0 for all a < x < b. By part (b) of the previous exercise, h(x)
is nondecreasing. So, for any a ≤ x ≤ b we have h(a) ≤ h(x). But
h(a) = f (a) − g(a) = 0. Thus, h(x) ≥ 0 or f (x) ≤ g(x).
(b) Now, for a ≤ x ≤ b we have h(x) ≤ h(b). But h(b) = f (b) − g(b) = 0.
Thus, h(x) ≤ 0 or f (x) ≥ g(x).

Recommended Problems (pp. 210 - 1): 1, 3, 7, 9, 11, 12, 13,

15, 19, 24, 26, 27.

168
32 Measuring The Distance Traveled
We have seen that the velocity of an object moving along the curve s(t) is
obtained by taking the average rate of change on smaller and samller intervals
, that is finding the derivative of s, i.e. v(t) = s0 (t). In this and the following
sections we want to go the opposite direction. That is, given the velocity
function v(t) we want to find the position function s(t).
To be more precise, suppose that we want to estimate the distance s traveled
by a car after 10 seconds of departure. Assume for example, that we are
given the velocity of the car every two seconds as shown in the table below
Time (sec) 0 2 4 6 8 10
Velocity (ft/sec) 20 30 38 44 48 50

Since we don’t know the instantaneous velocity of the car at every momentm
then we can not calculate the distance exactly. What we can do is to estimate
the distance traveled. For the first two seconds, the velocity is at least 20
miles per second so that the distance traveled is at least 20 × 2 = 40 feet.
Likewise, at least 30 × 2 = 60 feet has been traveled the next two seconds
and so on. Thus, we obtain a lower estimate to the exact distance traveled

20 × 2 + 30 × 2 + 38 × 2 + 44 × 2 + 48 × 2 = 360 f eet.

However, we can reason differently and get an overestimate to the total dis-
tance traveled as follows: For the first two seconds the car’s velocity is at
most 30 feet so that the car travels at most 30 × 2 = 60 feet. In the next two
seconds, it travels 38 × 2 = 76 feet and so on. So an upper estimate of the
total distance traveled is

30 × 2 + 38 × 2 + 44 × 2 + 48 × 2 + 50 × 2 = 420 f eet

Hence,
360 f eet ≤ T otal distance traveled ≤ 420 f eet.
Notice that the difference between the upper and lower estimates is 60 feet.
Figure 88 shows both the lower estimate and the upper estimate. The graph
of the velocity is obtained by plotting the points given in the above table and
then connect them with a smooth curve. The area of the lower rectangles
represent the lower estimate and the larger rectangles represent the upper
estimate.

169
 Figure 88

To visualize the difference between the upper and lower estimates, look at
the above figure, and imagine that all the unshaded rectangles are pushed to
the right and stacked on top of each other. This gives a rectangle of width 2
and height 30 so its area is the difference between the estimates.

Example 32.1
Suppose that the velocity of the car is given every second instead as shown
in the table below. Find the lower and upper estimates of the total distance

170
traveled. What is the difference between the lower and upper estimates? Do
you think that knowing the velocity at every second is a better estimate than
knowing the velocity every two seconds?

Time (sec) 0 1 2 3 4 5 6 7 8 9 10
Velocity (ft/sec) 20 26 30 35 38 42 44 46 48 49 50

Solution.
The lower estimate is

(20)(1) + (26)(1) + · · · + (48)(1) + (49)(1) = 378 f eet

and the upper estimate is

(26)(1) + (30)(1) + · · · + (49)(1) + (50)(1) = 408 f eet

Hence,
378 f eet ≤ T otal distance traveled ≤ 408 f eet.
So the difference between the upper and lower estimates is 408 − 378 =
30 f eet. This shows that by increasing the partition points we get better and
better estimate.

Remark 32.1
Once the upper estimate and the lower estimate are found then one can get
an even better estimate by taking the average of the two estimates.

The use of the average rate of change of the distance leads to finding the total
distance traveled. This same method can be used to find the total change
from the rate of change of other quantities.

Example 32.2
The following table gives world oil consumptions, in billions of barrels per
year. Estimate the total oil consumption during this 20-year period.
Year 1980 1985 1990 1995 2000
Oil (barrels/yr) 22.3 23.0 23.9 24.9 27.0

171
Solution.
We underestimate the total oil consumption as follows:

22.3 × 5 + 23.0 × 5 + 23.9 × 5 + 24.9 × 5 = 470.5 billion barrels.

The overestimate is

23.0 × 5 + 23.9 × 5 + 24.9 × 5 + 27.0 × 5 = 494 billion barrels.

A good single estimate of the total oil consumption is the average of the
above estimates. That is
470.5 + 494
T otal oil consumption ≈ = 482.25 billion barrels.
2
• Finding the Exact Distance Traveled
Suppose that we want to find the total distance traveled over the time interval
a ≤ t ≤ b. We take measurements of the velocity v(t) at equally spaced times,
a = t0 , t1 , t2 , · · · , tn = b. This means that we devide the interval [a, b] into n
equal pieces each of length ∆t = b−a n
. We first use the left-end point of each
interval [ti−1 , ti ] and construct the left-hand sum

L(v, n) = v(t0 )∆t + v(t1 )∆t + · · · + v(tn−1 )∆t.

Geometrically, this sum represents the sum of areas of rectangles constructed

by taking the height to be the value of the function at the left-endpoint of
each subinterval. See Figure 89.

172
Secondly, we use the right-end point of each interval [ti−1 , ti ] and construct
the right-hand sum

R(v, n) = v(t1 )∆t + v(t2 )∆t + · · · + v(tn )∆t.

Geometrically, this sum represents the sum of areas of rectangles constructed

by taking the height to be the value of the function at the right-endpoint of
each subinterval. See Figure 90.

Figure 90

Now, the exact distance traveled lies between the two estimates. As we have
seen earlier, by making the time interval smaller and smaller we can make the
difference between the two estimates as small as we like. This is equivalent
to letting n → ∞. If the function v(t) is continuous then the following two
limits are equal to the exact distance traveled from t = a to t = b.

T otal distance traveled = n→∞

lim L(v, n) = n→∞
lim R(v, n).

Geometrically, each of the above limit represents the area under the graph
of v(t) bounded by the lines t = a, t = b and the horizontal axis. See Figure
91.

173
 Figure 91

Remark 32.2
Notice that for an increasing function the left-hand sum is an underestimate
whereas the right-hand sum is an overestimate. This role is reversed for a
decreasing function.

Recommended Problems (pp. 227 - 8): 2, 3, 4, 5, 8, 10, 12, 13.

174
33 The Definite Integral
We discussed in the previous section how to find the exact distance traveled
by taking the limit of both the left-hand sum and the right-hand sum as
the number of subintervals increases without bound. In this section, we will
construct these sums for any continuous function on a closed interval [a, b].
We start by dividing the interval [a,b] into n subintervals each of length

b−a
∆x = .
n
Let a = x0 , x1 , · · · , xn−1 , xn = b be the endpoints of the subdivisions. We
construct the left-hand sum or the left Riemann sum
n−1
X
L(f, n) = f (x0 )∆x + f (x1 )∆x + · · · + f (xn−1 )∆x = f (xi )∆x
i=0

and the right-hand sum or the right Riemann sum

n
X
R(f, n) = f (x1 )∆x + f (x2 )∆x + · · · + f (xn )∆x = f (xi )∆x.
i=1

It is shown in advanced calculus that for a continuous function on a closed

interval [a, b] that as n → ∞ both the left-hand sum and the right-hand sum
exist and are equal. We denote the common value by the notation ab f (x)dx.
R

Thus, Rb
a f (x)dx = limn→∞ L(f, n)
= limn→∞ R(f, n).
We call ab f (x)dx the definite integral of f from x = a to x = b. We call
R

a the lower limit and b the upper limit. The function f is called the
integrand.

Example 33.1
(a) On a sketch of y = ln x, represent the left Riemann sum with n = 2
approximating 12 ln xdx. Write out the terms in the sum, but do not evaluate.
R

(b) On a another sketch of y = ln x, represent the right Riemann sum with

n = 2 approximating 12 ln xdx. Write out the terms in the sum, but do not
R

evaluate.
(c) Which sum is an underestimate? Which sum is an overestimate?

175
Solution.
(a) The left Riemann sum is the sum

L(ln x, 2) = ln 1(0.5) + ln (1.5)(0.5) = 0.5 ln (1.5).

The Left Riemann sum is shown in Figure 92.

Figure 92

(b) The right Riemann sum is the sum

R(ln x, 2) = ln (1.5)(0.5) + ln 2(0.5) = 0.5 ln (2)(1.5) = 0.5 ln 3.

The Right Riemann sum is shown in Figure 92.

(c) L(ln x, 2) < 12 ln xdx < R(ln x, 2).
R

Geometrical Significance of ab f (x)dx

R

Case 1:f (x) ≥ 0

Looking closely to either the left Riemann sum or the right Riemann sum
we see that if f (x) ≥ 0 then a term of the form f (x)∆x represents the area
of a rectangle. As n increases without bound, that is, the width ∆x of the
rectangles approaches zero, the rectangles fit the curve of the graph more
exactly, and the sum of their areas gets closer and closer to the area under
the graph, bounded by the vertical lines x = a and x = b and the x-axis.
Thus, Z b
f (x)dx = Area under graph of f between a and b.
a

176
 Figure 93

Example 33.2 R1 √
Consider the integral −1 1 − x2 dx.

(a) Interpret the integral as an area, and find its exact value.
(b) Estimate the integral using a calculator.

Solution.
(a) Note that the equation of a circle centered at the origin √and with ra-
2 2 2
√ by x + y = 1. Solving for y we find y = ± 1 − x . The
dius 1 is given
2
√ y = 1 − x corresponds to the upper semicircle and the function
function
y = − 1 − x2 corresponds to the lower semicircle. See Figure 94.

Figure 94

177
It follows that the given integral represents the area of the upper semicircle
and therefore is equal to π2 . That is,
Z 1 √ π
1 − x2 dx = .
−1 2
(b) Using a TI-83 calculator we find
√
f nInt( 1 − x2 , x, −1, 1) ≈ 1.571.

Case 2:f (x) ≤ 0

In this case, since each product of the form f (x)∆x is less than or equal to
zero then the area gets counted negatively. That is, the absolute value of the
integral gives the area above the curve between x = a and x = b.

Example 33.3
Find the area above the graph of y = x2 − 1 from x = −1 to x = 1.

Solution.
The graph of y = x2 − 1 is shown in Figure 95.

Figure 95

R1 2
The area is given by | −1 (x − 1)dx| ≈ | − 1.33| = 1.33.

Case 3:f changes sign

In this case, the integral is the sum of the areas above the x-axis, counted

178
positively, and the areas below the x-axis, counted negatively. If the integral
is positive then the region above the x-axis has larger area than the region
below the x-axis. If the integral is negative then the region below the x-axis
has a larger area then the region above the x-axis.

Example 33.4
Find the area between the graph of y = x3 and the x-axis from x = −1 to
x = 1.

Solution.
The area is shown in Figure 96.

Figure 96

It follows that the area is given by

0 1
Z  Z
3
x3 dx = 0.5.
 

 x dx +
−1 0

Recommended Problems (pp. 234 - 6): 2, 3, 7, 8, 9, 11, 15, 17, 20,

22, 24, 27, 30, 31, 32.

179
34 Interpretations of the Definite Integral
We start this section by showing that the definite integral of a rate of change
gives the total change of the function. We define the total change of a function
F (t) from t = a to t = b to be the difference F (b) − F (a). Suppose that F (t)
is continuous in [a,b] and differentiable in (a, b). Divide the interval [a, b] into
n equal subintervals each of length ∆t = b−a n
. Let a = t0 , t1 , · · · , tn = b be the
partition points of the subdivision. Then on the interval [t0 , t1 ] the change
in F can be estimated by the formula

F (t0 + ∆t) − F (t0 )

F 0 (t0 ) ≈
∆t
or
F (t0 + ∆t) − F (t0 ) ≈ F 0 (t0 )∆t
that is
F (t1 ) − F (t0 ) ≈ F 0 (t0 )∆t
On the interval [t1 , t2 ] we get the estimation

F (t2 ) − F (t1 ) = F 0 (t1 )∆t

Continuing in this fashion we find that on the interval [tn−1 , tn ] we have

F (tn−1 ) − F (tn ) ≈ F 0 (tn−1 )∆t.

Adding all these approximations we find that

n−1
F 0 (ti )∆t
X
F (tn ) − F (t0 ) ≈
i=0

Letting n → ∞ we see that

Z b
F (b) − F (a) = F 0 (t)dt.
a

Example 34.1
The amount of waste a company produces, W, in metric tons per week, is
approximated by W = 3.75e−0.008t , where t is in weeks since January 1, 2000.
Waste removal for the company costs $15/ton. How much does the company
pay for waste removal during the year 2000?

180
Solution.
The
R 52
amount of tons produced during the year 2000 is just the definite integral
0 W (t)dt. Using a calculator we find that
Z 52
T otal waste during the year = 3.75e−0.008t dt ≈ 159 tons.
0

The cost to remove this quantity is 159 × 15 = $2385.

Remark 34.1
When using ab f (x)dx in applications then its units is the product of the
R

units of f (x) with the units of x.

The Definite Integral as an Average

We know that the average of n given numbers is just the sum divided by n.
What is the average in the continuous case. That is, what is the average of
a continuous function on a closed interval [a, b]? Partition the interval into
n equal subintervals each of length ∆t = b−a
n
and let a = t0 , t1 , t2 , · · · , tn be
the division points. Then

f (t0 ) + f (t1 ) + · · · + f (tn−1 )

Average of f (t) on [a, b] ≈
n
b−a
But n = ∆t
so that
1
Average of f (t) on [a, b] ≈ b−a
(f (t0 ) + f (t1 ) + · · · + f (tn−1 ))∆t
1 Pn−1
= b−a i=0 f (ti )∆t

Letting n → ∞ we see that

1 Zb
Average of f (t) on [a, b] = f (x)dx.
b−a a
Example 34.2
A bar of metal is cooling from 1000◦ C to room temperature, 20◦ C. The
temperature, H, of the bar t minutes after it starts cooling is given by

H = 20 + 980e−0.1t .

Find the average temperature over the first hour.

181
Solution.
The average temperature is given by
1 Z 60
Average temperature f or the f irst hour = (20+980e−0.1t )dt ≈ 183◦ C.
60 0
Remark 34.2
From the definition of the average value we can write
Z b
(average value of f ) × (b − a) = f (x)dx
a

Geometrically, this says that the area of the rectangle of dimensions (average value of f )×
(b − a) is equal to the area under the graph of f (x) from x = a to x = b. See
Figure 97.

Figure 97

Recommended Problems (pp. 242 - 4): 3, 4, 6, 10, 12, 13, 14, 20,
21, 22, 23, 25, 31.

182
35 Theorems About Definite Integrals
The following result is considered among the most important result in calcu-
lus.

Theorem 35.1 (The Fundamental Theorem of Calculus)

If f (x) is a continuous function on [a,b] and F 0 (x) = f (x) then
Z b
f (x)dx = F (b) − F (a)
a

We call the function F (x) an antiderivative of f (x).

Proof.
Partition the interval [a,b] into n subintervals each of length ∆x = b−an
and
let {a = x0 , x1 , · · · , xn = b} be the partition points. Applying the Mean
Value Theorem on the interval [x0 , x1 ] we can find a number x0 < c1 < x1
such that
F (x1 ) − F (x0 ) = F 0 (c1 )∆x.
Continuing this process on the remaining intervals we find

F (x2 ) − F (x1 ) = F 0 (c1 )∆x

..
.
F (xn ) − F (xn−1 ) = F 0 (cn )∆x

Adding these equalities we find

n
X
F (xn ) − F (x0 ) = f (ci )∆x
i=1

Letting n → ∞ to obtain
Z b
F (b) − F (a) = f (x)dx
a

Example 35.1
Use FTC to compute 12 2xdx. Use a calculator to find the answer to the
R

integral and compare.

183
Solution.
Since the derivative of x2 is 2x then F (x) = x2 . Thus, by the FTC we have
Z 2
2xdx = F (2) − F (1) = 4 − 1 = 3.
1
R2
Using a calculator we find 1 2xdx = 3.

When a function is symmetric about the y-axis, i.e. the function is even,
then the y-axis divide the region under consideration into two equal subre-
gions. This leads to Z a Z a
f (x)dx = 2 f (x)dx.
−a 0

If the function is symmetric about the origin, i.e. the function is odd, then one
region in the reflection of the other about the x-axis followed by a reflection
about the y-axis so that

Z a Z 0 Z a Z 0 Z 0
f (x)dx = f (x)dx + f (x)dx = f (x)dx − f (x)dx = 0.
−a −a 0 −a −a

Example 35.2 R1
Without using a calculator, find the value of the integral −1 (x3 − x)1001 dx.

Solution. R1
Since the integrand is odd then −1 (x3 − x)1001 dx = 0.

Next, we discuss several properties of definite integrals. In what follows,

the functions f and g are continuous on [a,b] and c is a constant. We parti-
tion the interval [a,b] into n subintervals and we let {a = x0 , x1 , · · · , xn = b}.
Then
Z b n−1
X n−1
X Z b
cf (x) = n→∞
lim cf (xi )∆x = c n→∞
lim f (xi )∆x = c f (x)dx.
a i=0 i=0 a

Rb n−1
± g(x)]dx = [f (xi ) ± g(xi )]∆x
P
a [f (x) limn→∞ i=0
Pn−1 Pn−1
= limn→∞ i=0 f (xi )∆x ± limn→∞ i=0 g(xi )∆x
Rb Rb
Rb
= a f (x)dx ± a g(x)dx
Pn−1 b−a
a f (x)dx = lim n→∞ i=0 f (xi ) n R
= − ba f (x)dx
Pn−1
= − limn→∞ i=0 f (xi ) a−b n

184
 Ra
It Rfollows from the last equality that if we replace b by a then a f (x)dx =
a Ra
− a f (x)dx or 2 a f (x)dx = 0 and this implies that
Z a
f (x)dx = 0.
a

Now, if f (x) ≤ g(x) in [a, b] then

Rb
limn→∞ n−1
P
a f (x)dx = i=0 f (xi )∆x
g(xi )∆x = ab g(x)dx
Pn−1
≤ limn→∞ i=0
R

If f (x) satisfies m ≤ f (x) ≤ M on [a,b] then applying the last result and the
fact that ab dx = b − a to get
R

Z b
m(b − a) ≤ f (x)dx ≤ (b − a)M.
a

Example 35.3
R√ √
Without using a calculator, show that 0 π sin (x2 )dx ≤ π.
Solution. √
Since sin (x2 ) ≤ 1 for all x then by integrating both sides from 0 to π to
obtain Z √π Z √π
2
sin (x )dx ≤ dx
0 0
But the right-hand integral is the area of the rectangle of length 1 and height
√ R √π √
π so that 0 dx = π.
Example 35.4
Show the following
Z b Z c Z b
f (x)dx = f (x)dx + f (x)dx
a a c

Solution.
Let F (x) be a function such that F 0 (x) = f (x). Then by the Fundamental
Theorem of Calculus we have
Rc Rb
a f (x)dx + c f (x)dx = (F (c) − F (a)) + (F (b) − F (c))
= F (b) − F (a)
Rb
= a f (x)dx

Recommended Problems (pp. 250 - 2): 1, 3, 6, 8, 10, 13, 15,

17, 18, 19, 21, 22.

185
36 Finding Antiderivatives Graphically and
Numerically
Recall that if a function f (x) is given then the new function f 0 (x) is called
the derivative of f (x). For example, if f (x) = x2 then f 0 (x) = 2x. This
process is referred to as differentiation.
Now, if instead f 0 (x) is known then the process of finding f (x) is called inte-
gration or antidifferentiation. In this case, we call f (x) an antideriva-
tive of f 0 (x). In general, if f and F are two functions such that F 0 = f
then we say that F (x) is an antiderivative of f (x). For example, x2 is an
antiderivative of 2x since (x2 )0 = 2x. Note that, there are infinitely many
antiderivatives of 2x, namely, x2 + C where C is a constant. We call x2 + C
the general antiderivative ( or the indefinite integral) of x2 and we
represent this symbolically by
Z
2xdx = x2 + C.

Next, we see how to reconstruct the graph of f given the graph of its deriva-
tive f 0 .

Example 36.1
The graph of f 0 (x) is given in Figure 98.

Figure 98
Sketch a graph of the function f (x) satisfying f (0) = 1.

Solution.
Note that since f 0 (x) is always increasing then f 00 (x) > 0 so that the graph of

186
f (x) is always concave up. Since f 0 (x) < 0 for x < 0 then f (x) is decreasing
there. Similarly, since f 0 (x) > 0 for x > 0 then f (x) is increasing there.
Since f 0 (0) = 0 and f (x) is decreasing and then increasing we conclude that
x = 0 is a minimum. A graph of f (x) is given in Figure 99.

Figure 99

Example 36.2
2
The graph of f 0 (x) = e−x is given in Figure 100.

Figure 100
Sketch the graph of f (x) satisfying f (0) = 0.

Solution.
Since f 0 (x) is always positive then the graph of f (x) is always increasing.
Now, for x < 0, f 0 (x) is increasing so that f 00 (x) > 0 and therefore f (x) is
concave up. For x > 0 the function f 0 (x) is decreasing and so f 00 (x) < 0.
That is, f (x) is concave down there. Thus, x = 0 is a point of inflection.
Finally, since limx→±∞ f 0 (x) = 0 then the graph of f (x) levels off at both
ends. See Figure 101.

187
 Figure 101

Next, we will reconstruct numerically the antiderivative Rf by using the Fun-

damental Theorem of Calculus: If F 0 (x) = f (x) then ab f (x)dx = F (b) −
F (a). In particular, we have
Z b
f 0 (x)dx = f (b) − f (a).
a

Example 36.3
Suppose that f 0 (t) = t cos t and f (0) = 2. Find f (0.3).

Solution.
Let a = 0 in the FTC to obtain
Z b
f (b) − f (0) = t cos tdt.
0

But f (0) = 2 so the previous equation becomes

Z b
f (b) = 2 + t cos tdt.
0

Thus, Z 0.3
f (0.3) = 2 + t cos tdt.
0
R 0.3
Using the TI83 command f nInt(x∗cosx, x, 0, 0.3) we find that 0 t cos tdt ≈
0.044 so that f (0.3) ≈ 2.044.

188
Now, recall that for f (x) ≥ 0 the definite integral ab f (x)dx represents the
R

area under the graph of fR(x) between the lines x = a and x = b. If the region
is below the x-axis then ab f (x)dx is the negative of the area of that region.

Example 36.4
Figure 102 shows the graph of f 0 (x). Suppose that f (−1) = −2. Find
f (0), f (1), and f (3).

Figure 102

Solution.
By the FTC we have
Z b
f (b) = f (−1) + f 0 (x)dx.
−1

Thus,
0
f 0 (x)dx
R
f (0) = f (−1) + −1
1
= −2 + 2 (1 · 2) = −1
f (0) + 01 f 0 (x)dx
R
f (1) =
= −1 + 1 · 2 = 1
f (3) = f (1) + 13 f 0 (x)dx = 1 + 12 (2 · 2) = 3
R

where we compute at f 0 (x)dx by determining the area between f 0 and the

R

horizontal axis for a ≤ x ≤ t.

Recommended Problems (pp. 265 - 7): 1, 3, 5, 7, 9, 11, 13, 17, 18.

189
37 Analytical Construction of Antiderivatives
In this section we will find analytical expressions of antiderivatives. As dis-
cussed in the previous section, there are infinitely many antiderivatives of a
given function f (x). They all differ by a constant and the family of antideriva-
tives is represented by F (x) + C. The notation of the general antiderivative
is called an indefinite integral and is written
Z
f (x)dx = F (x) + C.
R
The symbol is the symbol of integration, f (x) is called the integrand and
C is called the constant of integration. Keep in mind the relationship
between f (x) and F (x) which is given by F 0 (x) = f (x).

Warning: The indefinite integral is a short-hand notation for a family of

functions F (x) + C with the property
Rb
F 0 (x) = f (x) for all x. It is not to be
confused with the definite integral a f (x)dx which is a real number.

Example 37.1
R
Show that 0dx = C.

Solution.
Since the derivative of a constant function is always zero then
Z
0dx = C.

Example 37.2
R
Show that kdx = kx + C where k is a constant.

Solution.
Since the derivative of kx is just k then
Z
kdx = kx + C.

Example 37.3
xn+1
Show that xn dx = + C for n 6= −1.
R
n+1

190
Solution.
xn+1
By the power rule, if F (x) = n+1
then F 0 (x) = xn . Thus,
Z
xn+1
xn dx = + C.
n+1
0
Note that this formula is valid only if n 6= −1 for if n = −1 we would have x0
which doesn’t make sense. The case n = −1 is treated in the next problem.

Example 37.4
Show that Z
dx
= ln |x| + C.
x
Solution.
Suppose first that x > 0 so that ln |x| = ln x. Then (ln |x|)0 = (ln x)0 =
1
x
. Now, if x < 0 then ln |x| = ln (−x) and by the chain rule (ln |x|)0 =
−1
(ln (−x))0 = −x = x1 . Thus, in both cases (ln |x|)0 = x1 .

Example 37.5
eax
Show that for a 6= 0, eax dx =
R
a
+ C.

Solution.
eax
If a is a nonzero constant and F (x) = a
then F 0 (x) = eax . This shows that
Z
eax
eax dx = + C.
a
Example 37.6
sin (ax)
+ C where a 6= 0.
R
Show that cos (ax)dx = a

Solution.
If a is a nonzero constant and F (x) = sin a(ax) then F 0 (x) = cos (ax). This
shows that Z
sin (ax)
cos (ax)dx = + C.
a
Example 37.7
Show that sin (ax)dx = − cos a(ax) + C where a 6= 0.
R

191
Solution.
If a is a nonzero constant and F (x) = − cos a(ax) then F 0 (x) = sin (ax). This
shows that Z
cos (ax)
sin (ax)dx = − + C.
a
Example 37.8
R 1
Show that √1−x 2 dx = arcsin x + C.

Solution.
Let F (x) = arcsin x. Then F 0 (x) = √ 1
1−x2
. Thus,
Z
1
√ dx = arcsin x + C.
1 − x2
Example 37.9
R 1
Show that 1+x 2 dx = arctan x + C.

Solution.
Let F (x) = arctan x. Then F 0 (x) = 1
1+x2
. Thus,
Z
1
dx = arctan x + C.
1 + x2
Properties of Indefinite Integrals
Z Z Z
[f (x) ± g(x)]dx = f (x)dx ± g(x)dx.
To see why this property is true, let F (x) be an antiderivative of f (x) and
G(x) be an antiderivative of g(x). The result follows from the fact that
d
dx
[F (x) ± G(x)] = f (x) ± g(x).
Z Z
cf (x)dx = c f (x)dx.
R
To see this, suppose that F (x) is an antiderivative of f (x). Then f (x)dx =
d
F (x) + C. But dx (cF (x)) = cf (x) so that cF (x) is an antiderivative of cf (x),
that is, cf (x)dx = cF (x) + C 0 . This implies
R

Z Z Z Z
cf (x)dx = cF (x)+C 0 = c( f (x)dx−C)+C 0 = c f (x)dx−cC+C 0 = c f (x)dx.

Note that the constant

R
−cC + C 0 is ignored since a constant of integration
will result from f (x)dx.

192
Example 37.10
Find Z
3 5
(sin (2x) − e−3x + − 3 )dx.
x x
Solution.
Using the linearity property of indefinite integrals together with the formulas
of integration obtained above we have

(sin (2x) − e−3x + 3

− 5
sin (2x)dx − e−3x dx + 3 dx − 5 x−3 dx
R R R R R
x x3
)dx = x
−3x
= − cos 2(2x) + e 3 + 3 ln |x| + 2x52 + C

Once we have found an antiderivative of f (x), computing definite integrals

is easy by the Fundamental Theorem of Calculus.

ExampleR37.11
Compute 12 3x2 dx.

Solution.
Since F (x) = x3 is an antiderivative of f (x) = 3x2 , then by FTC
Z 2
3x2 dx = x3 |21 = 23 − 13 = 7.
1

Finally, we end this section by pointing out that not every funtion has a
simple analytical antiderivative. For example, the function
Z x sin t
Si(x) = dt
0 t
is an antiderivative of the function sinx x . There is no way of expressing Si(x)
in terms of simple analytic formula.

Recommended Problems (pp. 271 - 3): 1, 5, 9, 15, 16, 18, 20,

22, 27, 29, 36, 39, 46, 53, 56, 58, 67, 69, 73, 75, 77, 81.

193
38 Applications of Antiderivatives
In this section we will discuss four basic applications of antidifferentiation.

• Antiderivatives and Differential Equations

Antidifferentiation can be used in finding the general solution of the dif-
ferential equation
dy
= f (x).
dx
Note that the general solution is basically the general antiderivative of f (x). A
particular solution is one that satisfies an initial condition such as y(x0 ) =
y0 . A differential equation together with an initial condition is called an
initial value problem.
Example 38.1
Find the general solution of the differential equation
dy
= sin x + 2
dx
and then graph the particular solution corresponding to the inital value prob-
lem y(3) = 5.
Solution.
The general solution is the general antiderivative of the function f (x) =
sin x+2 which is given by F (x) = − cos x+2x+C. The solution corresponding
to F (3) = 5 is given by F (x) = − cos x + 2x + cos 3 − 1. The graph is given
in Figure 103

194
 Figure 103

Motion along a Straight Line

Antidifferentiation can be used to find specific antiderivatives using initial
conditions, including applications to motion along a line.

Example 38.2
An egg is dropped from an appartment window 20 meters above the ground.
Assuming that the acceleration of the egg is −16m/sec2 find
(a) the velocity of the egg as a function of time t;
(b) the position function s(t);
(c) the time when the egg hits the ground and the corresponding velocity.

Solution
(a) The motion is described by the equation
dv
= −16m/sec2 .
dt
The general solution of this equation is given by

v(t) = −16t + C.

Since the initial velocity is v(0) = 0 then C = 0 and therefore

v(t) = −16t.

(b) The solution to the equation

ds
= −16t
dt
is the antiderivative
s(t) = −8t2 + C.
But s(0) is the initial distance of the egg from the ground, i.e s(0) = 20.
Thus, 20 = C and therefore s(t) = −8t2 + 20.
(c) The egg hits the ground when s(t) = 0 that is, −8t2 + 20 = 0. Solving for
t > 0 we find t ≈ 1.6sec. At that time, v = −16(1.6) = −25.6m/sec. (The
velocity is negative because we are considering up as positive and down as
negative.)

195
Example 38.3
An object is thrown vertically upward with a speed of 10m/sec from a height
of 2 meters. Assuming that the acceleration of the object is −9.8m/sec2 find
(a) the velocity of the object as a function of time t;
(b) the position function s(t);
(c) the highest point the object reaches.

Solution.
(a) The motion of the object is described by the equation
dv
= −9.8.
dt
Finding the general solution of this differential equation we have

v(t) = −9.8t + C.

Since v(0) = 10 then C = 10 so that v(t) = −9.8t + 10.

(b) Finding the general solution of the differential equation
ds
= −9.8t + 10
dt
we have
s(t) = −4.9t2 + 10t + C.
But s(0) = 2 so that C = 2. Hence, s(t) = −4.9t2 + 10t + 2.
(c) The object reaches its highest point when the velocity is zero.Solving
−9.8t + 10 = 0 we find t ≈ 1.02sec.

Antidifferentiation and Rate of Change

Antidifferentiation can also be used when dealing with rates of change.

Example 38.4
If the temperature is constant, then the rate of change of barometric pressure
p with respect to altitude h is proportional to p. If p = 40 in at sea level and
p = 32 in when h = 1500 f t, find the pressure at an altitude of 6000 f t.

Solution.
Since the rate of change of p with respect to h is proportional to p then
dp
= kp,
dh
196
where k is the constant of proportionality. To find the general solution of
this differential equation we seperate the variables to obtain
dp
= kdh.
p
Integrate both sides to obtain ln |p| = kh + C or |p| = ekh+C . Thus, p(h) =
Cekh . To find C we use the fact that P (0) = 40 so that C = 40. Hence, P (h) =
40ekh . Now, to find the constant k we use the fact that P (1500) = 32. That
ln ( 32 )
is, 32 = 40e1500k . Solving this equation for k we find k = 1500 40
≈ −0.000149.
Hence,
P (h) = 40e−0.000149h
and in particular P (6000) = 40e−0.000149(6000) ≈ 16.36 in.

Modeling Exponential Growth and Decay

Antidifferentiation can be used also to model exponential growth and de-
dy
cay of the form dx = ky. This differential equation says that the rate of
change of y with respect to x is proportional to y. The constant k is called
the constant of proportionality. Solving the equation by the method of
seperation of variables discussed earlier we find

y(x) = Cekx . (5)

Equation (5) represents an exponential growth for k > 0, and exponential

decay for k < 0. The constant C is the value of y when x = 0.

Example 38.5
A bank account earns interest continuously at a rate of 5% of the current
balance per year. Assume that the initial deposit is $1, 000 and that no other
deposits or withdrawels are made.

(a) Write the differential equation satisfied by the balance in the account.
(b) Solve the differential equation and graph the solution.

Solution.
(a) If A denotes the balance at time t then the differential equation that
describes the model is given by
dA
= 0.05A.
dt
197
(b) Since A(0) = $1, 000 then A(t) = 1, 000e0.05t . This function is shown in
Figure 104.

Figure 104

Recommended Problems (pp. 277 - 8): 2, 4, 7, 10, 11, 12, 13,

17.

198
39 The Second Fundamental Theorem of Cal-
culus
We have seen so far that most of the functions that we considered have
elementary antiderivatives, that is, antiderivatives that can be expressed as a
linear combination of constants, powers of x, sin x, cos x, ex and ln x. However,
not all functions have antiderivatives that can be expressed in simple analytic
formula and we already encountered an example of such functions, i.e. sinx x .
In this section, we will present a method for constructing antiderivatives.

Theorem 39.1 (Second Fundamental Theorem of Calculus)

Suppose that f is continuous on an interval and a is any point of that interval.
Then the funtion Z x
F (x) = f (t)dt
a
0
is an antiderivative of f (x). That is, F (x) = f (x).

Proof.
What we must show here is that F 0 (x) = f (x), i.e.

F (x + h) − F (x)
lim = f (x).
h→0 h
From the definition of F we have that
F (x + h) − F (x) = ax+h f (t)dt − ax f (t)dt
R R

= ax+h f (t)dt + xa f (t)dt

R R
Ra R x+h
= x f (t)dt + a f (t)dt
R x+h
= x f (t)dt

This is illustrated in Figure 105.

199
 Figure 105
Since f (x) is continuous on the interval [x, x + h] then it is bounded there.
This means that there exist two constants M and m such that
m ≤ f (t) ≤ M
for all t in [x, x + h]. See Figure 105.
Integrate this inequality from x to x + h we find
Z x+h Z x+h Z x+h
m dt ≤ f (t)dt ≤ M dt
x x x
or
mh ≤ F (x + h) − F (x) ≤ M h.
Assume first that h > 0. (The case h < 0 is similar.) Dividing through by h
to obtain
F (x + h) − F (x)
m≤ ≤ M.
h
Since f is continuous at x then limh→0 m = limh→0 M = f (x). Hence, by the
squeeze law we have F 0 (x) = f (x).

Example 39.1
sin x
According to the above theorem an antiderivative of the function f (x) = x
is given by Z x
sin t
Si(x) = dt.
0 t
(a) Find the values of Si(x) for x = 1, 2, 3.
(b) Find the derivative of xSi(x).

Solution.
(a) Using a TI 83 calculator with the command f nInt( sinx x , x, 0, 1) we find
Si(1) = 0.95. Similarly, we find Si(2) = 1.61 and Si(3) = 1.85.
(b) Applying the product rule,
d
dx
(xSi(x)) = (x)0 Si(x) + x(Si(x))0
= Si(x) + x sinx x
= Si(x) + sin x

Recommended Problems (p. 281): 1, 2, 4, 5, 7, 8, 12, 16, 18, 19,

22, 25, 26.

200