Chapter 4: Definite Integrals

National High School of Mathematics

CONTENTS Definite Integrals

Contents

1 Definite Integrals 180

1.1 The Riemann Integral . . . . . . . . . . . . . . . . . . . . . . . . . . . . 180
1.1.1 Step functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 180
1.1.2 Types of Riemann sums . . . . . . . . . . . . . . . . . . . . . . . 181
1.2 Properties of the Riemann Integral . . . . . . . . . . . . . . . . . . . . . 191
1.3 The Fundamental Theorems of Calculus . . . . . . . . . . . . . . . . . . 197
1.3.1 The Substitution Theorem and Integration by Parts . . . . . . . . 202

I
1 DEFINITE INTEGRALS Definite Integrals

1. Definite Integrals

1.1 The Riemann Integral

Now, we give the definition of the Riemann integral of a function.

Definition 9.1
Let [a, b] be a closed interval.

1. The set P = {x0 , x1 , x2 , . . . , xn } of finite points a = x0 , x1 , x2 , . . . , xn = b with

a = x0 < x1 < x2 < · · · < xn = b

is called a partition of [a, b];

2. For every i = 1, 2, . . . , n, a closed interval Ii = [xi−1 , xi ] is called the subinter-

val of [a, b];

3. For every i = 1, 2, . . . , n, the length of Ii is defined by ∆xi = xi − xi−1 ,

4. ℘[a, b] is denoted by the set of all partitions of [a, b],

5. The width of the largest sub-interval in a partition is called the norm of the
partition.

1.1.1 Step functions

Definition 9.2
A function f : [a, b] → R is called a step function if it is piecewise constant, i.e. if
there are numbers a = x0 < x1 < x2 < · · · < xN −1 < xN = b such that f is constant
on each half open interval [xi−1 , xi ) with 1 ≤ i ≤ N . For a step function we define
Rb PN
the integral to be a f (x)dx = i=1 f (xi−1 )(xi − xi−1 ). The collection of numbers
x0 , x1 , x2 , . . . , xN are called a partition for the step function f .
In this definition we have to take into account that a step function f might be
defined using different partitions, and show that the integral does not depend on
which partition is used to compute it.

180
1 DEFINITE INTEGRALS Definite Integrals

Lemma 9.1
If f and g are step functions on an interval [a, b] with f (x) ≤ g(x) for all x ∈ [a, b],
Rb Rb
then a f (x)dx ≤ a g(x)dx.

Definition 9.3
Let f : [a, b] → R be a bounded function and P = {a = x0 , x1 , x2 , . . . , xn = b} be
a partition of [a, b]. For every subinterval Ii = [xi−1 , xi ]. We define Riemann sum
Pn
S(f, P ) = i=1 f (x⋆i )∆xi , where ∆xi = xi − xi−1 and x⋆i ∈ [xi−1 , xi ].

Remark 9.1
One might produce different Riemann sums depending on which x⋆i ’s are chosen. In
the end this will not matter, if the function is Riemann integrable, when the difference
or width of the summands ∆xi approaches zero.

1.1.2 Types of Riemann sums

Specific choices of x⋆i give different types of Riemann sums:

1. If x⋆i = xi−1 for all i, the method is the left rule and gives a left Riemann sum.

2. If x⋆i = xi for all i, the method is the right rule and gives Riemann sum.

3. If x⋆i = (xi + xi−1 )/2 for all i, the method is the midpoint rule and gives a middle
Riemann sum.

4. If f (x⋆i ) = sup f ([xi−1 , xi ]) (that is, the supremum of f over [xi−1 , xi ]), the method
is the upper rule and gives an upper Riemann sum or upper Darboux sum.

5. If f (x⋆i ) = inf f ([xi−1 , xi ]) (that is, the infimum of f over [xi−1 , xi ]), the method is
the lower rule and gives a lower Riemann sum or lower Darboux sum.

Definition 9.4
Let f : [a, b] → R be a bounded function and P = {a = x0 , x1 , x2 , . . . , xn = b} be a
partition of [a, b]. For every subinterval Ii = [xi−1 , xi ], put

Mi = sup{f (x) : x ∈ [xi−1 , xi ]}, mi = inf{f (x) : x ∈ [xi−1 , xi ]}

Then, for every P ∈ ℘[a, b], define the upper Riemann sum U (f, P ) and the lower

181
1 DEFINITE INTEGRALS Definite Integrals

Riemann sum L(f, P ) as follows, respectively:

n
X n
X
U (f, P ) = Mi ∆xi , L(f, P ) = mi ∆xi
i=1 i=1

Example 9.1
Let f : [a, b] → R be a function defined by f (x) = c, where c ∈ R is a constant.
Prove that, for every P = {a = x0 , x1 , x2 , . . . , xn = b} ∈ ℘[a, b]

U (f, P ) = L(f, P ) = c(b − a)

Solution. Let ∆xi = xi − xi−1 for every i = 1, 2, . . . , n. Since Mi = mi = c, we have

the following:
n
X n
X n
X
U (f, P ) = Mi ∆xi = c∆xi = c ∆xi = c(b − a)
i=1 i=1 i=1

and
n
X n
X n
X
L(f, P ) = mi ∆xi = c∆xi = c ∆xi = c(b − a).
i=1 i=1 i=1

Thus, we have
U (f, P ) = L(f, P ) = c(b − a)

Example 9.2
Let f : [0, 1] → R be a function defined by

 1, if x ∈ Q


f (x) =
/ Qc

 0, if x ∈


Then, calculate U (f, P ) and L(f, P ).

Solution. Let P = {0 = x0 , x1 , x2 , . . . , xn = 1} ∈ ℘[0, 1]. Each subinterval Ii =
[xi−1 , xi ] for every i = 1, 2, . . . , n has infinitely many rational numbers and irrational
numbers, and so, for every i = 1, 2, . . . , n,

Mi = sup{f (x) : x ∈ [xi−1 , xi ]} = 1

and
mi = inf{f (x) : x ∈ [xi−1 , xi ]} = 0.

Hence, we have L(f, P ) = 0 and U (f, P ) = 1.

182
1 DEFINITE INTEGRALS Definite Integrals

Lemma 9.2
Let f : [a, b] → R be a bounded function and P = {a = x0 , x1 , x2 , . . . , xn = b}
be a partition of [a, b]. Then, we have the following: (1) L(f, P ) ≤ U (f, P ); (2)
m(b − a) ≤ L(f, P ) ≤ U (f, P ) ≤ M (b − a), where m, M ∈ R are numbers such that

m ≤ f (x) ≤ M for every x ∈ [a, b]

Definition 9.5
Let [a, b] be a closed interval.

1. For every P, Q ∈ ℘[a, b], Q is called the refinement of P if

P ⊂Q

2. The partition R = P ∪ Q is called the common refinement of P and Q.

Lemma 9.3
Let f : [a, b] → R be a bounded function. For every P, Q ∈ ℘[a, b], if Q is a refinement
of P , then we have the following: (1) L(f, P ) ≤ L(f, Q)
(2) U (f, Q) ≤ U (f, P ).

• Proof
(1) First, let
P = {a = x0 , x1 , x2 , . . . , xn = b}

and
Q = {a = x0 , x1 , x2 , . . . , xj−1 , x⋆ , xj , . . . , xn = b}

that is, Q = P ∪ {x⋆ }. Then, Q is a refinement of P , that is, P ⊂ Q. Next, put

m′j = inf{f (x) : x ∈ [xj−1 , x⋆ ]}, m′′j = inf{f (x) : x ∈ [x⋆ , xj ]}

and
mj = min{m′j , m′′j }

Then, we have
n
X n
X
L(f, P ) = mi ∆xi = mi ∆xi + mj ∆xj
i=1 i=1,i̸=j

183
1 DEFINITE INTEGRALS Definite Integrals

and
j−1 n
mi ∆xi + m′j (x⋆ − xj−1 ) + m′′j (xj − x⋆ ) +
X X
L(f, Q) = mi ∆xi .
i=1 i=j+1

Thus, since mj ≤ m′j , m′′j , we have

mj ∆xj = mj (x⋆ − xj−1 ) + mj (xj − x⋆ ) ≤ m′j (x⋆ − xj−1 ) + m′′j (xj − x⋆ )

and so
n n n n n
mj (x⋆ −xj−1 )+ mj (xj −x⋆ ) ≤ m′j (x⋆ −xj−1 )+ m′′j (xj −x⋆ ),
X X X X X
mj ∆xj =
i=1,i̸=j i=1,i̸=j i=1,i̸=j i=1,i̸=j i=1,i̸=j

which implies that L(f, P ) ≤ L(f, Q). Next, let Q be the refinement adjoining a finite
points to P , and then, repeating the proceeding process, we can prove L(f, P ) ≤
L(f, Q). (2) By the same argument as in (1), we can prove U (f, Q) ≤ U (f, P ). This
completes the proof. □

Theorem 9.4
Let f : [a, b] → R be a bounded function. For every P, Q ∈ ℘[a, b], (Q isn’t a
refinement of P ), we have L(f, P ) ≤ U (f, Q).

• Proof
Let T = P ∪ Q. Then, T is a refinement of P and Q. Thus, by Lemma (9.3), we
have
L(f, P ) ≤ L(f, T ), U (f, T ) ≤ U (f, Q).

Therefore, since L(f, P ) ≤ U (f, P ), we have

L(f, P ) ≤ U (f, Q).

This completes the proof. □

Corollary 9.1
Let f : [a, b] → R be a bounded function. For every P, Q ∈ ℘[a, b], if L(f, P ) ≤
U (f, Q), then

sup{L(f, P ) : P ∈ ℘[a, b]}, inf{U (f, P ) : P ∈ ℘[a, b]}

exist.

184
1 DEFINITE INTEGRALS Definite Integrals

• Proof
Let Q ∈ ℘[a, b] be fixed. Then, for every P ∈ ℘[a, b], we have

L(f, P ) ≤ U (f, Q), L(f, Q) ≤ U (f, P ),

which means that U (f, Q) and L(f, Q) are upper bound and lower bound of the
following sets:

{L(f, P ) : P ∈ ℘[a, b]}, {U (f, P ) : P ∈ ℘[a, b]},

respectively. Therefore, by the Completeness Axiom of R, it follows that

sup{L(f, P ) : P ∈ ℘[a, b]}, inf{U (f, P ) : P ∈ ℘[a, b]}

exist. This completes the proof. □

Corollary 9.2
Let f, g : [a, b] → R be bounded functions and c ∈ R. For every P ∈ ℘[a, b], we have
the following:

1. U (f + g, P ) ≤ U (f, P ) + U (g, P )

2. L(f + g, P ) ≥ L(f, P ) + L(g, P );

3. If c ≥ 0, then we have

U (cf, P ) = cU (f, P ), L(cf, P ) = cL(f, P )

4. If c < 0, then we have

U (cf, P ) = cL(f, P ), L(cf, P ) = cU (f, P ).

• Proof
Let P = {a = x0 , x1 , x2 , . . . , xn = b} ∈ ℘[a, b]. (1) We have
n
X
U (f + g, P ) = sup{f (x) + g(x) : x ∈ [xi−1 , xi ]}∆xi
i=1
Xn
≤ [sup{f (x) : x ∈ [xi−1 , xi ]} + sup{g(x) : x ∈ [xi−1 , xi ]}]∆xi
i=1

= U (f, P ) + U (g, P )

185
1 DEFINITE INTEGRALS Definite Integrals

(2) As in the proof of (1), we have L(f + g, P ) ≥ L(f, P ) + L(g, P ). (3) Let c ≥ 0.
Then, we have
n
X n
X
U (cf, P ) = sup{cf (x) : x ∈ [xi−1 , xi ]}∆xi = c sup{f (x) : x ∈ [xi−1 , xi ]}∆xi = cU (f, P )
i=1 i=1

(4) Let c < 0. Then, we have

n
X
U (cf, P ) = sup{cf (x) : x ∈ [xi−1 , xi ]}∆xi
i=1
n
X
=c inf{f (x) : x ∈ [xi−1 , xi ]}∆xi
i=1

= cL(f, P ).

This completes the proof. □

Definition 9.6
Let f : [a, b] → R be a bounded function.

1. The upper Riemann integral of f on [a, b] is the number

Z b
f (x)dx = inf{U (f, P ) : P ∈ ℘[a, b]}
a

2. The lower Riemann integral of f on [a, b] is the number

Z b
f (x)dx = sup{L(f, P ) : P ∈ ℘[a, b]}
a

3. f is said to be Riemann integrable on [a, b] if

Z b Z b
f (x)dx = f (x)dx
a a

4. In this case, the Riemann integral of f is defined to be the number

Z b Z b Z b
f (x)dx = f (x)dx = f (x)dx
a a a
Rb
which is denoted by a f (x)dx.

5. In addition, we define
Z a Z b Z a
f (x)dx = − f (x)dx, f (x)dx = 0
b a a

6. f is said to be not Riemann integrable on [a, b] if

Z b Z b
f (x)dx ̸= f (x)dx
a a

186
1 DEFINITE INTEGRALS Definite Integrals

Remark 9.2
Note that ℜ[a, b] denotes the set of Riemann integrable functions on [a, b]. From
the definition of the Riemann integral, it follows that, for every P, Q ∈ ℘[a, b] with
P ⊂Q
Z b
L(f, P ) ≤ L(f, Q) ≤ f (x)dx ≤ U (f, Q) ≤ U (f, P )
a

Example 9.3
Let f : [a, b] → R be the function defined by

 1, if x ∈ Q


f (x) =
/ Qc

 0, if x ∈


Then, we have L(f, P ) = 0 and U (f, P ) = 1 and so

Z b
f (x)dx = inf{U (f, P ) : P ∈ ℘[a, b]} = 1
a

and
Z b
f (x)dx = sup{L(f, P ) : P ∈ ℘[a, b]} = 0.
a
Rb Rb
Therefore, since a f (x)dx = 1 ̸= 0 = a f (x)dx, f is not Riemann integrable on [0, 1].

Theorem 9.5 (The Riemann Integral Test)

Let f : [a, b] → R be a bounded function. Then, the following are equivalent: (1)
f ∈ ℜ[a, b]
(2) For every ε > 0, there exists P ∈ ℘[a, b] such that

U (f, P ) − L(f, P ) < ε

• Proof
(1) =⇒ (2) Suppose that f ∈ ℜ[a, b]. Since
Z b
f (x)dx = inf{U (f, P ) : P ∈ ℘[a, b]}
a

Rb ε
for every ε > 0, a f (x)dx + 2
is not a lower bound of {U (f, P ) : P ∈ ℘[a, b]}, and so
there exists P1 ∈ ℘[a, b] such that
Z b
ε
U (f, P1 ) < f (x)dx +
a 2

187
1 DEFINITE INTEGRALS Definite Integrals

Also, since
Z b
f (x)dx = sup{L(f, P ) : P ∈ ℘[a, b]}
a
Rb ε
a f (x)dx − 2
is not an upper bound of {L(f, P ) : P ∈ ℘[a, b]}, and so there exists
P2 ∈ ℘[a, b] such that
Z b
ε
f (x)dx − < L(f, P2 )
a 2
Letting P = P1 ∪ P2 , we have

L(f, P2 ) ≤ L(f, P ) ≤ U (f, P ) ≤ U (f, P1 )

Thus, we have

U (f, P ) − L(f, P ) ≤ U (f, P1 ) − L(f, P2 )

Z b ! Z b !
ε ε
< f (x)dx + − f (x)dx − = ε.
a 2 a 2
(2) =⇒ (1) Suppose that, for every ε > 0, there exists P ∈ ℘[a, b] such that

U (f, P ) − L(f, P ) < ε

Rb Rb
Since a f (x)dx ≤ U (f, P ) and L(f, P ) ≤ a f (x)dx, we have
Z b Z b
0≤ f (x)dx − f (x)dx ≤ U (f, P ) − L(f, P ) < ε
a a

and so
Z b Z b Z b Z b
f (x)dx − f (x)dx = f (x)dx − f (x)dx < ε.
a a a a

Therefore, we have
Z b Z b
f (x)dx = f (x)dx
a a

and f ∈ ℜ[a, b]. This completes the proof □

Example 9.4
Let f : [a, b] → R be the function defined by f (x) = x for every x ∈ [a, b]. By using
the definition of the Riemann integral, show that
Z b
1
f (x)dx = (b2 − a2 )
a 2
Solution. For every n ≥ 1, let Pn be a partition of [a, b] given by
( )
b−a 2(b − a) n(b − a)
Pn = a, a + ,a + ,...,a + =b .
n n n

188
1 DEFINITE INTEGRALS Definite Integrals

h i
(i−1)(b−a) i(b−a)
Since f (x) = x is increasing on [a, b], at each subinterval a + n
,a + n
,f
(i−1)(b−a) i(b−a)
has the minimum value mi = a + n
and the maximum value Mi = a + n
.
Therefore, we have
n n
! !
(i − 1)(b − a) b−a (b − a)2 1
X X  
L(f, Pn ) = mi ∆xi = a+ = a(b−a)+ 1− .
i=1 i=1 n n 2 n

Similarly, we have
n n
! !
i(b − a) b−a (b − a)2 1
X X  
U (f, Pn ) = Mi ∆xi = a+ = a(b − a) + 1+
i=1 i=1 n n 2 n

which implies that f (x) = x is Riemann integrable on [a, b] and

Z b
1
f (x)dx = (b2 − a2 )
a 2

Corollary 9.3
Let f : [a, b] → R be a bounded function. Then, the following are equivalent: (1)
f ∈ ℜ[a, b];
(2) There exists a sequence {Pn } of partitions of [a, b]

lim (U (f, Pn ) − L(f, Pn )) = 0.

n→∞

In this case, we have

Z b
lim L(f, Pn ) = f (x)dx = lim U (f, Pn ).
n→∞ a n→∞

Example 9.5
Let f : [0, 1] → R be the function defined by f (x) = x for every x ∈ [0, 1]. By using
the Riemann Integral Test, prove that f (x) = x is Riemann integrable on [0, 1].
Solution. Let Pn ∈ ℘[0, 1] as in Example 3.4 be a partition of [a, b]. Then, we have
n−1 n+1
L(f, Pn ) = , U (f, Pn ) = .
2n 2n
For every ε > 0, if we take a positive integer n0 such that
1
< n0
ε
then we have
1
U (f, Pn0 ) − L(f, Pn0 ) = < ε.
n0

189
1 DEFINITE INTEGRALS Definite Integrals

Therefore, by Theorem (the Riemann Integral Test), f is Riemann integrable on

[0, 1].

Theorem 9.6
Every monotone function f : [a, b] → R is Riemann integrable.

• Proof
Assume that f is increasing on [a, b]. Since f (a) ≤ f (x) ≤ f (b) for every x ∈ [a, b],
f is clearly bounded on [a, b]. Let ε > 0, and select a positive integer n such that

(f (b) − f (a))(b − a)
< ε.
n

For a partition
P = {a = x0 < x1 < · · · < xn−1 < xn = b},
b−a
where ∆xi = xi − xi−1 = n
for every i ≥ 1, we have
n n
X X b−a
U (f, P ) = Mi ∆xi = f (xi )
i=1 i=1 n

and
n n
X X b−a
L(f, P ) = mi ∆xi = f (xi−1 ) .
i=1 i=1 n
Thus, since f is increasing on [a, b], we have
n
b−aX b−a
U (f, P ) − L(f, P ) = |f (xi ) − f (xi−1 )| = (f (b) − f (a)) < ε.
n i=1 n

Therefore, by the Riemann Integral Test, f is Riemann integrable on [a, b]. Similarly,
we can prove this theorem when f is decreasing on [a, b]. This completes the proof.
□

Theorem 9.7
Let f : [a, b] → R be a continuous function. Then, f ∈ ℜ[a, b], that is, f is Riemann
integrable on [a, b].

• Proof
Since f is a continuous function on [a, b], f is uniformly continuous on [a, b]. Thus,
for every ε > 0, there exists δ > 0 such that

ε
|x − y| < δ, x, y ∈ [a, b] =⇒ |f (x) − f (y)| <
b−a

190
1 DEFINITE INTEGRALS Definite Integrals

Now, select a sufficiently large positive integer n, and consider a partition P =

{x0 , x1 , x2 , . . . , xn } of [a, b] with
b−a
∆x1 = ∆x2 = · · · = ∆xn0 =
n
Further, for each subinterval Ii = [xi−1 , xi ] of the partition Pn (1 ≤ i ≤ n), there exist
ti , ui ∈ Ii such that

Mi = sup{f (x) : x ∈ [xi−1 , xi ]} = f (ti )

and
mi = inf{f (x) : x ∈ [xi−1 , xi ]} = f (ui )

Thus, we have
n
X
0 ≤ U (f, Pn ) − L(f, Pn ) = (Mi − mi )∆xi
i=1
Xn
= (f (ti ) − f (ui ))∆xi
i=1
n0
!
X b−a
= |f (ti ) − f (ui )|
i=1 n
ε
< (b − a) = ε
b−a
Therefore, by the Riemann Integral Test, f is Riemann integrable on [a, b]. This
completes the proof. □

1.2 Properties of the Riemann Integral

In this section, we give the Riemann integrals of the composition and the product of
two Riemann integrable functions and some basic properties of the Riemann integral
including some algebraic properties of the Riemann integral.
Theorem 9.8
Let f : [a, b] → R be Riemann integrable on [a, b] and g : [c, d] → R be a continuous
function on [c, d] with f ([a, b]) ⊂ [c, d]. Then, the composition g ◦ f is Riemann
integrable on [a, b].

• Proof
Let ε > 0. Since g is continuous on [c, d], let

M = sup{|g(x)| : x ∈ [c, d]}

191
1 DEFINITE INTEGRALS Definite Integrals

and let
ε
ε′ = .
2M + (b − a)
Also, g is uniformly continuous on [c, d], and there exists δ > 0 such that 0 < δ < ε′
and
y, z ∈ [c, d], |y − z| ≤ δ =⇒ |g(y) − g(z)| < ε′ .

On the other hand, since f is Riemann integrable on [a, b], there exists a partition
P = {a = x0 , x1 , x2 , . . . , xn = b} of [a, b] such that

U (f, P ) − L(f, P ) < δ 2 .

Now, we show that, for this partition P ,

U (g ◦ f, P ) − L(g ◦ f, P ) ≤ ε

and then g ◦ f is Riemann integrable on [a, b]. For every i = 1, 2, . . . , n, put

Mi = sup{f (x) : x ∈ [xi−1 , xi ]},

mi = inf{f (x) : x ∈ [xi−1 , xi ]},

Mi′ = sup{(g ◦ f )(x) : x ∈ [xi−1 , xi ]},

m′i = inf{(g ◦ f )(x) : x ∈ [xi−1 , xi ]}.

Now, to be convenient, we separate the indices of the partition P into two disjoint
subsets X and Y as follows:

X = {k : Mi − mi < δ}, Y = {i : Mi − mi ≥ δ}.

Now, if i ∈ X, that is, Mi − mi < δ, then, for every x, x′ ∈ [xi−1 , xi ], we have

|f (x) − f (x′ )| < δ

and so
|(g ◦ f )(x) − (g ◦ f )(x′ )| < ε′

which implies that

Mi′ − m′i ≤ ε′

Thus, we have
(Mi′ − m′i )(xi − xi−1 ) ≤ ε′ (b − a).
X
(1)
i∈X

192
1 DEFINITE INTEGRALS Definite Integrals

But, if i ∈ Y , then we have δ ≤ Mi − mi and so

X 1X
(xi − xi−1 ) ≤ (Mi − mi )(xi − xi−1 )
i∈Y δ i∈Y
1
≤ (U (f, P ) − L(f, P ))
δ
< δ < ε′

Therefore, we have
(Mi′ − m′i )(xi − xi−1 ) ≤ 2M ε′ .
X
(2)
i∈Y

If we combine (1) and (2) then we have

(Mi′ − m′i )(xi − xi−1 ) + (Mi′ − m′i )(xi − xi−1 )

X X
U (g ◦ f, P ) − L(g ◦ f, P ) =
i∈X i∈Y

≤ ε′ (b − a) + 2M ε′ = ε.

Therefore, by the Riemann Integral Test, g ◦ f is Riemann integrable on [a, b]. This
completes the proof. □

Theorem 9.9
Let f, g : [a, b] → R be Riemann integrable on [a, b], that is, f, g ∈ ℜ[a, b], and let
c ∈ R. Then, we have the following: (1) f + g ∈ ℜ[a, b] and
Z b Z b Z b
(f + g)(x)dx = f (x)dx + g(x)dx
a a a

(2) cf ∈ ℜ[a, b] and

Z b Z b
cf (x)dx = c f (x)dx
a a

(3) f g ∈ ℜ[a, b] (4) If f (x) ≤ g(x) for every x ∈ [a, b], then
Z b Z b
f (x)dx ≤ g(x)dx
a a

Rb Rb
(5) If |f | ∈ ℜ[a, b] then a f (x)dx ≤ a |f (x)|dx

• Proof
(1) Since f, g ∈ ℜ[a, b], for every ε > 0, there exist P1 , P2 ∈ ℘[a, b] such that

ε ε
U (f, P1 ) − L(f, P1 ) < , U (g, P2 ) − L(g, P2 ) <
2 2

193
1 DEFINITE INTEGRALS Definite Integrals

Put P = P1 ∪ P2 . Then, we have P ∈ ℘[a, b] and

U (f + g, P ) − L(f + g, P ) ≤ U (f, P ) + U (g, P ) − L(f, P ) − L(g, P )

= U (f, P ) − L(f, P ) + U (g, P ) − L(g, P )

≤ U (f, P1 ) − L(f, P1 ) + U (g, P2 ) − L(g, P2 )

ε ε
< + = ε.
2 2
Therefore, by the Riemann Integral Test, we have f + g ∈ ℜ[a, b]. On the other hand,
it follows that, for every P ∈ ℘[a, b],
Z b
L(f, P ) + L(g, P ) ≤ L(f + g, P ) ≤ (f + g)(x)dx
a (3)
≤ U (f + g, P ) ≤ U (f, P ) + U (g, P ).

Observe that
ε Zb ε
U (f, P ) ≤ U (f, P1 ) < L(f, P1 ) + ≤ f (x)dx +
2 a 2
and
ε Zb ε
U (g, P ) ≤ U (g, P2 ) < L(g, P2 ) + ≤ g(x)dx +
2 a 2
Thus, by (3), we have
Z b Z b Z b
(f + g)(x)dx ≤ U (f + g, P ) ≤ U (f, P ) + U (g, P ) < f (x)dx + g(x)dx + ε
a a a

Since ε > 0 is arbitrary, we have

Z b Z b Z b
(f + g)(x)dx ≤ f (x)dx + g(x)dx (4)
a a a

Similarly, we have
Z b
ε ε Zb Z b
(f +g)(x)dx ≥ L(f +g, P ) ≥ L(f, P )+L(g, P ) ≥ U (f, P )− +U (g, P )− > f (x)dx+ g(x)dx−ε
a 2 2 a a

Which leads to
Z b Z b Z b
(f + g)(x)dx ≥ f (x)dx + g(x)dx (5)
a a a

Therefore, from (4) and (5), it follows that

Z b Z b Z b
(f + g)(x)dx = f (x)dx + g(x)dx
a a a

(2) It is easy to prove (2) by the definition of the Riemann integral. (3) Let f ∈ ℜ[a, b]
and h(x) = x2 for all x ∈ [a, b]. Then, we have

h ◦ f = f 2 ∈ ℜ[a, b].

194
1 DEFINITE INTEGRALS Definite Integrals

Next, let f, g ∈ ℜ[a, b]. Then, by (1) and (2), we have

−g, f + g, f − g, (f + g)2 , (f − g)2 ∈ ℜ[a, b].

Therefore, since
1h i
fg =(f + g)2 − (f − g)2
4
we have f g ∈ ℜ[a, b]. (4) Just consider H(x) = g(x) − f (x) ≥ 0. (5) Since g(x) = |x|
is continuous on [a, b], we have |f | = g ◦ f ∈ ℜ[a, b]. On the other hand, let c ∈ R
(c = ±1) be such that
Z b Z b
c f (x)dx = f (x)dx
a a

Then, since cf (x) ≤ |f (x)| for every x ∈ [a, b], we have

Z b Z b
cf (x)dx ≤ |f (x)|dx
a a

Therefore, by (2), we have

Z b Z b Z b Z b
f (x)dx = c f (x)dx = cf (x)dx ≤ |f (x)|dx.
a a a a

This completes the proof. □

From the theorem above, we know that

ℜ[a, b] = {f : [a, b] → R : a Riemann integrable function on [a, b]}

is a vector space.
Theorem 9.10
Let f : [a, b] → R be a bounded function on [a, b], and let a number c with a < c < b.
Then, the following are equivalent:

1. f is Riemann integrable on [a, b];

2. f is Riemann integrable on both [a, c] and [c, b]. In this case, we have
Z b Z c Z b
f (x)dx = f (x)dx + f (x)dx
a a c

• Proof
(2) =⇒ (1) Suppose that f is Riemann integrable on both [a, c] and [c, b]. Then,
by the Riemann Integral Test, for every ε > 0, there exist P1 ∈ ℘[a, c] and P2 ∈ ℘[c, b]
such that
ε ε
U (f, P1 ) − L(f, P1 ) < and U (f, P2 ) − L(f, P2 ) <
2 2
195
1 DEFINITE INTEGRALS Definite Integrals

Let P = P1 ∪ P2 . Then, we have

U (f, P )−L(f, P ) = [U (f, P1 )+U (f, P2 )]−[L(f, P1 )+L(f, P2 )] = [U (f, P1 )−L(f, P1 )]+[U (f, P2 )−L(f, P2

Since ε > 0 is arbitrary, by the Riemann Integral Test, f is Riemann integrable on

[a, b]. (1) =⇒ (2) Suppose that f is Riemann integrable on [a, b]. Then, by the
Riemann Integral Test, for every ε > 0, there exists P1 ∈ ℘[a, b] such that

U (f, P1 ) − L(f, P1 ) < ε

If P = P1 ∪ {c}, then P is a refinement of P1 and so

U (f, P ) − L(f, P ) ≤ U (f, P1 ) − L(f, P1 ) < ε

Let P2 = P ∩ [a, c] and P3 = P ∩ [c, b]. Then, for P = P2 ∪ P3 , we have

U (f, P ) = U (f, P2 ) + U (f, P3 ), L(f, P ) = L(f, P2 ) + L(f, P3 )

and so, from (9.6), it follows that

[U (f, P2 ) − L(f, P2 )] + [U (f, P3 ) − L(f, P3 )] < ε

Thus, we have

U (f, P2 ) − L(f, P2 ) < ε, U (f, P3 ) − L(f, P3 ) < ε

Since ε > 0 is arbitrary, by the Riemann Integral Test, f is Riemann integrable on

both [a, c] and [c, b]. Finally, we prove
Z b Z c Z b
f (x)dx = f (x)dx + f (x)dx
a a c

In fact, from (9.7), it follows that

Z b Z c Z b
f (x)dx ≤ U (f, P ) = U (f, P2 )+U (f, P3 ) < L(f, P2 )+L(f, P3 )+2ε ≤ f (x)dx+ f (x)dx+2ε
a a c

Also, from from (9.7), it follows that

Z c Z b Z b
f (x)dx+ f (x)dx ≤ U (f, P2 )+U (f, P3 ) < L(f, P2 )+L(f, P3 )+2ε = L(f, P )+2ε ≤ f (x)dx+2ε
a c a

Since ε > 0 is arbitrary, it follows from (9.8) and (9.9) that

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

This completes the proof. □

196
1 DEFINITE INTEGRALS Definite Integrals

1.3 The Fundamental Theorems of Calculus

In this section, we prove the Fundamental Theorem of Calculus for a correction between
the derivative and the Riemann integral.
Theorem 9.11
Let f : [a, b] → R be Riemann integrable on [a, b], and define a function F : [a, b] → R
by
Z x
F (x) = f (t)dt for every x ∈ [a, b]
a

Then, we have the following:

1. F is uniformly continuous on [a, b];

2. If f is continuous at x0 ∈ [a, b], then F is differentiable at x0 and F ′ (x0 ) = f (x0 )

• Proof
(1) Since f is bounded on [a, b], there exists a constant M > 0 such that

|f (t)| ≤ M for every t ∈ [a, b]

Further, for every x, y ∈ [a, b], we have

Z y Z y
|F (x) − F (y)| = f (t)dt ≤ |f (t)|dt ≤ M |y − x|.
x x

ε
Therefore, for every ε > 0, if δ = M
, then

|y − x| < δ, x, y ∈ [a, b] =⇒ |F (y) − F (x)| < ε

which implies that F is uniformly continuous on [a, b]. (2) Let ε > 0. Since f is
continuous at x = x0 , there exists δ > 0 such that

|x − x0 | < δ, x ∈ [a, b] =⇒ |f (x) − f (x0 )| < ε

Thus, if 0 < |x − x0 | < δ, then we have

F (x) − F (x0 ) 1 1
Z x Z x0  Z x
− f (x0 ) = f (t)dt − f (t)dt − f (x0 ) ≤ |f (t)−f (x0 )|dt < ε
x − x0 x − x0 0 0 |x − x0 | x0

Therefore, F is differentiable at x0 , and F ′ (x0 ) = f (x0 ). This completes the proof. □

197
1 DEFINITE INTEGRALS Definite Integrals

Theorem 9.12 (The Fundamental Theorem of Calculus I)

Let f : [a, b] → R be a continuous function on [a, b], and define a function F : [a, b] →
R by
Z x
F (x) = f (t)dt for every x ∈ [a, b].
a

Then, F is differentiable on [a, b], and F ′ (x) = f (x).

• Proof
Since f is continuous on [a, b], for every ε > 0, there exists δ > 0 such that

|t − x| < δ =⇒ |f (t) − f (x)| < ε.

Let h be a number with 0 < h < δ. Then, we have

Z x+h Z x Z x+h
F (x + h) − F (x) = f (t)dt − f (t)dt = f (t)dt
a a x

and so

F (x + h) − F (x) 1 Z x+h 1 Z x+h 1 Z x+h

− f (x) = f (t)dt − f (x)dt ≤ |f (t)−f (x)|dt < ε.
h h x h x h x

Therefore, we have

F (x + h) − F (x)
F+′ (x) = lim+ = f (x).
h→0 h

Similarly, for −δ < h < 0, we have

F (x + h) − F (x)
F−′ (x) = lim− = f (x).
h→0 h

Since F+′ (x) = F−′ (x) = f (x), F is differentiable on [a, b] and, for all x ∈ [a, b],
F ′ (x) = f (x). This completes the proof. □

Theorem 9.13 (The Fundamental Theorem of Calculus II)

Let f : [a, b] → R be Riemann integrable on [a, b] and F be differentiable on [a, b]. If
F ′ (x) = f (x), then we have
Z b
f (x)dx = [F (x)]ba = F (b) − F (a)
a

• Proof
Let P = {a = x0 , x1 , x2 , . . . , xn = b} be a partition of [a, b]. Then, if we apply
Mean Value Theorem to F at each subinterval Ii = [xi−1 , xi ] for every i = 1, 2, . . . , n,

198
1 DEFINITE INTEGRALS Definite Integrals

then it follows that there exists ti ∈ [xi−1 , xi ] such that

F (xi) − F (xi−1 ) = f (ti )(xi − xi−1 )

Thus, we have
n
X n
X
F (b) − F (a) = (F (xi ) − F (xi−1 )) = f (ti )(xi − xi−1 ).
i=1 i=1

Put
Mi = sup{f (x) : x ∈ [xi−1 , xi ]}, mi = inf{f (x) : x ∈ [xi−1 , xi ]}.

Then, we have mi ≤ f (ti ) ≤ Mi for every i = 1, 2, . . . , n and so

n
X n
X
mi (xi − xi−1 ) ≤ F (b) − F (a) ≤ Mi (xi − xi−1 ).
i=1 i=1

Thus, it follows that, for every P ∈ ℘[a, b],

L(f, P ) ≤ F (b) − F (a) ≤ U (f, P )

and so, since

sup{L(f, P ) : P ∈ ℘[a, b]} ≤ F (b) − F (a) ≤ inf{U (f, P ) : P ∈ ℘[a, b]}

and f ∈ ℜ[a, b], we have

Z b
f (x)dx = F (b) − F (a).
a

This completes the proof. □

Example 9.6
Find an example in which if F is not differentiable on [a, b], then Theorem (9.13) is
not true.
Solution. Define two functions f, F : [0, 1] → R by f (x) = 1 for every x ∈ [0, 1] and

 x, if x ∈ [0, 1),


F (x) = 
 0,

if x = 1,
R1
respectively. Then, f is Riemann integrable on [0, 1] and 0 f (x)dx = 1. On the
other hand, we have F (1) = F (0) = 0, and, for every x ∈ [0, 1), F ′ (x) = f (x), but F
is not differentiable at x = 1. Thus, the Theorem (9.13) is not true.

199
1 DEFINITE INTEGRALS Definite Integrals

Theorem 9.14 (The Generalized Mean Value Theorem for Integral)

Let f : [a, b] → R be a continuous function on [a, b] and g : [a, b] → R be Riemann
integrable on [a, b] with g(x) ≥ 0 for every x ∈ [a, b]. Then, there exists c ∈ [a, b]
such that
Z b Z b
f (x)g(x)dx = f (c) g(x)dx.
a a

• Proof
Since f : [a, b] → R is continuous on [a, b], let

M = sup{f (x) : x ∈ [a, b]}, m = inf{f (x) : x ∈ [a, b]}

Then, it follows that

mg(x) ≤ f (x)g(x) ≤ M g(x) for every x ∈ [a, b]

and f g is Riemann integrable on [a, b] and so

Z b Z b Z b
m g(x)dx ≤ f (x)g(x)dx ≤ M g(x)dx
a a a
Rb Rb
If a g(x)dx = 0, then the conclusion follows easily. If a g(x)dx ̸= 0, then it follows
from (9.10) that
Rb
a f (x)g(x)dx
m≤ Rb ≤M
a g(x)dx
and so, since f is continuous on [a, b], by the Intermediate Value Theorem, there exists
c ∈ [a, b] such that
Rb
a f (x)g(x)dx
f (c) = Rb
a g(x)dx
that is,
Z b Z b
f (x)g(x)dx = f (c) g(x)dx
a a

This completes the proof. □

Example 9.7
Define two functions f, g : [−1, 1] → R by

f (x) = x, g(x) = ex for every x ∈ [−1, 1]

Then, apply Theorem (9.14) to the functions f and g on [−1, 1].

Solution. Since f and g are continuous on [−1, 1], they are Riemann integrable on

200
1 DEFINITE INTEGRALS Definite Integrals

[−1, 1]. Further, g(x) > 0 for every x ∈ [−1, 1] and

Z 1
2 Z 1
e2 − 1
f (x)g(x)dx = , g(x)dx =
−1 e −1 e

and so R1
2 −1 f (x)g(x)dx
= R1
e −1
2
−1 g(x)dx

Since f is continuous on [−1, 1] and

2
f (−1) = −1 < < 1 = f (1)
e2 −1

by the Intermediate Value Theorem, there exists c ∈ [−1, 1] such that

2
c = f (c) =
e2 −1

Therefore, we have
Z 1
2 e2 − 1 2 Z1
f (c) g(x)dx = · = = f (x)g(x)dx.
−1 e2 − 1 e e −1

Corollary 9.4 (The Mean Value Theorem for Integral)

Let f : [a, b] → R be a continuous function on [a, b]. Then, there exists c ∈ [a, b] such
that
Z b
f (x)dx = f (c)(b − a)
a

• Proof
In Theorem (9.14), if g(x) = 1 for every x ∈ [a, b], then we have
Z b
f (x)dx = f (c)(b − a)
a

Example 9.8
By using the Mean Value Theorem for Integral, prove the following inequalities:

x
< ln(1 + x) < x for every x > 0
1+x

Solution. Let f (x) = 1

1+x
for every x > 0. Then, we have
Z x
1
dt = ln(1 + x)
0 1+t

201
1 DEFINITE INTEGRALS Definite Integrals

Thus, by Corollary (9.4), there exists c ∈ [0, x] such that

Z x
1
dt = f (c)(x − 0)
0 1+t

Since f is decreasing on [0, x] and

inf{f (t) : t ∈ [0, x]} ≤ f (c) ≤ sup{f (t) : t ∈ [0, x]}

we have
1 ln(1 + x)
< < 1 for every x > 0
1+x x
which implies that

x
< ln(1 + x) < x for every x > 0
1+x

1.3.1 The Substitution Theorem and Integration by Parts

In this section, we prove the Substitution Theorem and the Integration by Parts as
techniques of integration, which are based on the Fundamental Theorems of Calculus.
Theorem 9.15 (The Substitution Theorem)
Let f : [a, b] → R be a continuous function on [a, b] and g : [c, d] → [a, b] be
differentiable on [c, d]. If g(c) = a and g(d) = b, then we have
Z b Z d
f (x)dx = f (g(t))g ′ (t)dt
a c

• Proof
For every x ∈ [a, b], define a function F : [a, b] → R by
Z x
F (x) = f (t)dt
a

Then, by Theorem (The Fundamental Theorem of Calculus I), F is differentiable on

[a, b]. Thus, we have

(F (g(t)))′ = F ′ (g(t))g ′ (t) = f (g(t))g ′ (t).

Therefore, by Theorem (The Fundamental Theorem of Calculus II), we have

Z d Z b
f (g(t))g ′ (t)dt = F (g(d)) − F (g(c)) = F (b) − F (a) = f (x)dx
c a

This completes the proof. □

202
1 DEFINITE INTEGRALS Definite Integrals

Example 9.9
Let f : [0, 1] → R be a continuous function on [0, 1]. Evaluate the following:
Z 1
xf (x2 )dx
−1

Solution. Let ϕ(x) = x2 for every x ∈ [−1, 1]. Then, since we have ϕ([−1, 1]) = [0, 1]
and f is continuous on [0, 1], we have
Z 1
2 1Z 1 ′ 1Z 1
f (x )dx = f (ϕ(x))ϕ (x)dx = f (t)dt = 0
−1 2 −1 2 1

Theorem 9.16 (The Integration by Parts)

Let f, g : [a, b] → R be differentiable on [a, b] and f ′ , g ′ be Riemann integrable on
[a, b]. Then, we have
Z b Z b
′
f (x)g (x)dx = f (b)g(b) − f (a)g(a) − f ′ (x)g(x)dx
a a

• Proof
Since f and g are differentiable on [a, b], f g is also differentiable on [a, b] and so,
for every x ∈ [a, b],

(f (x)g(x))′ = f ′ (x)g(x) + f (x)g ′ (x).

Since f and g are continuous on [a, b], they are Riemann integrable on [a, b]. Since f ′
and g ′ are Riemann integrable on [a, b],(f g)′ = f ′ g + f g ′ is also Riemann integrable
on [a, b]. Therefore, by Theorem (The Fundamental Theorem of Calculus I), we have
Z b
(f (x)g(x))′ dx = f (b)g(b) − f (a)g(a)
a

and so
Z b Z b
f ′ (x)g(x)dx + f (x)g ′ (x)dx = f (b)g(b) − f (a)g(a),
a a

which implies that

Z b Z b
′
f (x)g (x)dx = f (b)g(b) − f (a)g(a) − f ′ (x)g(x)dx.
a a

This completes the proof. □

203
1 DEFINITE INTEGRALS Definite Integrals

Example 9.10
Evaluate the following:
Z π/2
cos3 xdx
0

Solution. Now, we have

Z π/2 Z π/2
3
cos xdx = cos2 x(cos xdx)
0 0
Z π/2
= (1 − sin2 x)(cos xdx).
0

Let u = sin x. Then, we have du = cos xdx. If x = 0, then u = 0. If x = π2 , then

u = 1. Therefore, we have
Z π/2 Z 1
cos3 xdx = (1 − u2 )du
0 0
1 1
 
= u − u3
3 0
1 2
=1− = .
3 3

Theorem 9.17 (Luxemburg Monotone Convergence Theorem)

Let {fn (x)} be a decreasing sequence of bounded functions on [a, b], with a < b,
which converges pointwise to 0 on [a, b]. Then
Z b
lim fn (x)dx = 0.
n→∞ a

• Proof
The lower integral of any function is well defined provided the function is bounded.
Therefore for any bounded function f (x), and by definition of the lower integral, for
any ε > 0, there exists a step function L(x) ≤ f (x) such that
Z b
ε
(f (x) − L(x))dx <
a 2
Also note the existence of a continuous function c(x) ≤ L(x) such that
Z b
ε
(L(x) − c(x))dx <
a 2
Putting all this together, we conclude that for any bounded function f (x), and for
any ε > 0, there exists a continuous function c(x) ≤ f (x) such that
Z b
(f (x) − c(x))dx < ε
a

204
1 DEFINITE INTEGRALS Definite Integrals

Note that if f (x) ≥ 0, then the construction of c(x) will be done to have c(x) ≥ 0 as
well. Back to our claim. Let ε > 0. Then there exists a positive continuous function
c1 (x) ≤ f1 (x) such that
Z b
ε
(f1 (x) − c1 (x))dx <
a 22
Since min(c1 (x), f2 (x)) is bounded and positive, there exists a positive continuous
function c2 (x) ≤ min(c1 (x), f2 (x)) such that
Z b
ε
(min(c1 (x), f2 (x)) − c2 (x))dx <
a 23
Since f2 (x)−c2 (x) ≤ f2 (x)−min(c1 (x), f2 (x))+min(c1 (x), f2 (x))−c2 (x), and f2 (x) ≤
f1 (x), we get
Z b Z b Z b
ε ε
(f2 (x) − c2 (x))dx ≤ (f1 (x) − c1 (x))dx + (min(c1 (x), f2 (x)) − c2 (x))dx < 2
+ 3.
a a a 2 2
By the induction argument, a similar construction will lead to the existence of a
decreasing sequence of positive continuous functions {cn (x)} such that cn (x) ≤ fn (x)
and
Z b
ε ε ε ε
(fn (x) − cn (x))dx < 2
+ 3 + · · · + n+1 < .
a 2 2 2 2
Since {fn (x)} converges pointwise to 0 on [a, b] this will force the sequence {cn (x)} to
also converge pointwise to 0 on [a, b]. Dini’s theorem will imply that {cn (x)} converges
uniformly to 0 on [a, b]. So there exists n0 ≥ 1 such that for any n ≥ n0 we have
ε
cn (x) ≤ 2(b−a)
for any x ∈ [a, b]. Hence
Z b Z b Z b
ε ε
fn (x)dx ≤ (fn (x) − cn (x))dx + cn (x)dx < + (b − a) = ε
a a a 2 2(b − a)
whenever n ≥ n0 . This finishes the proof of our claim. □

Theorem 9.18 (Monotone Convergence Theorem)

Let {fn (x)} be a decreasing sequence of Riemann integrable functions on [a, b] which
converges pointwise to a Riemann integrable function f (x). Then
Z b Z b
lim
n→∞ a
fn (x)dx = f (x)dx
a

• Proof
Since f (x) is Riemann integrable, {fn (x) − f (x)} is a sequence of Riemann in-
tegrable functions which decreases to 0 . Obviously they are all bounded functions.

205
1 DEFINITE INTEGRALS Definite Integrals

The Luxemburg Monotone Convergence Theorem will imply

Z b
lim (fn (x) − f (x))dx = 0.
n→∞ a

Rb Rb
But a (fn (x) − f (x))dx = a (fn (x) − f (x))dx, hence
Z b
lim (fn (x) − f (x))dx = 0
n→∞ a

which implies
Z b Z b
lim
n→∞ a
fn (x)dx = f (x)dx
a

Theorem 9.19 (Arzelà Theorem)

Let {fn (x)} be a sequence of Riemann integrable functions on [a, b] which converges
pointwise to a Riemann integrable function f (x). Assume that there exists M > 0
such that |fn (x)| ≤ M for any x ∈ [a, b] and n ≥ 1. Then
Z b Z b
lim fn (x)dx = f (x)dx
n→∞ a a

• Proof
Since f (x) is Riemann integrable, {fn (x) − f (x)} is a sequence of Riemann in-
tegrable functions which decreases to 0 . Obviously they are all bounded functions.
The Luxemburg Monotone Convergence Theorem will imply
Z b
lim (fn (x) − f (x))dx = 0.
n→∞ a

Rb Rb
But a (fn (x) − f (x))dx = a (fn (x) − f (x))dx, hence
Z b
lim (fn (x) − f (x))dx = 0
n→∞ a

which implies
Z b Z b
lim
n→∞
fn (x) = f (x)dx
a a

206
1 DEFINITE INTEGRALS Definite Integrals

Lemma 9.20 (Fatou)

Let {fn (x)} be a sequence of Riemann integrable functions on [a, b] which converges
pointwise to a Riemann integrable function f (x). Then
Z b Z b
f (x)dx ≤ lim inf
n→∞
fn (x)dx
a a

• Proof
Rb
If lim inf n→∞ a fn (x)dx = ∞, then the conclusion is obvious. Assume that
Rb
lim inf n→∞ a fn (x)dx < ∞. Then there exists a subsequence {fnk } of {fn } such
that
Z b Z b
lim
nk →∞ a
fnk (x)dx = lim inf
n→∞
fn (x)dx
a

Clearly the subsequence {fnk } also converges pointwise to f (x). Set hni (x) = inf nk ≥ni fnk (x).
Then {hnk (x)} also converges pointwise to f (x) and is increasing. It is easy to see that
this sequence is bounded. The Luxemburg Monotone Convergence Theorem applied
to {f (x) − hnk (x)} will easily imply that
Z b Z b
lim hnk (x)dx = f (x)dx
n→∞ a a

Since hnk (x) ≤ fnk (x), we get

Z b Z b Z b
f (x)dx ≤ n→∞
lim fnk (x)dx = lim inf
n→∞
fn (x)dx
a a a

207