Chapter 6

Constructing measures

So far we have been taking measures for granted; except for a few, almost
trivial examples we have not shown that they exist. In this chapter we shall
develop a powerful technique for constructing measures with the properties
we want, e.g. the Lebesgue measures and the coin tossing measure described
in Section 5.1.
When we construct a measure, we usually start by knowing how we want
it to behave on a family of simple sets: For the Lebesgue measure on R, we
want the intervals (a, b) to have measure b − a, and for the coin tossing
measure we know what we want the measure of a cylinder set to be. The
art is to extend such “pre-measures” to full-blown measures.
We shall use a three step procedure to construct measures. We start
with a collection R of subsets of our space X and a function ρ : R → R+ .
The idea is that the sets R in R are the sets we “know” the size ρ(R) of – if,
e.g., we want to construct the Lebesgue measure, R could be the collection
of all open intervals, and ρ would then be given by ρ((a, b)) = b − a. From
R and ρ, we first construct an “outer measure” µ∗ which assigns a “size”
µ∗ (A) to all subsets A of X. The problem with µ∗ is that it usually fails to
be countably additive; i.e. the crucial equality

[ ∞
X
∗
µ ( An ) = µ∗ (An )
n∈N n=1

doesn’t hold for all disjoint sequences {An } of subsets of X. The second step
in our procedure is therefore to identify a σ-algebra A such that countable
additivity holds when the disjoint sets An belong to A. The restriction of µ∗
to A will then be our desired measure µ. The final step in the procedure is
to check that µ really is an extension of ρ, i.e. that R ⊂ A and µ(R) = ρ(R)
for all R ∈ R. This is not always the case, but requires special properties of
R and ρ.
We begin by constructing outer measures.

1
2 CHAPTER 6. CONSTRUCTING MEASURES

6.1 Outer measure

To construct outer measures, we don’t need to require much of R and ρ:

Definition 6.1.1 Assume that X is a nonempty set and that R is a collec-

tion of subsets of X such that ∅, X ∈ R. Throughout this section, we assume
that ρ : R → R+ is a function such that ρ(∅) = 0.

Assume that B is a subset of X. A covering of X is a countable collection

C = {Cn }n∈N of sets from R such that
[
B⊂ Cn
n∈N

We define the size of the covering to be

∞
X
|C| = ρ(Cn )
n=1

We are now ready to define the outer measure µ∗ generated by R and ρ: For
all B ⊂ X, we set

µ∗ (B) = inf{|C| : C is a covering of B}

We see why µ∗ is called an outer measure; it is obtained by approximating

sets from the outside by unions of sets in R.
The essential properties of outer measures are not hard to establish:

Proposition 6.1.2 The outer measure µ∗ satisfies:

(i) µ∗ (∅) = 0.

(ii) (Monotonicity) If B ⊂ C, then µ∗ (B) ≤ µ∗ (C).

(iii) (Countable subadditivity) If {Bn }n∈N is a sequence of subsets of Rd ,

then
∞
[ ∞
X
µ∗ ( Bn ) ≤ µ∗ (Bn )
n=1 n=1

Proof: (i) Since C = {∅, ∅, ∅, . . .} is a covering of ∅ and ρ(∅) = 0, we get

µ∗ (∅) = 0.

(ii) Since any covering of C is a covering of B, we have µ∗ (B) ≤ µ∗ (C).

6.1. OUTER MEASURE 3

(iii) If µ∗ (Bn ) = ∞ for some n ∈ N, there is nothing to prove, and we

may hence assume that µ∗ (Bn ) < ∞ for all n. Let  > 0 be given. For each
(n) (n)
n ∈ N, we can find a covering C1 , C2 , . . . of Bn such that
∞
X (n) 
ρ(Ck ) < µ∗ (Bn ) +
2n
k=1

(n)
S {Ck }k,n∈N of all sets in all the coverings is a countable
The collection
covering of ∞ n=1 Bn , and

∞ ∞ ∞  ∞
!
X (n)
X X (n)
X
∗   X ∗
ρ(Ck ) = ρ(Ck ) ≤ µ (Bn ) + n = µ (Bn ) + 
2
k,n∈N n=1 k=1 n=1 n=1

(if you feel unsure about these manipulations, take a look at Exercise 5).
This means that
[ ∞
X
µ∗ ( Bn ) ≤ µ∗ (Bn ) + 
n∈N n=1

and since  is an arbitrary, positive number, we must have

[ ∞
X
µ∗ ( Bn ) ≤ µ∗ (Bn )
n∈N n=1

Remark: Note that property (iii) in the proposition above also holds for
finite sums; i.e.
N
[ XN
µ∗ ( Bn ) ≤ µ∗ (Bn )
n=1 n=1

(to see this, just apply (iii) to the sequence B1 , B2 , . . . , BN , ∅, ∅, . . .). In par-
ticular, we always have µ∗ (A ∪ B) ≤ µ∗ (A) + µ∗ (B).

We have now completed the first part of our program: We have con-
structed the outer measure and described its fundamental properties. The
next step is to define the measurable sets, to prove that they form a σ-
algebra, and show that µ∗ is a measure when we restrict it to this σ-algebra.

Exercises for Section 6.1

1. Show that µ∗ (R) ≤ ρ(R) for all R ∈ R.
2. Let X = {1, 2}, R = {∅, {1}, {1, 2}}, and define ρ : R → R by ρ(∅) = 0,
ρ({1}) = 2, ρ({1, 2}) = 1. Show that µ∗ ({1}) < ρ({1}).
3. Assume that X = R, and let R consist of ∅, X, plus all open intervals (a, b),
where a, b ∈ R. Define ρ : R → R by ρ(∅) = 0, ρ(R) = ∞, ρ((a, b)) = b − a.
4 CHAPTER 6. CONSTRUCTING MEASURES

a) Show that if I = [c, d] is a closed and bounded interval, and C = {Cn }

is a covering of I, then there is a finite number Ci1 , Ci2 , . . . , Cin of sets
from C that covers I (i.e., such that I ⊂ Ci1 ∪ Ci2 ∪ . . . ∪ Cin ). (Hint:
Compactness.)
b) Show that µ∗ ([c, d]) = ρ([c, d]) = d − c for all closed and bounded
intervals.

4. Assume that R is a σ-algebra and that ρ is a measure on R. Let (X, R̄, µ̄)
be the completion of (X, R, µ). Show that µ∗ (A) = µ̄(A) for all A ∈ R̄.
5. Let {an,k }n,k∈N be a collection of nonnegative, real numbers, and let A be
the supremum over all finite sums of distinct elements in this collection, i.e.
I
X
A = sup{ ani ,ki : I ∈ N and all pairs (n1 , k1 ), . . . , (nI , kI ) are different}
i=1

a) Assume that {bm }m∈N is a sequence whichPcontains each element in the

∞
set {an,k }n,k∈N exactly ones. Show that m=1 bm = A.
P∞ P∞
b) Show that n=1 ( k=1 an,k ) = A.
c) Comment on the proof of Proposition 6.1.2(iii).

6.2 Measurable sets

The definition of a measurable set is not difficult, but it is rather myste-
rious in the sense that it is not easy to see why it captures the essence of
measurability.

Definition 6.2.1 A subset E of X is called measurable if

µ∗ (A ∩ E) + µ∗ (A ∩ E c ) = µ∗ (A)

for all A ⊂ X. The collection of all measurable sets is denoted by M.

This approach to measurability was introduced by the Greek mathematician

Constantin Carathéodory (1873-1950) and replaced more cumbersome (but
easier to understand) earlier approaches (see Exercise 6.3.4 for one such).
As already mentioned, it is not at all easy to explain why it captures the
intuitive notion of measurability. The best explanation I can offer, is that the
reason why some sets are impossible to measure (and hence non-measurable),
is that they have very irregular boundaries. The definition above says that
a set is measurable if we can use it to cut any other set in two parts without
introducing any further irregularities – hence all parts of its boundary must
be reasonably regular. I admit that this explanation is rather vague, and a
better argument may simply be to show that the definition works. So let us
get started.
6.2. MEASURABLE SETS 5

Let us first of all make a very simple observation. Since A = (A ∩ E) ∪

(A ∩ E c ), subadditivity (recall Proposition 6.1.2(iii)) tells us that we always
have
µ∗ (A ∩ E) + µ∗ (A ∩ E c ) ≥ µ∗ (A)
Hence to prove that a set is measurable, we only need to prove that

µ∗ (A ∩ E) + µ∗ (A ∩ E c ) ≤ µ∗ (A)

Our first result is easy.

Lemma 6.2.2 If E has outer measure 0, then E is measurable. In partic-

ular, ∅ ∈ M.

Proof: If E has outer measure 0, so has A ∩ E since A ∩ E ⊂ E. Hence

µ∗ (A ∩ E) + µ∗ (A ∩ E c ) = µ∗ (A ∩ E c ) ≤ µ∗ (A)

for all A ⊂ X. 2

Next we have:

Proposition 6.2.3 M satisfies:1

(i) ∅ ∈ M.

(ii) If E ∈ M, then E c ∈ M.

(iii) If E1 , E2 , . . . , En ∈ M, then E1 ∪ E2 ∪ . . . ∪ En ∈ M.

(iv) If E1 , E2 , . . . , En ∈ M, then E1 ∩ E2 ∩ . . . ∩ En ∈ M.

Proof: We have already proved (i), and (ii) is obvious from the definition
of measurable sets. Since E1 ∪ E2 ∪ . . . ∪ En = (E1c ∩ E2c ∩ . . . ∩ Enc )c by
De Morgan’s laws, (iii) follows from (ii) and (iv). Hence it only remains to
prove (iv).
To prove (iv) is suffices to prove that if E1 , E2 ∈ M, then E1 ∩ E2 ∈ M
as we can then add more sets by induction. If we first use the measurability
of E1 , we see that for any set A ⊂ Rd

µ∗ (A) = µ∗ (A ∩ E1 ) + µ∗ (A ∩ E1c )

Using the measurability of E2 , we get

µ∗ (A ∩ E1 ) = µ∗ ((A ∩ E1 ) ∩ E2 ) + µ∗ ((A ∩ E1 ) ∩ E2c )

1
As you probably know from Chapter 5, a family of sets satisfying (i)-(iii) is usually
called an algebra. As (iv) is a consequence of (ii) and (iii) using one of De Morgan’s laws,
the proposition simply states that M is an algebra, but in the present context (iv) is in
many ways the most important property.
6 CHAPTER 6. CONSTRUCTING MEASURES

Combining these two expressions, we have

µ∗ (A) = µ∗ ((A ∩ (E1 ∩ E2 )) + µ∗ ((A ∩ E1 ) ∩ E2c ) + µ∗ (A ∩ E1c )

Observe that (draw a picture!)

(A ∩ E1 ∩ E2c ) ∪ (A ∩ E1c ) = A ∩ (E1 ∩ E2 )c

and hence

µ∗ (A ∩ E1 ∩ E2c ) + µ∗ (A ∩ E1c ) ≥ µ∗ (A ∩ (E1 ∩ E2 )c )

Putting this into the expression for µ∗ (A) above, we get

µ∗ (A) ≥ µ∗ ((A ∩ (E1 ∩ E2 )) + µ∗ (A ∩ (E1 ∩ E2 )c )

which means that E1 ∩ E2 ∈ M. 2

We would like to extend parts (iii) and (iv) in the proposition above to
countable unions and intersection. For this we need the following lemma:

Lemma 6.2.4 If E1 , E2 , . . . , En is a disjoint collection of measurable sets,

then

µ∗ (A ∩ (E1 ∪ E2 ∪ . . . ∪ En )) = µ∗ (A ∩ E1 ) + µ∗ (A ∩ E2 ) + . . . + µ∗ (A ∩ En )

Proof: It suffices to prove the lemma for two sets E1 and E2 as we can then
extend it by induction. Using the measurability of E1 , we see that

µ∗ (A ∩ (E1 ∪ E2 )) = µ∗ ((A ∩ (E1 ∪ E2 )) ∩ E1 ) + µ∗ (A ∩ (E1 ∪ E2 )) ∩ E1c ) =

= µ∗ (A ∩ E1 ) + µ∗ (A ∩ E2 )
2

We can now prove that M is closed under countable unions.

S
Lemma 6.2.5 If An ∈ M for each n ∈ N, then n∈N An ∈ M.

Proof: Define a new sequence {En } of sets by E1 = A1 and

En = An ∩ (E1 ∪ E2 ∪ . . . ∪ En−1 )c

for n > 1, and note that En ∈ M since M is an algebra. The sets {En } are
same union as {An } (make a drawing!), and hence it
disjoint and have the S
suffices to prove that n∈N En ∈ M, i.e.
∞ ∞
[ [
c

µ∗ (A) ≥ µ∗ A ∩ En + µ∗ A ∩


En
n=1 n=1
6.2. MEASURABLE SETS 7

SN
Since n=1 En ∈ M for all N ∈ N, we have:

N N
∗ ∗
[
∗
[
c

µ (A) = µ A ∩ En + µ A ∩ En ≥
n=1 n=1

N ∞
X
∗ ∗
[
c

≥ µ (A ∩ En ) + µ A ∩ En
n=1 n=1

where Swe in the
last step

SN have
cused the lemma above plus the observation
∞ c
that n=1 En ⊂ n=1 En . Since this inequality holds for all N ∈ N,
we get
∞ ∞
∗
X
∗ ∗
[
c

µ (A) ≥ µ (A ∩ En ) + µ A ∩ En
n=1 n=1
P∞ ∗ ∗
S∞ ∗
By
S∞sublinearity, we have n=1 µ (A ∩ En ) ≥ µ ( n=1 (A ∩ En )) = µ (A ∩
n=1 (En )), and hence

∞ ∞
∗ ∗
[
∗
[
c

µ (A) ≥ µ A ∩ En + µ A ∩ En
n=1 n=1

Lemmas 6.2.3 and 6.2.5 tell us that M is a σ-algebra. We may restrict

µ∗ to M to get a function
µ : M → R+
defined by2
µ(A) = µ∗ (A) for all A ∈ M
We can now complete the second part of our program:

Theorem 6.2.6 M is a σ-algebra, and µ is a complete measure on M.

Proof: We already know that M is a σ-algebra, and if we prove that µ is

a measure, the completeness will follow from Lemma 6.2.2. As we already
know that µ(∅) = µ∗ (∅) = 0, we only need to prove that
∞
[ ∞
X
µ( En ) = µ(En ) (6.2.1)
n=1 n=1

for all disjoint sequences {En } of sets from M

Since µ is just the restriction of µ∗ to a smaller domain, it may seem a luxury to intro-
2

duce a new symbol for it, but in some situations it is important to be able to distinguish
quickly between µ and µ∗ .
8 CHAPTER 6. CONSTRUCTING MEASURES

By Proposition 6.1.2(iii), we already know that

∞
[ ∞
X
µ( En ) ≤ µ(En )
n=1 n=1

To get the opposite inequality, we use Lemma 6.2.4 with A = X to see that
N
X N
[ ∞
[
µ(En ) = µ( En ) ≤ µ( En )
n=1 n=1 n=1

Since this holds for all N ∈ N, we must have

∞
X ∞
[
µ(En ) ≤ µ( En )
n=1 n=1

Hence we have both inequalities, and (6.2.1) is proved. 2

We have now completed the second part of our program — we have

shown how we can turn an outer measure µ∗ into a measure µ by restricting
it to the measurable sets. This is in an interesting result in itself, but we still
have some work to do – we need to compare the measure µ to the original
function ρ.

Exercises for Section 6.2

1. Explain in detail why 6.2.3(iii) follows from (ii) and (iv).
2. Carry out the induction step in the proof of Proposition 6.2.3(iv).
3. Explain the equality (A ∩ E1 ∩ E2c ) ∪ (A ∩ E1c ) = A ∩ (E1 ∩ E2 )c in the proof
of Lemma 6.2.3.
4. Carry out the induction step in the proof of Lemma 6.2.4.
5. Explain why the sets En in the proof of Lemma 6.2.5 are disjoint and have
the same union as the sets An . Explain in detail why the sets En belong to
M.

6.3 Carathéodory’s Theorem

In the generality we have have been working so far, there isn’t much that can
be said about the relationship between the original set function ρ and the
measure µ generated via the outer measure construction. Since R, ∅, ∅, . . . is
a covering of R, we always have µ∗ (R) ≤ ρ(R) for all R ∈ R, but if we want
equality instead of inequality, we need to impose conditions on R and ρ:

Definition 6.3.1 R is called an algebra of sets on X if the following con-

ditions are satisfied:
6.3. CARATHÉODORY’S THEOREM 9

(i) ∅ ∈ R

(ii) If R ∈ R, then Rc ∈ R.

(iii) If R, S ∈ R, then R ∪ S ∈ R

If R is an algebra of sets, the function ρ : R → R+ is called a premeasure

if the following conditions are satisfied:

(iv) ρ(∅) = 0

(v) If {Rn }n∈N is a disjoint sequence of sets in R whose union happens to

belong to R, then3
[ X∞
ρ( Rn ) = ρ(Rn )
n∈N n=1

We are now ready for the main result of this section.

Theorem 6.3.2 (Carathéodory’s Extension Theorem) Assume that R

is an algebra and that ρ is a premeasure on R. Then there exists a com-
plete measure µ extending ρ, i.e., a measure µ defined on a σ-algebra M
containing R such that µ(R) = ρ(R) for all R ∈ R.

Proof: We know that the outer measure construction generates a σ-algebra

M of measurable sets and a complete measure µ on M, and it suffices to
show that all sets in R are measurable and that µ(R) = ρ(R) for all R ∈ R.
Let us first prove that any set R ∈ R is measurable, i.e. that

µ∗ (A) ≥ µ∗ (A ∩ R) + µ∗ (A ∩ Rc )

for any setPA ⊂ X. Given  > 0, we can find a covering C = {Cn }n∈N of A
such that ∞ ∗ c
n=1 ρ(Cn ) < µ (A) + . Note that {Cn ∩ R} and {Cn ∩ R } are
coverings of A ∩ R and A ∩ Rc , respectively, and hence
∞
X ∞
X
∗
ρ(Cn ∩ R) + ρ(Cn ∩ Rc ) =


µ (A) +  > ρ(Cn ) =
n=1 n=1

∞
X ∞
X
= ρ(Cn ∩ R) + ρ(Cn ∩ Rc ) ≥ µ∗ (A ∩ R) + µ(A ∩ Rc )
n=1 n=1

Since  < 0 is arbitrary, this means that µ∗ (A) ≥ µ∗ (A ∩ R) + µ∗ (A ∩ Rc ),

and hence R is measurable.
It remains to prove that µ(R) = µ∗ (R) = ρ(R) for all R ∈ R. As we
already know that µ∗ (R) ≤ ρ(R), it suffices to prove the opposite inequality.
3
In general, there is no reason why such a union should belong to R, but it does happen,
e.g. when all but a finite number of the sets Rn equals ∅
10 CHAPTER 6. CONSTRUCTING MEASURES

P∞
For any  > 0, there is a covering C = {C n }n∈N of R such that n=1 ρ(Cn ) <
µ∗ (R) + . The sets Cn0 = R ∩ (Cn \ n−1
S
k=1 C k ) are disjoint elements of R
whose union is R, and hence by condition (v) in the definition above
∞
X ∞
X
ρ(R) = ρ(Cn0 ) ≤ ρ(Cn ) < µ∗ (R) + 
n=1 n=1

Since  > 0 is arbitrary, this means that ρ(R) ≤ µ∗ (R), and since we already
have the opposite inequality, we have proved that µ∗ (R) = ρ(R). 2

In general, there is more than one measure extending a given premeasure

ρ, but for most spaces that occur in practice, there isn’t too much freedom
(but see Exercise 2 for an extreme case):

Proposition 6.3.3 Let ρ be a premeasure on an algebra R and let (X, M, µ)

be the measure space obtained by the outer measure construction. Assume
that B is a σ-algebra containing A and that ν is a measure on B such that
ν(R) = ρ(R) for all R ∈ R. Then ν(A) ≤ µ(A) for all A ∈ M ∩ B, with
equality whenever µ(A) < ∞.

Proof: Assume A ∈ M ∩ B and let C = {Cn } be a covering of A. Then

[ ∞
X ∞
X
ν(A) ≤ ν( Cn ) ≤ ν(Cn ) = ρ(Cn ) = |C|
n∈N n=1 n=1

and since µ(A) = µ∗ (A)

= inf{|C| : C is a covering of A}, we see that
ν(A) ≤ µ(A).
Now assume that µ(A) < ∞. There clearly exists a covering C = {Cn }
of A such that |C| < ∞. Replacing Cn by Cn \ (C1 ∪ . . . Cn−1 ) if necessary,
we
P∞may assume that the sets Cn are disjoint, and hence µ(C) = ν(C) =
n=1 ρ(Cn ) < ∞. We thus have

ν(A) + ν(C \ A) = ν(C) = µ(C) = µ(A) + µ(C \ A)

and since we already know that ν(A) ≤ µ(A) and ν(C \ A) ≤ µ(C \ A), this
is only possible if ν(A) = µ(A) (and ν(C \ A) = µ(C \ A)). 2

Theoretically, Carathéodory’s theorem is a very natural and satisfactory

result, but it is a little inconvenient to use in practice as we seldom start with
a premeasure defined on an algebra of sets – experience shows that what we
usually start from is something slightly weaker called a semi-algebra. We
shall now extend Carathéodory’s result to deal with this situation, and we
begin with the definition of a semi-algebra.

Definition 6.3.4 Let X be a non-empty set and R a non-empty collection

of subsets of X. We call R a semi-algebra if the following conditions are
satisfied:
6.3. CARATHÉODORY’S THEOREM 11

(i) If R, S ∈ R, then R ∩ S ∈ R.

(ii) If R ∈ R, then Rc is a a disjoint union of sets in R.

Starting with a semi-algebra, it is not hard to build an algebra:

Lemma 6.3.5 Assume that R is a semi-algebra on a set X and let A consist

of ∅ plus all finite, disjoint unions of sets in R. Then A is the algebra
generated by R.

Proof: As all sets in A clearly have to be in any algebra containing R, we

only need to show that A is an algebra, and for this it suffices to show that
A is closed under complements and finite intersections.
Let us start with the intersections. Assume that A, B are two nonempty
sets in A, i.e. that
A = R1 ∪ R2 ∪ . . . ∪ Rn
B = Q1 ∪ Q2 ∪ . . . ∪ Qm
are disjoint unions of sets from R. Then
[
A ∩ B = (Ri ∩ Qj )
i,j

is clearly a disjoint, finite union of sets in R, and hence A ∩ B ∈ A. By

induction, we see that any finite intersection of sets in A is in A.
Turning to complements, we first observe that X = ∅c belongs to A.
To see this, pick a nonempty set R ∈ R. Since R is a semi-algebra, Rc =
R1 ∪ . . . ∪ Rn for disjoint sets R1 , . . . , Rn ∈ R, and hence

X = R ∪ R1 ∪ . . . ∪ Rn

which proves that X ∈ A. Assume next that A = R1 ∪ R2 ∪ . . . ∪ Rn is a

disjoint union of sets in R. Then by one of De Morgan’s laws

Ac = R1c ∩ R2c ∩ . . . ∩ Rnc

Since R is a semi-algebra, each Ric is a disjoint union of sets in R and hence

belongs to A. Since we have already proved that A is closed under finite
intersections, Ac ∈ A. 2

Here is the promised extension of Carathéodory’s theorem to semi-algebras:

Theorem 6.3.6 (Carathéodory’s Theorem for Semi-Algebras) Let R

be a semi-algebra of subsets of X and assume that λ : R → R+ is a function
such that
12 CHAPTER 6. CONSTRUCTING MEASURES

Sn
(i) if a set R ∈ R is a disjoint, finite union R = i=1 Ri of sets in R,
then
Xn
λ(R) = λ(Ri )
i=1
S
(ii) If a set R ∈ R is a disjoint, countable union R = i∈N Ri of sets in
R, then
X∞
λ(R) = λ(Ri )
i=1

Then λ has an extension to a complete measure on a σ-algebra containing

R.4

Before we turn to the proof of the theorem, it is convenient to prove the

following lemma.

Lemma 6.3.7 Assume that R and λ are as in the theorem above.

(i) Assume that a set A ⊂ X can be written
Sn as disjoint, finite
Sm unions of
sets in R in two different ways: A = i=1 Ri and A = j=1 Sj . Then
n
X m
X
λ(Ri ) = λ(Sj )
i=1 j=1

(ii) Assume that a set A ⊂ X can be written as

Sndisjoint unions of
S sets in
R in a finite and a countable way: A = i=1 Ri and A = j∈N Sj .
Then
Xn ∞
X
λ(Ri ) = λ(Sj )
i=1 j=1

Proof: We only prove (ii). The proof of (i) is similar, the only difference be-
ing that we in one place use condition (i) of the theorem instead of condition
(ii).
To prove (ii), observe that since R is a semi-algebra, the intersections
Ri ∩ Sj belong to R, and hence by condition (ii) in the theorem
∞
X
λ(Ri ) = λ(Ri ∩ Sj )
j=1

Similarly by condition (i) in the theorem

n
X
λ(Sj ) = λ(Ri ∩ Sj )
i=1
4
If ∅ ∈ R and λ(∅) = 0, (i) is just a special case of (ii), but since we haven’t assumed
that ∅ ∈ R, we have to treat finite and infinite unions separately.
6.3. CARATHÉODORY’S THEOREM 13

This means that

n
X X ∞
n X ∞ X
X n ∞
X
λ(Ri ) = λ(Ri ∩ Sj ) = λ(Ri ∩ Sj ) = λ(Sj )
i=1 i=1 j=1 j=1 i=1 j=1

which is what we wanted to prove. 2

Proof of Carathéodory’s theorem for semi-algebras: Let A be the algebra

generated by R. The plan is to extend λ to a premeasure ρ on A, and then
use the version of Carathéodory’s theorem that we have already proved.
Since any non-empty set A in A is a finite, disjoint union A = ni=1 Ri
S
of sets in R, we define ρ : A → R+ by putting ρ(∅) = 0 and
n
X
ρ(A) = λ(Ri )
i=1

= m
S
(ρ is well-defined as part (i) of the lemma tells us that if A P j=1 Sj is
n
another
Pm way of writing A as a disjoint union of sets in R, then i=1 λ(Ri ) =
j=1 λ(S j )). To use Carathéodory’s Extension Theorem 6.3.2, we need to
show that ρ is a premeasure, i.e. that
∞
X
ρ(A) = ρ(An )
n=1

whenever A and A1 , A2 , . . . are in A and A is the disjoint union of A1 , A2 , . . ..

Since A and the An ’s are in A, they can be written as finite, disjoint unions
of sets in R:
[M
A= Rj
j=i

and
Nn
[
An = Sn,k
k=1
S S
Since A = n∈N An = n,k Sn,k , part (ii) of the lemma tells us that

M
X X
λ(Rj ) = λ(Sn,k )
j=1 n,k

The rest is easy: By definition of ρ

M
X X ∞ X
X Nn ∞
X
ρ(A) = λ(Rj ) = λ(Sn,k ) = λ(Sn,k ) = ρ(An )
j=1 n,k n=1 k=1 n=1
14 CHAPTER 6. CONSTRUCTING MEASURES

We have now proved that ρ is a premeasure on the algebra A, and by

the original Carathéodory’s theorem 6.3.2, there is a complete measure on
a σ-algebra containing A which extends ρ (and hence λ). 2

We now have the machinery we need to construct measures, and in the

next section we shall use it to construct Lebesgue measure on R.

Exercises for Section 6.3

1. Prove part (i) of Lemma 6.3.7.
2. A measure space (X, A, µ) is called σ-finite
S if there there is a disjoint family
{An }n∈N of sets in A such that X = n∈N An and µ(An ) < ∞ for all n ∈ N.
a) Show that (X, A, µ) is σ-finite if and only
S if there is an increasing family
{En }n∈N of sets in A such that X = n∈N En and µ(En ) < ∞ for all
n ∈ N.
b) Show that the Lebesgue measure is σ-finite.
c) Show that if the measure µ in Proposition 6.3.3 is σ-finite, then ν(A) =
µ(A) for all A ∈ M∩B. Hence µ has a unique extension to the σ-algebra
generated by R.
3. Let X = Q and let R be the collection of subsets of X consisting of
(i) All bounded, half-open, rational intervals (r, s]Q = {q ∈ Q | r < q ≤ s}
where r, s ∈ Q.
(ii) All unbounded, half-open, rational intervals (−∞, s]Q = {q ∈ Q | q ≤ s}
and (r, ∞)Q = {q ∈ Q | r < q} where r, s ∈ Q.

a) Show that R is a semi-algebra.

b) Show that the σ-algebra M generated by R is the collection of all subset
of X.

Define ρ : R → R+ by ρ(∅) = 0, ρ(R) = ∞ otherwise.

c) Show that ρ is a satisfies the conditions of Carathéodory’s theorem for
semi-algebras.
d) Show that there are infinitely many extensions of ρ to a measure ν on
M. (Hint: Which values may ν({q}) have?)
4. In this problem we shall take a look at a different (and perhaps more intuitive)
approach to measurability. We assume that R is an algebra and that ρ is a
premeasure on R. We also assume that ρ(X) < ∞. Define the inner measure
of a subset E of X by

µ∗ (E) = µ∗ (X) − µ∗ (E c )

We call a set E ∗-measurable if µ∗ (E) = µ∗ (E). (Why is this a natural

condition?)
a) Show that µ∗ (E) ≤ µ∗ (E) for all E ⊂ X.
6.4. LEBESGUE MEASURE ON R 15

b) Show that if E is measurable, then E is ∗-measurable. (Hint: Use

Definition 6.2.1 with A = X.)
c) Show that if E is ∗-measurable, then for every  > 0 there are measur-
able sets D, F such that D ⊂ E ⊂ F and
 
µ∗ (D) > µ∗ (E) − and µ∗ (F ) < µ∗ (E) +
2 2

d) Show that µ∗ (F \D) <  and explain that µ∗ (F \E) <  and µ∗ (E\D) <
.
e) Explain that for any set A ⊂ X

µ∗ (A ∩ F ) = µ∗ (A ∩ D) + µ∗ (A ∩ (F \ D)) < µ∗ (A ∩ D) + 

and

µ∗ (A ∩ Dc ) = µ∗ (A ∩ F c ) + µ∗ (A ∩ (F \ D)) < µ∗ (A ∩ F c ) + 

and use this to show that µ∗ (A ∩ D) > µ∗ (A ∩ E) −  and µ∗ (A ∩ F c ) >

µ∗ (A ∩ E c ) − .
f) Explain that for any set A ⊂ X

µ∗ (A) ≥ µ∗ (A ∩ (F c ∪ D)) =

= µ∗ (A ∩ F c ) + µ∗ (A ∩ D) ≥ µ∗ (A ∩ E c ) + µ∗ (A ∩ E) − 2
and use it to show that if E is ∗-measurable, then E is measurable. The
notions of measurable and ∗-measurable hence coincide when µ∗ (X) is
finite.

6.4 Lebesgue measure on R

In this section we shall use the theory in the previous section to construct
the Lebesgue measure on R. Essentially the same argument can be used to
construct Lebesgue measure on Rd for d > 1, but as the geometric consid-
erations become a little more complicated, we shall restrict ourselves to the
one dimensional case. In Section 6.6 we shall see how we can obtain higher
dimensional Lebesgue measure by a different method.
Recall that the one dimensional Lebesgue measure is a generalization of
the notion of length: We know how long intervals are and want to extend this
notion of size to a full-blown measure. For technical reasons, it is convenient
to work with half-open intervals (a, b].

Definition 6.4.1 In this section, R consists of the following subsets of R:

(i) All finite, half-open intervals (a, b], where a, b ∈ R, a < b.

(ii) All infinite intervals of the form (−∞, b] and (a, ∞) where a, b ∈ R.
16 CHAPTER 6. CONSTRUCTING MEASURES

The advantage of working with half-open intervals become clear when we

check what happens when we take intersections and complements:
Proposition 6.4.2 R is a semi-algebra of sets.
Proof: I leave it to the reader to check the various cases. The crucial obser-
vation is that the complement of a half-open interval is either a half-open
interval or the union of two half-open intervals. 2

We define λ : R → R+ simply by letting λ(R) be the length of the interval

R. To use Carathéodory’s theorem for semi-algebras, we first need to check
that the main condition is satisfied.
S
Lemma 6.4.3 If a set R ∈ R is a disjoint, countable union R = i∈N Ri
of sets in R, then
∞
X
λ(R) = λ(Ri )
i=1
The same holds for finite unions: If R ∈ R is a disjoint, finite union R =
R1 ∪ R2 ∪ . . . ∪ Rn of sets in R, then λ(R) = λ(R1 ) + λ(R2 ) + . . . λ(Rn ).
Proof: The case for finite unions is easy and left to the reader. For the
countable case, note that for any finite N , the nonoverlapping intervals
R1 , R2 , . . . , RN can be ordered from left Pto right, and obviously make up a
N
part of R in a such a way that λ(R) ≥ n=1 λ(Rn ). Since this holds for all
finite N , we must have λ(R) ≥ ∞
P
n=1 λ(R n ).
The opposite inequality is more subtle, and we need to use a compactness
argument. Let us first assume that R is a finite interval (a, b]. Given an
 > 0, we extend each interval Rn = (an , bn ] to an open interval R̂n =
(an , bn + 2n ). These open intervals cover the compact interval [a + , b],
and by Theorem 2.6.6 there is a finite subcover R̂n1 , R̂n2 , . . . , R̂nk . Since
Pk finite intervals cover an interval of length b − a − , we
these

clearly have
j=1 λ(R̂nj ) ≥ (b − a − ). But since λ(R̂n ) = λ(Rn ) + 2n , this means
P∞
that kj=1 λ(Rnj ) ≥ (b − a − 2). Consequently,
P
n=1 λ(Rn ) ≥ b − a −
2
P∞ = λ(R) − 2, and since  is an arbitrary, positive number, we must have
n=1 λ(R n ) ≥ λ(R).
It remains to see what happens if R is an interval of infinite length,
i.e. when λ(R) = ∞. For each n ∈ N, the intervals {Ri ∩ (−n, n]}i∈N
are disjoint and have P union R ∩ (−n, n]. By P∞what we have already proved,
λ(R ∩ (−n, n]) = ∞ i=1 λ(R i ∩ (−n, n]) ≤ i=1 λ(Ri ). Since limn→∞ λ(R ∩
(−n, n]) = ∞, we see that ∞ 2
P
i=1 λ(R i ) = ∞ = λ(R).

We have reached our goal: The lemma above tells us that we can apply
Carathéodory’s theorem for semi-algebras to the semi-algebra R and the
function λ. The resulting measure µ is the Lebesgue measure on R. Let us
sum up the essential points.
6.4. LEBESGUE MEASURE ON R 17

Theorem 6.4.4 The Lebesgue measure µ is the unique, completed Borel

measure on R such that µ(I) = |I| for all intervals I.

Proof: By construction, the µ-measure of all half-open intervals is equal

to their length, and by continuity of measure (Proposition 5.1.5), the same
must hold for open and closed intervals. By Proposition 6.3.3, if ν is any
other completed, Borel measure with the same property, µ(A) = ν(A) for
all Borel sets such that µ(A) < ∞. But the same must hold for a general
Borel set A as we can slice it up into pieces of finite measure:
X
X

µ(A) = µ A ∩ (n, n + 1] = ν A ∩ (n, n + 1] = ν(A)
n∈Z n∈Z

The next result tells us that measurable sets can be approximated arbi-
trarily well by open and closed set.

Proposition 6.4.5 Assume that A ⊂ R is a measurable set. For each  > 0,

there is an open set G ⊃ A such that µ(G \ A) < , and a closed set F ⊂ A
such that µ(A \ F ) < .

Proof: We begin with the open sets. Assume first that A has finite measure.
Then for every  > 0 there is a covering {Cn } of A by half-open rectangles
Cn = (an , bn ] such that
∞
X 
|Cn | < µ(A) +
2
n=1

If we replace the half-open intervals Cn = (aSn , bn ] by the open intervals


Bn = (an , bn + 2n+1 ), we get an open set G = ∞
n=1 Bn containing A with

∞ ∞ 
X X  
µ(G) ≤ µ(Bn ) = |Cn | + < µ(A) + 
2n+1
n=1 n=1

and hence
µ(G \ A) = µ(G) − µ(A) < 
by Lemma 5.1.4c).
If µ(A) is infinite, we slice A into pieces of finite measure An = A ∩
(n, n + 1] for all n ∈ Z, and use what we have already proved to find
S an open

set Gn such that An ⊂ Gn and µ(Gn \ An ) < 2n+2S. Then G = n∈Z Gn is
an open set which contains A, and since G \ A ⊂ n∈Z (Gn \ An ), we get
X X 
µ(G \ A) ≤ µ(Gn \ An ) < < ,
2n+2
n∈Z n∈Z
18 CHAPTER 6. CONSTRUCTING MEASURES

proving the statement about approximation by open sets.

To prove the statement about closed sets, just note that if we apply
the first part of the theorem to Ac , we get an open set G ⊃ Ac such that
µ(G \ Ac ) < . This means that F = Gc is a closed set such that F ⊂ A,
and since A \ F = G \ Ac , we have µ(A \ F ) < . 2

Another important property of the Lebesgue measure is that it is trans-

lation invariant — if we move a set a distance to the left or to the right, it
keeps its measure. To formulate this mathematically, let E be a subset of R
and a a real number, and write

E + a = {e + a | e ∈ E}

for the set we obtain by moving all points in E a distance a.

Proposition 6.4.6 If E ⊂ R is measurable, so is E + a for all a ∈ R, and

µ(E + a) = µ(E).

Proof: We shall leave this to the reader (see Exercise 7). The key observa-
tion is that µ∗ (E + a) = µ∗ (E) holds for outer measure since intervals keep
their length when we translate them. 2

One of the reasons why we had to develop the rather complicated ma-
chinery of σ-algebras, is that we can not in general expect to define a measure
on all subsets of our measure space X — some sets are just so complicated
that they are nonmeasurable. We shall now take a look at such a set. Before
we begin, we need to modify the notion of translation so that it works on
the interval [0, 1). If x, y ∈ [0, 1), we first define

.
 x+y if x + y ∈ [0, 1)
x+y=
x + y − 1 otherwise


If E ⊂ [0, 1) and y ∈ [0, 1), let

. .
E + y = {e + y | e ∈ E}
.
Note that E + y is the set obtained by first translating E by y and then
moving the part that sticks out to the right of [0, 1) one unit to left so that
it fills
.
up the empty part of [0, 1). It follows from translation
.
invariance that
E + y is measurable if E is, and that µ(E) = µ(E + y).

Example 1: A nonmeasurable set. We start by introducing an equiva-

lence relation ∼ on the interval [0, 1):

x ∼ y ⇐⇒ x − y is rational
6.4. LEBESGUE MEASURE ON R 19

Next, we let E be a set obtained by picking one .

element from each equiva-
lence class.5 We shall work with the sets E + q for all rational numbers q
in the interval [0, 1), i.e., for all q ∈ Q̂ = Q .∩ [0, 1). .
First observe that if q1 6= q2 , then (E + q1 ) ∩ (E .+ q2 ) = ∅. If not,. we
could write the common element x as . both x = .e1 + q1 and x = e2 + q2
for some e1 , e2 ∈ E. The equality e1 + q1 = e2 + q2 , implies that e1 − e2
is rational, and by definition of E this is only possible if e1 = e2 . Hence
q1 = q2 , contradicting the assumption that S q1 6= q2..
The next observation is that [0, 1) = q∈Q̂ (E + q). To see this, pick an
arbitrary x ∈ [0, 1). By definition, there is an e in E that belongs to the
same equivalence
.
class as x, i.e. such that q = x − e is rational. If q .∈ [0, 1),
then x ∈ E + q, if q < 0 (the onlySother possibility),
.
we have x ∈ E + (q +1)
(check this!). In either case, x ∈ q∈Q̂ (E + q).
Assume for contradiction
.
that E is Lebesgue measurable. .
Then, as al-
ready observed, E + q is Lebesgue measurable with µ(E + q) = µ(E) for
q ∈ Q̂. But by countable additivity
X .
µ([0, 1)) = µ(E + q)
q∈Q̂

and since µ([0, 1)) = 1, this is impossible — a sum of countable many, equal,
nonnegative numbers is either ∞ (if the numbers are positive) or 0 (if the
numbers are 0).
Note that this argument works not only for the Lebesgue measure, but
for any (non-zero) translation invariant measure on R. This means that it is
impossible to find a translation invariant measure on R that makes all sets
measurable. ♣

The existence of nonmeasurable sets is not a surprise – there is no reason

to expect that all sets should be measurable – but it is a nuisance which
complicates many arguments. As we have seen in this chapter, the hard
part is often to find the right class of measurable sets and prove that it is a
σ-algebra.

Exercises for Section 6.4

1. Complete the proof of Proposition 6.4.2.
2. Prove the “finite case” of Lemma 6.4.3.
P∞
3. In the proof of Lemma 6.4.3, explain why n=1 |In | ≥ b − a − 2 = |I| − 2.
c
4. Explain that A \ F = G \ A at the end of the proof of Proposition 6.4.5.
5
Here we are using a principle from set theory called the Axiom of Choice which allows
us to make a new set by picking one element from each set in an infinite family. It has
been proved that it is not possible to construct a nonmeasurable subset of R without using
a principle of this kind.
20 CHAPTER 6. CONSTRUCTING MEASURES

5. A subset of R is called a Gδ -set if it is the intersection of countably many

open sets, and it is called a Fσ -set if it is union of countably many closed set.
a) Explain why all Gδ - and Fσ -sets are measurable.
b) Show that if A ⊂ R is measurable, there is a Gδ -set G such that A ⊂ G
and µ(G \ A) = 0.
c) Show that if A ⊂ R is measurable, there is a Fσ -set F such that F ⊂ A
and µ(A \ F ) = 0.
6. Assume that A ⊂ R is a measurable set with finite measure. Show that for
every  > 0, there is a compact set K ⊂ A such that µ(A \ K) < .
7. Prove Proposition 6.4.6 by:
a) Showing that if E ⊂ R and a ∈ R, then µ∗ (E + a) = µ(E).
b) Showing that if E ⊂ R is measurable, then so is E + a for any a ∈ R.
c) Explaining how to obtain the proposition from a) and b).
8. Check that the equivalence relation in Example 1 really is an equivalence
relation.
9. If A is a subset of R and r is a positive, real number, let

rA = {ra | a ∈ A}

Show that if A is measurable, then so is rA and µ(rA) = rµ(A).

10. Use Proposition 6.4.6 to prove that that if E ⊂ [0, 1) is Lebesgue measurable,
. .
then E + y is Lebesgue measurable with µ(E + y) = µ(E) or all y ∈ [0, 1).

6.5 The coin tossing measure

As a second example of how to construct measures, we shall construct the
natural probability measure on the space of infinite sequences of unbiased
coin tosses (recall Example 8 of Section 5.1). Be aware that this section is
rather sketchy – it is more like a structured sequence of exercises than an
ordinary section. The results here will not be used in the sequel.
Let us recall the setting. The underlying space Ω consists of all infinite
sequences
ω = (ω1 , ω2 , . . . , ωn , . . .)
where each ωn is either H (for heads) or T (for tails). If a = a1 , a2 , . . . , an
is a finite sequence of H’s and T’s, we let

Ca = {ω ∈ Ω | ω1 = a1 , ω2 = a2 , . . . , ωn = an }

and call it the cylinder set generated by a. We call n the length of Ca . Let
R be the collection of all cylinder sets (of all possible lengths).

Lemma 6.5.1 R is a semi-algebra.

6.5. THE COIN TOSSING MEASURE 21

Proof: The intersection of two cylinder sets Ca and Cb is either equal to one
of them (if one of the sequences a, b is an extension of the other) or empty.
The complement of a cylinder set is the union of all other cylinder sets of
the same length. 2

We define a function λ : R → [0, 1] by putting

1
λ(Ca ) =
2n
where n is the length of Ca . (There are 2n cylinder sets of length n and they
correspond to 2n equally probable events).
We first check that λ behaves the right way under finite unions:
SN
Lemma 6.5.2 If A1 , A2 , . . . , AN are disjoint sets in R whose union n=1 An
belongs to R, then
[N XN
λ( An ) = λ(An )
n=1 n=1

Proof: Left to the reader (see Exercise 1). 2

To use Carathéodory’s theorem for semi-algebras we must extend the result

above to countable unions, i.e., we need to show that if {An } is a disjoint
sequence of cylinder sets whose union is a cylinder set, then
[ ∞
X
λ( An ) = λ(An )
n∈N i=1

The next lemma tells us that this condition is trivially satisfied because
the situation never occurs – a cylinder set is never a disjoint, countable
union of infinitely many cylinder sets! As this is the difficult part of the
construction, I spell out the details. The argument is actually a compactness
argument in disguise, and corresponds to Lemma 6.4.3 in the construction
of the Lebesgue measure.
Before we begin, we need some notation and terminology. For each
k ∈ N, we write ω ∼k ω̂ if ωi = ω̂i for i = 1, 2, . . . , k, i.e., if the first k coin
tosses in the sequences ω and ω̂ are equal. We say that a subset A of Ω is
determined at k ∈ N if whenever ω ∼k ω̂ then either both ω and ω̂ belong to
A or none of them do (intuitively, this means that you can decide whether ω
belongs to A by looking at the k first coin tosses). A cylinder set of length k
is obviously determined at time k. We say that a set A is finally determined
if it is determined at some k ∈ N.

Lemma 6.5.3 A cylinder set A is never an infinite, countable, disjoint

union of cylinder sets.
22 CHAPTER 6. CONSTRUCTING MEASURES

Proof: Assume for contradictionS that {An }n∈N is a disjoint sequence of

cylinder sets whose union A = n∈N An is also a cylinder set. Since the
An ’s are disjoint and nonempty, the set BN = A \ N
S
n=1 An is nonempty for
all N ∈ N. Note that the sets BN are finitely determined since A and the
An ’s are.
Since the sets BN are decreasing and nonempty, there must either be
strings ω = (ω1 , ω2 , . . .) starting with ω1 = H in all the sets BN or strings
starting with ω1 = T in all the sets BN (or both, in which case we just
choose H). Call the appropriate symbol ω̂1 . Arguing in the same way, we
see that there must either be strings starting with (ω̂1 , H) or strings start-
ing with (ω̂1 , T ) in all the sets BN (or both, in which case we just choose
H). Call the appropriate symbol ω̂2 . Continuing in this way, we get an
infinite sequence ω̂ = (ω̂1 , ω̂2 , ω̂3 , . . .) such that for each k ∈ N, there is a
sequence starting with (ω̂1 , ω̂2 , ω̂3 , . . . , ω̂k ) in each BN . Since BN is finitely
determined, this means that ω̂ ∈ BN for all N (just choose S k so large that
BN is determined at k). SBut this implies that ω̂ ∈ A \ n∈N An , which is a
contradiction since A = n∈N An by assumption. 2

We are now ready for the main result:

Theorem 6.5.4 There is a complete measure P on Ω such that P (A) =

λ(A) for all finitely determined sets A. The measure is unique on the σ-
algebra of measurable sets.

Proof: The three lemmas above give us exactly what we need to apply
Carathéodory’s theorem for semi-algebras. The uniqueness follows from
Proposition 6.3.3. The details are left to the reader. 2

Exercises for Section 6.5

1. Prove Lemma 6.5.2 (Hint: If the cylinder sets A1 , A2 , . . . , AN have length
K1 , K2 , . . . , KN , respectively, then this is really a statement about finite
sequences of length K = max{K1 , K2 , . . . , KN }.)
2. Fill in the details in the proof of Theorem 6.5.4.
3. Let {Bn }n∈N
T be a decreasing sequence of nonempty, finitely determined sets.
Show that n∈N Bn 6= ∅.
4. Let E = {ω ∈ Ω | ω contains infinitely many H’s}. Show that E is not
finitely determined.
5. Let H = 1, T = 0. Show that
n
1X 1
E = {ω ∈ Ω | lim ωi = }
n→∞ n 2
i=1

is not finitely determined.

6.5. THE COIN TOSSING MEASURE 23

6. Show that a set is in the algebra generated by R if and only if it is finitely

determined. Show that if E is a measurable set, then there is for any  > 0
a finitely determined set D such that P (E 4 D) < 

7. (You should do Exercise 6 before you attempt this one.) Assume that I is a
nonempty subset of N. Define an equivalence relation ∼I on Ω by

ω ∼I ω̂ ⇐⇒ ωi = ω̂i for all i ∈ I

We say that a set B ⊂ Ω is I-determined if whenever ω ∼I ω̂, then either

ω and ω̂ are both in B or both in B c (intuitively, this means that we can
determine whether ω belongs to B by looking at the coin tosses ωi where
i ∈ I).

a) Let A be the σ-algebra of all measurable sets, and define

AI = {A ∈ A | A is I-determined}

Show that AI is a σ-algebra.

b) Assume that A ∈ AI and that C is a finitely determined set such that
A ⊂ C. Show that there is a finitely determined Ĉ ∈ AI such that
A ⊂ Ĉ ⊂ C.
c) Assume that A ∈ AI . Show that for any  > 0, there is a finitely
determined B ∈ AI such that P (A 4 B) < .
d) Assume that I, J ⊂ N are disjoint. Show that if B ∈ AI and D ∈ AJ
are finitely generated, then P (B ∩ D) = P (B)P (D). In the language of
probability theory, B and D are independent events. (Hint: This is just
finite combinatorics and has nothing to do with measures. Note that
finitely determined sets are in the algebra generated by R, and hence
their measures are given directly in terms of λ.)
e) We still assume that I, J ⊂ N are disjoint. Show that if A ∈ AI and
C ∈ AJ , then P (A ∩ C) = P (A)P (C). (Hint: Combine c) and d).)
f) Let In = {n, n + 1, n + 2, . . .}. A set E ⊂ Ω is called a tail event if
E ∈ In for all n ∈ N. Show that the sets E in Exercises 4 and 5 are tail
events.
g) Assume that E is a tail event and that A is finitely generated. Show
that P (A ∩ E) = P (A)P (E).
h) Assume that E is a tail event, and let En be finitely determined sets
such that P (E 4 En ) < n1 (such sets exist by Exercise 6). Show that
P (E ∩ En ) → P (E) as n → ∞.
i) Show that on the other hand P (E ∩ En ) = P (En )P (E) → P (E)2 .
Conclude that P (E) = P (E)2 , which means that P (E) = 0 or P (E) =
1. We have proved Kolmogorov’s 0-1-law : A tail event can only have
probability 0 or 1.
24 CHAPTER 6. CONSTRUCTING MEASURES

6.6 Product measures

In calculus you learned how to compute double integrals by iterated inte-
gration: If R = [a, b] × [c, d] is a rectangle in the plane, then
ZZ Z d Z b  Z b Z d 
f (x, y) dxdy = f (x, y) dx dy = f (x, y) dy dx
R c a a c

There is a similar result in measure theory that we are now going to look
at. Starting with two measure spaces (X, A, µ) and (Y, B, ν), we shall first
construct a product measure space (X × Y, A ⊗ B, µ × ν) and then prove
that
Z Z Z  Z Z 
f d(µ × ν) = f (x, y) dµ(x) dν(y) = f (x, y) dν(y) dµ(x)

(I am putting in x’s and y’s to indicate which variables we are integrat-

ing with respect to). The guiding light in the construction of the product
measure µ × ν is that we want it to satisfy the natural product rule

µ × ν(A × B) = µ(A)ν(B)

for all A ∈ A and all B ∈ B (think of the formula for the area of an ordinary
rectangle).
As usual, we shall apply Carathéodory’s theorem for semi-algebras. If
we define measurable rectangles to be subsets of X × Y the form R = A × B,
where A ∈ A and B ∈ B, we first observe that the class of all such sets form
a semi-algebra.

Lemma 6.6.1 The collection R of measurable rectangles is a semi-algebra.

Proof: Observe first that since

(R1 × S1 ) ∩ (R2 × S2 ) = (R1 ∩ R2 ) × (S1 ∩ S2 )

the intersection of two measurable rectangles is a measurable rectangle.

Moreover, since

(A × B)c = (Ac × B c ) ∪ (Ac × B) ∪ (A × B c ) (6.6.1)

the complement of measurable rectangle is a finite, disjoint union of mea-

surable rectangles, and hence R is a semi-algebra. 2

The next step is to define a function λ : R → R+ by

λ(A × B) = µ(A)ν(B)
6.6. PRODUCT MEASURES 25

To use Carathéodory’s theorem for semi-algebras, we need to show that if

[
A×B = (Cn × Dn )
n∈N
P∞
is a disjoint union, then λ(A × B) = n=1 λ(Cn × Dn ), or in other words
∞
X
µ(A)ν(B) = µ(Cn )ν(Dn ) (6.6.2)
n=1

(note that in this case, ∅ ∈ R and there

R is no need to treat finite
R sums sepa-
rately). Observe that since µ(C) = 1C (x)dµ(x) and ν(D) = 1D (y)dν(y),
we have Z Z
µ(C)ν(D) = 1C (x)dµ(x) 1D (y)dν(x) =
Z Z  Z Z 
= 1C (x)1D (y)dµ(x) dν(x) = 1C×D (x, y)dµ(x) dν(x)
S
for any two sets C ∈ A and D ∈ B. If A × B = n∈N (Cn × Dn ) is a disjoint
union, the Monotone Convergence Theorem 5.5.6 thus tells us that
∞
X N
X
µ(Cn )ν(Dn ) = lim µ(Cn )ν(Dn ) =
N →∞
n=1 n=1

N Z Z
X 
= lim 1Cn ×Dn (x, y)dµ(x) dν(x)
N →∞
n=1
N
Z "Z X #
= lim 1Cn ×Dn (x, y)dµ(x) dν(x)
N →∞
n=1
Z " N
Z X #
= lim 1Cn ×Dn (x, y)dµ(x) dν(x)
N →∞
n=1
Z "Z N
#
X
= lim 1Cn ×Dn (x, y)dµ(x) dν(x)
N →∞
n=1
∞
Z "Z X #
= 1Cn ×Dn (x, y)dµ(x) dν(x)
n=1
Z Z 
= 1C×D (x, y)dµ(x) dν(x) = µ(C)ν(D)

which proves equation (6.6.2).

We are now ready to prove the main theorem. Remember that a measure
space is σ-finite if it is a countable union of sets of finite measure.
26 CHAPTER 6. CONSTRUCTING MEASURES

Theorem 6.6.2 Assume that (X, A, µ) and (Y, B, ν) are two measure spaces
and let A ⊗ B be the σ-algebra generated by the measurable rectangles A × B,
A ∈ A, B ∈ B. Then there exists a measure µ × ν on A ⊗ B such that
µ × ν(A × B) = µ(A)ν(B) for all A ∈ A, B ∈ B
If µ and ν are σ-finite, this measure is unique and is called the product
measure of µ and ν.

Proof: The existence follows from Theorem 6.3.6 as formula (6.6.2) guaran-
tees that conditions (i) and (ii) hold (as ∅ = ∅ × ∅ ∈ R, we need not treat
finite and countable sums separately in this case). The uniqueness follows
from Proposition 6.3.3 (see also Exercise 6.3.1c)). 2

Product measures can be used to construct Lebesgue measure in higher

dimension. If µ is Lebesgue measure on R, the completion of the product
µ × µ is the Lebesgue measure on R2 . To get Lebesgue measure on R3 , take
a new product (µ × µ) × µ and complete it. Continuing in this way, we get
Lebesgue measure in all dimensions.

Exercises for Section 6.6

1. Check that formula (R1 × S1 ) ∩ (R2 × S2 ) = (R1 ∩ R2 ) × (S1 ∩ S2 ) in the
proof of Lemma 6.6.1 is correct.
2. Check that formula (6.6.1) is correct and that the union is disjoint.
3. Assume that µ is the counting measure on N. Show that µ × µ is counting
measure on N2 .
4. Show that any open set in Rd is a countable union of open boxes of the form
(a1 , b1 ) × (a2 , b2 ) × . . . × (ad , bd )
where a1 < b1 , a2 < b2 , . . . , ad < bd (this can be used to show that the
Lebesgue measure on Rd is a completed Borel measure).
5. In this problem we shall generalize Proposition 6.4.5 from R to R2 . Let µ be
the Lebesgue integral on R and let λ = µ × µ.
a) Show that if D, E are open sets in R, then D × E is open in R2 .
b) Assume that A × B is a measurable rectangle with µ(A), µ(B) < ∞.
Show that for any  > 0 there are open sets D, E ⊂ R such that
A × B ⊂ D × E and λ(E × D) − λ(A × B) < .
c) Assume that Z ⊂ R2 is measurable with λ(Z) < ∞. Show that for any
 > 0, there is an open set G ⊂ R2 such that Z ⊂ G and λ(G)−λ(Z) < .
Explain why this mean that λ(G \ Z) < .
d) Assume that Z ⊂ R2 is measurable. Show that for any  > 0, there is
an open set G ⊂ R2 such that Z ⊂ G and λ(G \ Z) < .
e) Assume that Z ⊂ R2 is measurable. Show that for any  > 0, there is
a closed set F ⊂ R2 such that that Z ⊃ F and λ(Z \ F ) < .
6.7. FUBINI’S THEOREM 27

6.7 Fubini’s Theorem

In this section we shall see how we can integrate with respect to a product
measure; i.e. we shall prove the formulas
Z Z Z  Z Z 
f d(µ × ν) = f (x, y) dµ(x) dν(y) = f (x, y) dν(y) dµ(x) (6.7.1)

mentioned in the previous section. As one might expect, these formulas do

not hold for all measurable functions, and part of the challenge is to find
the right conditions. We shall prove two theorems; one (Tonelli’s Theorem)
which only works for nonnegative functions, but doesn’t need any additional
conditions; and one (Fubini’s Theorem) which works for functions taking
both signs, but which has integrability conditions that might be difficult to
check. Often the two theorems are used in combination – we use Tonelli’s
Theorem to show that the conditions for Fubini’s Theorem are satisfied.
We have to begin with some technicalities. For formula (6.7.1) to make
sense, the functions x 7→ f (x, y) and y 7→ f (x, y) we get by fixing one of
the variables in the function f (x, y), have to be measurable. To simplify
notation, we write f y (x) for the function x 7→ f (x, y) and fx (y) for the
function y 7→ f (x, y). Similarly for subsets of X × Y :
E y = {x ∈ X | (x, y) ∈ E} and Ex = {y ∈ Y | (x, y) ∈ E}
These sets and functions are called sections of f and E, respectively (make
a drawing).

Lemma 6.7.1 Assume that (X, A, µ) and (Y, B, ν) are two measure spaces,
and let A ⊗ B be the product σ-algebra.
(i) For any E ∈ A ⊗ B, we have E y ∈ A and Ex ∈ B for all y ∈ Y and
x ∈ X.
(ii) For any A ⊗ B-measurable f : X × Y → R, the sections f y and fx are
A- and B-measurable, respectively, for all y ∈ Y and x ∈ X.

Proof: I only prove the lemma for the y-sections, and leave the x-sections
to the readers.
(i) Since A ⊗ B is the smallest σ-algebra containing the measurable rect-
angles, it clearly suffices to show that
C = {E ⊂ X × Y | E y ∈ A}
is a σ-algebra containing the measurable rectangles. That the measurable
rectangles are in C, follows from the observation

 A if y ∈ B
(A × B)y =
∅ if y ∈/B

28 CHAPTER 6. CONSTRUCTING MEASURES

To show that C is closed under complements, just note that if E ∈ C, then

E y ∈ A, and hence (E c )y = (E y )c ∈ A (check this!) which means that
E c ∈ C. Similarly for countable
S unions: yIf forSeach n ∈y N, En ∈ C, then
y
(En ) ∈ A for all n,
S and hence n∈N (En ) = ( n∈N En ) ∈ A (check this!)
which means that n∈N En ∈ C.
(ii) We need to check that (f y )−1 (I) ∈ A for all intervals of the form
[−∞, r). But this follows from (i) and the measurability of f since

(f y )−1 (I) = (f −1 (I))y

(check this!) 2

There is another measurability problem in formula (6.7.1): We need to

know that the integrated sections
Z Z
y 7→ f (x, y) dµ(x) = f y (x) dµ(x)

and Z Z
x 7→ f (x, y) dν(y) = fx (y) dν(y)

are measurable. This is a more complicated question, and we shall need

a quite useful and subtle result known as the Monotone Class Theorem.
A monotone class of subsets of a set Z is a collection M of subsets of Z
that is closed under increasing countable unions and decreasing countable
interesections. More precisely:
S
(i) If E1 ⊂ E2 ⊂ . . . ⊂ En ⊂ . . . are in M, then n∈N ∈ M
T
(ii) If E1 ⊃ E2 ⊃ . . . ⊃ En ⊃ . . . are in M, then n∈N ∈ M

All σ-algebras are monotone classes, but a monotone class need not be a σ-
algebra. If R is a collection of subsets of Z, there is (by the usual argument)
a smallest monotone class containing R. It is called the monotone class
generated by R.

Theorem 6.7.2 (Monotone Class Theorem) Assume that Z is a nonempty

set and that A is an algebra of subsets of Z. Then σ-algebra and the mono-
tone class generated by A coincide.

Proof: Let C be the σ-algebra and M the monotone class generated by A.

Since all σ-algebras are monotone classes, we must have M ⊂ C.
To prove the opposite inclusion, we show that M is a σ-algebra. Observe
that it suffices to prove that M is an algebra as closure under countable
unions will then take care of itself: If {En } is a sequence from M, the
sets Fn = E1 ∪ E2 ∪ . . . ∪ En are in M since M is an algebra, and hence
6.7. FUBINI’S THEOREM 29

S S
n∈N En n∈N Fn ∈ M since M is closed under increasing countable
=
unions.
To prove that M is an algebra, we use a trick. For each M ∈ M define

M(M ) = {F ∈ M | F \ M, F \ M, F ∩ M ∈ M}

It is not hard to check that since M is a monotone class, so is M(M ). Note

also that by symmetry, N ∈ M(M ) ⇐⇒ M ∈ M(N ).
Our aim is to prove that M(M ) = M for all M ∈ M. This would mean
that the intersection and difference between any two sets in M are in M,
and since Z ∈ M (because Z ∈ A ⊂ M), we may conclude that M is an
algebra.
To show that M(M ) = M for all M ∈ M, pick a set A ∈ A. Since
A is an algebra, we have A ⊂ M(A). Since M is the smallest monotone
class containing A, this means that M(A) = M, and hence M ∈ M(A) for
all M ∈ M. By symmetry, A ∈ M(M ) for all M ∈ M. Since A was an
arbitrary element in A, we have A ⊂ M(M ) for all M ∈ M, and using the
minimality of M again, we see that M(M ) = M for all M ∈ M. 2

The advantage of the Monotone Class Theorem is that it is often much

easier to prove that families are closed under monotone unions and intersec-
tions than under arbitrary unions and intersections, especially when there’s
a measure involved in the definition. The next lemma is a typical case. Note
the σ-finiteness condition; Exercise 9 shows that the result does not always
hold without it.

Lemma 6.7.3 Let (X, A, µ) and (Y, B, ν) be two σ-finite measure spaces,
and assume that E ⊂ X × Y is A ⊗ B-measurable. Then the functions
x 7→ ν(Ex ) and y 7→ µ(E y ) are A- and B-measurable, respectively, and
Z Z
ν(Ex ) dµ(x) = µ(E y ) dν(y) = µ × ν(E)

Proof: We shall prove the part about the x-sections Ex and leave the (sim-
ilar) proof for y-sections to the readers. We shall first carry out the proof
for finite measure spaces, i.e. we assume that µ(X), ν(Y ) < ∞.
Let
R
C = {E ⊂ X×Y | x 7→ ν(Ex ) is A-measurable and ν(Ex ) dµ(x) = µ × ν(E)}

If we can show that C is a monotone class containing the algebra generated

by the measurable rectangles, Monotone Class Theorem will tell us that C
is a σ-algebra, and hence contains A ⊗ B (which is the smallest σ-algebra
containing the measurable rectangles). This obviously suffices to prove the
finite case of the theorem.
30 CHAPTER 6. CONSTRUCTING MEASURES

To show that any measurable rectangle E = A × B belongs to C, just

observe that 
 ν(B) if x ∈ A
ν(Ex ) =
0 if x ∈
/A

R R
and that ν(Ex ) dµ(x) = A ν(B) dµ = µ(A)ν(B) = µ × ν(E).
A set F in the algebra
Sn generated by the measurable rectangles, is a
disjoint
Pn union F = i=1 i of measurable rectangles, and since ν(Fx ) =
E
i=1 ν((Ei )x ), the function x 7→ ν(F
R x ) is A-measurable
R Pn (a sum of measur-
able
Pn functions is measurable) and ν(Fx ) dµ(x) = i=1 ν((Ei )x )dµ(x) =
i=1 µ × ν(Ei ) = µ × ν(F ). Hence F ∈ C.
To show that C is monotoneSclass, assume that {En } is anSincreasing
sequence of sets in C. Let E = n∈N En , and note that Ex = ∞ n=1 (En )x .
By continuity of measure, ν(Ex ) = limn→∞ ν((En )x ), and hence x 7→ ν(Ex )
is measurable as the limit of a sequence of measurable functions. Moreover,
by the Monotone Convergence Theorem and continuity of measures,
Z Z Z
ν(Ex ) dµ(x) = lim ν((En )x ) dµ(x)) = lim ν((En )x ) dµ(x)) =
n→∞ n→∞

= lim µ × ν(En ) = µ × ν(E)

n→∞

This means that E ∈ C.

We must also check monotone intersections. T Assume that {En } is a
decreasing
T∞ sequence of sets in C. Let E = n∈N n , and note that Ex =
E
n=1 (E )
n x . By continuity of measure (here we are using that ν is a finite
measure), ν(Ex ) = limn→∞ ν((En )x ), and hence x 7→ ν(Ex ) is measurable.
Moreover, using the Dominated Convergenge Theorem (since the measure
space is finite, we can use the function that is constant 1 as the dominating
function), we see that
Z Z Z
ν(Ex ) dµ(x) = lim ν((En )x ) dµ(x)) = lim ν((En )x ) dµ(x)) =
n→∞ n→∞

= lim µ × ν(En ) = µ × ν(E)

n→∞

and hence E ∈ C.
Since this shows that C is a monotone class and hence a σ-algebra con-
taining A×B, we have proved the lemma for finite measure spaces. To extend
it to σ-finite spaces, let {Xn } and {Yn } be increasing
S sequence ofSsubsets
of X and Y of finite measure such that X = n∈N Xn and Y = n∈N Yn .
If E is a A ⊗ B-measurable subset of X × Y , it follows from what we have
already proved that x 7→ ν((E ∩ (Xn × Yn )x ) is measurable and
Z
ν((E ∩ (Xn × Yn )x ) dµ(x) = µ × ν(E ∩ (Xn × Yn ))
6.7. FUBINI’S THEOREM 31

The lemma for σ-finite spaces now follows from the Monotone Convergence
Theorem and continuity of measures. 2

Remark: The proof above illustrates a useful technique. To prove that all
sets in the σ-algebra F generated by a family R satisfies a certain property
P , we prove
(i) All sets in R has property P .
(ii) The sets with property P form a σ-algebra G.
Then F ⊂ G (since F is the smallest σ-algebra containing R), and hence all
sets in F has property P .

We are now ready to prove the first of our main theorems.

Theorem 6.7.4 (Tonelli’s Theorem) Let (X, A, µ) and (Y, B, ν) be two

σ-finite measure spaces, and assume that f : X × Y → RR+ is a nonnegative,
A ⊗ RB-measurable function. Then the functions x 7→ f (x, y) dν(y) and
y 7→ f (x, y)dµ(x) are A- and B-measurable, respectively, and
Z Z Z  Z Z 
f d(µ × ν) = f (x, y) dν(y) dµ(x) = f (x, y) dµ(x) dν(y)

Proof: We prove the first equality and leave the second to the reader. Notice
first that if f = 1E is an indicator function, then by the lemma
Z Z Z Z 
1E d(µ × ν) = µ × ν(E) = ν(Ex ) dµ(x) = 1E (x, y) dν(y) dµ(x)

which proves the theorem for indicator functions. By linearity, it also holds
for nonnegative, simple functions.
For the general case, let {fn } be an increasing sequence ofR nonnegative
simple functions converging
R pointwise to f . The functions x 7→ fn (x, y) dν(y)
increase to x 7→ f (x, y) dν(y) by the Monotone Convergence Theorem.
Hence the latter function is measurable (as the limit of a sequence of mea-
surable functions), and using the Monotone Convergence Theorem again,
we get Z Z
f (x, y) d(µ × ν) = lim fn (x, y) d(µ × ν) =
n→∞
Z Z  Z  Z 
= lim fn (x, y) dν(y) dµ(x) = lim fn (x, y) dν(y) dµ(x) =
n→∞ n→∞
Z Z  Z Z 
= lim fn (x, y) dν(y) dµ(x) = f (x, y) dν(y) dµ(x)
n→∞
2

Fubini’s Theorem is now an easy application of Tonelli’s Theorem.

32 CHAPTER 6. CONSTRUCTING MEASURES

Theorem 6.7.5 (Fubini’s Theorem) Let (X, A, µ) and (Y, B, ν) be two

σ-finite measure spaces, and assume that f : X × Y → R is µ × ν-integrable.
Then the functions fx andR f y are integrable for almost
R all x and y, and the
integrated functions x 7→ f (x, y) dν(y) and y 7→ f (x, y)dµ(x) are µ- and
ν-integrable, respectively. Moreover,
Z Z Z  Z Z 
f d(µ × ν) = f (x, y) dν(y) dµ(x) = f (x, y) dµ(x) dν(y)

Proof: Since f is integrable, it splits as the difference f = f+ − f− between

two nonnegative, integrable functions. Applying Tonelli’s Theorem to |f |,
we get
Z Z  Z Z 
|f (x, y)| dν(y) dµ(x) = |f (x, y)| dµ(x) dν(y) =

Z
= |f | d(µ × ν) < ∞

whichR implies the integrability statements for fx , f y , x 7→ f (x, y) dν(y) and

R

y 7→ f (x, y)dµ(x). Applying Tonelli’s Theorem to f+ and f− separately,

we get
Z Z Z  Z Z 
f+ d(µ × ν) = f+ (x, y) dν(y) dµ(x) = f+ (x, y) dµ(x) dν(y)

and
Z Z Z  Z Z 
f− d(µ × ν) = f− (x, y) dν(y) dµ(x) = f− (x, y) dµ(x) dν(y)

and subtracting the second from the first, we get Fubini’s Theorem. 2

Remark: The integrability condition in Fubini’s Theorem is occasionally a

nuisance: The natural way to show that f is integrable, is by calculating the
iterated integrals, but this presupposes that f is integrable! The solution is
often first to apply Tonelli’s TheoremR to the absolute value |f |, and use the
iterated integrals there to show that |f | d(µ × ν is finite. This means that
f is integrable, and we are ready to apply Fubini’s Theorem.

Even when the original measures µ and ν are complete, the product
measure µ × ν rarely is. A natural response is to take it’s completion µ × ν
(as we did with higher dimensional Lebesgue measures), but the question
is if Fubini’s theorem still holds. This is not obvious since there are more
µ × ν-measurable functions than µ × ν-measurable ones, but fortunately the
answer is yes. I just state the result without proof (see Excercise 11).
6.7. FUBINI’S THEOREM 33

Theorem 6.7.6 Let (X, A, µ) and (Y, B, ν) be two complete, σ-finite mea-
sure spaces, and assume that f : X × Y → R is µ × ν-measurable, where
µ × ν is the completion of the product measure. Then the functions fx
and fRy are measurable for almost
R all x and y, and the integrated functions
x 7→ f (x, y) dν(y) and y 7→ f (x, y)dµ(x) are measurable as well. More-
over
(i) (Tonelli’s Theorem for Completed Measures) If f is nonnega-
tive,
Z Z Z  Z Z 
f dµ × ν = f (x, y) dν(y) dµ(x) = f (x, y) dµ(x) dν(y)

(ii) (Fubini’s Theorem for Completed Measures) If f is integrable,

the functions fx and f y areR integrable for almost allR x and y, and the
integrated functions x 7→ f (x, y) dν(y) and y 7→ f (x, y)dµ(x) are
µ- and ν-integrable, respectively. Moreover,
Z Z Z  Z Z 
f dµ × ν = f (x, y) dν(y) dµ(x) = f (x, y) dµ(x) dν(y)

Exercises for Section 6.7

1. Show that (E c )y = (E y )c (here c i referring to complements).
2. Show that ( n∈N En )y = n∈N Eny .
S S

3. Show that (f y )−1 (I) = (f −1 (I))y .

4. Show that
M = {M ⊂ R | 0 ∈ M }
is a monotone class, but not a σ-algebra.
5. Show that the sets M(M ) in the proof of the Monotone Class Theorem really
are monotone classes.
6. In this problem µ Lebesgue measure on R, while ν is counting measure on
N. Let λ = µ × ν be the product measure and let

f :R×N→R

be given by
1
f (x, n) =
1 + (2n x)2
1
R R
Compute f dλ. Remember that 1+u2 du = arctan u + C.
x−y
7. Let f : [0, 1] × [0, 1] → R be defined by f (x, y) = (x+y) 3 (the expression

doesn’t make sense for x = y = 0, and you may give the function whatever
value you want at that point). Show by computing the integrals that
Z 1 Z 1 
1
f (x, y) dx dy = −
0 0 2
34 CHAPTER 6. CONSTRUCTING MEASURES

and Z 1 Z 1 
1
f (x, y) dy dx =
0 0 2
1 2y
(you may want to use that f (x, y) = x2 +y 2 − (x+y)3 in the first integral and

argue by symmetry in the second one). Let µ be the Lebesgue measure on

[0, 1] and λ = µ × µ. Is f integrable with respect to λ?
2
(1+y 2 )
8. Define f : [0, ∞) × [0, ∞) by f (x, y) = xe−x .
a) Show by performing the integrations that
Z ∞ Z ∞ 
π
f (x, y) dx dy =
0 0 4
R ∞ R ∞  π
b) Use Tonelli’s theorem to show that 0 0
f (x, y) dy dx = 4
c) Make the substitution u = xy in the inner integral of
Z ∞ Z ∞  Z ∞ Z ∞ 
2 2
f (x, y) dy dx = xe−x (1+y ) ) dy dx
0 0 0 0

R ∞ R ∞ R 2
∞ 2
f (x, y) dy dx = 0 e−u du .

and show that 0 0
R∞ √
−u2 π
d) Conclude that 0
e du = 2 .

9. Let X = Y = [0, 1], let µ be the Lebesgue measure on X (this is just the
restriction of the Lebesgue measure on R to [0, 1]) and let ν be
R the counting
measure. Let E = {(x, y) ∈ X × Y |x = y}, and show that ν(Ex ) dµ(x),
µ(E y ) dν(y), and µ × ν(E) are all different (compare lemma 6.7.3).
R

10. Assume that X = Y = N and that µ = ν is the counting measure. Let

f : X × Y → R be defined by


 1 if x = y



f (x, y) = −1 if x = y + 1




0 otherwise


R R R 
ShowR that
R f dµ×ν = ∞, but that the iterated integrals f (x, y) dµ(x) dν(u)
and f (x, y) dν(y) dµ(x) are both finite, but unequal.
11. In this exercise, we shall sketch the proof of Theorem 6.7.6, and we assume
that (X, A, µ) and (Y, B, ν) are as in that theorem.
a) Assume that E ∈ A ⊗ B and that µ × ν(E) = 0. Show that µ(E y ) = 0
for ν-almost all y and that ν(Ex ) = 0 for µ-almost all x.
b) Assume that N is an µ × ν-null set, i.e. there is an E ∈ A ⊗ B such
that N ⊂ E and µ × ν(E) = 0. Show that for ν-almost all y, N y is
µ-measurable and µ(N y ) = 0. Show also that for µ-almost all x, Nx is
ν-measurable and ν(Nx ) = 0. (Here you need to use that the original
measure spaces are complete).
6.7. FUBINI’S THEOREM 35

c) Assume that D ⊂ X × Y is in the completion of A ⊗ B with respect to

µ × ν. Show that for ν-almost all y, Dy is µ-measurable, and that for
µ-almost all x, Dx is ν-measurable (use that by Theorem 5.2.5 D can
be written as a disjoint union D = E ∪ N , where E ∈ A ⊗ B and N is
a null set).
d) Let D be as above. Show that the functions x 7→ ν(Dx ) and y 7→
µ(Dy ) are A- and B-measurable, respectively (define ν(Dx ) and µ(Dy )
arbitrarily on the sets of measure zero where Dx and Dy fail to be
measurable), and show that
Z Z
ν(Dx ) dµ(x) = µ(Dy ) dν(y) = µ × ν(D)

e) Prove Theorem 6.7.6 (this is just checking that the arguments that got
us from Lemma 6.7.3 to Theorems 6.7.4 and 6.7.5, still works in the
new setting).