Solutions to Real Analysis 2nd Edition,

G.B.Folland
Chapter 1. Measures (Part 2)
Leun Kim
([email protected])
Abstract
This solution may include some mistakes. If you find some errors or have a question, send me
a mail as written above.

Exercise 17.SIf µ∗ is an outer measure on X and {Aj }∞ ∗

1 is a sequence of disjoint µ -measurable sets,
∞ ∞
then µ∗ (E ∩ 1 Aj ) = 1 µ∗ (E ∩ Aj ) for any E ⊂ X.
P
Theorem, we already have µ∗ (F ) = ∞
P ∗ ∗
S∞ c
Solution. By Caratheodory’s S j=1 µ (F ∩ Aj ) + µ (F ∩ ( j=1 ) )
for any F ⊂ X. Put F = E ∩ ∞ j=1 Aj ⊂ X, then

∞
! ∞ ∞
! ∞
! ∞
!c !
[ X [ [ [
µ∗ E ∩ Aj = µ∗ E ∩ Aj ∩ Aj + µ∗ E ∩ Aj ∩ Aj
j=1 j=1 j=1 j=1 j=1

∞
X ∞
X
∗ ∗
= µ (E ∩ Aj ) + µ (φ) = µ∗ (E ∩ Aj ).
j=1 j=1

Exercise 18. Let A ⊂ P(X) be an algebra, Aσ the collection of countable unions of sets in A , and
Aσδ the collection of countable intersections of sets in Aσ . Let µ0 be a premeasure on A and µ∗ the
induced outer measure.
a. For any E ⊂ X and  > 0 there exists A ⊂ Aσ with E ⊂ A and µ∗ (A) 6 µ∗ (E) + .
b. If µ∗ (E) < ∞, then E is µ∗ -measurable iff there exists B ∈ Aσδ with E ⊂ B and µ∗ (B \ E) = 0.
c. If µ0 is σ-finite, the restriction µ∗ (E) < ∞ in (b) is superfluous.
Solution. S∞ S∞
a. Let {Aj }∞1 ⊂ A be a countable collection such that E ⊂ j=1 Aj . Define A = j=1 Aj ∈ Aσ . Then
(∞  ∞
) ∞
X  [ X
µ∗ (A) = inf µ0 (Aj )A ⊂ Aj = µ0 (Aj )


j=1 j=1 j=1

and (∞ 
∞
)
X  [
µ∗ (E) = inf µ0 (Aj )E ⊂ Aj = inf {µ∗ (A)|E ⊂ A} .


j=1 j=1

Therefore, for any  > 0, there exists A such that µ∗ (A) < µ∗ (E) + .
b. (i) Assume that E is µ∗ -measurable and µ∗ (E) < ∞. By Exercise 18.a,T∞ there exists {Bj }1 ⊂ Aσ
∞

such that E ⊂ Bj and µ (Bj ) 6 µ (E) + 1/j for any j. Define B = j=1 Bj . Then B ∈ Aσδ and
∗ ∗

E ⊂ B. Thus we have
∞
!
\
µ∗ (E) 6 µ∗ (B) = µ∗ Bj 6 µ∗ (Bj ) 6 µ∗ (E) + 1/j
j=1

which means that µ∗ (E) = µ∗ (B). And because E is µ∗ -measurable,

µ∗ (B) = µ∗ (B ∩ E) + µ∗ (B ∩ E c ) = µ∗ (E) + µ∗ (B \ E).

 2

Because µ∗ (E) < ∞, we have µ∗ (B \ E) = µ∗ (B) − µ∗ (E) = 0.

(ii) Assume that there exists B ∈ Aσδ with E ⊂ B and µ∗ (B \E) = 0. By Proposition 1.13, every set in
A is µ∗ -measurable. But also the union of a countable collection of µ∗ -measurable sets is measurable
which means that every set in Aσ is µ∗ -measurable. Also the intersection of a countable collection
of µ∗ -measurable sets is µ∗ -measurable which means that every set in Aσδ is µ∗ -measurable. Thus
B ∈ Aσδ is µ∗ -measurable. Now choose any F ⊂ X. Then F \ E ⊂ (F \ B) ∪ (B \ E) so that
µ∗ (F ∩ E c ) 6 µ∗ (F \ B) + µ∗ (B \ E) = µ∗ (F ∩ B c ). Therefore
µ∗ (F ∩ E) + µ∗ (F ∩ E c ) 6 µ∗ (F ∩ B) + µ∗ (F ∩ B c ) = µ∗ (F )
which means that E is µ∗ -measurable.
c. Because µ0 is σ-finite, X = ∞ j=1 Xj where Xj ∈ A and µ0 (Xj ) < ∞ for all j. And S because Xj ∈ A ,
S
µ0 (Xj ) = µ (Xj ) < ∞. Define Ej = E ∩ Xj , then µ (Ej ) 6 µ (Xj ) < ∞. Note that ∞
∗ ∗ ∗
j=1 Ej = E. By
Exercise 18.a, for any  > 0, there exists Aj ∈ A such that Ej ⊂ Aj S and µ (Aj ) 6 µ∗ (Ej ) + /2j+1 .
∗

Because µ (Ej ) < ∞, we have µ (Aj \ Ej ) 6 /2 . Define B0 = ∞

∗ ∗ j+1
j=1 Aj so that B0 ∈ Aσ and
E ⊂ B0 . Thus we have
∞
! ∞ ∞ ∞
∗ ∗
[ X
∗
X
∗
X  
µ (B0 \ E) = µ (Aj \ E) 6 µ (Aj \ E) 6 µ (Aj \ Ej ) 6 j+1
= <
j=1 j=1 j=1 j=1
2 2

1 ∈ Aσ such that E ⊂ Bj and

which means that µ∗ (B0 \ E)T= 0. Thus we can take a sequence {Bj }∞
∞
µ (Bj \ E) = 0. Define B = j=1 Bj , then E ⊂ B and B ∈ Aσδ so that µ∗ (B \ E) 6 µ∗ (Bj \ E) = 0.
∗

Exercise 19. Let µ∗ be an outer measure on X induced from a finite premeasure µ0 . If E ⊂ X, define
the inner measure of E to be µ∗ (E) = µ0 (X) − µ∗ (E c ). Then E is µ∗ -measurable iff µ∗ (E) = µ∗ (E).
(Use Exercise 18.)
Solution.
nP Let A ⊂ P(X) be an algebra o and µ0 be an finite premeasure on A . Note that µ (E) =
∗

∞ ∞
j=1 µ0 (Aj )|Aj ∈ A , E ⊂
S
inf j=1 Aj .
(i) Assume that E is µ∗ -measurable. Then for any F ⊂ X, µ∗ (F ) = µ∗ (F ∩ E) + µ∗ (F
P∞ ∩ E c ). Particu-
∗ ∗ ∗ c ∗ c ∗ ∗ ∗
S∞= µ (E) + µ (E ) so that µ (E ) = µ (X) − µ (E). And µ (X) = inf{ j=1 µ0 (Aj )|Aj ∈
larly µ (X)
A , X ⊂ j=1 Aj } = µ0 (X) when we take one of Aj = X ∈ A . Therefore we have
µ∗ (E) = µ0 (X) − µ∗ (E c ) = µ∗ (X) − (µ∗ (X) − µ∗ (E)) = µ∗ (E).
(ii) Suppose that µ∗ (E) = µ∗ (E). Then µ0 (X) − µ∗ (E c ) = µ∗ (E), µ∗ (E) + µ∗ (E c ) = µ0 (X) < ∞. Thus
both µ∗ (E) and µ∗ (E c ) are finite. By Exercise 18.b, there exists B1 , B2 ∈ Aσδ with E ⊂ B1 and E c ⊂ B2
such that µ∗ (E) = µ∗ (B1 ) and µ∗ (E c ) = µ∗ (B2 ). Note that µ0 (X) = µ∗ (E) + µ∗ (E c ) = µ∗ (B1 ) + µ∗ (B2 )
so that µ∗ (B1 ) = µ0 (X) − µ∗ (B2 ) = µ∗ (B2c ). Because B2c ⊂ E ⊂ B1 , we have
µ∗ (B1 \ E) 6 µ∗ (B1 \ B2c ) = µ∗ (B1 ) − µ∗ (B2c ) = 0
which means that E is µ∗ -measurable by Exercise 18.b.

∗ ∗ ∗
 ∗
Exercise 20. Let µ be an outer measure on X, M the σ-algebra of µ -measurable sets, µ = µ  ,
M∗
and µ the outer measure induced by µ as in (1.12) (with µ and M replacing µ0 and A ).
+ ∗

a. If E ⊂ X, we have µ∗ (E) 6 µ+ (E), with equality iff there exists A ∈ M∗ with A ⊃ E and
µ∗ (A) = µ∗ (E).
b. If µ∗ is induced from a premeasure, then µ∗ = µ+ . (Use Exercise 18a.)
c. If X = {0, 1}, there exists anPouter measure µ∗ on X such that µ∗ 6= µ+ .
∞ ∞
Solution. Note that µ+ (E) = inf{ j=1 µ(Aj )|Aj ∈ M∗ and E ⊂ j=1 Aj }.
S
S∞
a. (i) µ∗ (E) 6 µ+ (E) for any E ⊂ X : Choose any {Aj }∞ ∗
1 ⊂ M such that E ⊂ j=1 Aj . Note that
such {Aj } exists because we can take Aj = X for all j to be the worst case. Then we have
∞ ∞ ∞
!
#1 X #2
X #3 [ #4
µ+ (E) > µ(Aj ) = µ∗ (Aj ) > µ∗ Aj > µ∗ (E) (1)
j=1 j=1 j=1
 3


Here, #1 holds by the definition of µ ; #2 holds because µ = µ  ; #3 holds because µ∗ is an outer
+ ∗

M∗
measure; and #4 holds because E ⊂ ∞
S
A
j=1 j .
(ii) Assume that there exists A ∈ M∗ such that E ⊂ A and µ∗ (A) = µ∗ (E). Our goal is to prove that
µ∗ (E) > µ∗ (E). But we have
µ∗ (E) = µ∗ (A) > µ+ (E)
Here, the inequality holds because E ⊂ A, A ∈ M∗ , and of the definition of µ+ . S∞
(iii) Assume µ∗ (E) > µ+ (E). Choose any {Aj }∞ ∗
1 ⊂ M such that E ⊂ A. Define A = j=1 Aj so
∗ ∗ +
that A ∈ M . But because µ (E) > µ (E), S all the inequalities in the equation (1) are thus equalities.
Therefore we immediately have µ (E) = µ ( ∞
∗ ∗ ∗
j=1 Aj ) = µ (A).
b. Because µ (E) 6 µ (E) for any E ⊂PX, we claim that µ+ (E) 6Sµ∗ (E). Let A be an algebra such
∗ +

that A ⊂ M∗ . Note that µ∗ (E) = inf{ ∞ j=1 µ0 (Aj )|Aj ∈ A , E ⊂

∞
j=1 Aj } for any E ⊂ X. Then by
Exercise 18.a, for any  > 0, there exists A ∈ Aσ ⊂ M with E ⊂ A and µ∗ (A) 6 µ∗ (E) + . Because
∗

A ∈ M∗ and E ⊂ A, by the definition of µ+ , we have

µ+ (E) 6 µ∗ (A) 6 µ∗ (E) + .
Here,  is arbitrary, µ+ (E) 6 µ∗ (E).
c. If X = {0, 1}, then P(X) = {φ, {0}, {1}, {0, 1}}. Define µ∗ : P(X) → [0, ∞] by µ∗ (φ) =
0, µ∗ ({0}) = 1/2, µ∗ ({1}) = µ∗ ({0, 1}) = 1. Then clearly µ∗ is an outer measure on X. And
P∞easily confirm that only φ and {0,
we can S1} are µ∗ -measurable. By the definition of µ+ , µ+ (E) =
∞
inf{ j=1 µ(Aj )|Aj ∈ {φ, {0, 1}} and E ⊂ j=1 Aj }. If we take E = {0} ⊂ X, then
µ+ (E) = µ({0, 1}) = µ∗ ({0, 1}) = 1 6= 1/2 = µ∗ (E)
which gives a suitable example.

Exercise 21. Let µ∗ be an outer measure induced from a premeasure and µ the restriction of µ∗
to the µ∗ -measurable sets. Then µ is saturated. (Use Exercise 18.)
Solution.
 Let A ⊂ P(X) be an algebra and M∗ be a collection of µ∗ -measurable sets. Then

µ = µ∗  . Clearly µ is a measure on M∗ so that (X, M∗ , µ) is a measure space. Let M g∗ be the
M∗

g∗ ⊂ M∗ .
collection of all locally measurable sets. Our claim is that M
Let E ∈ M g∗ . Then for any A ∈ M∗ with µ(A) = µ∗ (A) < ∞, E ∩ A ∈ M∗ . · · · (∗)
Now, we have to prove that E is µ∗ -measurable. By Exercise 18.a, for any F ⊂ X and  > 0, there
exists A0 ∈ Aσ with F ⊂ A0 and µ∗ (A0 ) 6 µ∗ (F ) + . · · · (∗∗)
Note that A0 ∈ M∗ and we don’t have to consider the case that µ∗ (F ) = ∞, so assume that µ∗ (F ) < ∞
so that µ∗ (A0 ) < ∞. Then by (∗) and (∗∗),
µ∗ (F ) +  > µ∗ (A0 )
= µ∗ (A0 ∩ (E ∩ A0 )) + µ∗ (A0 ∩ (E ∩ A0 )c )
= µ∗ (A0 ∩ E) + µ∗ (A0 ∩ E c )
> µ∗ (F ∩ E) + µ∗ (F ∩ E c )
Here  is arbitrary, we have µ∗ (F ) > µ∗ (F ∩ E) + µ∗ (F ∩ E c ) which means that E is µ∗ -measurable
where µ∗ (F ) < ∞ so that E ∈ M∗ .

Exercise 22. Let (X, M, µ) be a measure space, µ∗ the outer  measure induced by µ according to

(1.12), M∗ the σ-algebra of µ∗ -measurable sets, and µ = µ∗  .
M∗
a. If µ is σ-finite, then µ is the completion of µ. (Use Exercise 18.)
b. In general, µ is the saturation of the completion of µ. (See Exercises
P∞16 and 21.) S∞
1 ∗ ∗
Solution. First, note that µ : P(X) → [0, ∞] is defined by µ (E) = inf{ j=1 µ(Aj )|E ⊂ j=1 Aj , Aj ∈
1
Reference : G.Harris, Math 7200 Solution to Assignment 6, p.3
 4

M}. Define N = {N ∈ M|µ(N  ) = 0} and M = {E ∪ F |E ∈ M and F ⊂ N for some N ∈ N }. By


Proposition 1.13, we have µ∗  = µ which means that if E ∈ M, then µ∗ (E) = µ(E) and also E ∈ M∗ .
M
∗
a. (I) M = M.
(i) M ⊂ M∗ .
Let E ∪ F ∈ M, then E ∈ M, and F ⊂ N for some N ∈ N . Then for any A ⊂ X,

µ∗ (A ∩ (E ∪ F )) + µ∗ (A ∩ (E ∪ F )c ) = µ∗ ((A ∩ E) ∪ (A ∩ F )) + µ∗ (A ∩ E c ∩ F c )
6 µ∗ (A ∩ E) + µ∗ (A ∩ F ) + µ∗ (A ∩ E c )
6 µ∗ (A ∩ E) + µ∗ (A ∩ E c ) + µ∗ (F )
6 µ∗ (A ∩ E) + µ∗ (A ∩ E c ) + µ∗ (N ) (∵ F ⊂ N )
= µ∗ (A ∩ E) + µ∗ (A ∩ E c ) + µ(N ) (∵ N ∈ M)
= µ∗ (A ∩ E) + µ∗ (A ∩ E c )
= µ∗ (A) (∵ E ∈ M ⇒ E ∈ M∗ )

which means that E ∪ F is µ∗ -measurable so that E ∪ F ∈ M∗ .

(ii) M∗ ⊂ M.
Assume that G ∈ M∗ . Then Gc ∈ M∗ . Because µ is σ-finite, by Exercise 18.b, there exists
E ∈ Mσδ = M such that Gc ⊂ E and µ∗ (E \ Gc ) = 0. Here, we can write G as G = (G \ E) ∪ (E \ Gc ).
Now we only have to confirm that G \ E ∈ M and E \ Gc ∈ M so that µ(E \ Gc ) = µ∗ (E \ Gc ) = 0.
But this is clear because Gc ∈ M∗ , Gc ⊂ E, E ∈ M and M ⊂ M∗ . Thus Gc ∈ M. Because M is a
σ-algebra, G \ E = G ∩ E c ∈ M and E \ Gc = E ∩ G ∈ M. Therefore G ∈ M.

(II) µ(E ∪ F ) = µ(E) for any E ∪ F ∈ M(= M∗ ).

Because F \ E ⊂ F ⊂ N for some N ∈ N , µ∗ (F \ E) = 0. Thus

µ(E ∪ F ) = µ∗ (E ∪ F ) = µ∗ (E ∪ (F \ E)) = µ∗ (E) + µ∗ (F \ E) = µ∗ (E) = µ(E).

By (I), (II), µ is the completion of µ if µ is σ-finite.

b. Note that E ⊂ X is locally µ̂-measurable if for any A ∈ M with µ̂(A) < ∞, E ∩ A ∈ M.

Define M
f be the collection of all locally µ̂-measurable sets. Here µ̂ : M → [0, ∞] denotes a completion
of µ. And by Exercise 16.c, define µ on M
2
f by

µ̂(E) if E ∈ M
µ2 (E) =
∞ otherwise.

f → [0, ∞] and µ : M∗ → [0, ∞] are equivalent.

Our goal is that µ2 : M
f = M∗ .
(1) M
f ⊂ M∗ : From part a.I.i, we have M ⊂ M∗ and if E ∈ M then µ∗ (E) = µ̂(E). Also if E ∈ M∗
(i) M
with µ∗ (E) < ∞, then E ∈ M by part a.I.ii.
f Then if A ∈ M∗ with µ∗ (A) < ∞ then A ∈ M. Because E ∈ M,
Let E ∈ M. f for any A ∈ M ⊂ M∗
with µ̂(A) = µ∗ (A) < ∞, E ∩ A ∈ M ⊂ M∗ which means that E ∈ M
g∗ = M∗ .(Note that M
g∗ = M∗
by Exercise 21.)
(ii) M∗ ⊂ M f : Take E ∈ M∗ and let A ∈ M with µ̂(A) < ∞. Because E ∈ M∗ and A ∈
M ⊂ M∗ , E ∩ A ∈ M∗ with µ∗ (E ∩ A) 6 µ∗ (A) = µ̂(A) < ∞. Because E ∩ A ∈ M∗ and
µ∗ (E ∩ A) < ∞, E ∩ A ∈ M which means that E ∈ M.
f
(2) µ = µ2 on M∗ = M.
f
 5

If E ∈ M then µ2 (E) = µ̂(E) = µ∗ (E). If E ∈ M∗ \ M, then µ2 (E) = ∞ = µ∗ (E). (∵ if E ∈ M∗ and

µ∗ (E) < ∞ then E ∈ M)

Exercise 23. Let A be the collection of finite unions of sets of the form (a, b] ∩ Q where −∞ 6
a < b 6 ∞.
a. A is an algebra on Q. (Use Proposition 1.7.)
b. The σ-algebra generated by A is P(Q).
c. Define µ0 on A by µ0 (φ) = 0 and µ0 (A) = ∞ for A 6= φ. Then µ0 is a premeasure on A , and
there is more than one measure on P(Q) whose restriction to A is µ0 .
Solution.
a. Let E = {φ, (a, b] ∩ Q} be a collection of subsets of Q. It is enough to prove that the finite disjoint
unions of members of E is an algebra. Here, E is an elementary family because (i) φ ∈ E, (ii) if
(a1 , b1 ], (a2 , b2 ] ∈ E then (a1 , b1 ] ∩ (a2 , b2 ] ∈ E, (iii) if (a, b] ∈ E then (a, b]c = (−∞, a] ∪ (b, ∞] is a
finite disjoint union of members of E. Thus by Proposition 1.7, the collection of finite disjoint unions
of members of E is an algebra which is equivalent to A is an algebra on Q.
b. (1) M(A ) ⊂ P(Q) : Clearly every countable union of (a, b] ∩ Q ⊂ Q and (a, b]c ∩ Q ⊂ Q, M(A ) ⊂
P(Q).
(2) P(Q) ⊂ M(A T ) : We only have to prove that 5 cases as below for any −∞ 6 a < b 6 ∞.
(i) {a} ∩ Q = ∞ n=1 ((a − 1/n, a) ∩ Q) ∈ M(A ).
(ii) [a, b] ∩ Q = ((a, b] ∪ {a}) ∩ Q = ((a, b] ∩ Q) ∪ ({a} ∩ Q) ∈ M(A ).
(iii) (a, b] ∩ Q ∈ M(A ).
(iv) (a, b) ∩ Q = ((a, b] \ {b}) ∩ Q = ((a, b] ∩ {b}c ) ∩ Q = ((a, b] ∩ Q) ∩ ({b}c ∩ Q) ∈ M(A ).
(v) [a, b) ∩ Q = ((a, b) ∪ {a}) ∩ Q = ((a, b) ∩ Q) ∪ ({a} ∩ Q) ∈ M(A ).
Therefore M(A ) = P(Q).
c. (1) µ0 is a premeasure on A . S∞
P∞ any disjoint {Aj }1 ⊂ A such that j=1 Aj ∈ A .S∞
∞
then µ0 ( ∞
S
Take If all Aj = φ, P j=1 Aj ) = 0 =
∞
j=1 0µ (A j ). If at least one of A j ’s is not empty, then µ 0 ( A
j=1 j ) = ∞ = µ
j=1 0 (A j ). Thus µ0 is a
premeasure on A .
(2) There is more than one measure on P(Q) whose restriction to A is µ0 . 

Clearly µ1 on P(Q) by µ1 (φ) = 0 and µ1 (E) = ∞ for E 6= φ is a measure on P(Q) such that µ1  = µ0 .
A
Define µ2 : P(Q) → [0, ∞] by

 0 if E = φ
µ2 (E) = n if E has n elements where n is finite value
∞ otherwise.


∞
S∞
Let
P∞ {E j } 1 ⊂ P(Q) be disjoint.
S∞ If one of Ej ’s
P∞has infinitely many elements, then µ 2 ( j=1 Ej ) = ∞ =
j=1 µ2 (Ej ). Otherwise, µ2 ( j=1 Ej ) = n = j=1 µ2 (Ej ) for some n ∈ N ∪ {0}. Thus µ2 is a measure

on P(Q). Clearly µ2  = µ0 , because A has no finite set by the density of Q.

A

Exercise 24. Let µ be a finite measure on (X, M), and let µ∗ be the outer measure induced by
µ. Suppose that E ⊂ X satisfies µ∗ (E) = µ∗ (X)(but not that E ∈ M).
a. If A, B ∈ M and A ∩ E = B ∩ E, then µ(A) = µ(B).
b. Let ME = {A ∩ E : A ∈ M}, and define the function ν on ME defined by ν(A ∩ E) = µ(A)
(which makes sense by (a)). Then ME is a σ-algebra on E and ν is a measure on ME .
Solution.
a. Note that µ∗ (E c ) = 0 because µ∗ (E c ) = µ∗ (X \ E) = µ∗ (X) − µ∗ (E) = 0. Then we have
µ∗ (A ∩ B c ) 6 µ∗ (A ∩ B c ∩ E) + µ∗ (A ∩ B c ∩ E c )
6 µ∗ (A ∩ E ∩ B c ) + µ∗ (E c )
= µ∗ (B ∩ E ∩ B c ) = µ∗ (φ) = 0
 6

which means that µ∗ (A ∩ B c ) = 0. Similarly,

µ∗ (B ∩ Ac ) 6 µ∗ (B ∩ Ac ∩ E) + µ∗ (B ∩ Ac ∩ E c )
6 µ∗ (B ∩ E ∩ Ac ) + µ∗ (E c )
= µ∗ (A ∩ E ∩ Ac ) = µ∗ (φ) = 0

which means that µ∗ (B ∩ Ac ) = 0. And because A, B ∈ M, by Proposition 1.13, A and B are

µ∗ -measurable and µ(A) = µ∗ (A), µ(B) = µ∗ (B). Thus

µ(A) = µ∗ (A) = µ∗ (A ∩ B) + µ∗ (A ∩ B c ) = µ∗ (A ∩ B) = µ∗ (B ∩ A) + µ∗ (B ∩ Ac ) = µ∗ (B) = µ(B).

b. (1) ME is a σ-algebra on E.
c c ∞
Let A ∩ E ∈ ME thenS∞A ∩ E ∈ ME because if A ∈ M then A ∈ M. Let {Aj ∩ E}1 ⊂ ME . Then
all Aj ∈ M so that j=1 Aj ∈ M because M is a σ-algebra. Thus,
∞
[ ∞
[
(Aj ∩ E) = E ∩ Aj ∈ ME .
j=1 j=1

(2) ν is a measure on ME .
(i) ν(φ) = ν(φ ∩ E) = µ(φ) = 0.
(ii) Let {Aj ∩ E}∞
1 ⊂ ME be disjoint. Then,

∞
! ∞
! ∞
! ∞ ∞
[ [ [ X X
ν (Aj ∩ E) = ν E ∩ Aj = µ Aj = µ(Aj ) = ν(Aj ∩ E).
j=1 j=1 j=1 j=1 j=1