0% found this document useful (0 votes)
14 views

Tutorial 8

Uploaded by

bhavikaai1729
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
14 views

Tutorial 8

Uploaded by

bhavikaai1729
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 22

Problem 1: Every compact metrizable space is second-countable.

Answer (2023MAS7123):
Let (X, d) be a compact metrizable space. Compact space means every open cover
has a finite subcover. Metrizable space means we can define a metric on it which
induces the topology of the set.
To show such set is second countable we need to prove that it has a countable basis
for its topology.
Firstly, for each n ∈ N, let us define:

An = {B(x, 1/n) : x ∈ X}
where B(x, 1/n) denotes the open ball with center x and radius 1/n. Each such
collection is an open cover of X. Since X is compact S so for each n ∈ N, ∃ finite
collection which covers X,say Bn ⊆ An . Let B = ∞ n=1 Bn ; B is countable as it is a
countable union of finite sets.
We claim that B forms a basis for metric topology on X.
1st property: Trivially satisfied as each Bn is an open cover of X so for every x ∈ X
we can find a basis element i.e. open set containing x.
2nd property: Let B(x1 , 1/n1 ), B(x2 , 1/n2 ) be two basis elements.
Let x ∈ B(x1 , 1/n1 ) ∩ B(x2 , 1/n2 ).
Since B(x1 , 1/n1 ), B(x2 , 1/n2 ) are open sets so their intersection is also open.
⇒ ∃ ϵ > 0 such that B(x, ϵ) ⊆ B(x1 , 1/n1 ) ∩ B(x2 , 1/n2 ) for each x ∈ B(x1 , 1/n1 ) ∩
B(x2 , 1/n2 ).
Now we need to find a basis element, say B s.t. x ∈ B ⊆ B(x1 , 1/n1 ) ∩ B(x2 , 1/n2 )
to prove second property.
Choose N ∈ N s.t. 1/N < ϵ. Since B2N covers X so ∃x3 ∈ X s.t. x ∈ B(x3 , 1/2N )
Now we show that B(x3 , 1/2N ) ⊆ B(x1 , 1/n1 ) ∩ B(x2 , 1/n2 ).
1
Let y ∈ B(x3 , 1/2N ) ⇒ d(x3 , y) < 2N .
Now

d(x, y) ≤ d(x, x3 ) + d(x3 , y)


1 1 1
< + =
2N 2N N
< ϵ

⇒ d(x, y) < ϵ ⇒ y ∈ B(x, ϵ) ⊆ B(x1 , 1/n1 ) ∩ B(x2 , 1/n2 )


⇒ B(x3 , 1/2N ) ⊆ B(x1 , 1/n1 ) ∩ B(x2 , 1/n2 )
Taking B = B(x3 , 1/2N ) for some x3 ∈ X, we are able to find for every x ∈
B(x1 , 1/n1 ) ∩ B(x2 , 1/n2 ) an open set B s.t. x ∈ B ⊆ B(x1 , 1/n1 ) ∩ B(x2 , 1/n2 ).
This proves second property of basis.
Hence B forms a basis and is countable.

Problem 2: Show that if X has a countable basis {Bn : n ∈ N}, then every basis C
contains a countable basis for X.

1
Answer (2023MAS7125): We are given that B = {Bn : n ∈ N} is a countable
basis.
For every pair of indices n, m ∈ N such that Bn ⊆ Bm , for which it is possible, choose
one set Cn,m ∈ C such that Bn ⊆ Cn,m ⊆ Bm .

Define C ′ = {Cn,m : n ∈ N} ⊆ C.Then C ′ is countable.


Now we need to show C ′ forms a basis for the topology on X.

Let U ⊆ X be an open set and x ∈ U . Since B is a basis, there exists m ∈ N


such that x ∈ Bm ⊆ U .
Since C is also a basis and Bm is open, there exists C ∈ C such that x ∈ C ⊆ Bm .
Since B is also a basis and C is open, there exists n ∈ N such that x ∈ Bn ⊆ C.

Now we have that there exists C ∈ C such that Bn ⊆ C ⊆ Bm .


Therefore, by the definition of C ′ , at least one of them is in C ′ .Hence, there exists a
set Cn,m ∈ C ′ such that x ∈ Cn,m ⊆ U .

Since U and x was chosen arbitrarily, we get that C ′ is a basis.(It is because for
a topological space X,if C is a collection of open sets of X such that for each open
set U of X and each x in U ,there is an element C of C such that x ∈ C ⊆ U ,then C
is a basis for the topology of X. )

Therefore, if X has countable basis {Bn : n ∈ N}, then every basis C contains a
countable basis for X.

Problem 3: If X is second countable then for any uncountable set has an uncountable
number of limit points in X.

Answer (2023MAS7126): Proof by contradiction - find an injection from the set


of all isolated points of A into the countable basis.

Let B be a countable basis for the topology on X.

Suppose that there are only countably many limit points of A (which means that
the set A′ is at most countable). Since A is uncountable, it follows that A \ A′ is
uncountable.

However, A \ A′ is the set of isolated points, so A has uncountably many isolated


points.

Let a ∈ A\A′ be any isolated point. Then there exists an open neighborhood Ua ⊆ X
of a such that Ua ∩ (A \ {a}) = ∅.

However, since Ua is open and B is a basis, there exists Ba ∈ B such that a ∈ Ba ⊆ Ua .

2
Notice that for a, b ∈ A \ A′ , a ̸= b, since Ba ∩ Bb = ∅ and a ∈ Ba , it follows that
a∈
/ Bb .

Hence, the function f : A \ A′ → B, given by a 7→ Ba , is an injection from an


uncountable set into B, which is a contradiction since B is countable.

Problem 4: Show that a metrizable separable space is second countable.

Answer (2023MAS7127): Given X is a separable metric space.

Claim : X is second countable.

proof : Since X is separable, ∃ a countable dense subset D(⊆ X) such that D̄ = X,


where D̄ is closure of D.

Let B(x, r) be an open ball such that r ∈ Q+ , x ∈ D

and B(x, r) = {y ∈ X|d(x, y) < r}.

Let β = collection of all open balls in X with rational radius and centered at points
in D.

Clearly β is countable as D and Q+ both are countable.

Let U be any open set in X. For x ∈ U , since D̄ = X, ∃ a point yx in D such that x


is in closure of B(yx , rx ), where rx ∈ Q+ .

A = {B(yx , rx )|x ∈ U }, A is an open cover for U since for each x ∈ U , ∃x ∈ B(yx , rx ).

As each B(yx , rx ) is an element of β =⇒ A is countable.

Thus U = ∪x∈U B(yx , rx ).

This shows that any open set U of X can be written as the union of countable elements
of β.

=⇒ β is a countable basis for the topology on X.

Hence by definition, X is second countable.

Problem 5: Show that a metrizable Lindelöf space is second countable.

Answer (2023MAS7128):

3
To prove that a metrizable Lindelöf space is second countable, we need to show that
there exists a countable base for the topology of the space.

Let (X, T ) be a metrizable Lindelöf space. This means that there exists a metric d
on X such that X is Lindelöf with respect to the metric topology induced by d. Now
for every n>0 ,

A = {Bd (x, n) | x ∈ X}

is open cover of X , as X is Lindelöf ,then their exist x1 , x2 , x3 , . . . such that

A′ = {Bd (xi , n) | x ∈ X, n > 0}

be a countable subcover of A , Consider n=1/m for all n ∈ N then

A′′ = {Bd (xi , n) | x ∈ X, n = 1/m, m ∈ N}

be also open cover and also countable union of countable set is countable so A” is
countable

Claim: A” is basis

Let U be an open set in (X,T), Now for all x ∈ U ,their exist m ∈ N such that n=1/m,

x ∈ Bd (x, n) ⊆ U

Thus A” is countable basis

Therefore, X is second countable.

Problem 6: Show that Rl and Io2 are not metrizable.

Answer (2023MAS7129): We will use the following result.


“Suppose X is metrizable and A ⊆ X. Let x ∈ X. If x ∈ Ā then there exists a
sequence (xn )∞
n=1 in A that converges to x.”

To show-Rl is not metrizable.

Let X = Rl and A = [0, 1). Then, Ā = [0, 1]

4
Suppose X is metrizable. Clearly, 1 ∈ Ā.
Using the above stated result, we can say that there exists a sequence (xn )∞
n=1 in A
that converges to 1.

Since, xn ∈ A ∀ n ∈ N

=⇒ xn < 1 ∀ n ∈ N (1)

Also, xn → 1 =⇒ ∃ an open ball centred at 1, say U ⊆ X, and N ∈ N such that


xn ∈ U ∀ n ≥ N (2)

Now, open sets in lower limit topology are of the form [a, b), a < b, so U is of form
[1, 1 + r) for some r > 0.
So, (2) =⇒ xn ≥ 1 ∀ n ≥ N , which is a contradiction to (1).

Therefore, there does not exists any sequence in A that converges to 1.

Hence, Rl is not metrizable.

To show I02 is not metrizable.


Consider I02 with dictionary order topology.
The topological space I02 is compact and thus has the least upper bound property.
So, each closed interval in I02 is compact.
(Using the result that each closed interval in an ordered set, having the least upper
bound property, is compact.)

Thus, every closed interval [a × b, c × d] in I02 is compact, where a × b ∈ I02 , c × d ∈ I02


and

a × b <I02 c × d
Consider, A = [0, 1] × (0, 1)
Claim-A is not Lindelöf.
A is the union of the disjoint sets Ux = x × (0, 1), x ∈ [0, 1], each of which is open in
A.

This collection of sets is uncountable, and no proper subcollection covers A.


Hence, A is not Lindelöf.

We know that every second-countable space is Lindelöf. But since A is not Lindelöf,
so A is not second-countable either.

5
And every subspace of a second-countable space is also second-countable.
Since the subspace A = [0, 1] × (0, 1) of the topological space I02 is not second-
countable, therefore, we can conclude that I02 is not second-countable.

=⇒ I02 has no countable basis.


Thus I02 is a compact space having no countable basis.

But every compact metrizable space has a countable basis and I02 is a compact space
that has no countable basis.

Therefore we can conclude that I02 is not metrizable.

Problem 7: Show that a countable product of separable spaces is also separable.

Answer (2023MAS7130): Let {Xi }i∈N be a sequence of separable spaces.


For each i ∈ N, let Di be the countable dense subset of Xi .

Now, we need to show that ∞


Q
i=1 Xi = X is separable.
Then, it would be sufficient to show that X has a countable dense subset.
Fix a point xn ∈ Dn for each n ∈ N.
Define,
Y
Em = {y ∈ Dn | yn = xn ; ∀n ≥ m};
n∈N
Y Y
= Dn × {xn }; (m ∈ N)
1≤n<m n≥m

Since each Di is countable and the finite Cartesian product of countable sets is count-
able,
Q
⇒ 1≤n<m Dn is countable.
⇒ Em is a countable
S subset of X, for each m ∈ N.
Define E = m∈N Em .
Since the countable union of countable sets is countable,
⇒ E is a countable subset of X.

Claim: E is dense Qin X.


Let y = (yi )i∈N ∈ i∈N Xi = X, and U be any open set in X, containing y.
Q −1
Then, there exists a basis element of the form , B = ( N
Q
i=1 Vi ) × ( i≥N Xi ),
such that y ∈ B ⊂ U .
Now for 1 ≤ i ≤ N − 1, Vi is open in Xi and yi ∈ Vi .
Since Di is dense in Xi for all i,
⇒ DQ i ∩ Vi ̸= ∅ for 1
Q≤N −1
i ≤ N − 1.
⇒ ( 1≤i<N Di ) ∩ ( i=1 Vi ) ̸= ∅.

6
Q Q  QN −1 Q 
⇒ 1≤i<N D i × i≥N {x i } ∩ i=1 V i × i≥N X i ̸= ∅.
⇒ (EN ∩ B) ̸= ∅.
Since EN ⊂ E and B ⊂ U ,
⇒ (EN ∩ B) ⊂ (E ∩ U ) ̸= ∅.
⇒ For every open set Q U in X containing y, E ∩ U ̸= ∅.
⇒ E is dense in X = i∈N Xi .
Thus, EQ is a countable dense subset of X.
⇒ X = i∈N Xi is separable.
Thus, the countable product of separable spaces is also separable.

Problem 8: Let f : X → Y be continuous. Show that if X is separable (resp.,


Lindelöf) then f (X) is also separable (resp., Lindelöf).

Answer (2023MAS7131):

To prove that if X is separable ⇒ f (X) is also separable.

Sol. Given X is separable, then ∃ A ⊂ X such that A is a countable dense subset of


X then A = X.
We claim to show that f (A) is a countable dense, subset of f (X).

(a) A is countable and f is continuous. So,

f (A) ⊆ f (X) and f (A) is countable  


(b) To show f (A) is dense in f (X). Note that f (A) ⊂ f A ⊂ f (A)
 
consider y ∈ f A then by continuity of f ∃x ∈ A such that f (x) = y, x ∈ A

Assume, V ⊂ Y , open in Y containing y

f −1 (V ) ⊂ X, open in X containing x

i.e. f −1 (V ) ∩ A ̸= ϕ

and from above V ∩ f (A) ̸= ϕ

SO f (A) is dense in Y ⇒ f (X) is separable.

If X is Lindelöf ⇒ f (X) is also Lindelöf.

Sol. Given X is Lindelöf, then by definition, let A be an open cover of X. Then


A′ ⊂ A be a countable sub collection of A that covers X.

7
Claim: f (X) is Lindelöf.
Let B be an open cover of f (X) then the set {f −1 (B) |B ∈ B} is an open cover of
X.
Since X is Lindelöf ⇒ We consider a countable sub cover
{f −1 (Bn ) : n ∈ N } ⊆ {f −1 (B) |B ∈ B}
Each B is open in f (X) ⊆ Y and {Bn } ⊆ B, n ∈ N is a countable collection.
{Bn } is a countable sub cover of B.
Also {f −1 (Bn ) : n ∈ N } covers X and
f is continuous ⇒ Bn , n ∈ N covers f (x) also.
Bn ⊆ B, countable collection covering f (X).

Problem 9: Let f : X → Y be a continuous open map. Show that if X is first


(resp., second) countable then f (X) is also first (resp., second) countable.

Answer (2023MAS7133): Let A = (SA , τA ) and B = (SB , τB ) be topological


spaces.

Let ϕ : A → B be a surjective open mapping which is also continuous.

Claim : If A is first-countable, then B is also first-countable.

Assume ϕ be surjective, continuous and open.

and A be first countable.

Let b ∈ SB .

Since ϕ is surjective there exists a ∈ SA such that:

ϕ(a) = b

From the first-countability of A, there is a countable local basis B, say, of a.

Let B = {Vn : n ∈ N}.

We need to show that {ϕ[Vn ] : n ∈ N} is a local basis for b.

Let U be an open set of B that contains b.

As b = ϕ(a), we have that:

a ∈ ϕ−1 [U ]

From the continuity of ϕ, we have that ϕ−1 [U ] is open.

8
As B is a local basis, there is an open set Vn ⊆ ϕ−1 [U ] such that a ∈ Vn .

Since ϕ is surjective

ϕ[ϕ−1 [U ]] = U

So, applying ϕ to Vn , from Image of Subset under Mapping is Subset of Image we


obtain:

ϕ[Vn ] ⊆ U such that b ∈ ϕ[Vn ].

This means that {ϕ[Vn ] : n ∈ N} is a local basis for b which is clearly countable.

Thus, B is first countable.

Now for second countability:

Let TA = (SA , τA ) and TB = (SB , τB ) be topological spaces.

Let ϕ : TA → TB be a surjective open mapping which is also continuous.

Claim : If TA is second-countable, then TB is also second-countable.

Assume ϕ to be surjective, continuous, and open, and let TA be second-countable.

By definition of second-countability, TA has a countable basis, B, say.

Let B = {Vn : n ∈ N}.

We need to show that {ϕ[Vn ] : n ∈ N} is a basis for TB .

Let U be an open set of TB .

Since ϕ is continuous, ϕ−1 [U ] is open in TA .

As B is a basis for TA , there exists an open set Vn ⊆ ϕ−1 [U ].

Since ϕ is surjective, ϕ[ϕ−1 [U ]] = U .

So, applying ϕ to Vn , from Image of Subset under Mapping is Subset of Image, we


obtain:

ϕ[Vn ] ⊆ U .

This means that {ϕ[Vn ] : n ∈ N} is a basis for TB that is clearly countable.

9
Thus, TB is second-countable.

Problem 10: Show that if X is Lindelöf and Y is compact then X × Y is Lindelöf.

Answer (2023MAS7135):

Definition of Lindelof space: X is said to be Lindelof space if every open cover of


X has a countable subcover.

The Tube Lemma: Consider the product space X × Y , where Y is compact. If N


is an open set of X × Y containing the slice {x0 } × Y of X × Y , then N contains some
tube W × Y about {x0 } × Y , where W is a neighborhood of x0 in X.

Let X and Y be topological spaces such that X is Lindelöf and Y is compact. Now
let A be an open covering of the product space X × Y .

Let x ∈ X. Then let Ax be the subcollection of A such that Ax covers {x}×Y , which
is compact being homeomorphic with the compact space Y . Therefore, some finite
subcollection of Ax also covers {x} × Y ; let one such subcollection be Ax,1 , Ax,2 , . . .,
Ax,n and then let us put
[nx
Sx = Ax,k .
k=1

Then Sx is an open set of X × Y such that Sx covers {x} × Y , and since Y is compact,
therefore by the tube lemma, there exists an open set Ux of x such that x ∈ Ux and
such that Sx also covers the tube Ux × Y .

In this way, we obtain an open covering {Sx1 , Sx2 , . . .} covers X × Y , therefore using
the above containment we can conclude that the collection

[
{Axj ,k : k = 1, 2, . . . , nxj }
j=1

is a countable subcollection of A that covers X × Y .

Thus, every open covering A of X × Y has a countable subcollection that also covers
X × Y , hence X × Y is Lindelöf.

Problem 11: Show that if X is normal, then every pair of disjoint closed sets have
neighborhoods whose closures are disjoint.

Answer (2023MAS7137): Let A and B be any two disjoint closed sets in a Normal
space X. Then ∃ disjoint open sets U and V of X such that A ⊂ U and B ⊂ V .

10
Lemma: X is Normal if and only if given a closed set A and an open set V containing
A such that V ⊂ U .

Proof of Lemma: Suppose that X is Normal and suppose that the set A and the
neighborhood U containing A is given. Let B = X − U ; then B is a closed set. By
hypothesis, ∃ disjoint open sets V and W containing A and B respectively. The set
V is disjoint from B. ∵ if y ∈ B, then W is a neighborhood of y disjoint from V , ∴
V ⊂ U , as required.

Suppose the set A and the closed set B not containing A are given. Let U = X − B.
By hypothesis, ∃ a neighborhood V of A such that V ⊂ U . The open sets V and
X − V are disjoint open sets containing A and B respectively.

Thus, X is Normal.

Now, A is a closed set in the Normal space X and as U is an open set containing A.
Therefore, by the above Lemma, there exists an open set U ′ in X such that A ⊂ U ′
and closure of U ′ ⊂ U .

Similarly, as B is a closed set in the Normal space X and as V is an open set containing
B, ∃ an open set V ′ in X such that B ⊂ V ′ and closure of V ′ ⊂ V .

Finally, as U and V are disjoint and as closure of U ′ ⊂ U and closure of V ′ ⊂ V ,


therefore the closures of U ′ and V ′ are disjoint.

Problem 12: Show that every order topology is regular.

Answer (2023MAS7138): Let X be an ordered set. Let x ∈ X and let U be a


neighborhood of x, which may assume U = (a, b), −∞ ≤ a < b ≤ ∞. Set A = (a, x)
and B = (x, b). Using the criterion for regularity: ”X is regular if and only if given a
point x of X and neighborhood V of x such that V ⊂ U ,” there are four cases:

1. If u ∈ A and v ∈ B then x ∈ (u, v) ⊂ (u, v) ⊂ [u, v] ⊂ (a, b).


2. If A = B = ∅ then (a, b) = {x} is open and closed, since X is Hausdorff.
3. If A = ∅ and v ∈ B then x ∈ (a, v) ⊂ [x, v) ⊂ [x, v) ⊂ [x, v] ⊂ (a, b).
4. If u ∈ A and B = ∅ then x ∈ (u, b) ⊂ (u, x] ⊂ (u, x] ⊂ [u, x] ⊂ (a, b).

Thus X is regular.

Problem 13: Let f, g : X → Y be continuous and Y be Hausdorff. Show that


{x : f (x) = g(x)} is closed in X.

11
Answer (2023MAS7139): Given : Continuous function f, g : X → Y , Where Y is
Hausdorff’s space.
To show that the set {x : f (x) = g(x)} is closed in X.
Consider h : X → Y × Y defined by
h(x) = (f (x), g(x))
Since f,g is a continuous function
⇒ h(x) is also continuous.

To show :
S = {(y, y) : y ∈ Y } is closed set in the product space Y × Y .
i.e To show S ∁ = {(y1 , y2 ) : y1 ̸= y2 } is open
Given that Y is Hausdorff’s space
y1 ̸= y2 ( By Hausdorff’s property of Y)
∃ Uy1 and Uy2 such that
y1 ∈ Uy1 and y2 ∈ Uy2
(y1 , y2 ) ∈ Uy1 × Uy2
Uy1 × Uy2 contain T distinct point and S contain the identical point
so (Uy1 × Uy2 ) S = Φ
⇒ Uy1 × Uy2 ⊆ S ∁
⇒ (y1 , y2 ) ∈ Uy1 × Uy2 ⊆ S ∁
⇒ S ∁ is open
⇒ S is closed in Y × Y .
According to the property of continuous function inverse image of closed set under a
continuous function is closed.
h−1 (x) = {x ∈ X : h(x) ∈ S} is closed in X
= {x ∈ X : (f (x), g(x) ∈ S}
= {x ∈ X : f (x) = g(x)} is closed in X.

Problem 14: Show that a closed subspace of a normal space is normal.

Answer(2023MAS7140):

Let X be a normal space and Y be a closed subspace of X.

claim : Y is a normal space.

let A1 and A2 be closed sets in Y such that A1 ∩ A2 = ϕ

Then A1 = F1 ∩ Y and A2 = F2 ∩ Y , where F1 and F2 are closed sets in X.

Since Y is a closed subspace of X =⇒ A1 and A2 are closed sets in X.

Now X is a normal space =⇒ ∃ two open sets U and V in X such that A1 ⊂ U, A2 ⊂

12
V and U ∩ V = ϕ.

let G1 = U ∩ Y and G2 = V ∩ Y are open sets in Y .

As A1 ⊂ U, A2 ⊂ V and A1 , A2 ⊂ Y

=⇒ A1 ⊂ U ∩ Y = G1 and A2 ⊂ V ∩ Y = G2

and G1 ∩ G2 = (U ∩ Y ) ∩ (V ∩ Y ) = (U ∩ V ) ∩ Y = ϕ

Thus Y is a normal space.

Problem 15: Show that every regular Lindelöf space is normal.

Answer (2023MAS7141):

TST:
A, B ⊂ X are closed, disjoint ⇒ ∃U, W ⊂ X s.t. U ⊃ A, W ⊃ B and U, W are
open+disjoint.

Let X be a regular Lindelöf topological space and let two disjoint closed sets A and
B be given.

Defining U and W

Let a ∈ A =⇒ a ∈ / B. Thus, B is closed and doesn’t contain a. As X is a regular


space, =⇒ ∃ open set Ua such that a ∈ Ua and open set Va such that B ⊂ Va ,
such that Ua ∩ Va = ∅. Thus, {Ua | a ∈ A} is an open cover of A. We know a closed
subset of Lindelöf is Lindelöf, thus A is Lindelöf. Therefore, there exists a countable
subcover of A. Let {Ua1 , Ua2 , . . .} be that subcover.

Similarly, let b ∈ B =⇒ b ∈/ A. Thus, A is closed and doesn’t contain b. As X is


a regular space, =⇒ ∃ open set Wb such that b ∈ Wb and open set Zb such that
A ⊂ Zb , such that Wb ∩ Zb = ∅. Similarly, ∃ a countable subcover {Wb1 , Wb2 , . . .} of
B.

Define sets U1′ , U2′ , . . . and W1′ , W2′ , . . . as follows:

n
[ m
[
Un′ = Uan \ Wbi and Wm′ = Wbm \ Uaj (open \ closed = open)
i=1 j=1

13

[ ∞
[
Let U = Un′ and W = Wm′ .
n=1 m=1

TST: U and W are open.

As U and W are countable unions of open sets, both U and W are open.

TST: A ⊂ U and B ⊂ W .

Let a ∈ A, as {Ua1 , Ua2 , . . .} covers A, =⇒ A ⊂ ∞


S
i=1 Uai .
As a ∈ A =⇒ a ∈ Uar for some r ∈ N. Also a ∈ A ⊂ Zbi for all bi. As Zbi is
open and Wbi ∩ Zbi = ∅, ∴ Zbi is a neighborhood
Sr of a such that it is disjoint with Wbi
=⇒ a ∈ / Wbi for all i ∈ N,
Sr =⇒ a ∈
/ i=1 W bi . Sr ′
Thus,
S∞ a ∈ U ar and a ∈
/ i=1 W bi =⇒ a ∈ U ar \ i=1 Wbi =⇒ a ∈ Ur =⇒ a ∈ U
(= n=1 Un′ ). Therefore, A ⊂ U . Similarly, we can show B ⊂ W .

TST: U ∩ W = ∅

Let U ∩ W ̸= ∅ =⇒ ∃x ∈ U ∩ W :
x∈S U x ∈SW
x∈ ∞ ′
n=1 Un x∈ ∞ ′
m=1 Wm
Let
x ∈ Ul′ x ∈ Wk′

Let without loss of generality l < k :

we know that Wk′ = Wbk \ kj=1 Uaj


S

x ∈ Wk′ =⇒ x ∈ / kj=1 Uaj


S

=⇒ x ∈ / UalS=⇒ x ∈ / Ual
∴x∈ / Ual \ li=1 Wbl
x∈/ Ul′ =⇒ x ∈ /U
This is a contradiction, thus U ∩ W = ∅.

Hence, X is a normal space.

Q
Problem 16: Show that if Xα ̸= ∅ and Xα is Hausdorff/ regular/ normal then so
is Xα .

Answer (2023MAS7142):

14
Q
Suppose that X = α Xα is Hausdorff. Let β be any index. Let x, y ∈ Xβ be any
distinct points. Since all Xα are nonempty, ∃ points {x′ } and {y ′ } ∈ X such that
xβ = yβ for every β not equal to α and xα = x, yα = y. Because x and y are distinct,
then so are {x′ } and {y ′ }.

Because X is Hausdorff, ∃ open disjoint sets U, V ⊂ X such that {x′ } ∈ U and


{y ′ } ∈ V .

For β ̸= α, we have xβ = yβ ∈ Πβ (U ) ∩ Πβ (V ) hence Πβ (U ) ∩ Πβ (V ) is nonempty.


Because U and V are disjoint, then so are Πα (U ) and Πα (V ) are disjoint.

We also have x ∈ Πα (U ) and y ∈ Πα (V ) and since the projection maps are open and
continuous, it implies that Πα (U ) and Πα (V ) are open.

Hence it proves that x and y can be separated by open sets, therefore Xα is Hausdorff.
Q
Suppose that X = α Xα is Regular. Let β be any index then we have to prove that
Xα is Hausdorff and Regular.

Now since X is regular and X is Hausdorff ⇒ Xα is Hausdorff (satisfies T1 axiom).

Now further let F ⊆ Xα be a closed set and x ∈ Xα −F be a point then Πβ Fβ ( where


Fα = F and Fβ = Xβ for β ̸= α ) is a closed set in X because Πβ Fβc = Πβ Uβ ( where
Uα = F c , Uβ = Xβ for β ̸= α ) is an open set because it is a base element for the
product topology. Further since all Xβ are nonempty, there exists a point {x′ } ∈ X
such that xα = x hence {x′ } does not belong to Πβ Fβ .

Further since X is regular and satisfies the Hausdorff axiom, there exist open sets
U, V ⊆ X such that {x′ } ∈ U and Πβ Fβ ∈ V . Now for every β ̸= α we have
xβ ∈ Xβ = Πβ V , further since xβ ∈ Πβ U thus we have Πβ (U ) ∩ Πβ (V ) is nonempty,
and U ∩ V is empty. Also, x ∈ Πα U and F ⊆ Πα V , further since Πα U and Πα V are
open sets because Πα is an open map hence Xα is Regular.

Q
Suppose that X = α Xα is normal and let β be any index.

Because X is normal, ⇒ X is Hausdorff (or Regular) ⇒ Xα is Hausdorff (or regular)


and satisfies T1 axiom.

Now we have to prove only that Xα is normal,

15
Let if F, G ⊆ Xα are disjoint closed sets then Πβ Fβ and Πβ Gβ (where Fα = F ,
Gα = G, and Fβ = Gβ = Xβ for β ̸= α ), are disjoint closed sets in X. Since
X is normal hence it satisfies the normal axiom, and there exist disjoint open sets
U, V ⊆ X such that Πβ Fβ ⊆ U and Πβ Gβ ⊆ V then Πα U and Πα V are disjoint open
sets in Xα such that F ⊆ Πα U and G ⊆ Πα V .

Hence, Xα is Normal.

Problem 17: Let f, g : X → Y be continuous and Y be Hausdorff, and A be dense


in X. If f |A = g|A then f = g.

Answer (2023MAS7143): Suppose if possible it is not true, i.e. ∃x0 ∈ X such


that f (x0 ) ̸= g(x0 )

As Y is Hausdorff, ∃ open sets U and V such that f (x0 ) ∈ U, g(x0 ) ∈ V and U ∩V = ϕ.

Let W = f −1 (U ) ∩ g −1 (V ). W is open set because U and V are open sets and both
f, g are continuous functions.

Since A is a dense subset of X, A ∩ W ̸= ϕ.

Let a ∈ A ∩ W and f |A = g|A =⇒ f (a) = g(a)

As f (a) ∈ U, g(a) ∈ V =⇒ f (a) = g(a) ∈ U ∩ V , which is a contradiction.

Hence f = g.

Question(Section 32/Problem 5 (page 205)):

1. Is Rω normal in the product topology? In the uniform topology?

Answer (2023MAS7153):

• For the given question, we will use the result that every metrizable topo-
logical space is normal.

(i) Let d′ be the standard bounded metric on R defined by d′ (a, b) = min {|a − b|, 1}
Then, define D on Rω as

d′ (xi , yi )
 
D(x, y) = sup
i

16
where, x = (xi )i∈ω ∈ Rω and y = (yi )i∈ω ∈ Rω
Claim: D is a metric that induces the product topology on Rω .
Firstly, we will show that D is metric on Rω
Clearly, d′ (xi , yi ) is a non-negative quantity because ′
n ′ d isoa metric on R and i
is always a positive integer, so the supremum of d (xii ,yi ) is also non-negative
=⇒ D(x, y) ≥ 0
This shows that D(x, y) is non-negative.
Now,
D(x, y) = 0
 ′ 
d (xi , yi )
⇐⇒ sup =0
i
⇐⇒ d′ (xi , yi ) = 0 ∀ i∈ω
⇐⇒ xi = yi ∀ i∈ω
⇐⇒ x = y
Hence, D(x, y) = 0 ⇐⇒ x = y
Now, since d′ is a metric, so d′ (xi , yi ) = d′ (yi , xi )
n ′ o n ′ o
Therefore, D(x, y) = sup d (xii ,yi ) = sup d (yii ,xi ) = D(y, x)
Hence, D(x, y) = D(y, x)
Now, for triangle inequality, note that

d′ (xi , zi ) d′ (xi , yi ) d′ (yi , zi )


≤ + ≤ D(x, y) + D(y, z)
i i i

d′ (xi , zi )
 
⇐⇒ sup ≤ D(x, y) + D(y, z)
i

⇐⇒ D(x, z) ≤ D(x, y) + D(y, z)

All the above results together proves that D is a metric.


The fact that D gives the product topology requires a little more work. First,
let U be open in the metric topology and let x ∈ U ; we find an open set V in
the product topology such that x ∈ V ⊂ U . Choose an ϵ-ball BD (x, ϵ) lying in
U . Then choose N large enough that N1 < ϵ. Finally, let V be the basis element
for the product topology.
V = (x1 − ϵ, x1 + ϵ) × · · · × (xN − ϵ, xN + ϵ) × R × R × · · · .
We assert that V ⊂ BD (x, ϵ): Given any y in Rω
d′ (xi , yi ) ≤ 1
N
for i ≥ N .

17
Therefore,

1
D(x, y) ≤ max {d′ (x1 , y1 ), . . . , d′ (xN , yN )} ≤
N
If y is in V , this expression is less than ϵ, so that V ⊂ BD (x, ϵ), as desired.
Conversely, consider a basis element
Y
U= Ui
i∈Z+

for the product topology, where Ui is open in R for i = α1 , . . . , αn and Ui = R for


all other indices i. Given x ∈ U , we find an open set V of the metric topology
such that x ∈ V ⊂ U . Choose an interval (xi − ϵi , xi + ϵi ) in R centered about
xi and lying in Ui for i = α1 , . . . , αn ; choose each ϵi ≤ 1. Then define

ϵ = min{ϵi /i|i = α1 , . . . , αn }
We assert that
x ∈ B(x, ϵ) ⊂ U
Let y be a point of B(x, ϵ). Then for all i

d′ (xi , yi )
≤ D(x, y) < ϵ
i
≤ ϵi /i, so that d′ (xi , yi ) < ϵi ≤ 1; it follows that
Now if i = α1 , . . . , αn , then ϵ Q
|xi − yi | < ϵi . Therefore, y ∈ Ui , as desired.
Therefore, Rω in the product topology is metrizable and hence is normal.
(ii) We know that uniform topology is metrizable(by definition) because it is defined
to be the topology induced by the metric

d′′ (x, y) = sup {d′ (xi , yi )|i ∈ ω}

where, d′ is the standard bounded metric on R.


And since Rω in the uniform topology is metrizable and hence is normal.

Question 7. Show that if X has a countable dense subset, every collection of disjoint
open sets in X is countable .

Solution. Let U be a collection of disjoint open sets in X, and let A be a countable


dense subset of X.

Construct an injection from the collection of disjoint open sets into the countable
dense subset.

18
Since A is dense in X, every U ∈ U intersects A. Therefore, there exists a point
xU ∈ U ∩ A.

Let U1 , U2 ∈ U, U1 ̸= U2 . Then xU1 ̸= xU2 since U1 ∩ U2 = ∅.

Thus, the function U → A given by U 7→ xU is injective .Therefore, since A is


countable, it follows that U is countable.

Section 31/Problem 4 (page 199).

Let X and X ′ denote the single set under two topologies τ and τ ′ , respectively; assume
that τ ′ ⊃ τ . If one of the spaces is Hausdorff ( or regular, or normal), what does that
imply about other?

Answer (2023MAS7149):

Observe that τ ′ is finer than τ .

τ is Hausdorff. Given x, y ∈ X, there are disjoint neighborhoods Ux of x and Uy of y,


respectively, in τ . These neighborhoods are also open in τ ′ . Thus τ ′ is also Hausdorff.

τ ′ is Hausdorff. τ is not necessarily Hausdorff. The discrete topology is always Haus-


dorff and finer than the indiscrete topology. The indiscrete topology is not Hausdorff,
as the only neighbourhood of any point in X is X itself.

τ is regular. τ ′ is not necessarily regular. The standard topology of R is regular, but


RK is not regular.

To show: standard topology of R is regular. Given x ∈ R and a neighborhood U of x,


the set U contains the basis element (α, β) where α < x < β. There are γ, δ ∈ R where
[γ, δ] is contained in (α, β). The closed set [γ, δ] is the closure of the neighborhood
(γ, δ) of x. By the lemma ”Let X be a topological space. Let one-point sets in X be
closed. Then X is regular if and only if given a point x of X and a neighborhood U
of x, there is a neighborhood V of x such that V̄ ⊂ U ”, the standard topology of R
is regular.

To show : RK is not regular. The set K is closed in RK , and it does not contain
the point 0. Suppose that there exist disjoint open sets U and V containing 0 and
K, respectively. Choose a basis element containing 0 and lying in U . Then it must
be basis element of the form (a, b) \ K, since each basis element of the form (a, b)
1
containing 0 intersects K. Choose n large enough that ∈ (a, b). Then choose a
n
1
basis element about contained in V . It must be a basis element of the form (c, d).
n

19
 
1 1
Finally, choose z so that z < and z > max c, . Then z belongs to both U
n n+1
and V , so they are not disjoint. Hence RK is not regular.

τ ′ is regular. The topology τ is not necessarily regular. The discrete topology of R is


regular but RK is coarser and not regular.

τ is normal. The topology τ ′ is not necessarily normal. The space R is normal but
RK is not normal.

To show : R is normal. Let A and B are disjoint closed sets in R. For each x ∈ A,
let Bx = (a, b) be a neighborhood of x, that does not intersect B. There is such a
neighborhood because R \ B is open and since x is contained in that set, there is some
basis element (a, b) about x that is contained in R \ B. Similarly for each
S y ∈ B, let
Cy = (c, d) be a neighborhood of y, that does not intersect A. The sets x∈X Bx and
S
y∈Y Cy are disjoint open sets containing A and B, respectively.

τ ′ is normal. The topology τ is not necessarily normal. The discrete topology of R is


normal, but RK is coarser, and not normal.

Question (Section 31/Problem 6 (page 199)): Let p : X → Y be a closed


continuous surjective map. If X is normal, then show that Y is also normal.

Answer (2023MAS7151): Let X be a normal space and p be a closed, continuous,


surjective map.

Claim: Y is a normal space.

1. Since X is T1 (being normal) and p : X → Y is closed and onto, therefore Y is


also T1 .
(If y ∈ Y , then there exists x ∈ X such that f (x) = y. Since X is T1 , x is
closed in X, and hence f is a closed map, implying f (x) = y is closed in Y , so
Y is T1 ).
Since X is normal, X is a T4 space.

2. Let A, B be disjoint closed sets in Y . Then p : X → Y being continuous,


p−1 (A) and p−1 (B) are closed in X. Also, p−1 (A) ∩ p−1 (B) = ∅. This implies
p−1 (A) ∩ p−1 (B) = ∅. So, X being T4 , there exist open sets U and V in X such
that p−1 (A) ⊆ U and p−1 (B) ⊆ V . U ∩ V = ∅ implies P (X − U ), P (X − V )
closed in Y and hence (Y − p(X − U )) = U ′ , (Y − p(X − V )) = V ′ are open in
Y.
We show that A ⊆ U ′ , B ⊆ V ′ , and U ′ ∩ V ′ = ∅.

20
Note that p−1 (A) ⊆ U = p(X − U ) ⊆ X − p−1 (A) = p−1 (Y ) − p−1 (A) =
p−1 (Y − A), which implies p(X − U ) ⊆ p(p−1 (Y − A)). Since p is onto, A ⊆
(Y − P (X − U )) = U ′ .
Similarly, B ⊆ V ′ as p−1 (B) ⊆ V . Also, p−1 (U ′ ) = p−1 (Y − p(X − U )) =
p−1 (Y ) − p−1 p(X − U ) = X − p−1 p(X − U ) ⊆ U . And p−1 (V ′ ) = p−1 (Y − p(X −
U )) = p−1 (Y ) − p−1 p(X − U ) = X − p−1 p(X − U ) ⊆ V . Since p−1 (U ′ ) ⊆ U ,
p−1 (V ′ ) ⊆ V , and U ∩V = ∅, p−1 (U ′ )∩p−1 (V ′ ) = ∅. This implies p−1 (U ′ ∩V ′ ) =
∅. As p is onto, this implies p−1 p(U ′ − V ′ ) = ∅ = (U ′ ∩ V ′ ). Hence, U ′ is T4 .
Since Y is T1 and T4 , therefore Y is a normal space.

Section 32 / Problem 6 (page 205)

A space X is said to be completely normal if every subspace of X is normal. Show


that X is completely normal if and only if for every pair A, B of separated sets in X
( that is sets such that Ā ∩ B = ∅ and A ∩ B̄ = ∅ ), there exist disjoint open sets
containing them. [ Hint: If X is completely normal, consider X − (Ā ∩ B̄) ]

Answer (2023VST9531):

Let X be completely normal. Let A and B be a separated pair of subsets of X. Then


Y = X − (Ā ∩ B̄) is an open subset of X that contains both A and B.

Define ĀY := Ā ∩ Y and B̄Y := B̄ ∩ Y . Then

ĀY ∩ B̄Y = (Ā ∩ Y ) ∩ (B̄ ∩ Y ) (1)


= Y ∩ Ā ∩ B̄ (2)
=∅ (3)

Since X is completely normal, Y is normal, so there exist open disjoint neighborhoods


U and V containing ĀY and B̄Y respectively. We have that

A ⊂ Ā (4)
A⊂Y (5)
=⇒ A ⊂ Ā ∩ Y (6)

so A ⊂ ĀY , and similarly B ⊂ B̄Y , so U and V are open neighbourhoods containing


A and B respectively, but A and B were an arbitrary pair of separated subsets of X

21
so for every pair of A, B of separated sets in X, there exist disjoint open sets U , V
containing them.

Now we show the inverse relation.

Assume for every pair of separated subsets of X there exist disjoint open sets in X
containing them. Let Y be a subspace of X and let A and B be disjoint closed
subsets of Y . Since A and B are closed, A = Ā and B = B̄, and since they are
disjoint, A ∩ B = Ā ∩ B = A ∩ B̄ = ∅. So A and B are separated in X. By
assumption there then exist open neighbourhoods U and V in X containing A and
B respectively. Then UY = U ∩ Y and VY = V ∩ Y contain A and B respectively
and are open in Y . But A, B and Y were chosen arbitrarily, so for any subspace Y
of X, for any pair of disjoint closed subsets of Y there exist open neighbourhoods in
Y containing them, so any subspace Y of X is normal, and therefore X is completely
normal.

22

You might also like