CALCULUS OF SINGLE

VARIABLES - 2
III SEMESTER
(2019 Admission)

B Sc MATHEMATICS
Core Course (MTS3 B03)

UNIVERSITY OF CALICUT
School of Distance Education
Calicut University P.O., Malappuram, Kerala, India - 673 635

19557
UNIVERSITY OF CALICUT
School of Distance Education

Study Material

III SEMESTER
(2019 Admission)

B Sc MATHEMATICS
Core Course (MTS3 B03)

CALCULUS OF SINGLE
VARIABLES - 2

Prepared by:

Sri. Sachin Chandran,

Assistant Professor in Mathematics,
SDE, Calicut University.

Scrutinized by:

Smt. Sangeetha M.V,

Assistant Professor,
Department of Mathematics.
St. Joseph’s College,
Devagiri, Calicut – 8.

2
Contents

1 The Transcendental functions 9

1.1 The natural logarithmic function . . . . . . . . . . 9
1.2 Inverse Functions . . . . . . . . . . . . . . . . . . 15
1.3 Exponential functions : . . . . . . . . . . . . . . . 20
1.4 General exponential and logarithmic functions . . . 25
1.5 Inverse trigonometric functions : . . . . . . . . . . 31
1.6 Hyperbolic functions . . . . . . . . . . . . . . . . 38
1.7 Indeterminate forms and l’Hopital’s rule . . . . . . 47

2 Infinite sequences and series 54

2.1 Improper Integral . . . . . . . . . . . . . . . . . . 54
2.2 Sequences . . . . . . . . . . . . . . . . . . . . . . 57
2.3 Series . . . . . . . . . . . . . . . . . . . . . . . . 65
2.3.1 Geometric Series . . . . . . . . . . . . . . 67
2.3.2 The Harmonic Series . . . . . . . . . . . . 70
2.3.3 The Divergence Test . . . . . . . . . . . . 71
2.3.4 Properties of Convergent Series . . . . . . 72
2.4 The Integral Test . . . . . . . . . . . . . . . . . . 73
2.5 The Comparison Tests . . . . . . . . . . . . . . . 77

3
 2.5.1 The Comparison Test . . . . . . . . . . . . 77
2.5.2 The Limit Comparison Test . . . . . . . . 79
2.6 Alternating Series . . . . . . . . . . . . . . . . . . 81
2.6.1 Approximating the Sum of an Alternating
Series by Sn . . . . . . . . . . . . . . . . 84
2.7 Absolute Convergence; the Ratio and Root Tests . . 86
2.7.1 Absolute Convergence . . . . . . . . . . . 86
2.7.2 Ratio Test . . . . . . . . . . . . . . . . . . 87
2.7.3 The Root Test . . . . . . . . . . . . . . . . 90
2.7.4 Rearrangement of Series . . . . . . . . . . 91

3 Power series, plane curves and polar coordinates 96

3.1 Power series . . . . . . . . . . . . . . . . . . . . . 96
3.2 Taylor and Maclaurin series . . . . . . . . . . . . . 103
3.3 Plane curves and Parametric equations . . . . . . . 112
3.4 The calculus of Parametric equations . . . . . . . . 117
3.5 Polar coordinates . . . . . . . . . . . . . . . . . . 120
3.6 Areas and Arc lengths in Polar coordinates . . . . . 126
4 Geometry of Space and Vector-valued function 134
4.1 Equations of Lines in Space . . . . . . . . . . . . 134
4.2 Surfaces in space . . . . . . . . . . . . . . . . . . 140
4.3 Cylindrical and spherical coordinates . . . . . . . . 147
4.4 Vector-valued functions and space curves . . . . . 150
4.5 Differentiation and integration of vector-valued func-
tions . . . . . . . . . . . . . . . . . . . . . . . . . 154
4.6 Arc length and curvature . . . . . . . . . . . . . . 159
4.7 Velocity and acceleration . . . . . . . . . . . . . . 167

4
4.8 Tangential and normal components of acceleration 171

5
6
 Course overview

Using the idea of definite integral developed in previous semester,

the natural logarithm function is defined and its properties are ex-
amined. This allows us to define its inverse function namely the
natural exponential function and also the general exponential func-
tion. Exponential functions model a wide variety of phenomenon of
interest in science, engineering, mathematics and economics. They
arise naturally when we model the growth of a biological popula-
tion, the spread of a disease, the radioactive decay of atoms, and
the study of heat transfer problems and so on. We also consider
certain combinations of exponential functions namely hyperbolic
functions that also arise very frequently in applications such as the
study of shapes of cables hanging under their own weight. After
this, the students are introduced to the idea of improper integrals,
their convergence and evaluation. This enables to study a related
notion of convergence of a series, which is practically done by ap-
plying several different tests such as integral test, comparison test
and so on. As a special case, a study on power series- their region
of convergence, differentiation and integration etc.,- is also done.
A detailed study of plane and space curves is then taken up. The
students get the idea of parametrization of curves, they learn how
to calculate the arc length, curvature etc. using parametrization and
also the area of surface of revolution of a parametrized plane curve.
Students are introduced into other coordinate systems which often
simplify the equation of curves and surfaces and the relationship

7
between various coordinate systems are also taught. This enables
them to directly calculate the arc length and surface areas of rev-
olution of a curve whose equation is in polar form. At the end of
the course, the students will be able to handle vectors in dealing
with the problems involving geometry of lines, curves, planes and
surfaces in space and have acquired the ability to sketch curves in
plane and space given in vector valued form.

8
Module 1

The Transcendental functions

1.1 The natural logarithmic function

Definition 1. The natural logarithmic function, denoted by ln, is
the function defined by
Z x
1
lnx = dt
1 t

for all x > 0.

Definition 2. (Derivative of lnx) Using Fundamental theorem of

Calculus, we get that
Z x
d d 1 1
lnx = dt = x>0
dx dx 1 t x

Laws of logarithms : Let x and y be positive numbers and let r

be a rational. Then
a. lnx = 0

9
b. lnxy = lnx + lny
c. ln xy = lnx − lny
d. lnxr = rlnx

2
Example : Expand the expression ln x√+1
x
.
Solution :
x2 + 1 x2 + 1
√ = 1/2
x x
= ln(x2 + 1) − ln(x1/2 )
1
= ln(x2 + 1) − lnx
2
Graph of the natural logarithmic function : f (x) = lnx has the
following properties:
1. The domain of f is (0, ∞), by definition.
2. f is continuous on (0, ∞), since it is differentiable there.
3. f is increasing on (0, ∞), since f 0 (x) = x1 > 0 on (0, ∞).
4. The graph of f is concave downward on (0, ∞) since f 00 (x) =
− x12 < 0 on (0, ∞).
Using these properties and the results limx→0+ lnx = −∞ and
limx→ inf ty lnx = ∞
we sketch the graph of f (x) = lnx, as shown below.

Theorem 1. (Derivative of the Natural logarithmic function) Let

u be a differential function of x. then

10
 d
a. dx
ln|x| = x1 x 6= 0
d
b. dx
ln|u| = u1 . du
dx
u 6= 0

Proof. a. If x > 0, then as we have have already seen earlier

d d 1
ln|x| = lnx =
dx dx x

If x < 0, then |x| = −x. So we have

d d 1
ln|x| = ln(−x) =
dx dx x

b. This follows from the Chain Rule.

Example : Find the derivative of

a. f (x) = ln(2x2 + 1) b. g(x) = x2 ln2x
Solution :
a. f 0 (x) = dx
d
f (x) = ln(2x2 + 1) = 2x21+1 dx
d
(2x2 + 1) = 4x
2x2 +1
b.

d 2 d d
g 0 (x) = (x ln2x) = x2 (ln2x) + (ln2x) (x2 )
dx dx dx
1
= x2 ( )(2) + (ln2x)(2x) = x(1 + 2ln2x)
2x

Logarithmic Differentiation : We have seen how the laws of log-

arithms can help to simplify the work involved in differentiating
logarithmic expressions.We now look at a procedure that takes ad-
vantage of these same laws to help us differentiate functions that
at first blush do not necessarily involve logarithms. This method,
called logarithmic differentiation, is especially useful for differen-
tiating functions involving products, quotients, and/or powers that

11
can be simplified by using logarithms.

dy
Steps to finding dx by Logarithmic Differentiation : Suppose
dy
that we are given the equation y = f (x) . To compute dx :
1. Take the logarithm of both sides of the equation, and use the
laws of logarithms to simplify the resulting equation.
2. Differentiate implicitly with respect to x.
dy
3. Solve the equation found in Step 2 for dx .
4. Substitute for y.

3
Example : Find the derivative of y = (2x−1)
√
3x+1
.
Solution : We begin by taking the natural logarithm on both sides
of the equation,
(2x − 1)3
lny = ln √
3x + 1
or
1
lny = 3ln(2x − 1) − ln(3x + 1)
2
Now diferentiating with respect to x

1 0 3 1
(y ) = (2) − (3)
y 2x − 1 2(3x + 1)
6 3
= −
2x − 1 2(3x + 1)
6.2(3x + 1) − 3(2x − 1)
=
2(2x − 1)(3x + 1)

12
This gives,

6.2(3x + 1) − 3(2x − 1)
y0 = .y
2(2x − 1)(3x + 1)
6.2(3x + 1) − 3(2x − 1) (2x − 1)3
= .√
2(2x − 1)(3x + 1) 3x + 1
2
15(2x + 1)(2x − 1)
=
2(3x + 1)3/2

Integration involving Logarithmic functions : By reversing the

rule
d 1 du
ln|u| =
dx u dx
we obtain the following rule of integration.

Theorem 2. (Rule for integrating u1 ) Let u = g(x) , where g is

differentiable, and suppose that g(x) 6= 0. Then
Z
1
du = ln|u| + C
u

Example : Evaluate the following

R 1
a. 2x+1 dx
R
b. tanxdx
R
c. sexxdx
Solution : a. Let u = 2x + 1, so that du = 2dx or dx = 21 du.
Making these substitutions, we get
Z Z Z
1 1 1 1
dx = du = ln|u| + C
2x + 1 2 u 2
1
= ln|2x + 1| + C
2

13
b. Since tanx = sinx
cosx
, we use the substitution u = cosx, so that
du = −sinxdx = or sinxdx = −du. This gives
Z Z Z
sinx 1
tanxdx = dx = − du
cosx u
= −ln|u| + C = −ln|cosx| + C
= ln|secx| + C

c. Multiplying both the numerator and denominator of the inte-

grand by secx + tanx gives

sec2 x + secx.tanx
Z Z Z
secx + tanx
secxdx = secx dx =
secx + tanx secx + tanx

Now, we use the substitution u = secx + tanx, so that we get

du = (secx.tanx + sec2 x)dx. This gives,
Z Z
1
secxdx = du = ln|u| + C = ln|secx + tanx| + C
u

We can use the above technique to find the integral of other

trigonometric functions, the results of which are summarized be-
low.

Theorem 3. ( Integrals of Trigonometric Functions)

R
a. tanudu = ln|secu| + C
R
b. cotudu = ln|sinu| + C
R
c. secudu = ln|secu + tanu| + C
R
d. cscudu = ln|cscu − cotu| + C
R
Example : Find xsexx2 dx.
Solution : Let u = x2 , so that du = 2xdx or xdx = 12 du. Making

14
these substitutions, we obtain
Z Z
2 1
xsexx dx = secudu
2
1
= ln|secu + tanu| + C
2
1
= ln|secx2 + tanx2 | + C
2
Theorem 4. a. limx→∞ lnx = ∞
b. limx→0+ lnx = −∞

Proof. a. By Law of logarithms, we have ln2n = nln2 for any

positive integer n. Since ln2 > 0, we see that ln2n → ∞ as
n → ∞. But lnx is an increasing function, so

lim lnx = ∞
x→∞

b. Let t = x1 . Then t → ∞ as x → 0+ . Therefore, using part (a),

we have

1
lim+ lnx = lim ln( ) = lim (−lnt) = ∞
x→0 t→∞ t t→∞

1.2 Inverse Functions

A function that undoes, or inverts, the effect of a function f is
called the inverse of f . Many common functions, though not all,
are paired with an inverse. Important inverse functions often show
up in applications. Inverse functions also play a key role in the de-

15
velopment and properties of the exponential functions. To have an
inverse, a function must possess a special property over its domain.

Definition 3. A function f (x) is one-to-one on a domain D if f (x1 ) 6=

f (x2 ) whenever x1 6= x2 in D.

Examples :

√
1. f (x) = x is one-to-one on any domain of non negative num-
√ √
bers because x1 6= x2 whenever x1 6= x2 .

2. g(x) = sin(x) is not one-to-one on the interval [0, π] because

sin(π/6) = sin(5π/6). In fact, for each element x1 in the sub
interval [0, π/2) there is a corresponding element x2 in the sub in-
terval (π/2, π] satisfying sin(x1 ) = sin(x2 ). The sine function is
one-to-one on [0, π/2] , however, because it is an increasing func-
tion on [0, π/2] and therefore gives distinct outputs for distinct in-
puts in that interval.

The Horizontal Line Test for One-to-One Functions : A func-

tion y = f (x) is one-to-one if and only if its graph intersects each
horizontal line at most once.

16
Definition 4. Suppose that f is a one-to-one function on a domain
D with range R. The inverse function f −1 is defined by f −1 (b) = a
if f (a) = b. The domain of f −1 is R and the range of f −1 is D.

Example : Suppose a one-to-one function y = f (x) is given by a

table of values

x 1 2 3 4 5 6 7 8
f(x) 3 4.5 7 10.5 15 20.5 27 34.5

Solution : Then the table of values of x = f −1 (y) is as given below

y 3 4.5 7 10.5 15 20.5 27 34.5

f −1 (y) 1 2 3 4 5 6 7 8

17
 Composing a function and its inverse has the same effect as do-
ing nothing i.e
(f −1 ◦ f )(x) = x, for all x in the domain of f
(f ◦ f −1 )(y) = y, for all y in the domain of f −1 (or range of f ).
Keep in mind that only a one-to-one function can have an inverse.
The reason is that if f (x1 ) = y and f (x2 ) = y for two distinct in-
puts x1 and x2 , then there is no way to assign a value to f −1 (y)
that satisfies both f −1 (f (x1 )) = x1 and f −1 (f (x2 )) = x2 . A
function that is increasing on an interval satisfies the inequality
f (x2 ) > f (x1 ) when x2 > x1 , so it is one-to-one and has an
inverse. A function that is decreasing on an interval also has an
inverse. Functions that are neither increasing nor decreasing may
still be one- to-one and have an inverse. An example is the function
f (x) = 1/x for x 6= 0 and f (0) = 0, defined on (−∞, ∞) and
passing the horizontal line test.

Finding the Inverse :

We want to set up the graph of f −1 so that its input values lie along
the x-axis, as is usually done for functions, rather than on the y-
axis. To achieve this we interchange the x- and y-axes by reflecting
across the 45 degree line y = x. After this reflection we have a
new graph that represents f −1 . The value of f −1 (x) can now be
read from the graph in the usual way, by starting with a point x on
the x-axis, going vertically to the graph, and then horizontally to
the y-axis to get the value of f −1 (x).
The process of passing from f to f −1 can be summarized as a two-
step procedure.

18
1. Solve the equation y = f (x) for x. This gives a formula
x = f −1 (y) where x is expressed as a function of y.
2. Interchange x and y, obtaining a formula y = f −1 (x) where f −1
is expressed in the conventional format with x as the independent
variable and y as the dependent variable.

Example : Find the inverse of y = 21 x + 1, expressed as a function

of x.
Solution : 1. Solve for x in terms of y:
y = 21 x + 1
2y = x + 2
x = 2y − 2.
2. Interchange x and y: y = 2x − 2.
The inverse function of f is f −1 (x) = 2x − 2.

Theorem 5. (The Derivative Rule for Inverses) If f has an interval

I as domain and f 0 (x) exists and is never zero on I, then f −1 is
differentiable at every point in its domain (the range of f ). The
value of (f −1 )0 at a point b in the domain of f −1 is the reciprocal
of the value of f 0 at the point a = f −1 (b):

1
(f −1 )0 (b) =
f 0 (f −1 (b))

or
df −1 1
|x=b = df
dx | −1
dx x=f (b)

Theorem 1 makes two assertions. The first of these has to

do with the conditions under which f −1 is differentiable; the sec-
ond assertion is a formula for the derivative of f −1 when it ex-

19
ists. The second assertion can be easily proved (hint : start with
f (f −1 (x)) = x and take the derivative wrt x).

Example : Let f (x) = x3 − 1, for x > 0.(x). Find the value

of df −1 /dx at x = 6 = f (2) without finding the formula for f −1 .

Solution : We apply Theorem 1 to obtain the value of the derivative

of f −1 at x = 6:
df
|x=2 = 3x2 |x=2 = 12
dx
df −1 1 1
|x=f (2) = df
=
dx |
dx x=2
12

1.3 Exponential functions :

We saw that the natural logarithm function defined by y = lnx is
continuous and increasing on the interval (0, ∞). Also, lnx is one-
to-one on (0, ∞) and, hence, has an inverse. This inverse function
is called the natural exponential function and is defined as follows.

Definition 5. The natural exponential function, denoted by exp, is

the function satisfying the conditions:
1. ln(expx) = x for all x ∈ (∞, ∞)
2. exp(lnx) = x for all x ∈ (0, ∞)
Equivalently, exp(x) = y if and only if lny = x.

That the domain of exp is (−∞, ∞) and its range is (0, ∞) fol-
lows because the range of ln is (−∞, ∞) and its domain is (0, ∞).
The graph of y = exp(x) can be obtained by reflecting the graph
of y = lnx about the line y = x.

20
 Recall that the natural logarithmic function ln is continuous and
one-to-one and that its range is (∞, ∞). Therefore, by the Interme-
diate Value Theorem there must be a unique real number x0 such
that lnx0 = 1. Let’s denote x0 by e. We will formally define e as
follows.

Definition 6. (The number e)The number e is the number such that

Z e
1
lne = dt = 1
t t

21
 The graph above gives a geometric representation of the number
e. It should be noted that e is an irrational number and has the
approximate value of 2.718281828.
Natural exponential function : For a real number x, we have

lnex = xlne = x(1) = x

Given this, let define natural exponential function.

Definition 7. The natural exponential function, exp, is defined by

the rule
exp(x) = ex

In view of this, we have the following theorem, which gives

us another way of expressing the fact that exp and ln are inverse
functions.

22
Theorem 6. a. lnex = x, for x ∈ (−∞, ∞)
b. elnx = x, for x ∈ (0.∞)
Properties of natural exponential function :
1. The domain of f (x) = ex is (−∞, ∞), and its range is (0, ∞).
2. The function f (x) = ex is continuous and increasing on (−∞, ∞).
3. The graph of f (x) = ex is concave upward on (−∞, ∞).
4. limx→−∞ ex = 0 and limx→∞ ex = ∞.

Theorem 7. (Laws of exponents) Let x and y be real numbers and

r be a rational number. Then
a. ex ey = ex+y
x
b. eey = ex−y
c. (ex )r = erx
Proof. We will prove Law (a). The proofs of the other two laws are
similar and is left yo the reader. We have

ln(ex ey ) = lnex + lney = x + y = lnex+y

Since the natural logarithmic function is one-to-one, we see that

ex ey = ex+y

Theorem 8. (The derivatives of exponential functions) Let u be a

differentiable function of x. Then
d x
a. dx e = ex
d u
b. dx e = eu du
dx

23
Proof. a. Let y = ex , so that lny = x. Differentiating both sides of
the last equation implicitly with respect to x gives

1 dy dy
= 1 or = y = ex
y dx dx

b. This follows from part (a) by using the Chain Rule.

Example : Find the derivative of

2
a. f (x) = e−x
b. y = ln(e2x + e−2x ).
d −x2 2 d 2
Solution : a. f 0 (x) = dx e = e−x dx (−x2 ) = −2xe−x
b.
dy d
= ln(e2x + e−2x )
dx dx
1 d 2x
= 2x −2x
(e + e−2x )
e +e dx
1
= 2x (2e2x − 2e−2x )
e + e−2x
2(e2x − e−2x )
= 2x
e + e−2x

Theorem 9. (Integration of natural exponential function) Since

the derivative of the natural exponential function is the function
itself, the following theorem is immediate. Let u be a differentiable
function of x. Then
Z
eu du = eu + C

Example : Find
R
a. e5x dx
R 1 ex
b. 0 1+e x dx

24
Solution : a. Let u = 5x, so that du = 5dx, or dx = 51 du. Making
these substitutions, we obtain
Z Z
5x 1 1 1
e dx = eu du = eu + C = e5x + C
5 5 5

b. Let u = 1 + ex , so that du = ex dx. If x = 0, then u = 2; and

if x = 1, then u = 1 + e. This gives the lower and upper limits of
integration with respect to u. We have
1 1+e
ex
Z Z
1
dx = du = [lnu]1+e
2 = ln(1 + e) − ln2 ≈ 0.620
0 1 + ex 2 u

1.4 General exponential and logarithmic func-

tions
Exponential functions with base a : The natural exponential func-
tion defined by f (x) = ex has base e. We will now consider expo-
nential functions that have bases other than e.

Definition 8. Let a be a positive real number with a 6= 1. The

exponential function with base a is the function f defined by

f (x) = ax = exlna

Theorem 10. Let a and b be positive numbers. If x and y are real

numbers, then
a. ax ay = ax+y
b. (ax )y = axy
c. (ab)x = ax bx

25
 x
d. aay = ax−y
x
e. ( ab )x = abx

Proof. We will prove the first law and leave the proofs of the other
laws as exercises.
ax ay = exlna eylna
= exlna+ylna
= ex+y lna
= ax+y

Theorem 11. (Derivatives of ax and au ) Let a be a positive number

with a 6= 1, and let u be a differentiable function of x. Then
d x
a. dx a = (lna)ax
d u
b. dx a = (lna)au dudx

d x d xlna d
Proof. a. dx a = dx e = exlna dx (xlna) = exlna (lna) = (lna)ax
b. Follows from the chain rule.

Example : Find the derivative of

a. f (x) = 2x
b. y = 10cos2x
Solution : a. f 0 (x) = dx
d x
2 = (ln2)2x
b.
dy d cos2x
= 10
dx dx
d
= (ln10)10cos2x cos2x
dx
cos2x
= (ln10)10 (−sin2x)(2)
= −2(ln10)(sin2x)10cos2x

26
Graphs of y = ax

If a > 1, then lna > 0, and therefore,

d x
(a ) = ax lna > 0
dx

This shows that the graph of y = ax is rising on (−∞, ∞). If

0 < a < 1, then lna < 0, and

d x
(a ) = ax lna < 0
dx

This implies that if 0 < a < 1, the graph of y = ax is falling on

(= ∞, ∞). The general shape of the graphs of y = ax are shown
below.

Example : Find the derivative of f (x) = xx .

Solution : Let y = xx . Taking the natural logarithm on both sides,
we obtain
lny = lnxx = xlnx

Differentiating both sides of this equation with respect to x, we

27
obtain

y0 d d d
= (xlnx) = x (lnx) + (lnx) (x)
y dx dx dx

Therefore, upon multiplying both sides by y, we obtain

y 0 = (1 + lnx)y = (1 + lnx)xx

Integrating ax : The formula for integrating an exponential func-

tion with base a follows from reversing the differentiation formula.
Thus, we have

ax
Z
ax dx = + C a > 0 and a 6= 1
lna

Logarithmic functions with base a : If a is a positive real

number with a 6= 1, then the function f defined by f (x) = ax
is one-to-one on (−∞, ∞), and its range is (0, ∞). Therefore, it
has an inverse on (0, ∞). This function is called the logarithmic
function with base a and is denoted by loga .

Definition 9. The logarithmic function with base a, denoted by

loga , is the function satisfying the relationship

y = loga x if and only if x = ay

Change of base formula :

lnx
loga x = a > 0 and a 6= 1
lna

28
Theorem 12. (The power rule) If n is a real number, then

d n
(x ) = nxn−1
dx

Proof. Let y = xn and consider the equation

|y| = |xn | = |x|n x 6= 0

Taking the natural logarithm on both sides of the equation, we ob-

tain
ln|y| = nln|x|

which, upon differentiation with respect to x, yields

y0 n
=
y x

or
ny nxn
y0 = = = nxn−1
x x

Theorem 13. (Derivatives of the logarithmic function with base

a) Let u be a differentiable function of x. Then
d 1
a. dx loga |x| = xlna x 6= 0
d 1 du
b. dx loga |u| = ulna . dx u 6= 0

Example : Find the derivative of f (x) = x2 log(e2x + 1).

29
Solution : Using the product rule, we obtain

d 2
f 0 (x) = [x log(e2x + 1)]
dx
d d
= [ (x2 )]log(e2x + 1) + x2 log(e2x + 1)
dx dx
2
x d
= 2xloh(e2x + 1) + 2x . (e2x + 1)
(e + 1)ln10 dx
2x2x
= 2xlon(e2x + 1) + 2x
(e + 1)ln10

The definition of the number e as a limit : If we use the definition

of the derivative as a limit to compute f 0 (1), where f (x) = lnx ,
we obtain

f (1 + h) − f (1)
f 0 (1) = lim
h→0 h
ln(1 + h) − ln(1) ln(1 + h)
= lim = lim
h→0 h h→0 h
= lim ln(1 + h)1/h
h→0

= ln[lim (1 + h)1/h ]
h→0

But
d 1
f 0 (1) = [ lnx]x=1 = [ ]x=1 = 1
dx x
Thus,
ln[lim (1 + h)1/h ] = 1
h→0

or
lim (1 + h)1/h = e
h→0

Above equation is sometimes used to define the number e. Another

30
equivalent definition of e is:

1 n
lim (1 + ) =e
n→∞ n

1.5 Inverse trigonometric functions :

Inverse trigonometric functions arise when we want to calculate an-

gles from side measurements in triangles. They also provide useful
anti derivatives and appear frequently in the solutions of differen-
tial equations.

The six basic trigonometric functions are not one-to-one since

their values repeat periodically. However, we can restrict their do-
mains to intervals on which they are one-to-one. The sine function
increases from −1 at x = −π/2 to +1 at x = π/2. By restrict-
ing domain to [−π/2, π/2] we make it one-one, so it has an inverse
which trigonometric is called arcsinx. Similar domain restrictions
can be applied to all six trigonometric functions.

31
Since these restricted functions are not one-one, we cam defined
their inverses and denoted as follows:

y = sin−1 (x) or y = arcsin(x)

y = cos−1 (x) or y = arccos(x)

y = tan−1 (x) or y = arctan(x)

y = csc−1 (x) or y = arcsc(x)

y = sec−1 (x) or y = arcsec(x)

y = cot−1 (x) or y = arccot(x)

Graphs of basic inverse trigonometric functions:

32
Definition 10. y = arcsinx is the number in [−π/2, π/2] for
which siny = x.
y = arccosx is the number in [0, π] for which cosy = x.

The graph of y = arcsinx as shown in the figure above, is sym-

metric about the origin (it lies along the graph of x = siny). The
arcsine is therefore an odd function i.e. arcsin(−x) = −arcsinx.
The graph of y = arccosx has no such symmetry.

Example : Evaluate
√
a. arcsin( 3/2)
b. arccos(−1/2)
√
Solution : a. sin(π/3) = 3/2 and π/3 belongs to the range
√
[−π/2, π/2] of the arcsine function. Therefore, arcsin( 3/2) =
π/3.
b. cos(2π/3) = −1/2 and 2π/3 belongs to the range [0, π] of the

33
arccosine function. Therefore, arccos(−1/2) = 2π/3.

Using the procedure as above, we can find common values for

arcsin and arccos functions.

Example : During a 240 mi airplane flight from Chicago to St.

Louis, after flying 180 mi the navigator determines that the plane
is 12 mi off course, as shown in Figure 7.26. Find the angle a for
a course parallel to the original correct course, the angle b, and the
drift correction angle c = a + b.

34
Solution : Using the Pythagorean theorem, we compute an approx-
imate hypothetical flight distance of 179 mi, had the plane been fly-
ing along the original correct course. Knowing the flight distance
from Chicago to St. Louis, we next calculate the remaining leg of
the original course to be 61 mi. Applying the Pythagorean theorem
again then gives an approximate distance of 62 mi from the posi-
tion of the plane to St. Louis. Finally, we see that 180sina = 12
and 62sinb = 12, so

a = arcsin(12/180) ≈ 0.067radian ≈ 3.8◦

b = arcsin(12/62) ≈ 0.195radian ≈ 11.2◦

Thus, c = a + b ≈ 15◦ .

Identities Involving Inverse Trigonometric functions :

1. arccosx + arccos(−x) = π
or
arccos(−x) = π − arccosx.
2. arccotx = π/2 − arctanx.
3. arccscx = π/2 − arcsecx.
4. arcsinx + arccosx = π/2.

Definition 11. 1. y = arctanx is the number in (−π/2, π/2) for

which tany = x.
2. y = arccotx is the number in (0, π) for which coty = x.
3. y = arcsecx is the number in (0, π/2) ∪ (π/2, π) for which

35
secy = x.
4. y = arccscx is the number in (−π/2, 0) ∪ (0, π/2) for which
cscy = x.

Derivative of Inverse trigonomnetric functions

Derivative of y = arcsin(x).

y = arcsin(x)

x = sin(y)

Derivating wrt x,
dy
1 = cos(y)
dx
dy 1
=
dx cos(y)
√
We have x = sin(y), that implies cos(y) = 1 − x2 . Substituting
into the above equation, we get

dy 1
=√
dx 1 − x2

. Derivative of y = arctan(x).

y = arctan(x)

x = tan(y)

Derivating wrt x,
dy
1 = sec2 (y)
dx

36
x = tan(y), that implies sec2 (y) = 1 + x2 . Substituting into the
above equation, we get

dy 1
=
dx 1 + x2

We can find the derivatives of remaining inverse trigonometric func-

tions in similar fashion. The table below summarises the deriva-
tives,

Integration formulas

The derivative formulas above yield three useful integration formu-

las in the table below. The formulas are readily verified by differ-
entiating the functions on the right-hand sides.

37
Example :
a. √
Z 3 √
2 dx 3

√ √ = sin−1 (x)| √22

2
2 1 − x2 2
√ √
3 2
= sin−1 ( ) − sin−1 ( )
2 2
π π
= −
3 4
π
=
12
b.
Z
dx 1
Z
du √
√ = √ . set a = 3, u = 2x
3 − 4x2 2 a2 − u 2
1 u
= sin−1 ( ) + C
2 a
1 −1 2x
= sin ( √ ) + C
2 3

1.6 Hyperbolic functions

The analysis of many problems in engineering and mathematics
involves combinations of exponential functions of the form ecx and
ecx , where c is a constant. Because combinations of these functions
arise so frequently in mathematics and its applications, they have
been given special names. These combinations—the hyperbolic
sine, the hyperbolic cosine, the hyperbolic tangent, and so on—are
referred to as hyperbolic functions and are so called because they
have many properties in common with the trigonometric functions.

Definition 12. (Hyperbolic functions)

38
 ex − e−x ex + e−x
sinh(x) = cosh(x) =
2 2

sinh(x) 1
tanh(x) = csch(x) = , x 6= 0
cosh(x) sinh(x)

1 cosh(x)
sech(x) = coth(x) = , x 6= 0
cosh(x) sinh(x)
The graphs of the Hyperbolic functions : The graph of y =
sinh(x) can be drawn by first sketching the graphs of y = 1/2ex
and y = −1/2e− x and then adding the y-coordinates of the points
on these graphs corresponding to each x to obtain the y-coordinates
of the points on y = sinh(x). Similarly, the graph of y = cosh(x)
can be drawn by first sketching the graphs of y = 1/2ex and y =
1/2e−x and then adding the y-coordinates of the points on these
graphs corresponding to each x to obtain the y-coordinates of the
points on y = cosh(x).

The graphs of the other four hyperbolic functions are shown below.

39
 Hyperbolic identities : The hyperbolic functions satisfy cer-
tain identities that look very much like those satisfied by trigono-
metric functions. The list of frequently used hyperbolic identities
is given below.

Theorem 14. Hyperbolic identities

a. sinh(−x) = − sinh(x)
b. cosh(−x) = cosh(x)
c. cosh2 (x) − sinh2 (x) = 1
d. sech2 (x) = 1 − tanh2 (x)
e. sinh(x + y) = sinh(x) cosh(y) + cosh(x) sinh(y)
f. cosh(x + y) = cosh(x) cosh(y) + sinh(x) sinh(y)

40
g. sinh(2x) = 2 sinh(x) cosh(x)
h. cosh(2x) = cosh2 (x) + sinh2 (x)
i. cosh2 (x) = 12 (1 + cosh(2x))
j. sinh2 (x) = 21 (−1 + cosh(2x))

Proof. We will discuss the proof of (a) and (c). rest is left to the
reader as exercise (Hint: use the definition of hyperbolic functions!)
−x −(−x) −x x x −x
a. sinh(−x) = e −e2 = e 2−x = − e −e 2
= − sinh(x)
c.

ex + e−x 2 ex − e−x 2
cosh2 (x) − sinh2 (x) = ( ) −( )
2 2
e2x + 2 + e−2x e2x − 2 + e−2x
= −
4 4
=1

Derivatives and integrals of hyperbolic functions : Since the hy-

perbolic functions are defined in terms of ex and e−x , their deriva-
tives are easily computed. For example,

d d ex − e−x ex + e−x
(sinh(x)) = ( )= = cosh(x)
dx dx 2 2

. The table below summarises the differentiation formulas together

with the corresponding integration formulas for the six hyperbolic
functions.

41
Example : Find the derivative of cosh2 (ln2x).
Solution :

d d
cosh2 (ln2x) = 2 cosh(ln2x) cosh(ln2x)
dx dx
d
= 2 cosh(ln2x) sinh(ln2x) ln2x
dx
2
= cosh(ln2x) sinh(ln2x)
x

Example : Find cosh2 (3x) sinh(3x)dx.

R

Solution : Let u = 3x so that du = 3dx or dx = 13 du. Then

Z Z
2 1
cosh (3x) sinh(3x)dx = cosh2 (u) sinh(u)du
3

Next, let v = cosh(u) so that dv = sinh(u)du. Then

Z Z
1 2 1 1
cosh (u) sinh(u)du = v 2 dv = v 3 + C
3 3 9

42
So, Z
1
cosh2 (3x) sinh(3x)dx = cosh3 (3x) + C
9

Inverse Hyperbolic functions : Notice that both sinh(x) and tanh(x)

are one-to-one on (−∞, ∞) and hence have inverse functions that
we denote by sinh−1 (x) and tanh−1 (x) respectively. Also, cosh(x)
is one-to-one on [0, ∞), so, if restricted to this domain, it has an in-
verse, cosh−1 (x). By examining the graphs of the other hyperbolic
functions and making the necessary restrictions on their domains,
we are able to define the other inverse hyperbolic functions.

The graphs of y = sinh−1 (x), y cosh−1 (x), and y = tanh−1 (x) are
shown below.

43
Example : Show that sinh−1 (x) = ln(x +
p
(x2 + 1)).
Solution : Let y = sinh−1 (x). Then

ey − e−y
x = sinh(y) =
2

or
ey − 2x − e−y = 0

On multiplying both sides of this equation by ey , we obtain

e2y − 2xey − 1 = 0

which is a quadratic in ey . Using the quadratic formula, we have

√
2x ± 4x2 + 4 √
ey = = x ± x2 + 1
2
√ √
Only the root x + x2 + 1 is admissible since x − x2 + 1 < 0
but ey > 0. Therefore, we have
√
ey = x + x2 + 1

or
√
y = ln(x + x2 + 1)

that is,
√
sinh−1 (x) = ln(x + x2 + 1)

In this similar manner, we can find out the representations of the

other inverse hyperbolic functions.

44
Representations of Inverse Hyperbolic Functions in Terms of
Logarithmic Functions
√
sinh−1 (x) = ln(x + x2 + 1), x ∈ (−∞, ∞)
√
cosh−1 (x) = ln((x + x2 − 1), x ∈ [1, ∞)
1 1+x
tanh−1 (x) = ln( ), x ∈ (−1, 1)
2 1−x

Derivatives of Inverse Hyperbolic functions : The derivatives of

the inverse hyperbolic functions can be found by differentiating the
function in question directly. For example,

d d √
sinh−1 (x) = ln(x + x2 + 1)
dx dx
1 1
= √ [1 + (x2 + 1)−1/2 (2x)]
x+ x +1 2 2
√
1 x + x2 + 1
= √ . √
x + x2 + 1 x2 + 1
1
=√
2
x +1

Alternatively, y = sinh−1 (x) if and only if x = sinh(y)

Differentiating this last equation implicitly with respect to x, we
obtain
d d
(x) = (sinh(y))
dx dx
dy
1 = (cosh(y))
dx
or
dy 1 1 1
= =√ =√
dx cosh(y) 2
sinh (y) + 1 2
x +1

45
 Using techniques above, we obtain the following formulas for
differentiating the inverse hyperbolic functions.

Example : A power line is suspended between two towers as de-

picted in Figure below. The shape of the cable is a catenary with
equation
x
y = 80 cosh( ) − 100 ≤ x ≤ 100
80
where x is measured in feet. Find the length of the cable.

46
Solution : Taking advantage of the symmetry of the situation, we
see that the required length is given by
Z 100
r
dy 2
L=2 1+( ) dx
0 dx

But

dy d x x d x x
= [80 cosh( )] = 80 sinh( ). ( ) = sinh( )
dx dx 80 80 dx 80 80

So,
r r r
dy x x
1 + ( )2 = 1 + sinh ( ) = 1 + cosh2 ( ) − 1
2
dx 80 80
r
x x
= cosh2 ( ) = cosh( )
80 80

Therefore, Z 100
x
L=2 cosh( )dx
0 80
x 100
= 2[80 sinh( )]
80 0
100 5
= 160 sinh( ) = 160 sinh( )
80 4
≈ 256f t.

1.7 Indeterminate forms and l’Hopital’s rule

If the limx→a f (x) = 0 and and limx→a g(x) = 0, then the limit

f (x)
lim
x→a g(x)

47
is called an indeterminate form of type 0/0. The undefined expres-
sion 0/0 does not provide us with a definitive answer concerning
the existence of the limit or its value, if the limit exists.So, given
an indeterminate form of the type 0/0, we want to see if there is a
more general and efficient method for resolving whether the limit

f (x)
lim
x→a g(x)

exists, and if so, what is the limit?

The indeterminate forms 0/0 and ∞/∞

Theorem 15. (l’Hopital’s Rule) Suppose that f and g are differ-

entiable on an open interval I that contains a, with the possible
exception of a itself, and g 0 (x) 6= 0 for all x in I. If limx→a fg(x)
(x)

indeterminate form of the type 0/0 or ∞/∞ , then

f (x) f 0 (x)
lim = lim 0
x→a g(x) x→a g (x)

provided that the limit on the right-hand side exists or is infinite.

Notes :
1. l’Hopital’s Rule is also valid for one-sided limits as well as limits
at infinity or negative infinity; that is, we can replace x → a by any
of the symbols x → a+ , x → a− , x → ∞, or x → −∞.
2. Before applying l’Hopital’s Rule, check to see that the limit does
have one of the indeterminate forms. For example, cos(x) → 1 as
x → 0+ , so
cos(x)
lim+ =∞
x→0 x

48
If we had applied l’Hopital’s Rule to evaluate the limit without first
ascertaining that it had an indeterminate form, we would have ob-
tained the erroneous result

cos(x) −sin(x)
lim+ = lim+ =0
x→0 x x→0 1

Example : Evaluate limx→1+ sin(πx)

√
x−1
Solution : We have an indeterminate form of the type 0¿0. Apply-
ing l’Hopital’s Rule, we obtain

sin(πx) πcos(πx)
lim+ √ = lim+ 1
x→1 x−1 x→1
2
(x − 1)−1/2
√
= lim+ 2π(cos(πx)) x − 1
x→1

=0

The Indeterminate forms ∞ − ∞ and 0.∞

If limx→a f (x) = ∞ and limx→a g(x) = ∞, then the limit

lim [f (x0 − g(x)]

x→a

is said to be an indeterminate form of the type ∞ − ∞. An indeter-

minate form of this type can often be expressed as one of the type
0/0 or ∞/∞ by algebraic manipulation.

Example : Evaluate limx→0+ ( x1 − ex1−1 ).

Solution : We have an indeterminate form of the type ∞ − ∞ . By
writing the expression as a single fraction, we obtain the indetermi-
nate form of the type 0/0. This enables us to evaluate the resulting

49
expression using l’Hopital’s Rule

1 1 ex − x − 1
lim+ ( − x = lim+
x→0 x e − 1 x→0 x(ex − 1)
ex − 1
= lim+ x
x→0 e − 1 + xex )
ex 1
= lim+ x
=
x→0 (x + 2)e 2

If limx→a f (x) = 0 and limx→a g(x)±∞, then limx→a f (x)g(x)

is said to be an indeterminate form of the type 0.∞. An indetermi-
nate form of this type also can be expressed as one of the type 0/0
or ∞/∞ by algebraic manipulation.

Example : Evaluate limx→)+ lnx.

Solution : We have an indeterminate form of the type 0.∞. By
writing
lna
xlnx = 1
x

the given limit can be cast in an indeterminate form of the type

∞/∞. Then, applying l’Hopital’s Rule, we obtain

lna
lim+ xlnx = lim+ 1
x→0 x→0
x
1
x
= lim+ = lim+ (−x) = 0
x→0 − x12 x→0

The Indeterminate forms 00 , ∞0 , and 1∞

50
The limit
lim [f (x)]g(x)
x→a

is said to be an indeterminate form of the type

1. 00 if limx→a f (x) = 0 and limx→a g(x) = 0.
2. ∞0 if limx→a f (x) = ∞ and limx→a g(x) = 0.
3. 1∞ if limx→a f (x) = 1 and limx→a g(x) = ±∞.

Example : Evaluate limx→0+ xx .

Solution : We have an indeterminate form of the type 00 . Let

y = xx

Then
lny = lnxx = xlnx

Finally, using the identity y = elny and the continuity of the expo-
nential function, we have

lim xx = lim+ y = lim+ elny = elimx→0+ lny = e0 = 1

x→0+ x→0 x→0

Example : Evaluate limx→0+ ( x1 )sin(x) .

Solution : We have an indeterminate form of the type ∞0 . Let

1
y = ( )sin(x)
x

Then
1 1
lny = ln( )sin(x) = (sin(x))ln( )
x x

51
and
1
lim+ lny = lim+ (sin(x))ln
x→0 x→0 x
This last limit is an indeterminate form of the type 0.∞ . By writing

1 ln x1
(sin(x))ln( ) = 1
x sin(x)

we can transform it into an indeterminate form of the type ∞/∞

and hence use l’Hopital’s Rule. We have

ln x1
lim+ lny = lim+ 1
x→0 x→0
sin(x)
lnx
= lim+ 1
x→0
sin(x)
−1
x sin2 (x)
= lim+ −cos(x)
= lim+
x→0 x→0 xcos(x)
sin2 (x)
sin(x)
= lim+ ( )(tan(x)) = 0
x→0 x

Therefore,

1
lim+ ( )sin(x) = lim+ y = lim+ elny = elimx→0+ lny = e0 = 1
x→0 x x→0 x→0

Example : Evaluate limx→∞ (1 + x1 )x .

Solution : We have an indeterminate form of the type 1∞ . Let
y = (1 + x1 )x Then

1 1
lny = ln(1 + )x = xln(1 + )
x x

52
so,
1
lim lny = lim xln(1 + )
x→∞ x→∞ x
has an indeterminate form of the type 0.∞. Rewriting and using
l’Hopital’s Rule, we obtain

1
lim lny = lim xln(1 + )
x→∞ x→∞ x
ln(1 + x1 )
= lim 1
x→∞
x
1
( 1+ 1 )(− x12 )
x
= lim [ −1 ]
x→∞
x2
1
= lim 1 =1
x→∞ 1 +
x

Therefore,

1
lim (1 + )x = lim y = lim elny = elimx→∞ lny = e1 = e
x→∞ x x→∞ x→∞

53
Module 2

Infinite sequences and series

2.1 Improper Integral

Definition 13. Integrals with infinite limits of integration are im-
proper integrals of Type I.
1. If f (x) is continuous on [a, ∞), then
Z ∞ Z b
f (x)dx = lim f (x)dx.
a b→∞ a

2. If f (x) is continuous on (−∞, b] , then

Z b Z b
(x)dx = lim f (x)dx.
∞ a→−∞ a

3. If f (x) is continuous on (−∞, ∞), then

Z ∞ Z c Z ∞
f (x)dx = f (x)dx + f (x)dx,
∞ ∞ c

where c is any real number .

54
In each case, if the limit exists and is finite, we say that the improper
integral converges and that the limit is the value of the improper
integral. If the limit fails to exist, the improper integral diverges
and we say the area under the curve is infinite.

R∞
Example : Evaluate 1 lnx x2
dx
Solution : Integrating by parts,
Z b Z b
lnx 1 1 1
2
dx = [(lnx)(− )]b1 − (− )( )dx
1 x x 1 x x
lnb 1
=− − [ ]b1
b x
lnb 1
=− − +1
b b
Z ∞ Z b
lnx lnx
dx = lim dx
1 x2 b→∞ 1 x
2

lnb 1
= lim [− − + 1]
b→∞ b b
lnb
= −[ lim ]−0+1
b→∞ b
1/b
= −[ lim ]+1=0+1=1
b→∞ 1
R∞ dx
Example : Evaluate −∞ 1+x2
.

55
 R∞ dx
R0 dx
R∞ dx
Solution : −∞ 1+x2
= −∞ 1+x2
+ 0 1+x2

Z 0 Z 0
dx dx
= lim
−∞ 1 + x2 a→−∞ a 1 + x2
= lim tan−1 x|0a
a→−∞

= lim (tan−1 0 − tan−1 a)

a→−∞
π
=
2
Z ∞ Z b
dx dx
= lim
0 1 + x2 b→∞ 0 1 + x2
= lim tan−1 x|b0
b→∞

= lim (tan−1 b − tan−1 0)

b→∞
π
=
2
R∞ dx
R0 dx
R ∞ dx
Thus, −∞ 1+x2
= −∞ 1+x 2 + 0 1+x2
= π2 + π2 = π.

R∞ dx
Exercise : Examine for what values of p does the integral 1 xp
converge?

Definition 14. Integrals of functions that become infinite at a point

within the interval of integration are improper integrals of Type II.
1. If f (x) is continuous on (a, b] and discontinuous at a, then
Z b Z b
f (x)dx = lim+ f (x)dx.
a c→a c

2. If f (x) is continuous on [a, b) and discontinuous at b, then

56
 Z b Z c
f (x)dx = lim− f (x)dx.
a c→b a

3. If f (x) is discontinuous at c, where a < c < b, and continuous

on [a, c) ∪ (c, b] , then
Z b Z c Z b
f (x)dx = f (x)dx + f (x)dx.
a a c

In each case, if the limit exists and is finite, we say the improper
integral converges and that the limit is the value of the improper
integral. If the limit does not exist, the integral diverges. In Part 3 of
the definition, the integral on the left side of the equation converges
if both integrals on the right side converge; otherwise it diverges.
R 1 dx
Example : Evaluate 0 1−x .
1
Solution : Notice that f (x) = 1−x is continuous at [0, 1) but is
discontinuous at x = 1.
Z b
dx
lim− = lim [−ln|1 − x|]b0
b→1 0 1 − x b→1−
= lim−1 [ln(1 − b) + 0] = ∞.
b→1

The limit is infinite, so the integral diverges.

2.2 Sequences
Definition 15. A sequence {an } is a function whose domain is the
set of positive integers. The functional values a1 , a2 , a3 , an ,.... are

57
the terms of the sequence, and the term an is called the a nth term
of the sequence.
Remark: 1. The sequence {an } is also denoted by {an }∞ n=1 .
2. Sometimes it is convenient to begin a sequence with ak . In this
case the sequence is {an }∞ n=1 , and its terms are ak , ak+1 , ak+2 ,..,
an ,... .
Example : List the terms of the sequence.
√
a.{ n+1n
} b.{(−1)n n − 2}∞ n=2 c.{sin nπ }∞
3 n=0
Solution:
n
a. Here an = f (n) = n+1 . Thus,
1
a1 = f (1) = 1+1 = 12 , a2 = f (2) = 2+1 2
= 32 , a3 = f (3) = 3+1
3
=
3
4
,... and we see that the given sequence can be written as

n 1 2 3 4 n
{ } = { , , , , ..., , ...}
n+1 2 3 4 5 n+1

√ √ √ √
b. {(−1)n n − 2}∞
n=2 = {(−1)
2
0, (−1)3 1, (−1)4 2,
√ √
(−1)5 3, ..., (−1)n n − 2, ...}
√ √ √ √
= {0, − 1, 2, − 3, ..., (−1)n n − 2, ...}

Note that n starts from 2 in this example. Refer to Remark: 2.

 
nπ π 2π 3π 4π nπ
c. {sin }∞ = sin0, sin , sin , sin , sin , ..., sin , ...
3 n=0 3 3 3 3 3
( √ √ √ √ )
3 3 3 3 nπ
= 0, , , 0, − ,− , ..., sin , ...
2 2 2 2 3

Note that n starts from 0 in this example. Refer to Remark: 2.

58
Example: Find an expression for the n th term of each sequence.
a. {2, √32 , √43 , √54 , ...} b. {1, − 21 , 13 , − 14 , ...}
Solution: a. The terms of the sequence may be written in the form

1+1 2+1 3+1 4+1

a1 = √ , a2 = √ , a3 = √ , a4 = √ ,
1 2 3 4

Thus, which we see that an = n+1√ .

n
b. Note that (−1)r is equal to 1 if r is an even integer and -1 if r is
an odd integer. Using this result, we obtain

(−1)0 (−1)1 (−1)2 (−1)3

a1 = , a2 = , a3 = , a4 = ,
1 1 3 4
(−1)n−1
Hence, we can conclude that the nth term is an = n
.

Recursively defined sequences: The sequence is defined by speci-

fying the first term or the first few terms of the sequence and a rule
for calculating any other term of the sequence from the preceding
term(s).

Definition 16. Limit of a Sequence

A sequence an has the limit L ,written

lim an = L
n→∞

if an can be made as close to L as we please by taking n sufficiently

large. If limn→∞ an exists, we say that the sequence converges.
Otherwise, we say that the sequence diverges.

59
 We say that a sequence {an } converges and has the limit L,
written
lim an = L
n→∞

if for every  > 0 there exists a positive integer N such that |an −
L| <  whenever n > N .

Theorem 16. If lim f (x) = L and {an } is a sequence defined by

x→∞
an = f (n), where n is a positive integer, then lim an = L.
n→∞

Remark: The converse of the above theorem is false. Consider the

sequence {sinnπ} = {0}. Note that this sequence converges to 0.
However, lim sinπx does not exist.
x→∞

Theorem 17. Limit Laws for Sequences

Suppose that lim an = Land lim bn = M and that c is a constant.
n→∞ n→∞
Then

• lim can = cL
n→∞

• lim (an ± bn ) = L ± M
n→∞

• lim an bn = LM
n→∞

an L
• lim = , provided that bn 6= 0 and M 6= 0
n→∞ bn M

• lim apn = Lp , if p > 0 and an > 0

n→∞

Theorem 18. Squeeze Theorem for Sequences

If there exists some integer N such that an ≤ bn ≤ cn for all n ≥ N
and lim an = lim cn = L, then lim bn = L.
n→∞ n→∞ n→∞

60
Figure 2.1: The sequence {bn } is squeezed between the sequences
{an } and {cn }
.

Example: For n! defined as n! = n(n − 1)(n − 2)...1; find

n!
lim .
n→∞ nn

Solution: Let an = n!/nn . Then, the first three terms are given by

1! 2! 2·1 3! 3·2·1
a1 = = 1, a2 = = , a3 = =
1 2 2·2 3 3·3·3

and the nth term is

n! n(n − 1) · ... · 3 · 2 · 1
a3 = =
n n · n · ... · n ·n · 
n   
n n − 1 3 2 1
= · ... ·
n n n n n
 
1
≤
n

61
 Therefore,
1
0 < an ≤ .
n
1
Since lim = 0, the Squeeze Theorem gives us
n→∞ n

n!
lim an = lim =0
n→∞ n→∞ nn

Theorem 19. If lim |an | = 0, then lim an = 0

n→∞ n→∞

Remark: The above theorem is an immediate consequence of the

Squeeze Theorem.

Theorem 20. If lim an = L and the function f is continuous at L

n→∞
, then
lim f (an ) = f ( lim an ) = f (L)
n→∞ n→∞

Definition 17. Monotonic Sequences

A sequence {an } is increasing if

a1 < a2 < a3 < ... < an < an+1 < ...

and decreasing if

a1 > a2 > a3 > ... > an > an+1 > ...

A sequence is monotonic if it is either increasing or decreasing.

Definition 18. Bounded Sequence

A sequence {an } is bounded above if there exists a number M such
that
an ≤ M for all n ≥ 1.

62
A sequence is bounded below if there exists a number m such that

m ≤ an for all n ≥ 1.

A sequence is bounded if it is both bounded above and bounded

below.

Remarks:

a. A bounded need not be convergent. Note that, the se-

quence (−1)n is bounded since −1 ≤ (−1)n ≤ 1; however,
it is evidently divergent.

b. A monotonic sequence need not be convergent. Consider

the sequence {n}. This sequence is clearly increasing, yet
divergent.

Theorem 21. Monotone Convergence Theorem for Sequences

Every bounded, monotonic sequence is convergent.
 n
Example: Show that 2n! is convergent and find its limit.
Solution: Here, an = 2n /n!. Therefore, the first few terms of the
sequence are
a1 = 2, a2 = 2, a3 ≈ 1.333333, a4 ≈ 0.066667, a5 = 0.266667,
a6 ≈ 0.0888889, ...a10 ≈ 0.000282
This suggests that the sequence is decreasing from n = 2 onward.
To prove this, we compute

2n+1
an+1 (n+1)! 2n+1 n! 2 · n! 2
= 2n = n
= =
an n!
2 (n + 1)! (n + 1)n! n+1

63
So, we have
an+1
≤ 1 if n ≥ 1.
an
Thus, an+1 ≤ an if n ≥ 1 and hence we have proved the assertion.
Since all the terms in the sequence are positive, {an } is bounded
below by 0. Therefore, the sequence is decreasing and bounded be-
low, and the Monotone Convergence Theorem for Sequences guar-
antees that it converges to a non-negative limit L.
Now to find L, consider:

2
an+1 = an
n+1

Note that since lim an+1 = L, we also have lim an = L. Taking

n→∞ n→∞
the limit on both sides of the above equation and using Law (3) for
limits of sequences, we obtain
 
2 2
L = lim an+1 = lim an = lim · lim an = 0·L = 0
n→∞ n→∞ n+1 n→∞ n + 1 n→∞

Thus, we can conclude that lim 2n /n! = 0.

n→∞

64
2.3 Series
Definition 19. In general, an expression of the form

a1 + a2 + a3 + ... + an + ...

is called an infinite series or, more simply, a series. The numbers

a1 , a2 , a3 , ... are called the terms of the series; an is called the nth
term, or general term, of the series; and the series itself is denoted
by the symbol
X∞
an
n=1
P
or simply an .

Definition 20. Convergence of Infinite Series

Given an infinite series
∞
X
an = a1 + a2 + a3 + ... + an + ...
n=1

the nth partial sum of the series is

n
X
Sk = ak = a1 + a2 + a3 + ... + an
k=1

If the sequence of partial sums {Sn } converges to the number S,

P
that is, if lim Sn = S, then the series an converges and has
n→∞
sum S, written
∞
X
an = a1 + a2 + a3 + ... + an + ... = S
n=1

65
 P
If {Sn } diverges, then the series an diverges.

Example: Determine whether the series converges. If the series

converges, find its sum.
∞ ∞  
X X 1 1
a. n b. −
n=1 n=1
n n+1

Solution : a. The nth partial sum of the series is

n(n + 1)
Sn = 1 + 2 + 3 + ... + n =
2

Since
n(n + 1)
lim Sn = lim =∞
n→∞ n→∞ 2
Thus, we can conclude that the limit does not exist and ∞
P
n=1 n
diverges.
b. The nth partial sum of the series is
       
1 1 1 1 1 1 1
Sn = 1 − + − + ... + − + −
2 2 3 n−1 n n n+1
1
=1−
n+1

Since  
1
lim Sn = lim 1− =1
n→∞ n→∞ n+1
we can conclude that
∞  
X 1 1
− =1
n=1
n n+1

Remark: The series in part b. of Example is called a telescoping

66
series.
Example: Show that the series ∞ 4
P
n=1 4n2 −1 is convergent, and find
its sum.
Solution: First, we use partial fraction decomposition to rewrite
the general term an = 4/(4n2 − 1):

4 4 2 2
an = = = −
4n2 −1 (2n − 1)(2n + 1) 2n − 1 2n + 1

Then we write the nth partial sum of the series as

n n  
X 4 X 2 2
Sn = 2−1
= −
k=1
4k 2k − 1 2k + 1
   k=1     
2 2 2 2 2 2 2 2
= − + − + − + ... + −
1 3 3 5 5 7 2n − 1 2n + 1
2
=2−
2n + 1

Note that this is also a telescoping series. Now, since

 
2
lim Sn = lim 2− =2
n→∞ n→∞ 2n + 1

we can conclude that the given series is convergent and has sum 2.

2.3.1 Geometric Series

Definition 21. Geometric Series
A series of the form
∞
X
arn−1 = a + ar + ar2 + ... + arn−1 + ...a 6= 0
n=1

67
is called a geometric series with common ratio r .

Theorem 22. If |r| < 1, then the geometric series

∞
X
arn−1 = a + ar + ar2 + ... + arn−1 + ...
n=1

P∞ a
converges, and its sum is n=1 arn−1 = 1−r
. The series diverges
if |r| ≥ 1.

Proof. The nth partial sum of the series is

∞
X
Sn = arn−1 = a + ar + ar2 + ... + arn−1
n=1

Multiplying both sides of this equation by r gives

rSn = ar + ar2 + ar3 + ... + arn

Subtracting the second equation from the first then yields (1 −

r)Sn = a(1 − rn ). If r 6= 1, we can solve for Sn , obtaining

a(1 − rn )
Sn =
1−r

Recall that lim rn = 0 if |r| < 1. Then,

n→∞

a(1 − rn ) a
lim Sn = lim =
n→∞ n→∞ 1−r 1−r

68
This implies that
∞
X a
arn−1 = |r| < 1
n=1
1−r

When |r| > 1, the sequence {rn } diverges. So, lim Sn does not
n→∞
exist. Thus, the geometric series diverges. Verify that {Sn } di-
verges if r = ±1, implying that the series also diverges for these
values of r.

Example: Express the number 3.214 = 3.2141414... as a rational

number.
Solution: We rewrite the number as

14 14 14
3.2141414... = 3.2 + 3
+ 5 + 7 + ...
10  10 10 
32 14 1 1
= + 3 1 + 2 + 4 + ...
10 10 10 10
∞    n−1
32 X 14 1
= +
10 n=1
103 102

Note that the expression after the first term is a geometric series
14 1
with a = 1000 and r = 100 . Using the theorem above, we have

14
32 32 14 3182
3.2141414... = + 10001 = + =
10 1 − 100 10 990 990

69
2.3.2 The Harmonic Series

Definition 22. The series

∞
X 1 1
=1+ 1 + ...
n=1
n 2+ 3

is called the harmonic series.

For showing that this series is divergent, firstly note that: If

a sequence {bn } is convergent, then any subsequence obtained by
deleting any number of terms from the parent sequence {bn } must
also converge to the same limit. Therefore, to show that a sequence
is divergent, it suffices to produce a subsequence of the parent se-
quence that is divergent.
Thus, it would be sufficient to show that the subsequence
S2 , S4 , S8 , S16 , ..., S2n , ...
of the sequence {Sn } of partial sums of the harmonic series is di-
vergent.

70
 1
S2 = 1 +
2      
1 1 1 1 1 1 1
S4 = 1 + + + >1+ + + =1+2
2 3 4 2 4 4 2
   
1 1 1 1 1 1 1
S8 = 1 + + + + + + +
2 3 4 5 6 7 8
     
1 1 1 1 1 1 1 1
>1+ + + + + + + =1+3
2 4 4 8 8 8 8 2
     
1 1 1 1 1 1 1
S16 = 1 + + + + + ... + + + ... +
2 3 4 5 8 9 16
     
1 1 1 1 1 1 1
>1+ + + + + ... + + + ... +
2 4 4 8 8 16 16
 
1
=1+4
2

Thus, S2n > 1 + n 12 . Therefore, lim S2n = ∞. So, {Sn } is

n→∞
divergent. This proves that the harmonic series is divergent.

2.3.3 The Divergence Test

P∞
Theorem 23. If n=1 an converges, then lim an = 0.
n→∞

Proof. We can write the partial sum as Sn = a1 + a2 + ... + an−1 +

an = Sn−1 + an , so an = Sn − Sn−1 . Since ∞
P
n=1 an is convergent,
the sequence {Sn } is convergent. Let lim Sn = S. Then
n→∞

lim an = lim (Sn − Sn−1 ) = lim Sn − lim Sn−1 = S − S = 0

n→∞ n→∞ n→∞ n→∞

71
Theorem 24. The Divergence Test
If lim an does not exist or lim an 6= 0, then ∞
P
n→∞ n→∞ n=1 an diverges.

Remark: The Divergence Test does not say that if lim an =

n→∞
0, then ∞
P
n=1 a n must converge. In other words, the converse of
2.3.3is not true in general.
Example: Show that the following series are divergent.
∞ ∞
X X 2n2 + 1
a. (−1)n−1 b.
n=1 n=1
3n2 − 1

Solution: a. Observe that lim an = lim (−1)n−1 does not exist.

n→∞ n→∞
Thus, by divergence test the series diverges.
b.
2n2 + 1 2 + n12 2
lim an = lim = lim 1 = 6= 0
n→∞ n→∞ 3n2 − 1 n→∞ 3 − 2 3
n

Thus, by the Divergence Test, the series diverges.

2.3.4 Properties of Convergent Series

Theorem 25. Properties of Convergent Series
If ∞
P P∞
n=1 an = A and n=1 bn = B are convergent and c is any real
number, then n=1 can and ∞
P∞ P
n=1 (an ± bn ) are also convergent,
and
P∞ P∞
a. n=1 can = c n=1 an = cA
P∞ P∞ P∞
b. n=1 (an ± bn ) = n=1 an ± n=1 bn = A ± B
h i
Example: Show that the series ∞ 2 4
P
n=1 n(n+1) − 3n is convergent,
and find its sum.
Solution: Consider ∞
P
n=1 1/[n(n + 1)]. Using partial fraction

72
decomposition, and previous example,
∞ ∞  
X 1 X 1 1
1= = −
n=1
[n(n + 1)] n=1 n n+1

Observe that ∞ 4 4 1
P
n=1 3n is a geometric series with a = 3 and r = 3 .
4
Thus, ∞ 4
P
n=1 3n = 1− 1 = 2. Thus, by Properties of Convergent
3
3
Series the given series is convergent and
∞   ∞ ∞
X 2 4 X 1 X 4
− n =2 − = 2·1−2 = 0
n=1
n(n + 1) 3 n=1
n(n + 1) n=1 3n

2.4 The Integral Test

Theorem 26. The Integral Test Suppose that f is a continuous,
positive, and decreasing function on [1, ∞) if f (n) = an for n ≥ 1,
then ∞ Z ∞
X
an and f (x)dx
n=1 1

either both converge or both diverge.

Proof. In Figure (a) given, observe that the height of the first rect-
angle is a2 = f (2). Since this rectangle has width 1, the area of
the rectangle is also a2 = f (2).Similarly, the area of the second
rectangle is a3 , and so on. Comparing the sum of the areas of the
first (n − 1) inscribed rectangles with the area of the region under
the graph of f over the interval [1, n] , we see that
Z n
a2 + a3 + ... + an ≤ f (x)dx
1

73
which gives us the following result. The second inequality follows
Rn
if we consider 1 f (x)dx is convergent and has value L.
Z n
Sn = a1 + a2 + a3 + ... + an ≤ a1 + f (x)dx ≤ a1 + L
1

This shows that {Sn } is bounded above. Also, since an+1 = f (n +

1) ≥ 0, we have that Sn+1 = Sn + an+1 ≥ Sn which proves
that {Sn } is increasing as well. Therefore, by Monotone Conver-
gence Theorem for Sequences, {Sn } is convergent. Consequently,
P∞
n=1 an is convergent.
Next, by examining Figure(b) we can see that
Z n
f (x)dx ≤ a1 + a2 + a3 + ... + an−1 = Sn−1
1

Rn
Thus, if 1 f (x)dx diverges (to ∞ since f (x) ≥ 0),then lim Sn−1 =
n→∞
lim Sn = ∞, and ∞
P
n=1 a n is divergent.
n→∞

Example: Use the Integral Test to determine whether the fol-

74
lowing converges or diverges.
∞ ∞
X 1 X ln n
a. 2
b.
n=1
n +1 n=1
n

Solution: a. Here an = f (n) = 1/(n2 +1), giving f (x) = 1/(x2 +

1). Since f is continuous, positive, and decreasing on [1, ∞), we
may use the Integral Test.
Z ∞ Z b
1 1
2
dx = lim dx
1 x +1 b→∞ 1 x2 +1
= lim [tan−1 x]b1
b→∞

= lim (tan−1 b − tan−1 1)

b→∞
π π
= −
2 4
π
=
4
R∞
Since 1 x21+1 dx converges, so does ∞ 1
P
n=1 n2 +1 .
b. Here an = (ln n)/n. Thus, f (x) = (ln x)/x. Observe that f is
continuous and positive on [1, ∞). Now to see if it is decreasing or
not. Consider
1


x − ln x 1 − ln x
f 0 (x) = x
2
=
x x2

Note that f 0 < 0 when ln x > 1, i.e., if x > e. Hence, f is

75
decreasing on [3, ∞). Therefore, we can use the Integral Test:
Z ∞ Z b
ln x ln x
dx = lim dx
3 x b→∞ 3 x
 b
1 2
= lim (ln x)
b→∞ 2
3
1
= lim [(ln b)2 − (ln 3)2 ]
b→∞ 2

=∞
P∞ ln n
We can conclude that n=1 n
diverges.

The p -Series
Definition 23. The p -Series
A p -series is a series of the form
∞
X 1 1 1
p
= 1 + p + p + ...
n=1
n 2 3

where p is a constant.
Remark: Note that when p = 1, the p -series is just the harmonic
series.
Theorem 27. Convergence of the p -Series
The p -series a p converges if p > 1 and diverges if p ≤ 1.
1 1
Proof. When p < 0, lim p = ∞ and if p = 0, then lim np = 1.
n→∞ n n→∞
In either case, lim 1p 6= 0, so the p-series diverges by the Di-
n→∞ n R∞
vergence Test. Recall that 1 1/xp dx converges if p > 1 and di-
verges if p ≤ 1. Using this and by the Integral Test, we have that
P∞ p
n=1 1/n converges if p > 1 and diverges if p ≤ 1.

76
Example: Determine whether the given series converges or di-
verges.
∞ ∞ ∞
X 1 X 1 X
a. b. √ c. n−1.001
n=1
n2 n=1
n n=1

Solution: a. This is a p-series with p = 2 > 1. Thus it converges

by theorem.
b. This is a p-series with p = 1/2 < 1. Thus it diverges by theorem.
c. This is a p-series with p = 1.001 > 1. Thus it converges by
theorem.

2.5 The Comparison Tests

2.5.1 The Comparison Test

P P
Theorem 28. Suppose that an and bn are series with positive
terms.
P P
a. If bn is convergent and an ≤ bn for all n , then an is also
convergent.
P P
b. If bn is divergent and an ≥ bn for all n , then an is also
divergent.

Proof. Let
n
X n
X
Sn = ak and Tn = bk
k=1 k=1
P P
be the nth terms of the sequence of partial sums of an and bn ,
respectively. Since both series have positive terms, {Sn } and {Tn }
are increasing.

77
a. If ∞
P
n=1 bn is convergent, then there exists a number L such that
lim Tn = L and Tn ≤ L for all n. Since an ≤ bn for all n, we have
n→∞
Sn ≤ Tn , and this implies that Sn ≤ L for all n. We have shown
that {Sn } is increasing and bounded above, so by the Monotone
P
Convergence Theorem for Sequences, an converges.
b. If ∞
P
lim Tn = ∞ since {Tn } increas-
n=1 bn is divergent, then n→∞
ing. But an ≥ bn for all n, we have Sn ≥ Tn , and this implies that
P
lim Sn = ∞. Therefore, an diverges.
n→∞

Example: Determine whether the following series are convergent

or divergent.
∞ ∞
X 1 X 1
a. b. √
n=1
3 + 2n n=2
n−1
1 n
Solution: a. Let an = 3+2 n . For large n, 3 + 2 behaves like 2n .
n
So, an behaves like bn = ( 21 ). Comparing
P P
an with bn , we
1
P
see that bn is a geometric series with r = 2 < 1, implying that
it is convergent. Observe that

1 1
an = < = bn n≥1
3 + 2n 2n

Thus by Comparison test, the given series is convergent.

Remark: Since the convergence or divergence of a series is not af-

fected by the omission of a finite number of terms of the series, the
condition an ≥ bn (or an ≤ bn ) for all n can be replaced by the con-
dition that these inequalities hold for all n ≥ N for some integer N .

√ √
b. Let an = √1 . For n large, n − 1 behaves like n, so an
n−1

78
 P∞ √
behaves like bn = √1n . Note that the series ∞
P
n=2 bn = n=2 n
is a p-series with p = 21 , 1, thus making it divergent. Since

1 1
an = √ > √ = bn for n≥2
n−1 n

we get that the given series id divergent by Comparison test.

2.5.2 The Limit Comparison Test

Theorem 29. The Limit Comparison Test
P P
Suppose that an and bn are series with positive terms and

an
lim =L
n→∞ bn

where L is a positive number. Then either both series converge or

both diverge.

Proof. Since lim abnn = L > 0, there exists an integer N such that
n→∞
n ≥ N implies that

an 1 1 an 3 1 3
− L < L =⇒ L < < L OR Lbn < an < Lbn
bn 2 2 bn 2 2 2
P P3
If bn converges, so does 2
Lbn . Therefore, the right side of
P
the last inequality implies that an converges by the Comparison
P P1
Test. On the other hand, if bn diverges, so does 2
Lbn , and the
left side of the last inequality implies by the Comparison Test that
P
an diverges as well.
√
Example: Determine whether the series ∞ n+ln n
P
n=1 n2 +1 converges
or diverges.

79
 √ √
Solution: Note that n large, n + ln n behaves like n. We can
√
verify this by computing the derivatives of f (x) = x and g(x) =
ln x:

1 1
f 0 ()x = √ and g 0 (x) =
2 x x

Observe that g 0 (x) approaches zero faster than f 0 (x) approaches

√
zero, as x −→ ∞. This shows that x grows faster than ln x.
Also, if n is large, n2 + 1 behaves like n2 . Therefore,
√ √
n + ln n n 1
an = = 3/2 = bn
n2 + 1 n 2 n

Next we compute,

an n1/2 + ln n n3/2 n2 + n3/2 ln n

lim = lim · = lim
n→∞ bn n→∞ n2 + 1 1 n→∞ n2 + 1
1 + nln1/2n
= lim
n→∞ 1 + 12
n

The last equality is obtained by dividing the numerator and denom-

inator by n2 . Now, note that by l’Hôpital’s Rule,
1
ln n x 2
lim = lim 1 = lim √ =0
x→∞ n 1/2 x→∞ x−1/2 x→∞ x
2

√
This result also supports the observation made earlier that x grows
faster than ln x. Using this, we see that

an 1 + nln1/2n
lim = lim =1
n→∞ bn n→∞ 1 + 12
n

80
Since 1/n3/2 converges (as it is a p-series with p = 32 ), the given
P

series converges, by the Limit Comparison Test.

2.6 Alternating Series

Theorem 30. The Alternating Series Test
If the alternating series
∞
X
(−1)n−1 an = a1 − a2 + a3 − a4 + a5 − a6 + ... an > 0
n=1

satisfies the conditions

1. an+1 ≤ an for all n
2. lim an = 0
n→∞
then the series converges.

Figure 2.2: Note that the terms of {Sn } oscillate in smaller and
smaller steps, and this suggests that lim Sn = S.
n→∞

81
Proof. Consider the subsequence {S2n } comprising the even terms
of the sequence of partial sums {Sn }. Now,

S 2 = a1 − a2 ≥ 0 Since a1 ≥ a2
S4 = S2 + (a3 − a4 ) ≥ S2 Since a3 ≥ a4
.
.
S2n+2 = S2n + (a2n+1 − a2n+2 ) ≥ S2n Since a2n+1 ≥ a2n+2

Thus, we have that 0 ≤ S2 ≤ S4 ≤ ... ≤ S2n ≤ .... That is, {S2n }

is increasing. Note that we can write S2n as

S2n = a1 − (a2 − a3 ) − (a4 − a5 ) − ... − (a2n−2 − a2n−1 ) − a2n

where every expression within the parenthesis is nonnegative. Thus,

S2n ≤ a1 for all n. This shows that the sequence {S2n } is bounded
above as well. Therefore, by the Monotone Convergence Theorem
for Sequences, the sequence {S2n } is convergent; that is, there ex-
ists a number S such that lim S2n = S.
n→∞
Now, consider the subsequence {S2n+1 } comprising the even terms
of {Sn }. Since S2n+1 = S2n + a2n+1 and lim a2n+1 = 0 by as-
n→∞
sumption, we have

lim S2n+1 = lim (S2n+1 +a2n+1 ) = lim S2n+1 + lim a2n+1 = S

n→∞ n→∞ n→∞ n→∞

Since the subsequences {S2n } and {S2n+1 } of the sequence of par-

tial sums {Sn } both converge to S, we have lim Sn = S, so the
n→∞
series converges.

82
Example : Determine whether the series converges or diverges.
∞ ∞
X 2n X 3n
a. (−1)n b. (−1)n
n=1
4n − 1 n=1
4n2 − 1

Solution: a. Here, an = 2n/(4n − 1). Now,

2n 1
lim = 6= 0
n→∞ 4n − 1 2

Observe that condition (2) of Alternating Series Test is not satis-

2n
fied. In fact, note that the lim (−1)n 4n−1 does not exist, and the
n→∞
divergence of the series follows from the Divergence Test.
b. Here an = 3n/(4n2 − 1). Consider

(4x2 − 1)(3) − (3x)(8x) −12x2 − 3

f 0 (x) = =
(4x2 − 1)2 (4x2 − 1)2

Thus, condition (1) that is, an ≥ an+1 for all n for the Alternat-
ing Series test is verified by showing that f (x) = 3x/(4x2 − 1) is
decreasing for x ≥ 0. Now, for condition (2).
3
3n n
lim an = lim = lim =0
n→∞ n→∞ 4n2 − 1 n→∞ 4 − 12
n

Therefore, both conditions of the Alternating Series Test are satis-

fied. We can conclude that the series is convergent.

83
2.6.1 Approximating the Sum of an Alternating Se-
ries by Sn

The sum of a convergent series can be approximated to any degree

of accuracy by its nth partial sum Sn , provided that n is taken large
enough. To measure the accuracy of the approximation, consider
the quantity
∞
X n
X ∞
X
Rn = S−Sn = ak = ak = ak = an+1 +an+2 +an+3 +...
k=1 k=1 k=n+1

called the remainder after n terms of the series ∞

P
n=1 ak . The
remainder measures the error incurred when S is approximated by
Sn .

Theorem 31. Error Estimate in Approximating an Alternating

Series
Suppose ∞ n−1
P
n=1 (−1) an is an alternating series satisfying
1. 0 ≤ an+1 ≤ an for all n
2. lim an = 0
n→∞
If S is the sum of the series, then

|Rn | = |S − Sn | ≤ an+1

In other words, the absolute value of the error incurred in approx-

imating S by Sn is no larger than an+1 , the first term omitted.

84
Proof. We have
∞
X n
X ∞
X
k−1 k−1
S − Sn = (−1) ak − (−1) ak = (−1)k−1 ak
k=1 k=1 k=n+1

= (−1)n an+1 + (−1)n+1 an+2 + (−1)n+2 an+3 + ...

= (−1)n (an+1 − an+2 + an+3 − ...)

Note that, since an+1 ≤ an for all n,

an+1 − an ≥ 0 for all n

and thus we have

|S − Sn | = an+1 − an+2 + an+3 − an+4 + an+5 − ...

= an+1 − (an+2 − an+3 ) − (an+4 − an+5 ) − ...
≤ an+1

Example: Show that the series ∞ n 1

P
n=1 (−1) n! is convergent, and
find its sum correct to three decimal places.
Solution: Note that for all n,

1 1 1
an+1 = = < = an and
(n + 1)! n!(n + 1) n!

1
lim an = lim =0
n→∞ n→∞ n!

Thus, we conclude that the series converges by the Alternating Se-

85
ries Test. Now to compute the sum consider the reminder term

1
|Rn | = |S − Sn | ≤ an+1 =
(n + 1)!

1
We need |Rn | < 0.0005, which is satisfied if (n+1)! < 0.0005 or
1
(n+1)! > 0.0005 = 2000. The smallest positive integer that satisfies
the last inequality is n = 6. Hence, the required approximation is

1 1 1 1 1 1 1
S ≈ S6 = − + − + − + ≈ 0.368
0! 1! 2! 3! 4! 5! 6!

2.7 Absolute Convergence; the Ratio and

Root Tests

2.7.1 Absolute Convergence

Definition 24. Absolutely Convergent Series
P P
A series an is absolutely convergent if the series |an | is con-
vergent.

Example: Show that the alternating harmonic series

∞
X (−1)n−1 1 1 1
=1− + − + ...
n=1
n 2 3 4

is not absolutely convergent.

Solution: Consider
∞ ∞
X (−1)n−1 X 1 1 1 1
= = 1 + + + + ...
n=1
n n=1
n 2 3 4

86
which is the divergent harmonic series. This shows that the series
is not absolutely convergent.

Definition 25. Conditionally Convergent Series

P
A series an is conditionally convergent if it is convergent but
not absolutely convergent.

P
Theorem 32. If a series an is absolutely convergent, then it is
convergent.

Proof. Using an absolute value property, we have −|an | ≤ an ≤

|an | and now adding |an | to both sides gives 0 ≤ an + |an | ≤ 2|an |.
P P
Let bn = an +|an |. If an is absolutely convergent, then |an | is
convergent, which in turn implies, by Theorem part(a) of Properties
P P
of Convergent Series that 2|an | is convergent. Therefore, bn
is convergent by the Comparison Test. Finally, since an = bn −|an |,
P P P
we see that an = bn − |an | is convergent by Theorem
part(b) of Properties of Convergent Series.

2.7.2 Ratio Test

Theorem 33. The Ratio Test
P
Let an be a series with nonzero terms.
P∞
a. If lim an+1 = L < 1, then n=1 an converges absolutely.
n→∞ an
P∞
b. If lim an+1 = L > 1, or lim an+1
= ∞, then n=1 an
n→∞ an n→∞ an
diverges.
c. If lim an+1 = 1, the test is inconclusive, and another test
n→∞ an
should be used.

87
Proof. Suppose that

an+1
lim =L<1
n→∞ an

Let r be any number such that 0 ≤ L < r < 1. Then there exists
an integer N such that

an+1
<r OR |an+1 | < |an |r for n ≥ N
an

Letting n take on the values N, N + 1, N + 2, ..., successively, we

obtain

|aN +1 | < |aN |r

|aN +2 | < |aN +1 |r < |aN |r2
|aN +3 | < |aN +2 |r < |aN |r3
.
.
In general, |aN +k | < |aN |rk for all k≥1

Observe that the series ∞ k

P
k=1 |aN |r (series(1)) is a convergent geo-
metric series with 0 < r < 1 and each term of the series ∞
P
k=1 |aN +k |
(series(2)) is less than the corresponding term of the geometric se-
ries (1). The Comparison Test then implies that series (2) is conver-
gent. Since convergence or divergence is unaffected by the omis-
sion of a finite number of terms, we see that the series ∞
P
n=1 |an | is
also convergent.

88
b. Suppose that
an+1
lim =L>1
n→∞ an
Let r be any number such that L > r > 1. Then there exists an
integer N such that

an+1
>r>1 whenever n≥N
an

This implies that |an+1 | < |an | when n ≥ N . Thus, lim an 6= 0,

P n→∞
and an is divergent by the Divergence Test.
c. Consider the series ∞
P P∞ 2
k=1 1/n and k=1 1/n . For the first se-
ries we have

an+1 1 n 1
lim = lim · = lim 1 =1
n→∞ an n→∞ n + 1 1 n→∞ 1 +
n

and for the second series,

an+1 1 n2 1
lim = lim · = lim =1
n→∞ an n→∞ (n + 1) 2 1 n→∞ (1 + 1 )2
n

The first series is the divergent harmonic series, whereas the second
series is a convergent p-series with p = 2. Thus, if L = 1, the series
may converge or diverge, and the Ratio Test is inconclusive.

P∞ n!
Example: Determine whether the series n=1 nn is convergent or
divergent.

89
Solution: Let an = n!/nn . Then,

an+1 (n + 1)! nn
lim = lim ·
n→∞ an n→∞ (n + 1)n+1 n!
(n + 1)n! nn
= lim ·
n→∞ (n + 1)(n + 1)n n!
 n
n 1
= lim = lim
n
n→∞ n+1 n→∞ 1 + 1
n
1 1
=
= <1
1 n
lim 1 + n e
n→∞

Therefore, the series converges, by the Ratio Test.

2.7.3 The Root Test

Theorem 34. The Root Test

Let ∞
P
n=1 an be a series
a. If lim n |an | = L < 1, then ∞
p P
n→∞ p n=1 an converges absolutely.

b. If lim n |an | = L > 1, or lim n |an | = ∞ then ∞

p P
n→∞ n→∞ n=1 an
diverges.
p
c. If lim n |an | = 1,the test is inconclusive, and another test
n→∞
should be used.

Example: Determine whether the series ∞ n−1 2n+3

P
n=1 (−1) (n+1)n
is
absolutely convergent, conditionally convergent, or divergent.
Solution: We apply the Root Test with an = (−1)n−1 2n+3 /(n +

90
1)n . We have
s
1/n
p 2n+3 2n+3
lim n |an | = lim n
(−1)n−1 = lim
n→∞ n→∞ (n + 1)n n→∞ (n + 1)n
21+3/n
= lim =0<1
n→∞ n+1

Thus, the series is absolutely convergent.

2.7.4 Rearrangement of Series

Example: Consider the alternating harmonic series that converges
to ln 2

1 1 1 1 1 1 1
1− + − + − + − + ... = ln2
2 3 4 5 6 7 8

If we rearrange the series so that every positive term is followed by

two negative terms, we obtain

1 1 1 1 1 1 1 1
1− − + − − + − − + ...
2 4 3 6 8 5 10 12  
1 1 1 1 1 1 1 1
= 1− − + − − + − − + ...
2 4 3 6 8 5 10 12
1 1 1 1 1 1
= − + − + − + ...
2  4 6 8 10 12 
1 1 1 1 1 1
= 1 − + − + − + ...
2 2 3 4 5 6
1
= ln2
2

Thus, rearrangement of the alternating harmonic series has a

sum that is one half that of the original series!

91
NOTE:

• Reimann proved that:

If x is any real number and ∞
P
n=1 an is conditionally conver-
gent, then there is a rearrangement of ∞
P
n=1 an that converges
to x.

Remark: Riemann’s result tells us that for conditionally

convergent series, we may not rearrange their terms, lest we
end up with a totally different series, that is, a series with
a different sum. In fact, for conditionally convergent series,
one can find rearrangements of the series that diverge to in-
finity, diverge to minus infinity, or oscillate between any two
prescribed real numbers!

• Q: What kind of convergent series will have rearrange-

ments that converge to the same sum as the original se-
ries?

If ∞
P P∞
n=1 an converges absolutely and n=1 bn is any rear-
rangement of n=1 an , then n=1 bn converges and ∞
P∞ P∞ P
n=1 an =
P∞
n=1 an .

• Since a convergent series with positive terms is absolutely

convergent, its terms can be written in any order, and the
resultant series will converge and have the same sum as the
original series.

92
Example: Indicate the test(s) that you would use to determine
whether the series converges or diverges.
∞ ∞   ∞  e
X 2n − 1 X 2 1 X 1
a. b. n
− c.
n=1
3n + 1 n=1
3 n(n + 1) n=1
n
∞ ∞ ∞ √
X 1 ln n
X X n3 + 2
d. √ e. f.
n=1 n ln n n=1
n2 n=1
n4 + 3n2 + 1
∞ √ ∞ ∞
X
n n X n X sin n
g. (−1) 2 h. i. √
n=1
n +1 n=1
2n n=1
n3 + 1

Solution:
a. Since
2n − 1 2
lim an = lim = 6= 0
n→∞ n→∞ 3n − 1 3
we use the Divergence Test.
b. The series is the difference of a geometric series and a tele-
scoping series, so we use the properties of these series to determine
convergence.

e
c. Here, an = n1 = n1e is a p-series, so we use the properties of a
p-series to study its convergence.
d.The function f (x) = x√1ln x is continuous, positive, and decreas-
ing on [3, ∞) and is integrable, so we choose the Integral Test.
e. Here, √
ln n n 1
an = 2 < 2 = 3/2 = bn
n n n
P
and we use the Comparison Test with the test series bn .
3 +2)1/2 (n3 )1/2 3/2
f. an = n(n4 +3n 2 +1 is positive and behaves like bn = n4
= nn4 =
1
n5/2
for large values of n, so we use the Limit Comparison Test
with test series ∞ 5/2
P
n=1 1/n .

93
g. This is an alternating series, and we use the Alternating Series
Test.  1/n n
h. Here, an = 2nn = n 2 involves the nth power, so the Root
√
n
Test is a candidate.In fact, here lim n |an | = lim 2n = 12 < 1
p
n→∞ n→∞
and the series converges.
i. The series involves both positive and negative terms and is not an
alternating series, so we use the test for absolute convergence.

94
95
Module 3

Power series, plane curves

and polar coordinates

3.1 Power series

Definition 26. Let x be a variable. A power series in x is a series
of the form
∞
X
an xn = a0 + a1 x + a2 x2 + .... + an xn + ....
n=0

where the an ’s are constants and are called the coefficients of the
series. More generally, a power series in (x − c), where c is a
constant, is a series of the form
∞
X
an (x−c)n = a0 +a1 (x−c)+a2 (x−c)2 +....+an (x−c)n +....
n=0

Notes :

96
1. A power series in (x − c) is also called a power series centered
at c or a power series about c. Thus, a power series in x is just a
series centered at the origin.
2. To simplify the notation used for a power series, we have adopted
the convention that (x − c)0 = 1, even when x = c.

We can view a power series as a function f defined by the rule

∞
X
f (x) = = an (x − c)n
n=0

The domain of f is the set of all x for which the power series
converges, and the range of f comprises the sums of the series ob-
tained by allowing x to take on all values in the domain of f . If a
function f is defined in this manner, we say that f is represented
by the power series f (x) = ∞ n
P
n=0 = an (x − c) .

Example : Consider the power series

∞
X
xn = 1 + x + x2 + .... + xn + ....
n=0

Notice that this is a geometric series with common ratio x, we see

that it converges for −1 ≤ x ≤ 1. Thus, the power series is a rule
for a function f with interval (−1, 1) as its domain; that is,
∞
X
f (x) = xn = 1 + x + x2 + .... + xn + ....
n=0

There is a simple formula for the sum of the geometric series,

97
namely, 1/(1 − x). and we see that the function represented by
the series is the function

1
f (x) = −1≤x≤1
1−x

Even though the domain of the function g(x) = 1/(1 − x) is the

set of all real numbers except x = 1, the power series represents
the function g(x) = 1/(1 − x) only in the interval of convergence
(−1, 1) of the series.
Theorem 35. (Convergence of a power series) Given a power se-
ries ∞ n
P
n=0 an (x − c) , exactly one of the following is true:
a. The series converges only at x = c.
b. The series converges for all x.
c. There is a number R > 0 such that the series converges for
|x − c| < R and diverges for |x − c| > R.
The number R referred in the theorem above is called the ra-
dius of convergence of the power series. The radius of conver-
gence is R = 0 in the case (a) and R = ∞ in the case (b). The
set of all values for which the power series converges is called the
interval of convergence of the power series. Thus, the theorem
tells us that the interval of convergence of a power series centered
at c is (a) just the single point c, (b) the interval (−∞, ∞), or (c)
the interval (c − R, c + R) . But in the last case, the theorem does
not tell us whether the endpoints x = c − R and x = c + R are
included in the interval of convergence. To determine whether they
are included, we simply replace x in the power series by c − R and
c + R in succession and use a convergence test on the resultant se-
ries.

98
Example : Find the radius of convergence and the interval of con-
vergence of ∞ n
P
n=0 n!x .
Solution : We can think of the given as ∞
P
n=0 un , where un =
n
n!x . Applying the Ratio test, we have

un+1 (n + 1)!xn+1
lim | | = lim | | = lim (n + 1)|x| = ∞
x→∞ un n→∞ n!xn x→∞

whenever x 6= 0, and we conclude that the series diverges when-

ever x 6= 0. Therefore, the series converges only when x = 0, and
its radius of convergence is accordingly R = 0.

Example : Find the radius of convergence and the interval of con-

vergence of
(−1)n x2n
∼∞n=0
(2n)!
Solution : Applying ratio test

(−1)n+1 x2n+2 (2n)! x2

lim | . | = lim =0<1
nto∞ (2n + 2)! (−1)n x2n n→∞ (2n + 1)(2n + 2)

for each fixed value of x, so by the Ratio Test, the given series con-
verges for all values of x. Therefore, the radius of convergence of
the series is R = ∞ , and its interval of convergence is (−∞, ∞).

Example : Find the radius of convergence and the interval of con-

vergence of
∞ xn
lim .
n=1 n

99
Solution : Let un = xn /n. Then

un+1 xn+1 n n
lim | | = lim | . | = lim ( )|x| = |x|
n→∞ un n→∞ n + 1 xn n→∞ n + 1

By the Ratio Test, the series converges if −1 < x < 1. Therefore,

the radius of convergence of the series is R = 1. To determine the
interval of convergence of the power series, we need to examine the
behavior of the series at the end-points x = −1 and x = 1. Now, if
x = −1, the series becomes
∞
X (−1)n
n=0
n

which is the convergent alternating harmonic series, and we see

that x = −1 is in the interval of convergence of the power series. If
x = 1, we obtain the harmonic series ∞
P
n=0 1/n which is divergent,
so x = 1 is not in the interval of convergence. We conclude that
the interval of convergence of the given power series is [−1, 1).

Theorem 36. (Differentiation and Integration of power series)

Suppose that the power series ∞ n
P
n=0 an (x − c) has a radius of
convergence R > 0. Then the function f defined by
∞
X
f (x) = an (x−c)n = a0 +a1 (x−c)+a2 (x−c)2 +....+an (x−c)n +....
n=0

for all x in (c − R, c + R) is both differentiable and integrable on

(c − R, c + R) . Moreover, the derivative of f and the indefinite
integral of f are
a. f 0 (x) = a1 = 2a2 (x − c) + 3a3 (x − c)2 + ... = ∞
P
n=1 nan (x −

100
c)n−1
2 3
f (x)dx = C + a0 (x − c) + a1 (x−c) + a2 (x−c)
R
b. 2 3
+ .... =
P∞ (x−c)n+1
n=0 n+1
+ C.

Notes
1. The series in parts (a) and (b) of above theorem have the same
radius of convergence, R, as the series ∞ n
P
n=0 an (x − c) . But the
interval of convergence may change. More specifically, you may
lose convergence at the endpoints when you differentiate and gain
convergence there when you integrate.
2. Above theorem implies that a function that is represented by a
power series in an interval (c − R, c + R) is continuous on that in-
terval.

Example : Find a power series representation for ln(1 − x) on

(−1, 1).
Solution : We start with the equation
∞
1 X
= 1 + x + x2 + x3 + .... = xn |x| < 1
1−x n=0

Integrating both sides of this equation with respect to x, we obtain

Z Z
1
dx = (1 + x + x2 + ...)dx
1−x

1 1
−ln(1 − x) = x + x2 = x3 + ... + C
2 3
To determine the value of C, we set x = 0 in this equation to obtain

101
−ln1 = 0 = C. Using this value of C, we see that

X xn ∞
1 1
ln(1 − x) = −x − x2 − x3 − .... = |x| < 1
2 3 n=1
n

Example : Find a power series representation for tan−1 (x) by in-

tegrating a power series representation of f (x) = 1/(1 − x2 .
Solution : Observe that we can obtain a power series representation
of f by replacing x with −x2 in the equation

1
= 1 + x + x2 + .... |x| < 1
1−x

Thus,

1 1
2
= = 1 + (−x2 ) + (−x2 )2 + (−x2 )3 ......
1+x 1 − (−x2 )
= 1 − x2 + x3 − x6 + ....
X∞
= (−1)n x2n
n=0

Since the geometric series converges for |x| < 1, we see that this
series converges for | − x2 | < 1, that is, x2 < 1 or |x| < 1. Finally,
integrating this equation, we have
Z Z
−1 1
tan (x) = 2
dx = (1 − x2 + x4 − x6 + ....)dx
1+x
x3 x5 x7
=C +x− + − + .....
3 5 7

To find C, we use the condition tan−1 (0) = 0 to obtain 0 = C.

102
Therefore,
∞
−1 x3 x5 x7 X x2n+1
tan (x) = C + x − + − + ..... = (−1)n
3 5 7 n=1
2n + 1

3.2 Taylor and Maclaurin series

In the previous section we saw that every power series represents a
function whose domain is precisely the interval of convergence of
the series. We now look at the general problem of finding power
series representations for functions, specifically on what form does
the power series representation of the function f take? (In other
words, what does an look like?)

Theorem 37. (Taylor series of f at c) If f has a power series rep-

resentation at c, that is, if
∞
X
f (x) = an (x − c)n |x − c| < R
n=0

then f (n) (c) exits for every positive integer n and

f (n) (c)
an =
n!

Thus,
∞
X f (n) (c)
f (x) = (x − c)n
n=0
n!
f 00 (c) f 000 (c)
= f (c) + f 0 (c)(x − c) + (x − c)2 + (x − c)3 + ....
2! 3!

103
 A series of this form is called the Taylor series of the function f
at c. In the special case in which c = 0, the Taylor series becomes
∞
X f (n) (0) f 00 (0) 2
f (x) = (x)n = f (0) + f 0 (0)(x) + (x) +
n=0
n! 2!
000
f (0)
(x − 0)3 + ....
3!

This series is just the Taylor series of f centered at the origin.

It is called the Maclaurin series of f .

Note : The above theorem states that if a function f has a power se-
ries representation at c, then the (unique) series must be the Taylor
series at c. The converse is not necessarily true. Given a function
f with derivatives of all orders at c, we can compute the Taylor
coefficients of f at c,

f (n) (c)
n = 0, 1, 2, ...
n!

and, therefore, the Taylor series of f at c. But the series that is

obtained formally in this fashion need not represent f .

Example : Let f (x) = ex . Find the Maclaurin series of f , and

determine its radius of convergence.
Solution : The derivatives of f (x) = ex are f 0 (x) = ex , f 00 (x) =
ex , and, in general, f (n) (x) = ex , where n ≥ 1. So

f (n) (0) = 1

104
Therefore, the Maclaurin series of f is
∞ ∞
X f (n) (0) n
X 1 n
x = x
n=0
n! n=0
n!

To determine the radius of convergence of the power series, we use

the ratio test with un = xn /n!. Since

un+1 xn+1 n! |x|

lim | | = lim | . n | = lim =0
n→∞ un n→∞ (n + 1)! x n→∞ n + 1

we conclude that the radius of convergence of the series is R = ∞.

Example : Find the Maclaurin series of f (x) = sinx, and de-

termine its interval of convergence.
Solution : To find the Maclaurin series of f (x) = sinx , we com-
pute the values of f and its derivatives at x = 0. We obtain
f (x) = sinx f (0) = 0
f 0 (x) = cosx f 0 (0) = 1
f 00 (x) = −sinx f 00 (0) = 0
f 000 (x) = −cosx f 000 (0) = −1
f (4) (x) = sinx f (4) (0) = 0
We need not go further, since it is clear that successive derivatives
of f follow this same pattern. Then, we obtain the Maclaurin series
of f (x) = sinx:

105
 ∞
X f (n) (0) f 00 (0) 2 f 000 (0) 3
= f (0) + f 0 (0)x + x + x + .....
n=0
n! 2! 3!
x3 x5 x7
=x− + − .....
3! 5! 7!
X (−1)n 2n+1
= ∞ x
n=0
(2n + 1)!

To find the interval of convergence of the series, we use the Ratio

Test

un+1 (−1)n+1 x2n+3 (2n + 1)!

lim | | = lim | . |
n→∞ un n→∞ (2n + 3)! (−1)n x2n+1
|x|2
= lim =0
n→∞ (2n + 2)(2n + 3)

we conclude that the interval of convergence of the series is (−∞, ∞).

Example : Find the Maclaurin series for f (x) = (1 + x)k , where

k is a real number.
Solution : We compute the values of f and its derivatives at x = 0,
obtaining
f (x) = (1 + x)k f (0) = 1
f 0 (x) = k(1 + x)k−1 f 0 (0) = k
f 00 (x) = k(k − 1)(1 + x)k−2 f 00 (0) = k(k − 1)
f 000 (x) = k(k − 1)(k − 1)(1 + x)k−3 f 000 (0) = k(k − 1)(k − 2)
Thus, we get
f (n) (x) = k(k − 1)....(k − n + 1)(1 + x)k−n f (n) (0) = k(k −
1)....(k − n + 1)

106
So the Maclaurin series of f (x) = (1 + x)k is
∞
X f (n) (0) f 00 (0) 2 f 000 (0) 3
xn = f (0) + f 0 (0)x + x + x + ....
n=0
n! 2! 3!
k(k − 1) 2 k(k − 1)(k − 2) 3
= 1 + kx + x + x + .....
2! 3!
∞
X k(k − 1)....(k − n + 1) n
= x
n=0
n!

Observe that if k is a positive integer, then the series is infinite (by

the Binomial Theorem), and so it converges for all x. If k is not a
positive integer, then we use the Ratio Test to find the interval of
convergence.

un+1 k(k − 1)(k − 2)....(k − n + 1)(k − n)xn+1

lim | | = lim | ×
n→∞ un n→∞ (n + 1)!
n!
|
k(k − 1)(k − 2)....(k − n + 1)
|k − n|
= lim |x|
n→∞ n + 1

| k − 1|
= lim n 1 |x|
n→∞ n +
n

= |x|

and we see that the series converges for x in the interval (−1, 1).

The series in example above is called the binomial series.

Definition 27. (Binomial series) If k is any real number and |x| <

107
1, then
∞  
k k(k − 1) 2 k(k − 1)(k − 2) 3 X k n
(1+x) = 1+kx+ x+ x +.... = x
2! 3! n=0
n

Notes :
1. The coefficients in the binomial series are referred to as binomial
coefficients and are denoted by
   
k k(k − 1)...(k − n + 1) k
= n ≥ 1, =1
n n! 0

2. If k is a positive integer and n > k, then the binomial coefficient

contains a factor (k −k), so nk = 0 for n > k. The binomial series

reduces to a polynomial of degree k:

k  
k
X k n
(1 + x) = x
n=0
n

In other words, the expression (1 + x)k can be represented by a

finite sum if k is a positive integer and by an infinite series if k is
not a positive integer. Thus, we can view the binomial series as an
extension of the Binomial Theorem to the case in which k is not a
positive integer.
3. Even though the binomial series always converges for −1 < x <
1, its convergence at the endpoints x = −1 or x = 1 depends on
the value of k. It can be shown that the series converges at x = 1 if
−1 < k < 0 and at both endpoints x = ±1 if k ≥ 0.

Techniques for Finding Taylor Series We have already seen how

108
to find Taylor series of any function using the equation described
above. But it is often easier to find the series by algebraic manip-
ulation, differentiation, or integration of some well-known series.
We now elaborate further on such techniques. First, we list some
common functions and their power series representations.

1
Example : Find the Taylor series representation of f (x) = 1+x at
x = 2.
Solution : We first rewrite f (x) so that it includes the expression
(x − 2) . Thus,

1 1 1 1 1
f (x) = = = x−2 = .
1+x 3 + (x − 2) 3[1 + ( 3 )] 3 1 + ( x−2
3
)

Then, using Formula in table above with x replaced by −(x − 2)/3,

109
we obtain
1 1
f (x) = [ ]
3 1 − (−( x−23
))
1 x−2 x−2 2 x−2 3
= [1 + (−( )) + (−( )) + (−( )) + ....
3 3 3 3
1 x−2 x−2 2 x−2 3
= [1 − ( )+( ) −( ) + ....]
3 3 3 3
1 1 1 1
= = 2 (x − 2) + 3 (x − 2)2 − 4 (x − 2)3 + ....
3 3 3 3
∞ n
X (x − 2)
= (−1)b n+1
n=0
3

The series converges for |(x − 2)/3| < 1, that is, |x − 2| < 3 or
−1 < x < 5.

Example : Find the Taylor series for f (x) = sinx at x = π/6.

Solution : We write
π π
f (x) = sinx = sin[(x − ) + ]
6 6
π π π π
= sin(x − )cos( ) + cos(x − )sin( )
√ 6 6 6 6
3 π 1 π
= sin(x − ) + cos(x − )
2 6 2 6

Then using Formulas 3 and 4 with x − (π/6) in place of x, we

obtain
√ ∞ ∞
3 X (−1)n π 1 X (−1)n π
f (x) = (x − )2n+1 + (x − )2n
2 n=0 (2n + 1)! 6 2 n=0 (2n)! 6

which converges for all x in (−∞, ∞).

110
The power series representations of certain functions can also be
found by adding, multiplying, or dividing the Maclaurin or Taylor
series of some familiar functions as the following examples show.

Example : Find the Maclaurin series representation for f (x) =

sinh(x).
Solution : We have
1 1 1
f (x) = −x
= ex − e−x
2(ex
−e ) 2 2
2 3
1 x x 1 x2 x3
= (1 + x + ) + + ....) − (1 − x + ) − + ....)
2 2! 3! 2 2! 3!
x3 x5
=x+ + + ......
3! 5!
∞
X x2n+1
=
n=0
(2n + 1)!

Since the Maclaurin series of both ex and e−x converge for x in

(−∞, ∞), we see that this representation of sinh(x) is also valid
for all values of x.

We can also use Taylor series to integrate functions whose anti

derivatives cannot be found in terms of elementary functions. Ex-
2
amples of such functions are e−x and sinx2 . In particular, the
use of Taylor series enables us to obtain approximations to definite
integrals involving such functions, as illustrated in the following
example.

2
e−x dx.
R
Example : Find

111
Example : Replacing x in Formula (2) in Table by −x2 gives
∞
−x2 x4 x6 2
X x2n
e =1−x + − + .... = (−1)n
2! 3! n=0
n!

Integrating both sides of this equation with respect to x, we obtain,

x4 x6
Z Z
−x2
e dx = (1 − x2 +
− + ....)dx
2! 3!
1 1 5 1 7
= C + x − x3 + x − x + ....
3 5.2! 7.3!
∞
X 1
=C+ (−1)n x2n+1
n=0
(2n + 1).n!

2
Since the power series representation of e−x converges for x in
(−∞, ∞), this result is valid for all values of x.

3.3 Plane curves and Parametric equations

Why we use parametric equations :

Figure below gives a bird’s-eye view of a proposed training course

for a yacht. In Figure we have introduced an xy-coordinate sys-
tem in the plane to describe the position of the yacht. With respect
to this coordinate system the position of the yacht is given by the
point P (x, y), and the course itself is the graph of the rectangular
equation 4x4 − 4x2 + y 2 = 0, which is called a lemniscate. But
representing the lemniscate in terms of a rectangular equation in
this instance has three major drawbacks.
First, the equation does not define y explicitly as a function of

112
x or x as a function of y. You can also convince yourself that this is
not the graph of a function by applying the vertical and horizontal
line tests to the curve in the figure. Because of this, we cannot
make direct use of many of the results for functions. Second, the
equation does not tell us when the yacht is at a given point (x, y).
Third, the equation gives no inkling as to the direction of motion
of the yacht. To overcome these drawbacks when we consider the
motion of an object in the plane or plane curves that are not graphs
of functions, we turn to the following representation. If (x, y) is a
point on a curve in the xy-plane, we write

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

where f and t are functions of an auxiliary variable t with (com-

mon) domain some interval I. These equations are called para-
metric equations, t is called a parameter, and the interval I is
called a parameter interval.
If we think of t on the closed interval [a, b] as representing time,
then we can interpret the parametric equations in terms of the mo-
tion of a particle as follows: At t = a the particle is at the initial
point (f (a), g(a)) of the curve or trajectory C. As t increases
from t = a to t = b, the particle traverses the curve in a specific

113
direction called the orientation of the curve, eventually ending up
at the terminal point (f (b), g(b)) of the curve. (See Figure below.)

Sketching curves defined by parametric equations

Definition 28. (Plane curve) A plane curve is a set C of ordered

pairs (x, y) defined by the parametric equations

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

where f and g are continuous functions on a parameter interval I.

Example : Sketch the curves represented by

√
a. x = t and y = t
b. x = t and y = t2
Solution : a. We eliminate the parameter t by squaring the first
equation to obtain x2 = t. Substituting this value of t into the
second equation, we obtain y = x2 , which is an equation of a
parabola. But note that the first parametric equation implies that
t ≥ 0, so x ≥ 0. Therefore, the desired curve is the right portion of
the parabola shown in Figure below. Finally, note that the param-
eter interval is [0, ∞), and as t increases from 0, the desired curve
starts at the initial point (0, 0) and moves away from it along the
parabola.

114
b. Substituting the first equation into the second yields y = x2 .
Although the rectangular equation is the same as that in part (a), the
curve described by the parametric equations here is different from
that of part (a), as we will now see. In this instance the parameter
interval is (−∞, ∞). Furthermore, as t increases from −∞ to ∞
, the curve runs along the parabola y = x2 from left to right, as
you can see by plotting the points corresponding to, say, t = −1, 0,
and 1. You can also see this by examining the parametric equation
x = t, which tells us that x increases as t increases. (See Figure
below.)

Example : Describe the curve represented by

x = 4cosθ and y = 3sinθ 0 ≤ θ ≤ 2π

Solution : Solving the first equation for cosθ and the second equa-

115
tion for sinθ gives
x
cosθ =
4
and
y
sinθ =
3
Squaring each equation and adding the resulting equations, we ob-
tain
x y
cos2 θ + sin2 θ = ( )2 + ( )2
4 3
Since cos2 θ + sin2 θ = 1, we end up with the rectangular equations

x2 y 2
+ =1
16 9

From this we see that the curve is contained in an ellipse centered

at the origin. If θ = 0, then x = 4 and y = 0, giving (4, 0) as the
initial point of the curve. As θ increases from 0 to 2π, the elliptical
curve is traced out in a counterclockwise direction, terminating at
(4, 0). (See Figure below.)

116
3.4 The calculus of Parametric equations
Tangent line to curves defined by parametric equation

Suppose that C is a smooth curve that is parametrized by the equa-

tions x = f (t) and y = g(t) with parameter interval I and we wish
to find the slope of the tangent line to the curve at the point P . (See
Figure below.) Let t0 be the point in I that corresponds to P , and
let (a, b) be the subinterval of I containing t0 corresponding to the
highlighted portion of the curve C in the figure. This subset of C
is the graph of a function of x, as you can verify using the Vertical
Line Test.

Let’s denote this function by F so that y = F (x), where f (a) <

x < f (b). Since x = f (t) and y = g(t), we may rewrite this
equation in the form
g(t) = F [f (t)]

Using the Chain Rule, we obtain

g 0 (t) = F 0 [f (t)]f 0 (t)

= F 0 (x)f 0 (t)

117
If f 0 (t) 6= 0, we can solve for F 0 (x), obtaining

g 0 (t)
F 0 (x) =
f 0 (x)

which we can write as

dy
dy dt dx
= dx
if 6= 0
dx dt
dt

Horizontal and vertical Tangents

A curve C represented by the parametric equations x = f (t) and

y = g(t) has a horizontal tangent at a point (x, y) on C where
dy/dt = 0 and dx/dt 6= 0 and a vertical tangent where dx/dt = 0
and dy/dt 6= 0, so that dy/dx is undefined there. Points where both
dy/dt and dx/dt are equal to zero are candidates for horizontal or
vertical tangents and may be investigated by using l’Hopital’s rule.

Example : A curve C is defined by the parametric equations x = t2

and y = t3 − 3t
a. Find the points on C where the tangent lines are horizontal or
vertical.
b. Find the x- and y-intercepts of C.
Solution :a. Setting dy/dt = 0 gives 3t2 − 3 = 0, or t = ±1.
Since dx/dt = 2t 6= 0 at these values of t, we conclude that C has
horizontal tangents at the points on C corresponding to t = ±1 ,
that is, at (1, −2) and (1, 2). Next, setting dx/dt = 0 gives 2t = 0,
or t = 0. Since dy/dt 6= 0 for this value of t, we conclude that
C has a vertical tangent at the point corresponding to t = 0, or at

118
(0, 0).
b. To find the x-intercepts, we set y = 0, which gives t3 − 3t =
√ √
t(t2 − 3) = 0, or t = − 3, 0, and 3. Substituting these values of
t into the expression for x gives 0 and 3 as the x-intercepts. Next,
setting x = 0 gives t = 0, which, when substituted into the expres-
sion for y, gives 0 as the y-intercept.

d2 y
Finding dx2
from parametric equations

Suppose that the parametric equations x = f (t) and y = g(t) define

y as a twice-differentiable function of x over some suitable interval.
Then d2 y/dx2 may be found from Equation of dy/dx with another
application of the Chain Rule.

d dy
d2 y d dy ( )
dt dx dx
2
= ( )= dx
if 6= 0
dx dx dx dt
dt

Higher-order derivatives are found in a similar manner.

The length of a smooth curve

Theorem 38. (Length of a smooth curve) Let C be a smooth curve

represented by the parametric equations x = f (t) and y = g(t)
with parameter interval [a, b]. If C does not intersect itself, except
possibly for t = a and t = b, then the length of C is
Z bp Z br
dx dy
L= [f 0 (t)]2 + [g 0 (t)]2 dt = ( )2 + ( )2 dt
a a dt dt

The area of a surface of revolution

119
Theorem 39. (Area of surface of revolution) Let C be a smooth
curve represented by the parametric equations x = f (t) and y =
g(t) with parameter interval [a, b], and suppose that C does not
intersect itself, except possibly for t = a and t = b. If g(t) ≥ 0 for
all t in [a, b], then the area S of the surface obtained by revolving
C about the x-axis is
Z b p Z b r
dx dy
S = 2π y [f 0 (t)]2 + [g 0 (t)]2 = 2π y ( )2 + ( )2 dt
a a dt dt

If f (t) ≥ 0 for all t in [a, b], then the area S of the surface that
is obtained by revolving C about the y-axis is
Z b Z b r
p
0 2 0 2
dx dy
S = 2π x [f (t)] + [g (t)] = 2π x ( )2 + ( )2 dt
a a dt dt

3.5 Polar coordinates

The polar coordinate system : To construct the polar coordinate
system, we fix a point O called the pole (or origin) and draw a ray
(half-line) emanating from O called the polar axis. Suppose that P
is any point in the plane, let r denote the distance from O to P , and
let θ denote the angle (in degrees or radians) between the polar axis
and the line segment OP . Then the point P is represented by the
ordered pair (r, θ), also written P (r, θ), where the numbers r and θ
are called the polar coordinates of P . The angular coordinate θ
is positive if it is measured in the counterclockwise direction from
the polar axis and negative if it is measured in the clockwise di-
rection. The radial coordinate r may assume positive as well as

120
negative values. If r > 0, then P (r, θ) is on the terminal side of θ
and at a distance r from the origin. If r < 0, then P (r, θ) lies on
the ray that is opposite the terminal side of θ and at a distance of
|r| = −r from the pole. Also, by convention the pole O is repre-
sented by the ordered pair (0, θ) for any value of θ. Finally, a plane
that is endowed with a polar coordinate system is referred to as an
rθ-plane.

Note : Unlike the representation of points in the rectangular sys-

tem, the representation of points using polar coordinates is not
unique. For example, the point (r, θ) can also be written as (r, θ +
2nπ) or (−r, θ + (2n + 1)π), where n is any integer.

Relationship between Polar and Rectangular coordinates

Suppose that a point P (other than the origin) has representation

(r, θ) in polar coordinates and (x, y) in rectangular coordinates.
Then
x = rcosθ and y = rsinθ
y
r2 = x2 + y 2 and thanθ = if x 6= 0
x

121
Example : The point (−1, 1) is given in rectangular coordinates.
Find its representation in polar coordinates.
Solution : Here, x = −1 and y = 1, Then, we have

r2 = x2 + y 2 = (−1)2 + 12 = 2

and
y
tanθ = = −1
x
√
Let’s choose r to be positive; that is, r = 2. Next, observe that
the point (−1, 1) lies in the second quadrant and so we choose
θ = 3π/4 (other choices are θ = (3π/4) ± 2nπ, where n is an inte-
√
ger). Therefore, one representation of the given point is ( 2, 3π
4
).

Graphs of polar equations

The graph of a polar equation r = f (θ) or, more generally, F (r, θ) =

0 is the set of all points (r, θ) whose coordinates satisfy the equa-
tion.

Example : Sketch the graphs of the polar equations, and reconcile

your results by finding the corresponding rectangular equations.
a. r = 2
b. θ = 2π
3
Solution : a. The graph of r = 2 consists of all points P (r, θ)
where r = 2 and θ can assume any value. Since r gives the dis-
tance between P and the pole O, we see that the graph consists of
all points that are located a distance of 2 units from the pole; in
other words, the graph of r = 2 is the circle of radius 2 centered

122
at the pole.(See Figure (a) below.) To find the corresponding rect-
angular equation, square both sides of the given equation obtaining
r2 = 4. Then, we have r2 = x2 + y 2 , and this gives the desired
equation x2 + y 2 = 4. Since this is a rectangular equation of a cir-
cle with center at the origin and radius 2, the result obtained earlier
has been confirmed.

b. The graph of θ = 2π/3 consists of all points P (r, θ) where

u = 2π/3 and r can assume any value. Since θ measures the an-
gle the line segment OP makes with the polar axis, we see that
the graph consists of all points that are located on the straight line
passing through the pole O and making an angle of 2π/3 radians
with the polar axis.(See Figure (b) above.) Observe that the half-
line in the second quadrant consists of points for which r > 0,
whereas the half-line in the fourth quadrant consists of points for
which r < 0. To find the corresponding rectangular equation, we
use the equation, tanθ = y/x, to obtain

2π y y √
tan = or =− 3
3 x x
√
or y = − 3x. This equation confirms that the graph of θ = 2π/3

123
 √
is a straight line with slope − 3.

Symmetry

Test for Symmetry

a. The graph of r = f (θ) is symmetric with respect to the polar
axis if the equation is unchanged when θ is replaced by −θ.
b. The graph of r = f (θ) is symmetric with respect to the vertical
line
theta = π/2 if the equation is unchanged when θ is replaced by
π − θ.
c. The graph of r = f (θ) is symmetric with respect to the pole if
the equation is unchanged when r is replaced by −r or when θ is
replaced by θ − π.

Tangent lines to graph of polar equations

To find the slope of the tangent line to the graph of r = f (θ) at

the point P (r, θ), let P (x, y) be the rectangular representation of
P . Then
x = rcosθ = f (θ)cosθ

124
 y = rsinθ = f (θ)sinθ

We can view these equations as parametric equations for the graph

of r = f (θ) with parameter θ. Then, we have

dy dy
dy dθ dθ
sinθ + rcosθ dx
= dx
= dr
if 6= 0
dx dθ dθ
cosθ − rsinθ dθ

and this gives the slope of the tangent line to the graph of r = f (θ)
at any point P (r, θ).

The horizontal tangent lines to the graph of r = f (θ) are lo-

cated at the points where dy/dθ = 0 and dx/dθ 6= 0. The ver-
tical tangent lines are located at the points where dx/dθ = 0 and
dy/dθ 6= 0 (so that dy/dx is undefined). Also, points where both
dy/dθ and dx/dθ are equal to zero are candidates for horizontal or
vertical tangent lines, respectively, and may be investigated using
l’Hopital’s Rule.

Example : Find the tangent lines of r = cos2θ at the origin.

Solution : Setting f (θ) = cos2θ = 0, we find that

π 3π 5π 7π
2θ = , , , or
2 2 2 2

or
π 3π 5π 7π
θ= , , , or
4 4 4 4
Next, we compute f (θ) = −2sin2θ. Since f 0 (θ) 6= 0 for each of
0

these values of θ, we see that θ = π/4 and θ = 3π/4 (that is, y = x

and y = −x) are tangent lines to the graph of r = cos2θ at the
pole.

125
3.6 Areas and Arc lengths in Polar coordi-
nates
Areas in Polar coordinates

Theorem 40. (Area bounded by a Polar curve) Let f be a contin-

uous, nonnegative function on [α, β] where 0 ≤ β − α < 2π. Then
the area A of the region bounded by the graphs of r = f (θ), θ = α,
and θ = β is given by
Z β Z β
1 1 2
A= [f (θ)]2 dθ = r dθ
α 2 α 2

Note : When you determine the limits of integration, keep in mind

that the region R is swept out in a counterclockwise direction by
the ray emanating from the origin, starting at the angle α and ter-
minating at the angle β.

Example : Find the area of the region enclosed by the cardioid

r = 1 + cosθ.
Solution :

The graph of the cardioid r = 1+cosθ is shown in the figure above.

Observe that the ray emanating from the origin sweeps out the re-

126
quired region exactly once as θ increases from 0 to 2π. Therefore,
the required area A is
Z 2π Z 2π
1 2 1
A= r dθ = (1 + cosθ)2 dθ
0 2 0 2
1 2π
Z
= (1 + 2cosθ + cos2 θ)dθ
2 0
1 2π
Z
1 + cos2θ
= (1 + 2cosθ + )dθ
2 0 2
1 2π 3
Z
1
= ( + 2cosθ + cos2θ)dθ
2 0 2 2
1 3 1 3
= [ + 2sinθ + sin2θ]2π0 = π
2 2 4 2

Area bounded by two graphs

Theorem 41. (Area bounded by two polar curves) Let f and g be

continuous on [α, β], where 0 ≤ g(θ) ≤ f (θ) and 0 ≤ β − α < 2π.
Then the area A of the region bounded by the graphs of r = g(θ) ,
r = f (θ) , θ = α, and θ = β is given by
Z β
1
A= ([f (θ)]2 − [g(θ)]2 )dθ
2 α

Example : Find the area of the region that lies outside the circle
r = 3 and inside the cardioid r = 2 + 2cosθ.
Solution :

127
We first sketch the circle r = 3 and the cardioid r = 2 + 2cosθ.
The required region is shown shaded in the figure above. To find the
points of intersection of the two curves, we solve the two equations
simultaneously. We have 2 + 2cosθ = 3 or cosθ = 1/2 , which
gives θ = ±π/3. Since the region of interest is swept out by the
ray emanating from the origin as θ varies from −π/3 to π/3, we
see that the required area is,
Z β
1
A= ([f (θ)]2 − [g(θ)]2 )dθ
2 α

where f (θ) = 2 + 2cosθ = 2(1 + cosθ), g(θ) = 3, α = −π/3, and

128
β = π/3. If we take advantage of symmetry, we can write

1 π/3
Z
A = 2( ([2(1 + cosθ)]2 − 32 )dθ
2 0
Z π/3
= (4 + 8cosθ + 4cos2 θ − 9)dθ
0
Z π/3
1 + cos2θ
= (−5 + 8cosθ + 4. )dθ
0 2
Z π/3
= (−3 + 8cosθ + 2cos2θ)dθ
0
π/3
= [−3θ + 8sinθ + sin2θ]0
√ √
3 3
= (−π + 8( )+ )
√ 2 2
9 3
= −π
2

Arc length in polar coordinates

Theorem 42. (Arc length) Let f be a function with a continuous

derivative on an interval [α, β]. If the graph C of r = f (θ) is traced
exactly once as θ increases from α to β, then the length L of C is
given by
Z β Z β
r
p dr 2
L= [f 0 (θ)]2 + [f (θ)]2 dθ = ( ) + r2 dθ
α α dθ

Area of surface of revolution

Theorem 43. (Area of Surface of revolution)Let f be a function

with a continuous derivative on an interval [α, β]. If the graph C
of r = f (θ) is traced exactly once as θ increases from α to β, then
the area of the surface obtained by revolving C about the indicated

129
line is given by q
Rβ dr 2
a.S = 2π α rsinθ ( dθ ) + r2 dθ (about the polar axis)
Rβ q
dr 2
b.S = 2π α rcosθ ( dθ ) + r2 dθ (about the lie θ = π/2)

Points of intersection of graphs in polar coordinates

In a previous example we were able to find the points of intersection

of two curves with representations in polar coordinates by solving
a system of two equations simultaneously. This is not always the
case. Consider for example, the graphs of the cardioid r = 1+cosθ
and the circle r = 3cosθ shown in figure below.

Solving the two equations simultaneously, we obtain

3cosθ = 1 + cosθ

1
cosθ =
2
or θ = π/3 and 5π/3. Therefore, the point of intersection are
(3/2, π/3) and (3/2, 5π/3). But one glance at the figure shows
the pole as a third point of intersection that is not revealed in our

130
calculation. To see how this can happen, think of the cardioid as
being traced by the point (r, θ) satisfying

r = f (θ) = 1 + cosθ 0 ≤ θ ≤ 2π

with θ as a parameter. If we think of θ as representing time, then as

θ runs from θ = 0 through θ = 2π, the point (r, θ) starts at (2, 0)
and traverses the cardioid in a counter-clockwise direction, eventu-
ally returning to the point (2, 0). Similarly, the circle is traced twice
in the counterclockwise direction, by the point (r, θ), where

r = g(θ) = 3cosθ 0 ≤ θ ≤ 2π

and the parameter θ, once again representing time, runs from θ = 0

through θ = 2π.
Observe that the point tracing the cardioid arrives at the point
(3/2, π/3) on the cardioid at precisely the same time the point trac-
ing the circle arrives at the point (3/2, π/3) on the circle. A sim-
ilar observation holds at the point (3/2, 5π/3) on each of the two
curves. These are the points of intersection found earlier.
Next, observe that the point tracing the cardioid arrives at the
origin when θ = π. But the point tracing the circle first arrives at
the origin when θ = π/2 and then again when u = 3π/2. In other
words, these two points arrive at the origin at different times, so
there is no (common) value of θ corresponding to the origin that
satisfies both equations simultaneously. Thus, although the origin
is a point of intersection of the two curves, this fact will not show
up in the solution of the system of equations. For this reason it is
recommended that we sketch the graphs of polar equations when

131
finding their points of intersection.

Example : Find the points of intersection of r = cosθ and r =

cos2θ.
Solution : We solve the system of equations

r = cosθ

r = cos2θ

We set cosθ = cos2θ and use the identity cos2θ = 2cos2θ − 1. We

obtain
2cos2 θ − cosθ = 0

(2cosθ + 1)(cosθ − 1) = 0

So
1
cosθ = − or cosθ = 1
2
that is,
2π 4π
θ= , , or 0
3 3

132
These values of θ give (−1/2, 2π/3), (−1/2, 4π/3), and (1, 0) as
the points of interaction. Since both graphs also pass through the
pole, we conclude that the pole is also a point of intersection.

133
Module 4

Geometry of Space and

Vector-valued function

4.1 Equations of Lines in Space

Definition 29. (Parametric equation of line) The parametric equa-
tions of the line passing through the point P0 (x0 , y0 , z0 ) and paral-
lel to the vector v = (a, b, c) are
x = x0 + at, y = y0 + bt, and z = z0 + ct

Example : Find parametric equations for the line passing through

the point P0 (−2, 1, 3) and parallel to the vector v = h1, 2, −2i.
Solution : Using the above equation with x0 = −2, y0 = 1, z0 = 3,
a = 1, b = 2, and c = −2 , we obtain
x = −2 + t, y = 1 = 2t, and z = 3 − 2t

Suppose that the vector v = ha, b, ci defines the direction of a line

L. Then the numbers a, b, and c are called the direction numbers

134
of L. Observe that if a line L is described by a set of parametric
equations, then the direction numbers of L are precisely the coeffi-
cients of t in each of the parametric equations.
There is another way of describing a line in space.

Definition 30. (Symmetric equations of a Line) The symmetric

equations of the line L passing through the point P0 (x0 , y0 , z0 ) and
parallel to the vector v = ha, b, ci are

x − x0 y − y0 z − z0
= =
a b c

Note: Suppose a = 0 and both b and c are not equal to zero, then
the parametric equations of the line take the form
x = x0 , y = y0 + bt, and z = z0 + ct
and the line lies in the plane x = x0 (parallel to the yz-plane).
Solving the second and third equations for t leads to

y − y0 z − z0
x = x0 , =
b c

which are the symmetric equations of the line.

Example : a. Find parametric equations and symmetric equa-

tions for the line L passing through the points P (−3, 3, −2) and
Q(2, −1, 4).
b. At what point does L intersect the xy-plane?
Solution : a. The direction of L is the same as that of the vector
P~Q = h5, −4, 6i. Since L passes through P (−3, 3, −2) , we can
use a = 5, b = −4, c = 6, x0 = −3, y0 = 3, and z0 = −2, to
obtain the parametric equations

135
x = −3 + 5t, y = 3 − 4t, and z = −2 + 6t
Thus, we obtain the following symmetric equations for L:

x+3 y−3 z+2

= =
5 −4 6

b. At the point where the line intersects the xy-plane, we have

z = 0. So setting z = 0 in the third parametric equation, we ob-
tain t = 1/3. Substituting this value of t into the other parametric
equations gives the required point as (−4/3, 5/3.0).

Equations of Planes in space:

A plane in space is uniquely determined by specifying a point P0 (x0 , y0 , z0 )

lying in the plane and a vector n =ha, b, ci that is normal (perpen-
dicular) to it. (See the figure below)

Definition 31. (The standard form of the equations of a plane)

The standard form of the equation of a plane containing the point
P0 (x0 , y0 , z0 ) and having the normal vector n = ha, b, ci is

a(x − x0 ) + b(y − y0 ) + c(z − z0 ) = 0

By expanding the equation above and regrouping the terms, we

136
obtain the general form of the equation of a plane in space,

ax + by + cz = d

where d = ax0 + by0 + cz0 . Conversely, given ax + by + cz = d

with a, b, and c not all equal to zero, we can choose numbers x0 , y0 ,
and z0 such that ax0 + by0 + cz0 = d.
An equation of the form ax + by + cz = d, with a, b, and c not
all zero, is called linear equation in the three variables x, y, and
z.

Theorem 44. Every plane in space can be represented by a linear

equation ax + by + cz = d, where a, b, and c are not all equal to
zero. Conversely, every linear equation ax+by +cz = d represents
a plane in space having a normal vector ha, b, ci.

Note: Notice that the coefficients of x, y, and z are precisely the

components of the normal vector n = ha, b, ci. Thus, we can write
a normal vector to a plane by simply inspecting its equation.

Example : Find an equation of the plane containing the points

P (3, −1, 1), Q(1, 4, 2), and R(0, 1, 4).
Solution : We need to find a vector normal to the plane in ques-
tion. Observe that both of the vectors P~Q = h−2, 5, 1i and P~R =
h−3, 2, 3i lie in the plane, so the vector P~Q × P~R is normal to the
plane. Denoting this vector by n, we have

i j k
n = P~Q × P~R = −2 5 1 = 13i + 3j + 11k
−3 2 3

137
Finally, using the point P (3, −1, 1) in the plane (any of the other
two points will also do) and the normal vector n just found, with
a = 13, b = 3, c = 11, x0 = 3, y0 = −1, and z0 = 1, gives

13(x − 3) + 3(y + 1) + 11(z − 1) = 0

or
13x + 3y + 11z = 47

Parallel and orthogonal planes

Two planes with normal vectors m and n are parallel to each other
if m and n are parallel; the planes are orthogonal if m and n are
orthogonal.

Example : Find an equation of the plane containing P (2, −1, 3)

and parallel to the plane defined by 2x − 3y + 4z = 6.
Solution : The normal vector of the given plane is n = h2, −3, 4i.
Since the required plane is parallel to the given plane, it also has n
as a normal vector. Therefore, we obtain

2(x − 2) − 3(y + 1) + 4(z − 3) = 0

or
2x − 3y + 4z = 19

as an equation of the plane.

138
The angle between two plane

Two distinct planes in space are either parallel to each other or in-
tersect in a straight line. If they do intersect, then the angle between
the two planes is defined to be the acute angle between their normal
vectors.

Example : Find the angle between the two planes defined by 3x −

y + 2z = 1 and 2x + 3y − z = 4.
Solution : The normal vectors of these planes are
n1 = h3, −1, 2i and n2 = h2, 3, −1i
Therefore, the angle θ between the planes is given by
n1 .n2
cosθ =
|n1 ||n2 |
h3, −1, 2i.h2, 3, −1i
=√ √
9+1+4 4+9+1
3(2) + (−1)(3) + 2(−1) 1
= √ √ =
14 14 14

or
1
θ = cos−1 ( ) ≈ 80◦
14

The distance between a point and a plane

Suppose that P1 (x1 , y1 , z1 ) is a point not lying in the plane ax +

by + cz = d. Let P0 (x0 , y0 , z0 ) be any point lying in the plane.
Then, the distance D between P1 and the plane is given by the
length of the vector projection of P0~P1 onto the normal vector n
=ha, b, ci of the plane. Equivalently, D is the absolute value of the

139
scalar component of P0~P1 along n. Thus, we obtain

|P0~P1 .n|
D=
|n|

But P0~P1 = hx1 − x0 , y1 − y0 , z1 − z0 i. Substituting into above

equation and simplifying, we get

|ax1 + by1 + cz1 − d|

D= √
a2 + b 2 + c 2

4.2 Surfaces in space

In the previous section we saw that the graph of a linear equation
in three variables is a plane in space. In general, the graph of an
equation in three variables, F (x, y, z) = 0, is a surface in 3-space.
In this section we will study surfaces called cylinders and quadric
surfaces.

Traces

The trace of a surface S in a plane is the intersection of the sur-

face and the plane. In particular, the traces of S in the xy-plane,

140
the yz-plane, and the xz-plane are called the xy-trace, the yz-trace
and the xz-trace, respectively. To find the xy-traces, we set z = 0
and sketch the graph of the resulting equation in the xy-plane. The
other traces are obtained in a similar manner. Of course, if the sur-
face does not intersect the plane, there is no trace in that plane.

Cylinders

Definition 32. (Cylinder) Let C be a curve in a plane, and let l be

a line that is not parallel to that plane. Then the set of all points
generated by letting a line traverse C while parallel to l at all times
is called a cylinder. The curve C is called the directrix of the cylin-
der, and each line through C parallel to l is called a ruling of the
cylinder.

Example : Sketch the graph of y = x2 − 4.

Solution : The given equation has the form f (x, y) = 0, where
f (x, y) = x2 − y − 4. Therefore, its graph is a cylinder with direc-
trix given by the graph of y = x2 − 4 in the xy-plane and rulings
parallel to the z-axis (corresponding to the variable missing in the
equation). The graph of y = x2 − 4 in the xy-plane is the parabola
shown in Figure (a) below. The required cylinder is shown in Fig-
ure (b). It is called a parabolic cylinder.

141
Quadric surfaces

The graph of the second degree equation

Axx + By 2 + Cz 2 + Dxy + Exz + F yz + Gx + Hy + Iz + J = 0

where A, B, C, ..., J are constants, is called a quadric surface.

Ellipsoid : The graph of the equation

x2 y 2 z 2
+ 2 + 2 =1
a2 b c

is an ellipsoid because its traces in the planes parallel to the coordi-

nate planes are ellipses. In fact, its trace in the plane z = k, where
−c < k < c, is the ellipse

x2 y 2 k2
+ = 1 −
a2 b2 c2

and, in particular, its trace in the xy-plane is the ellipse

x2 y 2
+ 2 =1
a2 b

142
Similarly, its traces in the planes x = k (−a < k < a) and y =
k(−b < k < b) are ellipses and, in particular, that its yz- and xz-
traces are the ellipses
y2 z2
+ 2 =1
b2 c
and
x2 z 2
+ 2 =1
a2 c
respectively.

Hyperboloid of One sheet : The graph of the equation

x2 y 2 z 2
+ 2 − 2 =1
a2 b c

is a hyperboloid of one sheet. The xy-trace of this surface is the

ellipse
x2 y 2
+ 2 =1
a2 b
whereas both the yz- and xz-traces are hyperbolas. The trace of the
surface in the plane z = k is an ellipse

x2 y 2 k2
+ = 1 +
a2 b2 c2

143
 As |k| increases, the ellipses grow larger and large. The z−axis
is called the axis of the hyperboloid. Note that the orientation of the
axis of the hyperboloid is associated with the term that has a minus
sign in front of it. Thus, if the minus sign had been in front of the
term involving x, then the surface would have been a hyperboloid
of one sheet with the x-axis as its axis.

Hyperboloid of Two sheets : The graph of the equation

x2 y 2 z 2
− − 2 + 2 =1
a2 b c

is a hyperboloid of two sheets. The xz- and yz-traces are the hy-
perbolas
x2 z 2
− 2 + 2 =1
a c
and
y2 z2
− 2 + 2 =1
b c
The trace of the surface in the plane z = k is an ellipse

x2 y 2 k2
+ = −1
a2 b2 c2
provided that |k| > c. There are no values of x and y that satisfy

144
the equation if |k| < c, so the surface is made up of two parts: one
part lying on or above the plane z = c and the other part lying on or
below the plane z = −c. The axis of the hyperboloid is the z-axis.
Observe that the sign associated with the variable z is positive. Had
the positive sign been in front of one of the other variables, then the
surface would have been a hyperboloid of two sheets with its axis
along the axis associated with that variable.

Cones : The graph of the equation

x2 y 2 z 2
+ 2 − 2 =0
a2 b c

is a double-napped cone. The xz- and yz-traces are the lines z =

±(c/a)x and z = ±(c/b)y , respectively. The trace in the plane
z = k is an ellipse,
x2 y 2 k2
+ =
a2 b2 c2
As |k| increases, so do the lengths of the axes of the resulting el-
lipses. The traces in planes parallel to the other two coordinate
planes are hyperbolas.

145
Paraboloids : The graph of the equation

x2 y 2
+ 2 = cz
a2 b

where c is a real number, is called an elliptic paraboloid because its

traces in planes parallel to the xy-coordinate plane are ellipses and
its traces in planes parallel to the other two coordinate planes are
parabolas. If a = b, the surface is called a circular paraboloid. The
graph of an elliptic paraboloid with c > 0 is sketched in Figure (a)
below. The axis of the paraboloid is the z-axis, and its vertex is the
origin.

146
Hyperbolic paraboloid : The graph of the equation

x2 y 2
− 2 = cz
a2 b

where c is a real number, is called a hyperbolic paraboloid because

the xz- and yz-traces are parabolas and the traces in planes par-
allel to the xy-plane are hyperbolas. The graph of a hyperbolic
paraboloid with c < 0 is shown in Figure (b) above.

4.3 Cylindrical and spherical coordinates

147
The cylindrical coordinate system is just an extension of the po-
lar coordinate system in the plane to a three-dimensional system
in space obtained by adding the (perpendicular) z-axis to the sys-
tem. A point P in this system is represented by the ordered triple
(r, θ, z), where r and θ are the polar coordinates of the projection
of P onto the xy-plane and z is the directed distance from (r, θ, 0)
to P .
The relationship between rectangular coordinates and cylindri-
cal coordinates can be seen by examining Figure. If P has rep-
resentation (x, y, z) in terms of rectangular coordinates, then we
have the following equations for converting cylindrical coordinates
to rectangular coordinates and vice versa.
Converting cylindrical to rectangular coordinates:

x = rcosθ y = rsinθ z = z

Converting rectangular to cylindrical coordinates:

y
r2 = x2 + y 2 tanθ = z=z
x

Example : The point (3, π/4, 3) is expressed in cylindrical co-

ordinates. Find its rectangular coordinates.
Solution : We are given that r = 3, θ = π/4, and z = 3. Using the
above equations, we have
√
π 3 2
x = rcosθ = 3cos =
4 2
√
π 3 2
x = rsinθ = 3sin =
4 2

148
 z=3
√ √
Therefore, the rectangular coordinates of the given point are ( 3 2 2 , 3 2 2 , 3).

The spherical coordinate system

In the spherical coordinate system a point P is represented by an

ordered triple (ρ, θ, φ), where ρ is the distance between P and the
origin, θ is the same angle as the one used in the cylindrical coordi-
nate system, and φ is the angle between the positive z-axis and the
line segment OP . Note that the spherical coordinates satisfy ρ ≥ 0
, 0 ≤ θ < 2π, and 0 ≤ φ ≤ π.
The relationship between rectangular coordinates and spherical
coordinates can be seen by examining the figure above. If P has
representation (x, y, z)in terms of rectangular coordinates, then
x = rcosθ and y = rsinθ
Since r = ρsinφ and z = ρcosφ we have the following equations
for converting spherical coordinates to rectangular coordinates and
vice versa.
Converting spherical to rectangular coordinates

x = ρsinφcosθ y = ρsinφsinθ z = ρcosφ

149
Converting rectangular to spherical coordinates

y z
ρ2 = x2 + y 2 + z 2 tanθ = cosφ =
x ρ

Example : Find an equation in spherical coordinates for the paraboloid

with rectangular equation 4z = x2 + y 2 .
Solution : Using the above equations, we obtain

4ρcosφ = ρ2 sin2 φcos2 θ + ρ2 sin2 φsin2 θ

= ρ2 sin2 φ(cos2 θ + sin2 θ)
= ρ2 sin2 φ

or
ρsin2 φ = 4cosφ

4.4 Vector-valued functions and space curves

Definition 33. (Vector function) A vector-valued function, or vec-
tor function, is a function r defined by

r(t) = f (t)i + g(t)j + h(t)k

where the component functions f, g, and h of r are real-valued func-

tions of the parameter t lying in a parameter interval I.

Example : Find the domain (parameter interval) of the vector func-

tion
1 √
r(t) = h , t − 1, lnti
t
√
The component functions of r are f (t) = 1/t, g(t) = t − 1,

150
and h(t) = lnt. Observe that f is defined for all values of t except
t = 0, g is defined for all t ≥ 1, and h is defined for all t > 0.
Therefore; f, g, and h are all defined if t ≥ 1, and we conclude that
the domain of r is [1, ∞).

Curves defined by vector functions

A plane or space curve is the curve traced out by the terminal point
of r(t) of a vector function r as t takes on all values in a parameter
interval.

Example : Sketch the curve defined by the vector function

r(t) = h3cost, −2sinti 0 ≤ t ≤ 2π

Solution : The parametric equations for the curve are

x = 3cost and y = −2sint Solving the first equation for cost and
the second equation for sint and using the identity cos2 t + sin2 t =
1, we obtain the rectangular equation

x2 y 2
+ =1
9 4

The curve described by this equation is the ellipse shown in Figure

below. As t increases from 0 to 2π, the terminal point of r traces
the ellipse in a clockwise direction.

151
Example : Sketch the curve defined by the vector function

r(t) = 2costi + 2sintj + tk 0 ≤ t ≤ 2π

Solution : The parametric equations for the curve are

x = 2cost y = 2sint z = t

From the first two equations we obtain

x y
( )2 + ( )2 = cos2 t + sin2 t = 1 or x2 + y 2 = 4
2 2

This says that the curve lies on the right circular cylinder of ra-
dius 2, whose axis is the z-axis. At t = 0 , r(0) = 2i, and this
gives (2, 0, 0) as the starting point of the curve. Since z = t,
the z-coordinate of the point on the curve increases (linearly) as
t increases, and the curve spirals upward around the cylinder in
a counterclockwise direction, terminating at the point (2, 0, 2π)
[r(2π) = 2i + 2πk]. The curve, called a helix.

152
Limits and continuity
Definition 34. (The limit of a Vector function) Let r be a function
defined by r(t) = f (t)i + g(t)j + h(t)k. Then

lim r(t) = [lim f (t)]i + [lim g(t)]j + [lim h(t)]k

t→a t→a t→a t→a

provided that the limits of the component functions exist.

√
Example : Find limt→0 r(t), where r(t) = t + 2i + tcos2tj +
e−t k.
Solution :
√
lim = [lim t + 2]i + [lim icos2t]j + [lim e−t ]k
t→0 t→0 t→0 t→0
√
= 2i + k

Definition 35. (Continuity of a vector function) A vector function

r is continuous at a if

lim r(t) = r(a)

t→a

153
A vector function r is continuous on an interval I if it is continuous
at every number in I.

Example : Find the interval(s) on which the vector function r de-

fined by
√ 1
r(t) = ti + ( 2 )j + lntk
t −1
is continuous.
√
Solution : The component functions of r are f (t) = t, g(t) =
1/(t2 − 1) , and h(t) = lnt. Observe that f is continuous for
t ≥ 0, g is continuous for all values of t except t = ±1, and h is
continuous for t > 0. Therefore, r is continuous on the intervals
(0, 1) and (1, ∞).

4.5 Differentiation and integration of vector-

valued functions
The derivative of a vector function

Definition 36. (Derivative of a vector function) The derivative of

a vector function r is the vector function r’ defined by

dr r(t + h) − r(t)
r0 (t) = = lim
dt h→0 h

provided that the limit exists.

The derivative r’ of the vector r at P may be interpreted as the

tangent vector to the curve defined by r at the point P , provided
that r’(t) 6= 0. If we divide r’(t) by its length, we obtain the unit

154
tangent vector
r0 (t)
T(t) =
|r0 (t)|
which has unit length and the direction of r’.

Theorem 45. (Differentiation of Vector functions) Let r(t) = f (t)i+

g(t)j + h(t)k, where f, g, and h are differentiable functions of t.
Then
r0 (t) = f 0 (t)i + g 0 (t)j + h0 (t)k

Proof. We compute
r0 (t)

r(t + δt) − r(t)

= lim
∆t→0 ∆t
f (t + ∆t)i + g(t + ∆t)j + h(t + ∆t)k
= lim [
∆t→0 ∆t
[f (t)i + g(t)j + h(t)k]
− ]
∆t
f (t + ∆t) − f (t) g(t + ∆t) − g(t) h(t + ∆t) − h(t)
= lim [ i+ j+ k]
∆t→0 ∆t ∆t ∆t
f (t + ∆t) − f (t) g(t + ∆t) − g(t)
= [ lim ]i + [ lim ]j
∆t→0 ∆t ∆t→0 ∆t
h(t + ∆t) − h(t)
+ [ lim ]k
∆t→0 ∆t
= f 0 (t)i + g 0 (t)j + h0 (t)k

Example : a. Find the derivative of r(t) = (t2 +1)i+e−t j−sin2tk.

b. Find the point of tangency and the unit tangent vector at the point
on the curve corresponding to t = 0.

155
Solution : Using the theorem above, we get

r0 (t) = 2ti − e−t j − 2cos2tk

b. Since r(0) = i + j, we see that the point of tangency is (1, 1, 0).

Next, since r0 (0) = −j − 2k, we find the unit tangent vector at
(1, 1, 0) to be

r0 (0) −j − 2k 1 2
T(0) = 0
= √ = −√ j − √ k
|r (0)| 1+4 5 5

Example : Find parametric equations for the tangent line to the

helix with parametric equations

x = 3cost y = 2sint z = t

at the point where t = π/6.

Solution : The vector function that describes the helix is

r(t) = 3costi + 2sintj + tk

The tangent vector at any point on the helix is

r0 (t) = −3sinti + 2costj + k

√
In particular, the tangent vector at the point ( 3 2 3 , 1, π6 ), where t =
π/6, is
π 3 √
r0 ( ) = − i + 3j + k
6 2
Finally, we the observe that the required tangent line passes
√
through the point ( 3 2 3 , 1, π6 ) and has the same direction that the

156
as required the tangent vector line r0 (π/6). Thus, the parametric
equations of this line are
√
3 3 3 √ π
x= − t, y = 1 + 3t, and z = + t
2 2 6

Higher order derivatives

Higher-order derivatives of vector functions are obtained by suc-

cessive differentiation of the lower-order derivatives of the func-
tion. For example, the second derivative of r(t) is

d 0
r00 (t) = r (t) = f 00 (t)i + g 00 (t)j + h00 (t)k
dt

Rules of differentiation

Theorem 46. (Rules of differentiation) Suppose thatu and v are

differentiable vector functions, f is a differentiable real-valued func-
tion, and c is a scalar. Then

Example : Suppose that v is a differentiable vector function of

constant length c. Show that v.v0 = 0. In other words, the vector v

157
and its tangent vector v’ must be orthogonal.
Solution : The condition on v implies that

v.v = |v|2 = c2

Differentiating both sides of this equation with respect to t, we ob-

tain
d d
(v.v) = v.v0 + v0 .v = (c2 ) = 0
dt dt
But v0 .v = v.v0 , so we have

2v.v0 = 0 or v.v0 = 0

Integration of vector functions

Theorem 47. (Integration of vector functions) Let r(t) = f (t)i +

g(t)j + h(t)k, where f, g and h are integrable. Then
1. The indefinite integral of r with respect to t is
Z Z Z Z
r(t)dt = [ f (t)dt]i + [ g(t)dt]j + [ h(t)dt]k

2. The definite integral of r over the interval [a, b] is

Z b Z b Z b Z b
r(t)dt = [ f (t)dt]i + [ g(t)dt]j + [ h(t)dt]k
a a a a

√
Example : Find the antiderivative of r0 (t) = costi + e−t j + tk
satisfying the initial condition r(0) = i + 2j + 3k.

158
Solution : We have
Z
r(t) = r0 (t)dt
Z
= (costi + e−t j + t1/2 k)dt
2
= sinti − e−t j + t3/2 k + C
3

where C is a constant (vector) of integration. To determine C, we

use the condition r(0) = i + 2j + 3k to obtain

r(0) = C = i + 2j + 3k

Therefore,

2
r(t) = (1 + sint)i + (3 − e−t )j + (3 + t3/2 )k
3

4.6 Arc length and curvature

Arc length

We saw that the length of the plane curve given by the paramet-
ric equations x = f (t) and y = g(t), where a ≤ t ≤ b, is
Z br Z bp
dx 2 dy 2
L= ( ) + ( ) dt = [f 0 (t)]2 + [g 0 (t)]2 dt
a dt dt a

159
Now, suppose that C is described by the vector function r(t) =
f (t)i + g(t)j instead. Then

r0 (t) = f 0 (t)i + g 0 (t)j

and
p
|r0 (t)| = [f 0 (t)]2 + [g 0 (t)]2

from which we see that L can also be written in the form

Z b|r0 |
L= (t)dt
a

Theorem 48. (Arc length of a space curve) Let C be a curve given

by the vector function

r(t) = f (t)i + g(t)j + h(t)k a ≤ t ≤ b

where f 0 , g 0 , and h0 are continuous. If C is traversed exactly once

as t increases from a to b, then its length is given by
Z bp Z b
L= [f 0 (t)]2 + [g 0 (t)]2 + [h0 (t)]2 dt = |r0 (t)|dt
a a

Example : Find the length of the arc of the helix C given by the
vector function r(t) = 2costi + 2sintj + tk, where 0 ≤ t ≤ 2π.
Solution : We first compute

r0 (t) = −2sinti + 2costj + k

160
Then, the length of the arc is
Z 2π
0
Z 2π √
L= |r (t)|dt = 4sin2 t + 4cos2 t + 1dt
0 0
Z 2π √ √
= 5dt = 2 5π
0

Smooth curve

A curve that is defined by a vector function r on a parameter in-

terval I is said to be smooth if r0 (t) is continuous and r0 (t) 6= 0 for
all t in I with the possible exception of the endpoints. For example,
the plane curve defined by r(t) = t3 i + t2 j is smooth everywhere
except at the point (0, 0) corresponding to t = 0. To see this, we
compute r0 (t) = 3t2 i + 2tj and note that r0 (0) = 0. The point (0, 0)
where the curve has a sharp corner is called a cusp.

Arc length parameter

The curve C described by the vector function r(t) with parame-

ter t in some parameter interval I is said to be parametrized by t. A
curve C can have more than one parametrization. For example, the
helix represented by the vector function

r1 (t) = 2costi + 3sintj + tk 2π ≤ t ≤ 4π

with parameter t is also represented by the function

r2 (t) = 2coseu i + 3sineu j + eu k ln2π ≤ u ≤ ln4π

161
with parameter u, where t and u are related by t = eu .
A useful parametrization of a curve C is obtained by using the
arc length of C as its parameter. To see how this is done, we need
the following definition.
Definition 37. (Arc length function) Suppose that C is a smooth
curve described by r(t) = f (t)i + g(t)j + h(t)k , where a ≤ t ≤ b.
Then the arc length function s is defined by
Z t
s(t) = |r0 (u)|du
a

Differentiating s(t) with respect to t and using the Fundamental

theorem of calculus, we obtain

ds
= |r0 (t)|
dt

or, in differential form,

ds = |r0 (t)|dt

The following example shows how to parametrize a curve in terms

of its arc length.

Example : Find the arc length function s(t) for the circle C in
the plane described by

r(t) = 2costi + 2sintj 0 ≤ t ≤ 2π

Then use the result to find a parametrization of C in terms of s.

Solution : We first compute r0 (t) = −2sinti + 2costj, and then

162
compute
√
|r0 (t)| = 4sin2 t + 4cos2 t = 2

Then, Z t
s(t) = 2du = 2t 0 ≤ t ≤ 2π
0

Writing s for s(t), we have s = 2t, where 0 ≤ t ≤ 2π, which when

solved for t, yields t = t(s) = s/2. Substituting this value of t into
the equation for r(t) gives

s s
r(t(s)) = 2cos( )i + 2sin( )j
2 2

Finally, since s(0) = 0 and s(2π) = 4π, we see that the parameter
interval for this parametrization by the arc length s is [0, 4π].

Note : One reason for using the arc length of a curve C as the
parameter stems from the fact that its tangent vector r0 (s) has unit
length; that is, r0 (s) is a unit tangent vector. Consider the circle of
example above. Here,

s s
r0 (s) = −sin( )i + cos( )j
2 2

so r
0 s s
|r (s)| = sin2 ( ) + cos2 ( ) = 1
2 2

Curvature
Definition 38. (Curvature) Let C be a smooth curve defined by
r(s) , where s is the arc length of the parameter. Then the curvature

163
of C at s is
dT
κ(s) = | | = |T0 (s)|
ds
where T is the unit tangent vector.

Although the use of the arc length parameter s provides us with

a natural way for defining the curvature of a curve, it is generally
easier to find the curvature in terms of the parameter t. Applying
chain rule dT
dt
= dT ds
ds dt
Then

dT | dT
dt
|
κ(s) = | | = ds
ds | dt |

Since ds/dt = |r0 (t)|, we are led to the following formula:

|T0 (t)|
κ(t) =
|r0 (t)|

Theorem 49. (Formula for finding curvature) Let C be a smooth

curve given by the vector function r. Then the curvature of C at
any point on C corresponding to t is given by

|r0 (t) × r00 (t)|

κ(t) =
|r0 (t)|3

Proof. We begin by recalling that

r0 (t)
T(t) = 0
|r (t)|

Since |r0 (t)| = ds/dt, we have

ds
r0 (t) = T(t)
dt

164
Differentiating both sides of this equation with respect to t

d2 s ds 0
r00 (t) = T(t) + T (t)
dt2 dt

Next, we use the fact that T × T = 0 to obtain

ds 2
r0 (t) × r00 (t) = ( ) (T(t) × T0 (t))
dt

Also, |T(t)| = 1 for all t implies that T(t) and T0 (t) are orthogo-
nal.( as seen in the example in the previous section). Therefore, we
have

ds 2 ds ds
|r0 (t)×r00 (t)| = ( ) |T(t)×T0 (t)| = ( )2 |T(t)||T0 (t)| = ( )2 |T0 (t)|
dt dt dt

Upon solving for |T0 (t)|, we obtain

|r0 (t) × r00 (t)| |r0 (t) × r00 (t)|

|T0 (t)| = =
( ds
dt
)2 |r0 (t)|2

from which we deduce that

|T0 (t)| |r0 (t) × r00 (t)|

κ(t) = =
r0 (t) |r0 (t)|3

Theorem 50. (Formula for the curvature of the graph of a func-

tion) If C is the graph of a twice differentiable function f , then the
curvature at the point (x, y) where y = f (x) is given by

|f 00 (x)| |y 00 |
κ(x) = =
[1 + [f 0 (x)]2 ]3/2 [1 + (y 0 )2 ]3/2

165
Proof. Using x as the parameter, we can represent C by the vector
function r(x) = xi + f (x)j + 0k. Differentiating r(x) with respect
to x successively, we obtain

r0 (x) = i + f 0 (x)j + 0k and r00 (x) = 0i + f 00 (x)j + 0k

from which we obtain

i j k
r (x) × r (x) = 1 f (x) 0 = f 00 (x)k
0 00 0

0 f 00 (x) 0

and
|r0 (t) × r00 (x)| = |f 00 (x)|

Also,
p
|r0 (x)| = 1 + [f 0 (x)]2

Therefore,

|r0 (x) × r00 (x)| |f 00 (x)|

κ(x) = =
|r0 (x)|3 [1 + [f 0 (x)]2 ]3/2

Radius of curvature

Suppose that C is a plane curve with curvature κ at the point P .

Then the reciprocal of the curvature, ρ = 1/κ, is called the radius
of curvature of C at P . The radius of curvature at any point P
on a curve C is the radius of the circle that best “fits” the curve at
that point. This circle, which lies on the concave side of the curve

166
and shares a common tangent line with the curve at P , is called the
circle of curvature or osculating circle.
The center of the circle is called the center of curvature. As
an example, the curvature of the parabola y = 1/4x2 at the point
(0, 0) is found to be 1/2. Therefore, the radius of curvature of the
parabola at (0, 0) is ρ = 2. The circle of curvature is shown in the
figure below. Its equation is x2 + (y − 2)2 = 4.

4.7 Velocity and acceleration

Velocity, acceleration, and speed
Definition 39. (Velocity, acceleration, and speed) Let r(t) = f (t)i+
g(t)j + h(t)k be the position vector of an object. If f, g, and h are
twice differentiable functions of t, then the velocity vector v(t), ac-
celeration vector a(t), and speed |v(t)| of the object at time t are
defined by

v(t) = r0 (t) = f 0 (t)i + g 0 (t)j + h0 (t)k

a(t) = r00 (t) = f 00 (t)i + g 00 (t)j + h00 (t)k

167
 p
|v(t)| = |r0 (t)| = [f 0 (t)]2 + [g 0 (t)]2 + [h0 (t)]2

Example : The position of an object moving in a plane is given by

r(t) = t2 i + tj t ≥ 0

Find its velocity, acceleration, and speed when t = 2.

Solution : The velocity and acceleration vectors of the object are

v(t) = r0 (t) = 2ti + j

and
a(t) = r00 (t) = 2i

Therefore, its velocity, acceleration, and speed when t = 2 are

v(2) = 4i + j

a(2) = 2i

and
√ √
|v(2)| = 16 + 1 = 17.

Motion of a particle

A projectile of mass m is fired from a height h with an initial ve-

locity v0 and an angle of elevation α. If we describe the position
of the projectile at any time t by the position vector r(t), then its

168
initial position may be described by the vector

r(0) = hj

and its initial velocity by the vector

v(0) = v0 = (v0 cosα)i + (v0 sinα)j v0 = |v0 |

If we assume that air resistance is negligible and that the only ex-
ternal force acting on the projectile is due to gravity, then the force
acting on the projectile during its flight is

F = −mgj

where g is the acceleration due to gravity. By Newton’s Second

Law of Motion this force is equal to ma, where a is the acceleration
of the projectile. Therefore,

ma = −mgj

169
giving the acceleration of the projectile as

a(t) = −gj

To find the velocity of the projectile at any time t, we integrate the

last equation with respect to t to obtain
Z
v(t) = −gjdt = −gtj + C

Setting t = 0 and using the initial condition v(0) = v0 , we obtain

v(0) = C = v0

Therefore, the velocity of the projectile at any time t is

v(t) = −gtj + v0

Integrating this equation then gives

Z
1
r(t) = (−gtj + v0 )dt = gt2 j + v0 t + D
2

Setting t = 0 and using the initial condition r(0) = hj , we obtain

r(0) = D = hj

Therefore, the position of the projectile at any time t is

1
r(t) = − gt2 j + v0 t + hj
2

170
or
1
r(t) = − gt2 j + [(v0 cosα)i + (v0 sinα)j]t + hj
2
1
= (v0 cosα)ti + [h + (v0 sinα)t − gt2 ]j
2

4.8 Tangential and normal components of

acceleration

The unit normal

Suppose that C is a smooth space curve described by the vector

function r(t). Then,

r0 (t)
T(t) = r0 (t) 6= 0
|r0 (t)|

is the unit tangent vector to the curve C at the point correspond-

ing to t. Since |T(t)| = 1 for every t, then the vector T0 (t) is
orthogonal to T(t)( as seen in an earlier example). Therefore, if r0
is also smooth, we can normalize T0 (t) to obtain a unit vector that
is orthogonal to T(t). This vector

T0 (t)
N(t) =
|T0 (t)|

is called the principal unit normal vector (or simply the unit nor-
mal) to the curve C at the point corresponding to t.

171
Example : Let C be the helix defined by

r(t) = 2costi + 2sintj + tk t≥0

Find T(t) and N(t).

Solution : Since

r0 (t) = −2sinti + 2costj + k

and
√ √
|r0 (t)| = 4sin2 t + 4cos2 t + 1 = 5

we have
1
T(t) = √ (−2sinti + 2costj + k)
5
Now, differentiating T

2
T0 (t) = − √ (costi + sintj)
5

and
2
|T0 (t)| = √
5
it follows that
N(t) = −(costi + sintj)

Tangential and normal components of acceleration

Let’s return to the study of the motion of an object moving along

the curve C described by the vector function r defined on the pa-
rameter interval I. Recall that the speed v of the object at any time

172
t is v = |v(t)| = |r0 (t)|. But

r0 (t)
T=
|r0 (t)|

so we can write

v(t) = r0 (t) = |r0 (t)|T = vT

which expresses the velocity of the object in terms of its speed and
direction.

The acceleration of the object at time t is

d
a = v0 = (vT) = v 0 T + vT0
dt

To obtain an expression for T0 , recall that

T0
N=
|T0 |

so T0 = |T0 |N . Now we need an expression for |T0 |. But

|T0 |
κ=
|r0 |

where κ is the curvature of C. This gives

|T0 | = κ|r0 | = κv

so T0 = |T0 |N = κvN.
Therefore,
a = v 0 T + κv 2 N)

173
 This result shows that the acceleration vector a can be resolved
into the sum of two vectors—one along the tangential direction and
the other along the normal direction. The magnitude of the acceler-
ation along the tangential direction is called the tangential scalar
component of acceleration and is denoted by aT , whereas the
magnitude of the acceleration along the normal direction is called
the normal scalar component of acceleration and is denoted by
aN . Thus,
a = aT T + aN N

where aT = v 0 and aN = κv 2 .
The following theorem gives formulas for calculating aT and
aN directly from r and its derivatives.

Theorem 51. (Tangential and normal components of accelera-

tion) Let r(t) be the position vector of an object moving along a
smooth curve C. Then

a = aT T + aN N

where
r0 (t).r00 (t)
aT =
|r0 (t)|
and
r0 (t) × r00 (t)
aN =
|r0 (t)|

Proof. If we take the dot product of v and a, we obtain

v.a = (vT).(v 0 T + κv 2 N)
= vv 0 T.T + κv 3 T.N

174
But T.T = |T|2 = 1, since T is a vector, and T.N = 0, since T
and N are orthogonal. Therefore,

v.a = vv 0

or
v.a r0 (t).r00 (t)
aT = v 0 = =
v |r0 (t)|
Now,

2 |r0 (t) × r00 (t)| 0 2 |r0 (t) × r00 (t)|

aN = κv = |r (t)| =
|r0 (t)|3 |r0 (t)|

Example : A particle moves along a curve described by the vector

function r(t) = ti+t2 j+t3 k. Find the tangential scalar and normal
scalar components of acceleration of the particle at any time t.
Solution : We begin by computing

r0 (t) = i + 2tj + 3t2 k

r00 (t) = 2j + 6tk

Then,
r0 (t).r00 (t) 4t + 18t3
aT = = √
|r0 (t)| 1 + 4t2 + 9t4
Next, we compute

i j k
0 00
r (t) × r (t) = 1 2t 3t2 = 6t2 i − 6tj + 2k
0 2 6t

175
Then, we have
√ r
r0 (t) × r00 (t) 36t4 + 36t2 + 4 9t4 + 9t2 + 1
aT = = √ = 2
|r0 (t)| 1 + 4t2 + 9t4 9t4 + 4t2 + 1

176