Solutions to Homework 4

1. Let X be a topological space and A ⊂ X. The closure of A, denoted

Ā, is the intersection of all closed sets containing A.

(a) Show that Ā is the smallest closed subset of X containing A, in

the following sense: if A ⊂ F ⊂ X and F is closed, then Ā ⊂ F .
Solution: By definition, Ā is the intersection of all closed sets
containing A: that is, x ∈ Ā if and only if x ∈ F for all closed
sets F ⊂ X with A ⊂ F . Thus if F ⊂ X is some closed set with
A ⊂ F , then for all x ∈ Ā we have x ∈ F ; thus Ā ⊂ F .
It may be worth remarking that Ā is a closed set containing A: it’s
an intersection of closed sets, hence is closed, and it’s an intersec-
tion of sets containing A, hence contains A. (If x ∈ A then x ∈ F
for all closed sets F ⊂ X with A ⊂ F , so x is in the intersection
of all such F , which is Ā.)

(b) Show that if A ⊂ B ⊂ X then Ā ⊂ B̄.

Solution: We have A ⊂ B ⊂ B̄, so B̄ is a closed set containing
A, so by part (a) we get Ā ⊂ B̄.

(c) Show that A ∪ B = Ā ∪ B̄.

Solution: We have A ⊂ A ∪ B, so Ā ⊂ A ∪ B by part (b), and
similarly B̄ ⊂ A ∪ B, so Ā ∪ B̄ ⊂ A ∪ B.
For the reverse inclusion, we have A ⊂ Ā ⊂ Ā ∪ B̄, and B ⊂ B̄ ⊂
Ā ∪ B̄, so A ∪ B ⊂ Ā ∪ B̄. But Ā ∪ B̄ is a union of two closed sets,
hence is closed, so by part (a) we get A ∪ B ⊂ Ā ∪ B̄.

1
 (d) Show that A ∩ B ⊂ Ā ∩ B̄. Give an example where the inclusion
is strict.
Solution: We have A ∩ B ⊂ A ⊂ Ā, and A ∩ B ⊂ B ⊂ B̄, so
A ∩ B ⊂ Ā ∩ B̄. The latter is an intersection of closed sets, hence
is closed, so by part (a) we get A ∩ B ⊂ Ā ∩ B̄.
For the counterexample, let X = R with the usual topology, let
A = (0, 1), and let B = (1, 2). Then A ∩ B = ∅, which is closed,
so A ∩ B = ∅. On the other hand, Ā = [0, 1], and B̄ = [1, 2], so
Ā ∩ B̄ = {1}.

10 . Let X be a topological space and A ⊂ X. The interior of A, denoted

int A, or sometimes A◦ , is the union of all open sets contained in A.

(a) Show that int A is the biggest open subset of A, in the following
sense: if U ⊂ A and U is open, then U ⊂ int A.
Solution: By definition, int A is the union of all open sets con-
tained A: that is, x ∈ int A if and only if x ∈ U for some open set
U ⊂ A. Thus if U ⊂ X is some open set with U ⊂ A, then for all
x ∈ U we have x ∈ int A; thus U ⊂ int A.
It may be worth remarking that int A is an open set contained in
A: it’s a union of open sets, hence is open, and it’s a union of sets
contained in A, hence is contained in A. (If x ∈ int A then x ∈ U
for some open set U ⊂ A, so x ∈ A.)
(b) Show that if A ⊂ B ⊂ X then int A ⊂ int B.
Solution: We have int A ⊂ A ⊂ B, so int A is an open set
contained in B, so by part (a) we get int A ⊂ int B.
(c) Show that int A ∩ int B = int(A ∩ B).
Solution: We have A ∩ B ⊂ A, so int(A ∩ B) ⊂ int A by part (b),
and similarly int(A ∩ B) ⊂ int B; thus int(A ∩ B) ⊂ int A ∩ int B.

For the reverse inclusion, we have int A ∩ int B ⊂ int A ⊂ A,

and int A ∩ int B ⊂ int B ⊂ B, so int A ∩ int B ⊂ A ∩ B. But
int A ∩ int B is an intersection of two open sets, hence is open, so
by part (a) we get int A ∩ int B ⊂ int(A ∩ B).

2
 (d) Show that int A ∪ int B ⊂ int(A ∪ B). Give an example where the
inclusion is strict.
Solution: We have int A ⊂ A ⊂ A ∪ B, and int B ⊂ B ⊂ A ∪ B,
so int A ∪ int B ⊂ A ∪ B. Now int A ∪ int B is a union of open sets,
hence is open, so by part (a) we get int A ∪ int B ⊂ int(A ∪ B).
For the counterexample, let X = R with the usual topology, let
A = (0, 1], and let B = [1, 2). Then int A = (0, 1), and int B =
(1, 2), so int A∪int B = (0, 1)∪(1, 2). On the other hand, A∪B =
(0, 2) is open, so int(A ∪ B) = (0, 2).

2. Let X be a topological space and A ⊂ X.

(a) Show that X \ Ā = int(X \ A), and X \ int A = X \ A.

Solution: It may be clearer to use the notation Ac := X \ A for
the complement.
We first remark that for any subsets B, C ⊂ X, we have B ⊂ C if
and only if C c ⊂ B c ; and (B c )c = B. You can prove these if you
want.
Next, we have A ⊂ Ā, so (Ā)c ⊂ Ac ; because Ā is closed, (Ā)c is
open, so
(Ā)c ⊂ int(Ac )
by 10 (a).
Replacing A with Ac , we get (Ac )c ⊂ int((Ac )c ) = int A, so
(int A)c ⊂ Ac .
Thus we’ve proved one inclusion for each of the two desired equal-
ities. For the reverse inclusions, we have int A ⊂ A, so Ac ⊂
(int A)c ; because int A is open, (int A)c is closed, so
Ac ⊂ (int A)c .
Replacing A with Ac , we get Ā ⊂ (int(Ac ))c , so
int(Ac ) ⊂ (Ā)c
by 1(a).
Thus both equalities are proved.

3
4. Let X be a set, and let T be the set of subsets U ⊂ X that such that
X \ U is finite, together with the empty set.

(a) Show that T is a topology. (It is called the “finite complement

topology.”)
Solution: We have ∅ ∈ T by definition. To see that X ∈ T , note
that X \ X = ∅ is finite.
Suppose that U1 , . . . , Un ∈ T . If some Ui = ∅ then U1 ∩ · · · ∩ Un =
∅ ∈ T . Otherwise X \ Ui is finite for all i, so X \ (U1 ∩ · · · ∩ Un ) =
(X \ U1 ) ∪ · · · ∪ (X \ Un ) is a finite union of finite sets, hence is
finite, so U1 ∩ · · · ∩ Un ∈ T .
S
Given an arbitrary subset S ⊂ST , we want to show that S ∈ T .
If S = ∅ or S = {∅} then S = ∅ ∈ T . Otherwise there S is
someTnon-empty U ∈ S, so X \ U is finite. Then U ⊂ S, so
S \ S ⊂ X ∩ U is a subset of a finite set, hence is finite, so
X
S ∈ T.
(Or it might be nicer to show that the closed sets do what they
should.)

(b) Find the closure, interior, and boundary of Z as a subset of R in

the finite complement topology.
Solution: A set is open if and only if its complement is finite, or
it is empty. Thus a set is closed if and only if it is finite, or is the
whole space R.
Now Z is infinite, so the only closed set containing Z is R, so
Z̄ = R. Similarly, R \ Z is infinite, so R \ Z = R, so int Z = ∅ by
3(a). Finally, ∂Z = Z̄ \ int Z = R \ ∅ = R.

5
5. Let T be the set of subsets U ⊂ R such that U contains 0, together
with the empty set.

(a) Show that T is a topology.

Solution: We have ∅ ∈ T by definition, and R ∈ T because
0 ∈ R. .
Suppose that U1 , . . . , Un ∈ T . If some Ui = ∅ then U1 ∩ · · · ∩ Un =
∅ ∈ T . Otherwise 0 ∈ Ui for all i, so 0 ∈ U1 ∩ · · · ∩ Un , so
U1 ∩ · · · ∩ Un ∈ T . (Indeed, we can see that T is closed under
arbitrary intersections, not just finite intersections.)
S
Given an arbitrary subset SS⊂ T , we want to show that S ∈ T .
If S = ∅ or S = {∅} then S = ∅ S ∈ T . Otherwise
S there is some
non-empty U ∈ S, so 0 ∈ U , so 0 ∈ S, so S ∈ T .

(b) Find the closure, interior, and boundary of the one-point subsets
{1} and {0}.
Solution: A set is open if and only if it either contains 0, or is
empty. Thus a set is closed if and only if it either does not contain
0, or is the whole space R.
Thus {1} is closed, and it contains no non-empty open set, so its
interior is ∅, its closure is {1}, and its boundary is {1}, just as in
the usual topology.
On the other hand, {0} is open, and it is not contained in any
closed set apart from the whole space R, so its interior is {0}, its
closure is R, and its boundary is (−∞, 0) ∪ (0, ∞), quite unlike
the usual topology.

6
6. Let T be the subsets of R of the form (a, ∞) for some a ∈ R, together
with the empty set and the whole set R.

(a) Show that T is a topology. (It is called the “lower semi-continuous

topology” and we discussed it in lecture on Friday.)
Solution: We have ∅ ∈ T and R ∈ T by definition.
For finite intersections, by induction it is enough to prove it for
two: if U, V ∈ T then U ∩ V ∈ T . Observe moreover that for any
two subsets we either have U ⊂ V , in which case U ∩ V = U ∈ T ,
or V ⊂ U , in which case U ∩ V = V ∈ T .
S
Given an arbitrary
S subset S ⊂ T , we want to show that S ∈ T.
If R ∈ S then S = R ∈ T . If not, write

S = {(a, ∞) : a ∈ A}

or
S = {(a, ∞) : a ∈ A} ∪ {∅}
for some set
S A of real numbers. If A is non-empty and bounded be-
low then SS= (inf A, ∞) ∈ T . If A is non-empty
S and unbounded
below then S = R ∈ T . If A is empty then S = ∅ ∈ T .

(b) Find the closure, interior, and boundary of the interval (0, 1) as a
subset of R in this topology.
Solution: The closed sets are of the form (−∞, a] for a ∈ R,
together with the whole set R and the empty set.
A closed set containing (0, 1) is either all of R, or of the form
(−∞, a] for some a ≥ 1. The closure of (0, 1) is the intersection of
all these, which is (−∞, 1]. The only open set contained in (0, 1)
is ∅, so its interior is ∅. Thus its boundary is (−∞, 1].