Mat 355E Topology Lecture Notes

Instructor: Ali Sait Demir

Spring 2020
Contents

Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . iii

1 Preliminaries 1
1.1 Set Theory and Logic . . . . . . . . . . . . . . . . . . . . . . . 1
1.2 Functions and Relations . . . . . . . . . . . . . . . . . . . . . 2
1.2.1 Relations . . . . . . . . . . . . . . . . . . . . . . . . . 3
1.3 Countability . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5

2 Metric Spaces 7
2.1 Definition and First Examples . . . . . . . . . . . . . . . . . . 7
2.2 ε-balls and continuity . . . . . . . . . . . . . . . . . . . . . . . 8
2.3 Open Sets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9

3 Topological Spaces 12
3.1 Definition and first examples . . . . . . . . . . . . . . . . . . . 12
3.2 Basis for a Topology . . . . . . . . . . . . . . . . . . . . . . . 13
3.3 Topologies on R . . . . . . . . . . . . . . . . . . . . . . . . . . 16

4 Constructing Topologies 18
4.1 Order Topology . . . . . . . . . . . . . . . . . . . . . . . . . . 18
4.2 Product Topology on X × Y . . . . . . . . . . . . . . . . . . . 19
4.3 Subspace Topology . . . . . . . . . . . . . . . . . . . . . . . . 21

5 Closed Sets and Limit Points 23

5.1 Closed Sets . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
5.1.1 Interior and Closure of a Set . . . . . . . . . . . . . . . 24
5.2 Limit Points . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25

6 Continuous Functions 29
6.1 Homeomorphisms . . . . . . . . . . . . . . . . . . . . . . . . . 32

i
7 Product and Metric Topologies 35
7.1 Two topologies on a Product Space . . . . . . . . . . . . . . . 35
7.2 Metric Topology . . . . . . . . . . . . . . . . . . . . . . . . . . 36

8 Connected Spaces 39
8.1 Constructing Connected Spaces . . . . . . . . . . . . . . . . . 40
8.2 Connected Subspaces of R . . . . . . . . . . . . . . . . . . . . 42
8.2.1 Path Connected Spaces . . . . . . . . . . . . . . . . . . 43
8.2.2 Connected Components . . . . . . . . . . . . . . . . . 44

9 Compact Spaces 46

10 Separation Axioms 50
10.1 Hausdorff Spaces . . . . . . . . . . . . . . . . . . . . . . . . . 51
10.1.1 Properties of Hausdorff Spaces . . . . . . . . . . . . . . 52

11 Countability Properties 53
11.1 Uncountability . . . . . . . . . . . . . . . . . . . . . . . . . . 55

12 Regular and Normal Spaces 56

12.1 Normal Spaces . . . . . . . . . . . . . . . . . . . . . . . . . . 58
12.1.1 Urysohn’s Lemma . . . . . . . . . . . . . . . . . . . . . 59

ii
Preface
These notes are based on the classical book ’Topology’ by Munkres which we
use as the main textbook in İTÜ. I also benefited from my own notes I took
as a student in METU.
Typing of these notes in LaTex was done by the students below who took
the Topology course MAT 355E in Spring semester of 2020. I would like to
thank each one of them for volunteering in this project during the Covid-19
pandemic.

ˆ Kaan Işıldak ˆ Yaren Gök

ˆ Sueda Naciye Kaba ˆ Kaan Corum

ˆ Aytaç İmişçi ˆ Ali Tolga Dinçer

ˆ Arda Buğra Akcabelen ˆ Muhammet Kubilay Sardal

ˆ Hilal Yıldız ˆ Hikmet Can Küfteoğlu

ˆ Yasemin Tuna ˆ Hilal Tuğba Uysal

ˆ Tuğba Can ˆ Ceren Taşkın

ˆ Tuğçe Kılınç ˆ Yusuf Sarıduran

ˆ Melih Cem Çanak ˆ Selçuk Ersöyleyen

ˆ Gamze Fulya Kapucu ˆ Fatih Altiparmakoğlu

ˆ Ravza Tekin ˆ Muhammed Arda

ˆ Dilara Demiralp ˆ Burak Şahin

iii
Chapter 1

Preliminaries

1.1 Set Theory and Logic

Let us recall some basic definitions and notations from set theory:

ˆ A ∪ B = {x|x ∈ A or x ∈ B} union of A and B.

ˆ A ∩ B = {x|x ∈ A and x ∈ B} intersection of A and B.

ˆ ∅ = {} empty set.

ˆ A and B are disjoint if A ∩ B = ∅.

ˆ p =⇒ q: if p is true then q is true.

ˆ ”not q” is called the negation of q.

ˆ contrapositive of ”p =⇒ q” is ”not q =⇒ not p”

ˆ A statement and its contrapositive are equivalent.

ˆ ”p ⇐⇒ q” or ”p if and only if q” means ”p =⇒ q and q =⇒ p”. In

this case:

* p =⇒ q is called the ”‘only if”’ part or ”sufficiency” (p is

sufficient for q)
** q =⇒ p or p ⇐ q is the ”‘if ”’ part or ”necessity.” (p is necessary
for q)

Example 1.1.1. ”For every x ∈ A, the statement P holds.”

Its negation:
”There is at least one x ∈ A such that the statement P does not hold.”

1
Example 1.1.2. ”Every cover has a finite subcover.”
Its negation:
”There is at least one cover that has no finite subcover.”

Definition 1.1.3. A \ B = A − B = {x|x ∈ A and x ∈

/ B} is called the set
difference of A and B.

1. A ∪ B = (A \ B) ∪ (B \ A) ∪ (A ∩ B)

2. A − (A − B) = A ∩ B

3. A ⊂ B ⇐⇒ A − B = ∅

4. A ∩ (B − C) = (A ∩ B) − (A ∩ C)
A 4 B = (A − B) ∪ (B − A) is symmetric difference of A and B.
= (A ∪ B) − (A ∩ B)

5. A − (B ∪ C) = (A − B) ∩ (A − C) (De Morgan’s Law)

6. A − (B ∩ C) = (A − B) ∪ (A − C) (De Morgan’s Law)

7. A ∪ (B ∩ C) = (A ∪ B) ∩ (A ∪ C)

8. A ∩ (B ∪ C) = (A ∩ B) ∪ (A ∩ C)

Remark 1.1.4. To prove an equality A = B of sets, we must prove both

A ⊆ B and B ⊆ A.

Proof. (of 5.) (⊆) Let x ∈ A − (B ∪ C). Then, x ∈ A and x ∈

/ (B ∪ C).
=⇒ x ∈ A and x ∈ / B and x ∈
/C

(⊇) Let x ∈ (A − B) ∩ (A − C).Then, x ∈ (A − B)and x ∈ (A − C)

=⇒ x ∈ A and x ∈ / B and x ∈
/ C =⇒ x ∈ A and x ∈ / (B ∪ C)
=⇒ x ∈ A − (B ∩ C).

1.2 Functions and Relations

Definition 1.2.1. Cartesian product of the sets A and B, A×B = {(a, b)|a ∈
A, b ∈ B}, is a set of ordered pairs.

Sometimes we use a × b instead of (a, b) to denote ordered pairs.

Definition 1.2.2. A function f : A → B is a rule which assigns a unique

element of B to every element of A.

2
 Functions are subsets of Cartesian products.i.e. {(a, f (a)|a ∈ A} ⊂ A×B
The uniqueness condition in the above definition is also expressed as f being
well-defined : a1 = a2 =⇒ f (a1 ) = f (a2 )
Definition 1.2.3. 1. If f : A → B and A0 ⊂ A then we define the
restriction of f to A0 as

f |A0 : A0 → B with the same rule as f : a 7→ f (a)

2. f : A → B is one-to-one (injective) if “f (a1 ) = f (a2 ) implies a1 = a2 ”

3. f : A → B is onto (surjective) if “for every b ∈ B there is a ∈ A such

that f (a) = b”

4. f is bijective if it is 1-1 and onto.

5. Let f : A → B, A0 ⊂ A, and B0 ⊂ B
f (A0 ) = {b|b = f (a) for some a ∈ A0 } : image of A0
f − 1(B0 ) = {a|f (a) ∈ B0 } : inverse image of B0
Remark 1.2.4. f need not be invertible in order the inverse image of B0 to
be defined.
Lemma 1.2.5. 1. A0 ⊂ f −1 (f (A0 ))

2. A0 = f −1 (f (A0 )) only if f is 1-1.

3. f (f −1 (B0 )) ⊂ B0

4. f (f −1 (B0 )) = B0 only if f is onto.

Example 1.2.6. Let f (x) = 3x2 + 2. Then f is not 1-1, and not onto as a
function f : R → R

f −1 (f ([0, 1])) = [−1, 1] implying that f is not 1-1

f (f −1 ([0, 5])) = [2, 5] implying that f is not onto.

1.2.1 Relations
A relation ∼ on a set A is a subset C of A × A
and x ∼ y means (x, y) ∈ C ⊂ A × A and we read “ x is related to y”.

ˆ A relation ∼ on A is an equivalence relation if

1. (Reflexive) x ∼ x for every x ∈ A.

3
 2. (Symmetry) If x ∼ y then y ∼ x for every x, y ∈ A.
3. (Transitivity) If x ∼ y and y ∼ z then x ∼ z.

ˆ If ∼ is an equivalence relation on A, then

E = {y|x ∼ y} is called the equivalence class of x.

Lemma 1.2.7. If E1 and E2 are equivalence classes of ∼, then either E1 =

E2 or E1 ∩ E2 = ∅. (Equivalence classes are either identical or disjoint)

Proof. Let x ∈ E1 \ E2 and y ∈ E1 ∩ E2 (WLOG)

Then, x ∈ E1 and y ∈ E1 implying that x ∼ y
Since y ∈ E2 and x ∼ y we have x ∈ E2 which is a contradiction.

Definition 1.2.8. A family of disjoint sets whose union gives A is said to

be a partition of A.

Example 1.2.9. A = {a, b, c, d}

P = {{a, b}, {c}, {d}} is a partition for A.

Lemma 1.2.10. If ∼ is an equivalence relation on A, then its equivalence

classes partition A.

Definition 1.2.11. A relation ⊂ on A is an ordering relation if

i) For every x, y ∈ A, ¨ x 6= y implies x ⊂ y or y ⊂ x ¨

ii) There is no x ∈ A such that x ⊂ x

iii) If x ⊂ y and y ⊂ z then x ⊂ z

Example 1.2.12. x < y on R is an ordering relation.

Definition 1.2.13. Let X be a set and < be an ordering relation on X.

If a < b, then (a, b) = {x|a < x < b} is called an open interval.

Definition 1.2.14. Let <A and <B be ordering relations on A and B re-
spectively.
On A × B define a1 × b1 < a2 × b2 if ¨ a1 <A a2 or a1 = a2 and b1 <B b2 ¨
as the dictionary order on A × B.

4
1.3 Countability
Definition 1.3.1 (Countabilty). A set A is said to be countable if there is
a 1 − 1 and onto function
f : Z+ → A
(or equivalently f : A → Z+ ) . Z+ = N = {1, 2, 3, ...}
A subset of a countable set is countable.
Theorem 1.3.2. For the set B 6= ∅ the following are equivalent:
1. There is an onto function f : Z+ → B

2. There is a 1 − 1 function g : B → Z+

3. B is countable.
Theorem 1.3.3. Z+ × Z+ is countable.
Proof. f : Z+ × Z+ → Z+ defined by (n, m) 7→ 2n 3m is a 1 − 1function:
Suppose we have two pairs with same the image (n, m) and (p, q) 2n 3m =
2p 3q .
If n < p, 3m = 2p−n 3q contradiction since one side is even. If n > p,
3m 2p−q = 3q contradiction since one side is even. So n = p. This gives
3m = 3q . If m < q, then 1 = 3q−m ⇒ q = m. So (n, m) = (p, q) and hence f
is 1 − 1. ⇒ Z+ × Z+ is countable.
Theorem 1.3.4. Countable union of countable sets is countable.
Proof. Let {Aα }α∈J be a family of countable sets where J is countable. For
all α ∈ J there is an onto function fα : Z+ → Aα and g : Z+ → J.
Define h : Z+ × Z+ → ∪ Aα
α∈J
h(n, m) = fg(n) (m) and show that h is onto.
Let x ∈ ∪ Aα then there is α0 ∈ J such that x ∈ Aα0 . There is a m ∈ Z+
α∈J
such that fα0 (m) = x and n ∈ Z+ such that g(n) = α0 . Thus h(n, m) =
fg(n) (m) = fα0 (m) = x.
So h is onto ⇒ ∪ Aα is countable.
α∈J

Theorem 1.3.5. Finite product of countable sets is countable.

Proof. Is by induction.
Definition 1.3.6. Let X ω denote the infinite product of X with itself

X ω = {(x1 , x2 , x3 , ..., xn , ...)|xi ∈ X, i ∈ Z}

5
Theorem 1.3.7. Let X = {0, 1}. X ω is uncountable.

Proof. There is no onto function g : Z+ → X ω

Let g(n) = (xn1 , xn2 , xn3 , ..., xnm , ...) where xij ’s are either 0 or 1. Define
(
0 xnn = 1
yn =
1 xnn = 0

If y = (y1 , y2 , y3 , ..., yn , ...) ∈ X ω then there is no integer m such that g(m) =

y.
g is not onto ⇒ X ω is uncountable.

6
Chapter 2

Metric Spaces

2.1 Definition and First Examples

We study metric spaces to develop the concept of continuity.

Definition 2.1.1. Let M be a set,

ρ:M ×M →R

be a function. Then (M, ρ) is a metric space if

i) ρ(x, y) ≥ 0, and

i*) ρ(x, y) = 0 if and only if x = y,

ii) ρ(x, y) = ρ(y, x),

iii) ρ(x, y) + ρ(y, z) ≥ ρ(x, z) (Triangle Inequality)

In this case ρ is said to be a metric on M . If ρ does not satisfy i*), then it

is called a pseudo-metric on M .

Example 2.1.2. M = R , ρ(x, y) = |x − y|

i) |x − y| ≥ 0 (by definition) and |x − y| = 0 if and only if x = y

ii) |x − y| = |y − x|

iii) |x − z| = |x − y + y − z| ≤ |x − y| + |y − z|

So,(R, |, |) is a metric space with the absolute value metric and called the
standard or usual Euclidean metric space.

7
 s
n
Example 2.1.3. M = Rn , ρ((x1 , ..., xn ), (y1 , ..., yn )) =
P
(xk − yk )2
k=1
p
For n = 2 : ρ((x1 , x2 ), (y1 , y2 )) = (x1 − y1 )2 + (x2 − y2 )
ρ is the Standard metric on R2 .

Example 2.1.4. On M = R2 the function ρ1 ((x1 , x2 ), (y1 , y2 )) = |x1 − y1 | +

|x2 − y2 | is called the Taxi metric on R2 .

Example 2.1.5. Let M = R2 with ρ2 ((x1 , x2 ), (y1 , y2 )) = max{|x1 −y1 |, |x2 −

y2 |} which is the square metric on R2 .

Example 2.1.6. The distance from the point (x, y) ∈ R to the origin with
respect to above metrics is

ˆ ρ((x, y), (0, 0)) = x2 + y 2

p

ˆ ρ1 ((x, y), (0, 0)) = |x| + |y|

ˆ ρ2 ((x, y), (0, 0)) = max{|x|, |y|}

Remark 2.1.7. If (M, ρ) is a metric space and A ⊂ M then (A, ρ|A×A ) is

also a metric space.

ρ|A×A : A × A → R
(a1 , a2 ) 7→ ρ(a1 , a2 ).

Example 2.1.8. X: set , ρ : X × X → R

(
0 x=y
ρ(x, y) =
1 x 6= y
ρ is called the discrete metric on X.

Example 2.1.9. For any set X, let ρ(x, y) = 0 for all x, y ∈ X. ρ is called
the trivial (pseudo)-metric on X.

2.2 ε-balls and continuity

Definition 2.2.1. Let (M, ρ) and (N, σ) be metric spaces. f : M → N is
continuous at x0 ∈ M if and only if, for every ε > 0 there is a δ > 0 such
that ρ(x, y) < δ implies σ(f (x), f (y)) < ε.

8
 This definition applied to R with the usual absolute value metric yields
the usual continuity definition from our calculus courses:
f : R → R is continuous at x0 ∈ R iff ∀ε > 0 ∃δ > 0 such that
|x − x0 | < δ ⇒ |f (x) − f (x0 )| < ε.
Definition 2.2.2. Let (M, ρ) be a metric space and x ∈ M . For ε > 0
B (x) = B(x, ) = Uρ (x, ) = {y ∈ M | ρ(x, y) < } is called the  − disc (or
-ball) centered at x. (the open -disc)
Example 2.2.3. Let M = R and ρ = | · | be the absolute value metric.
Uρ (x, ) = {y | |x − y| < } = (x − , x + ). That is, the -ball in the usual
real line is the open interval centered at x with radius .
2
Example p2.2.4. Let M = R and ρ be the standard metric. Then, Uρ ((x1 , x2 ), ) =
{(y1 , y2 )| (x1 − y1 )2 + (x2 − y2 )2 < } is the open disc centered at (x1 , x2 )
with radius .
Example 2.2.5. For the discrete metric on R we have
Uρ (x, 21 ) = {x} Uρ (x, 32 ) = R
Definition 2.2.6. Let E and F be subsets of the metric space (M, ρ). ρ(E, F ) =
inf {ρ(x, y)|x ∈ E y ∈ F } is the distance between E and F.
Example 2.2.7. Let E = (0, 1) and F = (2, 3) be subsets of (R, ρ = | · |).
Then ρ(E, F ) = 1.
Definition 2.2.8. Uρ (E, ε) = {y ∈ M |ρ(E, y) < ε} is the ε-disc at the
subspace E ⊂ M
Example 2.2.9. For E = (0, 1] and ε = 1
2
we have Uρ ((0, 1], 21 ) = ( −1 , 3 ).
2 2

Definition 2.2.10. f : (M, ρ) → (N, σ) is continuous at x ∈ M if for all

 > 0 there is ε > 0 such that f (Uρ (x, δ)) ⊂ Uσ (f (x), ε).

2.3 Open Sets

Definition 2.3.1. E ⊂ (M, ρ) is said to be open, if for every x ∈ E there is
ε > 0 such that Uρ (x, ε) ⊂ E.
Example 2.3.2. M = R2 ρ : standard metric. The sets A = {(x, y)|y 6=
f (x)}, B = {(x, y)|y < f (x)} are open for any function f : R → R.
Example 2.3.3. M = R ρ=|·| (a, b) = {x ∈ R|a < x < b}

Let ε = 21 min{|a − x|, |x − b|} , then Uρ (x, ε) = (x − ε, x + ε) ⊂ (a, b).

So (a, b) is an open set.

9
Definition 2.3.4. A set F ⊂ M is closed iff M − F is open.
⇐⇒ [y ∈ M − F ⇒ ∃ε > 0 such that Uρ (y, ε) ⊂ M − F ]
⇐⇒ [∀ε > 0 Uρ (y, ε) ∩ F 6= ∅ ⇒ y ∈ F ].
Example 2.3.5. Let F = (0, 1] ⊂ R. Then F is not closed. For any
ε > 0 B(0, ε) ∩ (0, 1] 6= ∅, but 0 ∈
/ F.
Example 2.3.6. Singletons (single element set) are closed in metric spaces.
Proof. Let x ∈ (M, ρ) and consider {x}. If y ∈ M − {x}, then y 6= x and let
ε = 12 ρ(x, y) > 0. Since B(y, ε) ⊂ M − {x}, M − {x} is an open set. So {x}
is closed.
Theorem 2.3.7. Let (M, ρ) be a metric space.
i) Every union of open sets is open.
ii) Finite intersection of open sets is open.
iii) φ and M are open
S
Proof. (i)S Let Aα be open for any α ∈ λ and consider x ∈ α∈λ Aα .
x ∈ α∈A Aα ⇒ there is α0 such that x ∈ Aα0 and Aα0 is an open set.
⇒ There is anSε−disc UρS (x, ε) ⊂ Aα0
⇒ Uρ (x, ε) ⊂ α∈λ ⇒ α∈λ Aα is also an open set.
(ii) Let A1 , ..., An be open sets. Let x ∈ ni=1 Ai .
T
Since Ai are open for all i, there’s εi > 0 such that Uρ (x, εi ) ⊂ Ai
Choose ε = min{ε n } then Uρ (x, ε) ⊂ Ai for all i.
T 1 , ε2 , ..., εT
⇒ Uρ (x, ε) ⊂ ni=1 Ai ⇒ ni=1 Ai is open.

Example 2.3.8. For M = R, ρ = |·| let An = ( −1 , 1 ). Then ∞

T
n n n=1 An = {0}
is not open. (Since for any ε > 0 Uρ (0, ε) 6⊂ {0} therefore {0} is not open.)
So infinite intersection of open sets might not be open.
Example 2.3.9. ε-discs are open in any metric space (M, ρ).
Let δ = ε − ρ(x, y) or δ = n1 ’s

y
x

Show that Uρ (y, δ) ⊂ Uρ (x, ε).

10
Example 2.3.10. Singletons are open in discrete metric spaces.
For any point x Udiscrete (x, 12 ) = {x} ⊂ {x}
⇒ Therefore in a discrete metric space every set is open. (Union of single-
tons)
Moreover, every set is closed because it’s complement is an open set.

Theorem 2.3.11. f : (M, ρ) → (N, σ) is continuous at x0 ∈ M iff

for any open set V with f (x0 ) ∈ V ⊂ N there is an open set
U with x0 ∈ U ⊂ M st. f (U ) ⊂ V .

Proof. HOMEWORK

11
Chapter 3

Topological Spaces

3.1 Definition and first examples

Let X be a set and τ be a family of subsets of X such that:
i) φ ∈ τ and X ∈ τ ,

ii) Any union of elements of τ is in τ ,

iii) Intersection of finite number of elements of τ is in τ .

(X, τ ) is called a topological space and τ is a topology on X. Elements of τ
are called open sets in (X, τ ). Then (X, τ ) being a topological space implies:
i) φ, X are open,

ii) Any union of open sets is open,

iii) Intersection of finitely many open sets is open.

Example 3.1.1. (Cofinite or finite complement topology)
X: set, τf = {U ⊂ X | X − U is finite or U = ∅}

1) φ ∈ τf and X ∈ τf since X − X = φ is finite.

Note: DeMorgan

X − (A ∩ B) = (X − A) ∪ (X − B)
X − (A ∪ B) = (X − A) ∩ (X − B)
T
2) Let {Uα } ∈Sτf . Claim:
T Uα ∈ τf
Since X − Uα = (X − Uα ) and intersection of finite sets is finite,
claim is true. Therefore any union of sets in τf is also in τf .

12
 Tn
3) Let U1 , U2 , ..., Un ∈ τf . Claim: i=1 Ui ∈ τf

Proof. X − ni=1 Ui = ni=1 (X − Ui ) and the fact that finite union of

T S
finite sets is also finite imply intersection of finitely many sets in τf is
also in τf .

Question. Show that τc = {U ⊂ X | X − U is countable or U = φ} is

a topology on any set X, called the cocountable or countable complement
topology.
0 0
Definition 3.1.2. Let τ and τ be topologies on X. If τ ⊂ τ then
0
τ is stronger / finer / larger than τ
0
τ is weaker / coarser / smaller than τ
0 0 0
If τ ⊂ τ or τ ⊂ τ then τ and τ are comparable.
Example 3.1.3. On X = Z, τf and τc are comparable. τf ⊂ τc
U ∈ τf ⇒ X − U finite ⇒ X − U is countable ⇒ U ∈ τc
Cocountable topology is stronger than cofinite topology.
Example 3.1.4. → On any set X, the discrete topology is the strongest
topology on X.
τ ⊂ 2X = τdiscrete
Example 3.1.5. → τtrivial = {φ, X} is the weakest topology on X.
Example 3.1.6. X = {a, b, c}
τ1 = {φ, X}
τ2 = {φ, X, {b}}
τ3 = {φ, X, {b}, {a, b}}
τ4 = {φ, X, {b}, {a, b}, {b, c}}
τ5 = discrete topology on X
τ1 ⊂ τ2 ⊂ τ3 ⊂ τ4 ⊂ τ5

3.2 Basis for a Topology

Definition 3.2.1. Let X be a set and β be a family of subsets of X.
i) For every x ∈ XSthere’s Bx ∈ β st x ∈ Bx
(or equivalently B∈β B = X)

ii) For any x ∈ B1 ∩ B2 for B1 , B2 ∈ β , then there is B3 ∈ β st x ∈ B3 ⊂

B1 ∩ B2

13
Then β is said to be a basis (of a topology) on X.

Definition 3.2.2. If β is a basis as above, then the topology τ generated by

β is defined as follows:
U ⊂ X is open in (X, τ ) iff for any x ∈ U there’s B ∈ β such that x ∈ B ⊂ U .
By definition, every basis element B ∈ β is open in the topology generated by
β.
τ = {U | ∀x ∈ U ∃B ∈ β s.t. x ∈ B ⊂ U }

Example 3.2.3. Let β be the set of all circular regions in R2 . Then β is a

basis on R2 .

B3 ⊂ B1 ∩ B2

(a, b) ∈ R2 (a, b) ∈ B = {(x, y) | (x − a)2 + (y − b)2 < }

U is open iff for every x ∈ U there’s a circular region around x that lies
completely in U.

0
Example 3.2.4. On R2 , let β be the set of all rectangular regions (excluding
the bounding rectangle)

14
 Note that the intersection of two basis elements is a basis element itself:
0
B3 = B1 ∩ B2 . β is a basis for R2 .
Next, we prove that the topology generated by a basis in Definition 3.2.2
is a topology.
Theorem 3.2.5. Let β be a basis on X, then
τ = {U | ∀x ∈ U ∃B ∈ β st x ∈ B ⊂ U } is a topology on X.
Proof. 1) φ and X ∈ τ
2) Let {Uα } ∈ τ for α ∈ ∧
S 0 0
For x ∈ α∈∧ Uα then x ∈ Uα for some α ∈ ∧
0 0 0 S
This means there’s B ∈ β st X ∈ B ⊂ Uα ⊂ α∈∧ Uα
S
⇒ α∈∧ Uα ∈ τ
3) Let U1 , U2 ∈ τ and x ∈ U1 ∩ U2 ⇒ x ∈ U1 and x ∈ U2
⇒ there is B1 , B2 ∈ β st x ∈ B1 ⊂ U1 and x ∈ B2 ⊂ U2
So x ∈ B1 ∩B2 . Since β is a basis there is B3 ∈ β st x ∈ B3 ⊂ B1 ∩B2 ⊂
U1 ∩ U2 . The rest follows by induction.
⇒ U1 ∩ U2 ∈ τ

Next, we see another description of open sets in a topology generated by

a basis.
Theorem 3.2.6. Let X be a set, β be a bais and τ be the topology generated
by β. The family of unions of elements of β equals τ .
Proof. Since β ⊂ τ any union of elements of β is in τ
Let U ∈ τSFor any x ∈ U there’s Bx ∈ β st x ∈ Bx ⊂ U
But U = x∈U Bx

15
Remark 3.2.7. Given a topology τ on X, can we construct a basis which
generates the same topology as τ ? A basis for a topology gives a shorter list of
open sets and hence a simpler understanding of the properties of that space.
A basis for a topology τ can be constructed as follows:
Lemma 3.2.8. Let (X, τ ) be a topological space and C be a family of open
sets in X such that for every open set U ∈ τ and every point x ∈ U there is
C ∈ C satisfying x ∈ C ⊂ U . Then C is a basis for τ .
Example 3.2.9. For any X, the family C = {{x}|x ∈ X} is a basis for the
discrete topology on X.
Theorem 3.2.10. Let B1 and B2 be bases for τ1 and τ2 respectively. Then,
τ1 ⊂ τ2 if and only if for every x ∈ X and B1 ∈ B1 there is B2 ∈ B2 such
that x ∈ B2 ⊂ B1 .
Proof. (⇒) Let U ∈ τ1 (must show U ∈ τ2 ). Then there is B1 ∈ B1 such
that x ∈ B1 ⊂ U for all x ∈ U . By assumption there is B2 ∈ B2 with
x ∈ B2 ⊂ B1 ⊂ U .
(⇐) Let x ∈ X and B1 ∈ B1 with x ∈ B1 . Since B1 is open in τ1 and since
τ1 ⊂ τ2 , we have B1 ⊂ τ2 . Since B2 generates τ2 and B1 ∈ τ2 there is B2 ∈ B2
such that B2 ⊂ B1 .

3.3 Topologies on R
Definition 3.3.1. (Standard Topology on R) The family of open intervals
(a, b) = {x | a < x < b} is a basis Bst on R and topology it generates Rst is
called the standard (or usual) topology on R.
[
R= (−n, n)
n∈N

Definition 3.3.2. Lower Limit Topology on R. The family of intervals

[a, b) = {x|a 6 x < b} is a basis Bl and it generates Rl .
[
R= [x, y)
x∈R

Theorem 3.3.3.
Rst ⊂ Rl
Proof. Let x ∈ (a, b). Then x ∈ [x, b) ⊂ (a, b) and [x, b) ∈ Bl and by Theorem
3.2.10 Rst ⊂ Rl .

16
2nd Way : For (a, b) ⊂open Rst
∞
[a + b−a
S
x ∈ (a, b) = 2n
, b)
n=1
Union of base elements in B. So it must be open in Rl .

Example 3.3.4. Let K = { n1 |n = 1, 2, 3...} and BK = Bst ∪ {(a, b) − K|a <

b}. Denote the topology generated by BK with RK (K-topology)

a. Show Rst ⊂ RK
Since for every x ∈ (a, b) ∈ Bst ⊂ BK (a, b) ∈ BK , this follows from
Theorem 3.2.10.
On the other hand, B = (−1, 1) − K ∈ BK
0 ∈ B and we can not find an open interval (a, b) such that 0 ∈ (a, b) ⊂
B. Thus, RK 6⊂ Rst

b. Rl and RK are not comparable.

[5, 7) is open in Rl , being a basis element.
It is not open in RK since there is no basis element B ∈ BK with
5 ∈ B ⊂ [5, 7).
(−1, 1) − K is open in RK . 0 ∈ (−1, 1) − K but there is no [a, b) with
0 ∈ [a, b) ⊂ (−1, 1) − K.

Definition 3.3.5. A family of subsets of X is a subbase if the union of

elements of this family is X. The topology generated by a subbase S consists
of unions of finite intersections of elements of S.

17
Chapter 4

Constructing Topologies

4.1 Order Topology

Let < be an order relation on X, i.e.

i. For all x, y ∈ X if x 6= y then x < y or y < x,

ii. There is no x ∈ X with x < x,

iii. If x < y and y < z then x < z.

Define (a, b) = {x | a < x < b}, [a, b) = (a, b) ∪ {a}, (a, b] = (a, b) ∪ {b},
and [a, b] = (a, b) ∪ {a, b}.

Definition 4.1.1. Let < be an order relation on X and suppose X has at

least two elements. The family consisting of sets of the form,

i. (a, b) ⊂ X (open intervals)

ii. If there is a least element a0 ∈ X, then intervals of type [a0 , b)

iii. If there is a greatest element b0 ∈ X, then intervals of type (a, b0 ]

is the basis of order topology on X.

Example 4.1.2. Rst is also the order topology on R.

Example 4.1.3. On R2 = R × R define the dictionary order topology with

basis consisting of elements of the form,
(a × b, c × d) = {x × y | a × b < x × y < c × d}
(a × b < c × d iff a < c or ” if a = c then b < d”)

18
Example 4.1.4. Order topology on N = {1, 2, 3, ...} is equal to discrete topol-
ogy
{n} = (n − 1, n + 1)
{1} = [1, 2)
i.e. single element sets are basis elements, hence they are open.

Example 4.1.5. Consider dictionary on X = {1, 2} × N. X has the least

element 1 × 1. Denote the elements of the form 1 × n and 2 × n with an and
bn , respectively.
a1 < a2 < ...... < b1 < b2 < ......
Question: Is this also discrete topology?
Answer: No
{a1 } = [a1 , a2 )
{an } = (an−1 , an+1 )
{b2 } = (b1 , b3 )
{b1 } is not a basis element.

4.2 Product Topology on X × Y

In this section we define a topology on the cartesian product X × Y =
{(x, y)|x ∈ X, y ∈ Y } of two topological spaces X and Y .

Definition 4.2.1. Let X and Y be topological spaces. The family of sets of

the form,
U ×V ⊂X ×Y
where U ⊂ X, V ⊂ Y is a basis β for the product topology on X × Y .
open open

β = {U × V | U ⊂ X, V ⊂ Y }
open open

Question. Is β a basis ?

1) X ⊂open X, Y ⊂open Y then X × Y ∈ β

2) Let U1 × V1 and U2 × V2 be in β. Then (U1 × V1 ) ∩ (U2 × V2 ) =

(U1 ∩ U2 ) × (V1 ∩ V2 )
Ui × Vi ∈ β ⇒ Ui ⊂open X Vi ⊂open Y i = 1, 2
⇒ U1 ∩ U2 ⊂open X and V1 ∩ V2 ⊂open Y
⇒U ×V ∈β

19
 Figure 4.1: (U1 × V1 ) ∩ (U2 × V2 ) = (U1 ∩ U2 ) × (V1 ∩ V2 )

Theorem 4.2.2. Let βx and βy be bases for the topological spaces X and Y
resp.
D = {A × C | A ∈ βx , C ∈ βy } is a basis for the product topology on
X ×Y.
2
Example 4.2.3. Rst × Rst = Rst standard topology on R2

2
By the above thm open rectangles form a basis for Rst
Definition 4.2.4. π1 : X × Y → X π2 : X × Y → Y
(x, y) → x (x, y) → y
are called projections onto X and Y .
* π1 and π2 are onto
* U ⊂open X then π1−1 (U ) = U × Y ⊂open X × Y
V ⊂open Y then π2−1 (V ) = X × V ⊂open X × Y

20
Theorem 4.2.5. S1 = {π1−1 (U ) | U ⊂open X} and S2 = {π2−1 (V ) |
V ⊂open Y }
Then S = S1 ∪ S2 is a subbase for the product topology on X × Y

Proof. Let τ : product topology on X × Y

0
τ : topology that S generates
Since S ⊂ τ and finite intersections of open sets and their unions is open
0
the topology generated by S is in τ . (i.e. τ ⊂ τ )
On the other side, since basis elements of τ are
0
U × V = π1−1 (U ) ∩ π2−1 (V ) ∈ S ⇒ U ×V ∈τ

4.3 Subspace Topology

Let (X, τ ) be a topological space and Y ⊂ X be a subset.
Then τY = {Y ∩ U | U ∈ τ } is a topology on Y called the subspace topology
on Y inherited from X.

* ∅ = Y ∩ ∅ and Y = Y ∩ X ⇒ ∅, Y ∈ τy

* (U1 ∩ Y ) ∩ (U2 ∩ Y ) ∩ ... ∩ (Un ∩ Y ) = (U1 ∩ U2 ∩ ... ∩ Un ) ∩ Y (intersection

of finitely many open sets)
S S
* α∈J (Uα ∩ Y ) = ( α∈J Uα ) ∩ Y ∈ τy (union of any number of open
sets)

Theorem 4.3.1. If β is a basis for (X, τ ) then βY = {B ∩ Y | B ∈ β} is a

basis for (Y, τY )

Definition 4.3.2. U is said to be open Y if U ∈ τY

Theorem 4.3.3. If Y is open in X and U is open in Y , then U is open in

X.
(U ⊂open Y, Y ⊂open X ⇒ U ⊂open X)

Proof. If U ⊂open Y then there’s V ⊂open X s.t. U = Y ∩ V : intersection of

two open sets in X
⇒ U ⊂open X

Theorem 4.3.4. Let A and B subspaces of X and Y respectively. The

product topology on A × B is equal to subspace topology from X × Y

21
Proof. Let’s show that these topologies have same bases.
Let U ⊂open X and V ⊂open Y be bases elements of X and Y resp.
Then U × V is a basis element for X × Y . Thus (U × V ) ∩ (A × B) is a
basis element for the subspace topology on A × B. Since (U × V ) ∩ (A × B) =
(U ∩ A) × (V ∩ B) which is a typical element of basis for the product topology
on A × B. See figure 4.1.

Example 4.3.5. X = {a, b, c, d, e}

τ = {∅, X, {a}, {c, d}, {a, c, d}, {b, c, d, e}}
Y = {a, d, e}
τY = {∅, Y, {a}, {d}, {a, d}, {d, e}}

Example 4.3.6. X = Rst Y = [0, 1]

βy = {(a, b) ∩ [0, 1] | a, b ∈ R}


 (a, b) a, b ∈ Y

[0, b) b∈Y a∈ /Y
(a, b) ∩ [0, 1] =


 (a, 1] a∈Y b∈ /Y
[0, 1] or ∅ a∈ / Y andb ∈

/Y
Such elements form a basis for both subspace topology and order topology
on Y
We see that Y = [0, 1], (x, 1], [0, x] are open in Y .

Example 4.3.7. X = Rst Y = [0, 1) ∪ {2}

{2} = Y ∩ (1, 3)
⇒ {2} is open in Y but not in X
Note that {2} is not open in Y if we have order topology on Y . Any base
element containing 2 is of the form
{x | x ∈ Y a < x ≤ 2} for a ∈ Y . All those base elements must include
elements that are less than 2.

22
Chapter 5

Closed Sets and Limit Points

5.1 Closed Sets

Definition 5.1.1. A ⊂ X is a closed set if X − A is open.
Example 5.1.2. X = Rst .
[a, b] is closed since R − [a, b] = (−∞, a) ∪ (b, ∞) is open. [a, ∞) and
(−∞, b] are closed.
Example 5.1.3. On a discrete topological space since every subset is open,
every subset must be closed (being the complement of some open set).
Example 5.1.4. On a cofinite topological space finite sets are closed (to-
gether with ∅, X)
Example 5.1.5. On R2st , let Y = {(x, y) | x ≥ 0, y ≥ 0}.
R2 \ Y = ((−∞, 0) × R) ∪ (R × (−∞, 0)) is open. So Y is closed.
Definition 5.1.6. A is closed in the subspace Y if Y − A is open in Y .
Example 5.1.7. X = Rst and let Y = [0, 1] ∪ (2, 3). Consider the subspace
topology on Y ,
[0, 1] = Y ∩ (−1, 3/2) is open in Y .
Y − [0, 1] = (2, 3) is closed in Y .
(2, 3) = Y ∩ (2, 3) is open in Y .
We can describe a topology using closed sets as well:
Theorem 5.1.8. If X is a topological space,then
i) ∅, X are closed.

ii) Any intersection of closed sets is closed.

23
 iii) Finite union of closed sets is closed.
Proof. ii) Let {Aα }α∈J be a family of closed sets.
X − ∩ Aα = ∪ (X − Aα )
α∈J α∈J

is open (any union of open sets is open).

iii) {Ai }ni=1 be a finite collection of closed sets.
n n
X − ∪ Ai = ∩ (X − Ai )
i=1 i=1

is open since finite intersection of open sets is open.

Theorem 5.1.9. Let Y be a subspace of X. A is closed in Y iff A = Y ∩ C

for some C ⊂ X.
closed

Proof. (⇐) Let C ⊂ X and A = C ∩ Y . Then X − C ⊂ X and

closed open
(X − C) ∩ Y ⊂ Y .
open
Y − A = (X − C) ∩ Y , A is closed in Y .
(⇒) Let A ⊂ Y the Y − A ⊂ Y ⇒ Y − A = Y ∩ U for some U ⊂ X.
closed open open
A = Y ∩ (X − U ) so if we let C = X − U then C is closed in X and
A = Y ∩ C.
Homework:A ⊂ Y , Y ⊂ X. Show that A ⊂ X
closed closed closed

5.1.1 Interior and Closure of a Set

Let X be a topological space and A ⊂ X. Interior of A is the union of all
open sets that A covers and is denoted by Ao .
[
Ao = { U | U is open in X and U ⊂ A}
Being a union of open sets, Ao is open. It is the largest open set that is
included in A.
The closure of A is the intersection of all closed sets that cover A, denoted
by A. \
A = { K | K is closed in X and A ⊂ K}
Being the intersection of closed sets A is closed. A is the smallest closed
set that covers A.
Observe: If A is closed,then A = A. If A is open, then Ao = A.
Ao ⊂ A ⊂ A

24
Theorem 5.1.10. Let X be a topological space, Y a subspace and A ⊂ Y .
The closure of A in Y equals A ∩ Y .

Proof. Let B denote the closure of A in Y . Since A ⊂ X then (A∩Y ) ⊂

closed closed
Y . But A ⊂ (A ∩ Y ) and hence B ⊂ (A ∩ Y ).On the other hand, since B is
closed in Y , there’s C ⊂ X such that B = C ∩ Y .
closed
A ⊂ B = C ∩ Y implies C is a closed set and A ⊂ C. ⇒ A ⊂ C ⇒
(A ∩ Y ) ⊂ (C ∩ Y ) = B

Theorem 5.1.11. Let X be a topological space and A ⊂ X;

a) x ∈ A iff every open set containing x intersects A (x ∈ U ⊂ X ⇒

open
U ∩ A 6= ∅).

b) x ∈ A iff every basis element containing x intersects A.

Example 5.1.12. X = Rst , A = (0, 1] then Ao = (0, 1), A = [0, 1].

Example 5.1.13. B = {1/n|n ∈ N} then B o = ∅, B = B ∪ {0}.

Example 5.1.14. Q = R.

Example 5.1.15. N = N.

Example 5.1.16. C = {0} ∪ (1, 2) ⇒ C = {0} ∪ [1, 2].

Example 5.1.17. X = R, Y = (0, 1]

A = (0, 1/2) ⇒ A = [0, 1/2]
⇒ A ∩ Y = (0, 1/2] is the closure of A in Y .

5.2 Limit Points

Let A be a subset of the topological space X and x ∈ X. Then x is a
limit point of A, if every open set U containing x intersects A at some
point other than x.
Equivalently, x is a limit point of A if it belongs to the closure of A − {x}.

Observe: x need not be in A to be a limit point of A.

25
Example 5.2.1. x and y are limit points of A.

Definition 5.2.2. An open set U containing x ∈ X is said to be a neigh-

borhood of x.

Example 5.2.3. X = R A = (0, 1]. Then any point x ∈ [0, 1] is a limit

point of A.

Example 5.2.4. B = n1 | n ∈ N . The only limit point of B is 0.



Example 5.2.5. C = {0} ∪ (1, 2). The set of limit points of C is [1, 2].

Example 5.2.6. The set of limit points of Q in R is R.

Example 5.2.7. Z has no limit points.

Notation: A0 denotes the set of limit points of A.

Theorem 5.2.8. A = A ∪ A0

Corollary 5.2.9. A is closed if and only if A0 ⊂ A.

26
Definition 5.2.10. The boundary of A ⊂ X is ∂A = A ∩ (X − A).

Theorem 5.2.11. i) ∂A and A◦ are disjoint.

ii) A = A◦ ∪ ∂A

iii) ∂A = ∅ iff A is both open and closed.

iv) A is open if and only if ∂A = A − A

Proof. i) Let x ∈ A◦ so that x has a neighborhood U contained in A. For

any y ∈ U we have y ∈ U ⊂ A hence y ∈ / X − A. ⇒ U ∩ (X − A) = ∅.
?(Recall that x ∈ K if and only if any neighbor of x intersects K.)
⇒x∈ / (X − A) ⇒ x ∈ / ∂A = A ∩ (X − A)

ii) ⊆: Let x ∈ A. If x ∈ A◦ then x ∈ A◦ ∪ ∂A. Suppose x ∈

/ A◦ (must
show: x ∈ ∂A = A ∩ (X − A)). Then for any neighborhood U of x,
U ∩ (X − A) 6= ∅. By ?, x ∈ (X − A).
⊇: Let x ∈ A◦ ∪ ∂A. If x ∈ A◦ then x ∈ A◦ ⊂ A ⊂ A. If x ∈ ∂A =
A ∩ (X − A), then x ∈ A.

Example 5.2.12. Let Dn = B n = {(x1 , x2 , ..., xn ) | x21 + x22 + ... + x2n ≤ 1}

denote the unit n-disk or the n-ball which is a subset of Rn .
The boundary of Dn : ∂Dn = S n−1 = {(x1 , x2 , ..., xn ) | x21 + x22 + ... + x2n = 1}
is called the (n-1)-sphere.
The interior of Dn : (Dn )◦ = {(x1 , x2 , ..., xn ) | x21 + x22 + ... + x2n < 1} is the
open n-disk.
n=1: D1 = [−1, 1] = {x | x2 ≤ 1} ∂D1 = S 0 = {−1, 1}.

n=2: D2 = {(x, y) | x2 + y 2 ≤ 1} ∂D2 = S 1 = {(x, y) | x2 + y 2 = 1}: unit

circle.

27
n=3: D3 = {(x, y, z) | x2 + y 2 + z 2 ≤ 1} ∂D3 = S 2 = unit sphere.

Example 5.2.13. ∂R = ∅ = R ∩ (R − R) = R ∩ ∅

Example 5.2.14. ∂Q = Q ∩ (R − Q) = R ∩ R = R
OR ∂Q = Q − Q◦ = R − ∅ = R

 The boundary of A is also defined as the set of boundary points. A

point p ∈ X is a boundary point of A if every neighborhood of p contains at
least one point of A and at least one point not of A.

28
Chapter 6

Continuous Functions

Let X and Y be topological spaces. A function f : X → Y is continuous

if the inverse image of every open set in Y is open in X i.e. ”V ⊂open Y ⇒
f −1 (V ) ⊂open X if and only if f is continuous.”

Pointwise : f is continuous at x ∈ X if and only if for any neighborhood V

of f (x), there is a neighborhood U of x such that f (U ) ⊂ V .

Remark: ε − δ definition of Calculus 1 is equivalent to this definition.

”f : R → R is continuous at x ∈ R ⇔ ∀ε > 0 ∃δ > 0 such that
|x − y| < δ ⇒ |f (x) − f (y)| < ε.”
For every neighborhood of f (x) (V = (f (x) − ε, f (x) + ε)) there is a neigh-
borhood of x (U = (x − δ, x + δ)) such that if y ∈ (x − δ, x + δ) then
f (y) ∈ (f (x) − ε, f (x) + ε).
i.e. f ((x − δ, x + δ)) ⊂ (f (x) − ε, f (x) + ε).

As usual, working with basis elements, rather than open sets is more conve-
nient.
Theorem 6.0.1. If B is a basis for the topology on Y and for every basis
element B ∈ B if f −1 (B) is open in X, then f : X → Y is continuous.

29
 −1
f −1 (Bα ) ⊂open X.
S S
Proof. V ⊂open Y ⇒ V = α∈J Bα ⇒ f (V ) = α∈J

Example 6.0.2. Let Rst and Rl denote standard and lower limit topologies
on R, respectively.The identity function

f : Rst → Rl
i) is not continuous but
x → x

f : Rl → Rst
ii) is continuous.
x → x
For

i) [a, b) ⊂ Rl but f −1 ([a, b)) = [a, b) is not open in Rst .

open

ii) (a, b) ⊂ Rst and f −1 ((a, b)) = (a, b) ⊂ Rl .

open open

Example 6.0.3. Any function f : X → Y is continuous if X has discrete

topology.

Example 6.0.4. Any function f : X → Y is continuous if Y has triv-

ial(indiscrete) topology.

Theorem 6.0.5. Let X and Y be topological spaces, and f : X → Y be a

function.
Then the following are equivalent:

1) f is continuous.

2) For any subset A of X, we have f (A) ⊂ f (A).

3) For every closed set B of Y , f −1 (B) is closed in X.

4) For each x ∈ X and each neighborhood V of f (x), there is a neighbor-

hood U of x such that f (U ) ⊂ V .

Proof. (1) ⇒ (2) ⇒ (3) ⇒ (1) and (1) ⇒ (4) ⇒ (1)

(1) ⇒ (2): Let f be continuous, A be a subset of X and x ∈ A.

(must show: f (x) ∈ f (A), i.e f (x) is a limit point off (A))
Let V be a neighborhood of f (x). Then f −1 (V ) ⊂ X must be a
open
neighborhood of x and hence it must intersect A, (because x is a limit
point of A), at some point y. Then V intersect f (A) in the point f (y),
so that f (x) is a limit point of f (A).
⇒ f (x) ∈ f (A)

30
(2) ⇒ (3): Let B be closed in Y , and set A = f −1 (B). (must show: A = A)
f (A) = f (f −1 (B))) ⊂ B (Worksheet-1)
Thus, if x ∈ A, then f (x) ∈ f (A) ⊂ f (A) ⊂ B = B
This means x ∈ f −1 (B) = A i.e A ⊂ A.

(3) ⇒ (1): For V ⊂ Y set B = Y − V which is closed.

open
Then f (B) = f −1 (Y −V ) = f −1 (Y )−f −1 (V ) = X −f −1 (V ) is closed
−1

by the assumption. Hence f −1 (V ) ⊂ X.

open

(1) ⇒ (4): Let x ∈ X and V be a neighborhood of f (x).

The set U = f −1 (V ) is a neighborhood of x such that f (U ) ⊂ V .
(f (U ) = f (f −1 (V )) ⊂ V )

(4) ⇒ (1): Let V be open in Y and x be a point in f −1 (V ).

Then f (x) ∈ V . By the assumption, there is a neighborhood Ux of
x such that f (Ux ) ⊂ V . Then Ux ⊂ f −1 (V ). This means f −1 (V ) =
∪
−1
Ux i.e union of open sets.
x∈f (V )
⇒ f −1 (V ) is open.

31
6.1 Homeomorphisms
Let X and Y be topological spaces, f : X → Y a bijection. If both f and
f −1 are continuous then f is called a homeomorphism. In this case X and
Y are said to be homeomorpic spaces.
Equivalently, a homeomorphism is a bijection f : X → Y such that
”f (U ) is open if and only if U is open”. This means: a homeomorphism is
not only a 1 − 1 correspondence between elements of X and Y , it is also a
1 − 1 correspondence between open sets of X and Y . Any property of X
defined through open sets (connected, compact, Haussdorff, ...) also holds
for Y .

Example 6.1.1. Let X = {x, y, z} and Y = {1, 2, 3},

and τX = {∅, X, {a}, {c}, {a, c}}, τY = {∅, X, {2}, {3}, {2, 3}} be topologies
of X and Y , respectively.
Is X and Y homeomorphism?
Yes. Let f : X → Y be defined with f (a) = 2, f (b) = 1, f (c) = 3. (1 − 1
and onto)
f takes open sets of X to open sets of Y .
∅ ↔ ∅g
X ↔ X
{a} ↔ {2}
{c} ↔ {3}
{a, c} ↔ {2, 3}

Example 6.1.2. Let X = {a, b, c} , τ1 = {φ, X, {a}, {b}, {a, b}} and
τ2 = {φ, X, {a}, {b}, {c}, {a, b}, {a, c}, {b, c}} : discrete topology. Is the
identity function (X, τ1 ) → (X, τ2 ) a homeomorphism?
Answer: f : X → X f (x) = x is clearly bijection and its inverse is itself.
But f is not continuous : f −1 ({c}) = {c} ∈ / τ1 =⇒ f is not a homeomor-
phism.
Show: There is no such homeomorphism (X, τ1 ) → (X, τ2 ).

Example 6.1.3. F : (−1, 1) → Rst (We consider X = (−1, 1) with

x
x 7→ 1−x 2

subspace topology from Rst )

0
F ≥ 0 =⇒ F is increasing =⇒ F is 1-1
F −1 = G(y) = √2y 2 is defined for all y ∈ R =⇒ F : onto.
1+ 1+4y
Both F and G are continuous =⇒ F is a homeomorphism.

Example 6.1.4. (Stereographic Projection)

32
S n − {1point} is homeomorphic to Rn
0
n = 1: f : S → R
x
(x, y) 7 → 1−y

0
f −1 : R →  S  are both continuous.
2x −1+x2
x 7→ ,
1+x2 1+x2

n = 2: f : S 2 − {northpole} → R2

{x, y, z} 7→ x
, y
1−z 1−z
 2 +b2

f −1 (a, b) = 1+a2a2 +b2 , 1+a2b2 +b2 , −1+a
1+a2 +b2
)

The result of this example is sometimes expressed as: Rn ∪ {∞} = S n

33
Theorem 6.1.5. Let X, Y, Z be topological spaces.
i) Constant functions are continuous.

ii) A ⊂ X i : A ,→ X the inclusion function i = id|A is continuous.

iii) If f : X → Y and g : Y → Z are continuous, so is g ◦ f : X → Z.

iv) If f : X → Y is continuous and A ⊂ X, then f |A is continuous.

Proof. i) Let f : X → Y constant i.e. f (x) = a ∈ Y for every x ∈ X.
For any open set V ⊂ Y f −1 (V ) = ∅ if a ∈/V
f −1 (V ) = X if a ∈ V =⇒ f is continuous.

iii) U ⊂open Z =⇒ (g ◦ f )−1 (U ) = f −1 (g −1 (U ))

((g −1 (U )) is open since g is constant, f −1 (g −1 (U )) is open since f is
continuous.)

ii) i : A → X let U ⊂open X. i−1 = U ∩ A ( U ∩ A open in A with subspace

topology)

iv) f|A = (f ◦ i) : A ,→ X → Y composition is continuous.

Theorem 6.1.6. (Pasting Lemma) Let A and B be closed in X s.t. X =

A ∪ B. If f : A → Y and g : B → Y are continuous and f (x) = g(x) for any
x ∈ A ∩ B,(then
f (x) if x ∈ A
h(x) = is continuous X → Y
g(x) if x ∈ B
(
x if x ≤ 0
Example 6.1.7. h(x) = x is continuous.
2
if x > 0

Theorem 6.1.8. f : A → X ×Y is continuous iff f1 and f2

a 7 → (f1 (a), f2 (a))
are continuous.
Proof. (=⇒) f1 = π1 (f ) and f2 = π2 (f ) are continuous, where πi is the
projection onto ith component.
(⇐=) Recall that basis elements in X × Y are of the form U × V , where
U and V are basis element for X and Y , respectively.
f −1 (U × V ) = f1−1 (U ) ∩ f2−1 (V ) open. (f1−1 (U ) and f2−1 (V ) are open since
f1 and f2 are continuous.)

34
Chapter 7

Product and Metric Topologies

7.1 Two topologies on a Product Space

We have seen the product topology on X × Y in section 4.2.
Qn Let’s generalize
this to products of the form: XQ= X1 × X2 × ... × Xn = i=1 Xi or
Y = X1 × X2 × X3 × ... = ∞ i=1 Xi .

Definition 7.1.1. (Box Topology) For open sets Ui ⊂ Xi the sets of the form

U1 × U2 × ... × Un and
U1 × U2 × U3 × ...
are bases elements of the box topologies on X and Y above, respectively.
i can be taken from any index set.
Definition 7.1.2. If πi : Xor Y → Xi is the ith projection then B =
{πi−1 (Ui )|Ui ⊂open Xi } is a subbasis for the product topology on X or Y .
Box topology and product topology are equal on the finitely many product
of spaces i.e on X = X1 × X2 × ... × Xn .
Q
Theorem 7.1.3. Let f : A → α∈j Xα be the function defined by f (a) =
(fα (a))α∈j for fα : A → Xα . f is continuous iff fα is continuous for all α ∈ j.
In this theorem we assume Π Xα has product topology. Product topology
is stronger than box topology on the product of infinitely many spaces.
Theorem 7.1.4.
Q (Comparison of the box and product
Q topologies) The box
topology on Xα has as basis all sets of the form Uα , where Uα is open
in Xα for each α. Q Q
The product topology Xα has as basis of the form Uα is open in Xα for
each α, and Uα equals Xα except for finitely many valuse of α.

35
Example 7.1.5. For Xn = R, let Rw =
Q
n∈N Xn = {(x1 , x2 , ..., xn , ...) | xn ∈
R, n ∈ N}

Define f : R → Rw
t 7→ (t, t, t, ...)
so that fn (t) = t is the nth coordinate function. f is continuous if Rw has
product topology by theorem 7.1.3,
but not if
it has the box topology.
Consider B = (−1, 1) × −1 , 1 × −1
2 2
, 1 × ... which is a basis element of
3 3
the box topology on Rw .
Claim. f −1 (B) is not open in R
Since (0, 0, 0, ..) ∈ B , 0 ∈ f −1 (B). If f −1 (B) is open then 0 must have a
neighborhood (−δ, δ) s.t.
f ((−δ, δ)) ⊂ B. But f ((−δ, δ)) = (−δ, δ)×(−δ, δ)×
... ⊂ (−1, 1) × −1 ,
2 2
1
× ... is impossible.

7.2 Metric Topology

Remember that a metric d on a set X is a (distance) function d : X × X → R
such that:

1) d(x, y) ≥ 0 for all y, x ∈ X and d(x, y) = 0 ⇐⇒ x = y,

2) d(x, y) = d(y, x) for all x, y ∈ X,
3) d(x, y) + d(y, z) ≥ d(x, z) for all x, y ∈ X.

In this case, Bd (x, ) = {y | d(x, y) < } is called the -ball centered at x.

Definition 7.2.1. The collection of -balls Bd (x, ) for x ∈ X and  > 0 is
a basis for a topology on X, called the metric topology induced by d.
Example S 7.2.2. Is B = {Bd (x, )|x ∈ X,  > 0} a basis ?
1- X = {Bd (x, ) | x ∈ X,  > 0}
2- Next, we need to show that for any couple of -balls B1 and B2 , and any
y ∈ B1 ∩ B2 , there is a third basis element B3 (another -ball) such that
B3 ⊂ B1 ∩ B2 as in Figure ?? (a)
Given an -ball B(x, ) and y 6= x in this ball, we can find B(y, δ) ⊂
B(x, ), by simply letting δ =  − d(x, y). If z ∈ B(y, δ), then since d(x, z) ≤
d(x, y) + d(y, z) ≤ d(x, y) + δ ≤ . This implies z ∈ B(x,) .(Figure ?? (b))
Then taking the smaller of such balls for points in the intersection, we can
fulfill item 2 above.
There are positive numbers δ1 and δ2 such that B(y, δ1 ) ⊂ B1 and B(y, δ2 ) ⊂
B2 . Now letting δ = min{δ1 , δ2 } we get B3 = B(y, δ) ⊂ B1 ∩ B2 .

36
 (a) (b)

Definition 7.2.3. A set U is open in this topology if for each y ∈ U there is

a δ > 0 such that Bd (y, δ) ∈ U . Obviously -balls are open.

1 for x 6= y
Example 7.2.4. Discrete metric d(x,y) =
0 for x=y
induces the discrete topology. Bd =(x,1)={x}
Example 7.2.5. Standard metric on R given by d(x,y)=|x − y| induces Rst
with its order topology. If x = (a + b)/2 then (a, b) = B(x, (b − a)/2) (i.e.
open intervals are -balls) and B(x, ) = (x − , x + ) (i.e. -balls are open
intervals).
Definition 7.2.6. If the topology on X is induced by a metric d, then (X, τ )
is called metrizable.
Theorem 7.2.7. Let d1 and d2 induce τ1 and τ2 on X respectively. Then,
τ1 ⊂ τ2 if and only if for every x ∈ X, and  > 0 there is δ > 0 such that
Bd2 (x, δ) ⊂ Bd1 (x, ).
Proof. This is theorem 3.2.10 with metric spaces and -balls as their bases
elements.
n
Theorem p 7.2.8. The topology on R induced by the euclidean metric
d(x,y)= (x1 − y1 )2 + ... + (xn − yn )2 and the square metric ρ(x, y) = max {|x1 − y1 |, ...|xn − yn |}
are the same as the product topology on Rn .
√
Proof. Observe that ρ(x, y) ≤ d(x, y) ≤ nρ(x, y). The first inequality
implies that Bd (x, ) ⊂ Bρ (x, ) .
Suppose y ∈ Bρ (x, √n ) ⇒ ρ(x, y) < √n
√
⇒ d(x, y) < n · √n = 
⇒ y ∈ Bd (x, /) and hence ⇒ Bρ (x, √n ) ⊂ Bd (x, ).
Thus, from the previous theorem τd ⊂ τρ ⊂ τd i.e. the topologies are
equal.

37
 To show that they are both equal to product topology, let B=((a1 , b1 )×...(an , bn ))
be a basis element and X=(x1 , ...xn ) ∈ B for all i=1,2,...,n. There is i s.t.
(xi − i , xi + i ) ⊂ (ai , bi ). Choose  = min {1 , ...n }, then Bρ (x, ) ⊂ B.
On the other hand, if y ∈ Bρ (x, ) then we need to find B such that y
∈ B ⊂ Bρ (s, ). But
Bρ (x, ) = {y ∈ Rn | max {|x1 − y1 |, ...|xn − yn |} < }
= {y ∈ Rn | |x1 − y1 | < , ...|xn − yn | < }
= {(y1 , ..., yn ) ∈ Rn | y1 ∈ (x1 − , x1 + ), ..., yn ∈ (xn − , xn + )}
= (x1 − , x1 + ) × ... × (xn − , xn + ) = B.

Definition 7.2.9. Let {xn } = {x1 , x2 , ..., xn , ... | xn ∈ X} be a sequence in

a topological space X. It converges to x ∈ X if for every neighborhood U of
x, there is n̄ ∈ N such that xn ∈ U , whenever n ≥ n̄.

Theorem 7.2.10. Let A be a subset of a topological space X and {an } ⊂ A

be a sequence converging to x. Then x ∈ Ā. The converse is true only if X
is metrizable.

Proof. Recall that x ∈ Ā iff any neighborhood of x intersects A. Since

liman = x, any neigborhood of x contains infinitely many terms of {an }
⊂ A.
For the converse, see worksheet 1, Q9.

Theorem 7.2.11. If f : X → Y is continuous and lim xn = x, then

lim f (xn ) = f (x). The converse is true if X is metrizable.

Proof. Let V ⊂ Y be a neighborhood of f (x). Since f is continuous, f −1 (V )

is a neighborhood of x. Therefore there is n̄ ∈ N such that n ≥ n̄ implies
xn ∈ f −1 (V ) or {λn }∞
n ⊂ f
−1
(V ).
On the other hand, let A ⊂ X and x ∈ Ā. By the above theorem, there
is a sequence {xn } ⊂ A such that lim xn = x. Since lim f (xn ) = f (x) and
f (xn ) ⊂ f (A) we have f (x) ∈ f (A). This gives f (Ā) ⊂ f (A) and hence f is
continuous.

Definition 7.2.12. Let fn : X → Y be sequence of functions from the set X

to the metric space (Y, d). {fn } is said to converge uniformly to f : X → Y ,
if given  > 0 there is an integer n̄ ∈ N such that d(fn (x), f (x)) <  for all
n > n̄ and x ∈ X.

Theorem 7.2.13. If {fn } and f are as in the above definition, then f is

continuous.

38
Chapter 8

Connected Spaces

The intermediate value theorem we have seen in Calculus courses relies on a

property that the closed interval [a, b] has.
Theorem 8.0.1 (Intermediate Value Theorem). If f : [a, b] → R is contin-
uous and r ∈ R is between f (a) and f (b), then there is c ∈ [a, b] such that
f (c) = r.
The property of [a, b] that makes IVT true is connectedness. Since f is
continuous, we will see that f ([a, b]) is also connected.
Definition 8.0.2. Let X be topological space. A separation of X is a pair
U, V of disjoint nonempty open subsets of X, whose union is X.
If X has no separation, then X is called connected.
Connectedness is a topological property, (A property is topological if it
is preserved under a homeomorphism) since it is defined through open sets.
If U, V is a separation of X, then U = X − V and V = X − U . This means
U and V are both open and closed. In other words, X is connected if and
only if ∅ and X are the only sets that are both open and closed.
Theorem 8.0.3. Let Y be a subspace of X and A and B be disjoint nonempty
sets of Y , with Y = A ∪ B. The pair A, B is a separation for Y (i.e. Y is
disconnected) if and only if neither A nor B contains a limit point of the
other.
Proof. (⇒) (Recall that K = K ∪ K 0 ). We need to show that closures of A
and B in Y are also disjoint. Since A and B form a separation, A and B
are both open and closed in Y . The closure of A in Y is A ∩ Y . Since A is
closed in Y we have A = A ∩ Y (similarly B = B ∩ Y is the closure of B in
Y ). Therefore closures of A and B are disjoint in Y .

39
 (⇐) Suppose A and B are disjoint nonempty sets whose union is Y ,
neither of which contains a limit point of the other. Then A ∩ B = ∅ and
A ∩ B = ∅. To show that A is closed in Y , note that A ∩ (Y − A) =
(A ∩ Y ) − (A ∩ A) = (A ∩ Y ) − A = ∅ ⇒ A = A ∩ Y . i.e. all limit points of
A are in A, implying that A is closed.
Similarly B = B ∩ Y . A and B are both closed and open.

Example 8.0.4. 1) X = {a, b} with the trivial topology τ = {∅, X}.

⇒ X is connected

2) Y= [−1, 0) ∪ (0, 1] = [−1, 1]\{0} ⊂ R is disconnected.

[−1, 0)and (0, 1]are both open and closed in Y, hence they form a separation.

3) Q is not connected. Only connected sbspaces of Q are the one point

sets If Y ⊂ Q has two points p and q, then one can choose an irrational
number a between p and q and write Y as
Y=[Y ∩ (∞, a)] ∪ [Y ∩ (a, ∞)] union of disjoint sets.

Example 8.0.5. X = {(x, y) | y = 0} ∪ {(x, y) | x > 0 and y = 1/x} is a

subset of R2 which is not connected.
Neither of the two subsets of X above contain limit points of the other, hence
they form a separation.

8.1 Constructing Connected Spaces

Lemma 8.1.1. If the sets C and D form a separation of X and Y is a
connected subspace of X, then Y lies entirely within C or D.

Proof. C and D are both open and closed in X. The sets C ∩ Y and D ∩ Y
are open in Y and
(C ∩ Y ) ∩ (D ∩ Y )) = (C ∩ D) ∩ Y = ∅ ∩ Y = ∅)
(C ∩ Y ) ∪ (D ∩ Y ) = Y ∩ (C ∪ D) = Y ∩ X = Y .
i.e. they form a separation for Y if both of them are nonempty. Since Y is
connected one of them must be empty.

Theorem 8.1.2. The union of a collection of connected subspaces of X that

have a point in common is connected.
T
Proof. Let {ASα }be a collection of connected subspaces of X, and p ∈ Aα
Claim : Y = Aα is connected.
Suppose not and assume Y = C ∪D is a separation of Y . By previous lemma,
p is in one of C or D. Let p ∈ C. Since Aα is connected for each α, it must

40
S in C or D and cannot lie in D because p ∈ C. Hence A ⊂ C for all α and
lie
Aα ⊂ C contradicting the fact that D 6= ∅

Theorem 8.1.3. Let A be a connected subspace of X. If A ⊂ B ⊂ A, then

B is also connected.

The theorem says if we add to a connected set some or all of its limit
points, then we still preserve connectedness.

Proof. Let A be connected and A ⊂ B ⊂ A. Suppose B is not connected

and B = C ∪ D is a seperation of B. By above lemma, A must lie entirely in
C or D. WLOG let A ⊂ C. Then A ⊂ C = C. Since C and D are disjoint
and A ⊂ B ⊂ A ⊂ C = C we have B ∩ D = ∅. D must be empty which is a
contradiction since C ⊂ D is a seperation of B.

Theorem 8.1.4. Continuous image of a connected set is connected. i.e. If

f : X → Y is continuos and X is connected, then f (X) is connected.

Proof. Any map on to its image is onto. Let’s consider f : X → Y as

f : X → f (X) = Z which becomes onto. (must show: Z is connected.)
Suppose Z = A ⊂ B is a seperation into two disjoint nonempty sets open in
Z. Then f −1 (A) and f −1 (B) forms a separation for X:

ˆ both open, since f is cont and A, B open.

ˆ disjoint, since A, B are disjoint.

ˆ nonempty, since f is onto and A and B are nonempty.

This contradicts to the fact that A is connected.

Theorem 8.1.5. Finite product of connected spaces is connected.

Proof. Let X and Y be connected spaces and a × b ∈ X × Y . X × b is

homeomorphic to X and x × Y is homeomorphic to Y . Therefore there are
also connected. The set Tx =( X × b )∪(x × Y ) is the union of two connected
a common point x × b. This implies Tx is also connected. But
sets with S
X × Y = x∈X Tx so X × Y is also a union of connected spaces and must as
well be connected.
the proof for X1 × · · · × Xn = (X1 × · · · × Xn−1 ) × Xn is by induction.

Remark 8.1.6. Product of infinitely many connected spaces may or may not
be connected depending on the topology.

41
Example 8.1.7. Rw = RN = { (x1 , x2 , . . . , xn , . . . )|xi ∈ R i ∈ N } can be
seen as the space of sequences. With box topology on it Rw is not connected.
A = {{an } | an ∈ R and {an }is a bounded sequence.}
B = {{bn } | bn ∈ R, {bn }is unbounded}
Clearly A ∩ B = ∅. A and B are open: Let a ∈ Rw .
a = {a1 , a2 , . . . } ∈ U = (a1 − 1, a1 + 1) × (a2 − 1, a2 + 1) × · · · ⊂ A if a is
bounded,
a = {a1 , a2 , . . . } ∈ U = (a1 − 1, a1 + 1) × (a2 − 1, a2 + 1) × · · · ⊂ B if a is
unbounded.
Thus A and B form a separation for Rw with box topology.

Example 8.1.8. Consider Rw with product topology.

Let R̃n = {(x1 , x2 , . . . , xn , 0, 0, . . . ) | xi ∈ R} ⊂ Rw which is homeomor-
phic to Rn . Assuming that R is connected, by previous theorem Rn and hence
R̃n are connected.
Note that 0 = (0, . . . , 0, 0, . . . ) ∈ R̃n for any n ∈ N, therefore the set R∞ =
n
S
n∈N R̃ is also connected.

Claim. R∞ = Rw (and hence Rw also connected.)

Proof. We need to show that any a ∈ Rw is a limit point of R∞ i.e. any base
∞
element inQthe product topology containing a intersects R .
Let U = Ui be a basis element containing a = (a1 , a2 , . . . ).
Then there is N ∈ N such that Ui = R for i > N (see Theorem 7.1.4)
The point x= (a1 , a2 , . . . , an , 0, . . . ) ∈ R∞ ∩ U since
ai ∈ Ui for all i and 0 ∈ Ui for i > N

Question. If τ1 ⊂ τ2 are two topologies on X are the following true or false?

i. (X, τ1 ) connected =⇒ (X, τ2 ) connected.

ii. (X, τ2 ) connected =⇒ (X, τ1 ) connected. environment.

8.2 Connected Subspaces of R

A set X is simply ordered if there is a relation ≤ such that:

i. a ≤ b and b ≤ a imply a = b (antisymmetry)

ii. a ≤ b and b ≤ c imply a ≤ c (transitivity)

iii. a ≤ b or b ≤ a (comparability)

42
(One can also add reflexivity a ≤ a)

Definition 8.2.1. A simply ordered set L having more than one element is
a linear continuum if

i. L has the least upper bound property

ii. If x < y there exists z ∈ L s.t. x < z < y.

Recall that a set has the l.u.b property if every nonempty subset with an upper
bound has a least upper bound. (supremum)

Theorem 8.2.2. If L is a linear continuum in the order topology then L is

connected and so are intervals and rays in L.

Proof. see Munkres.

Corollary 8.2.3. R is connected and so are intervals and rays in R

Theorem 8.2.4. (Intermediate Value Theorem) Let X be a connected space

and Y have order topology. If f : X −→ Y is continuous, a, b ∈ X and r is
between f (a) and f (b), then there is c ∈ X such that f (c) = r.

Proof. The sets A = f (X) ∩ (−∞, r) and B = f (X) ∩ (r, ∞) are disjoint
and nonempty because one contains f (a) and the other contains f (b). A and
B are open in f (X) because they are intersections of open rays with f (X).
If there is no c ∈ X with f (c) = r, then f (X) = A ∪ B is a separation for
f (X). This contradicts to the fact that continuous image of a connected set
is connected.

8.2.1 Path Connected Spaces

Definition 8.2.5. Given points x and y of X, a path in X from x to y is a
continuous function f : [a, b] → X of some closed interval of R into X such
that f (a) = x and f (b) = y. A space X is path-connected if every pair of
points in X can be joined by a path in X.

Observe : P ath connected ⇒ connected. Suppose not and let A ∪ B be

a separation for the path connected space X. Let f : [a, b] → X be any path
in X. Since f is continuous and [a, b] is connected, f ([a, b]) is also connected.
So it must either lie entirely in A or entirely in B. This means we cannot
find a path from a point of A into a point of B which is a contradiction.
The converse is not true. The ”topologist’s sine curve” is connected but
not path-connected.

43
 Figure 8.1: Topologist’s sine curve

Example 8.2.6. Let S = {(x, y) ∩ R2 | 0 < x ≤ 1 , y = sin( x1 )}

Since f (x) = sin( x1 ) is continuous on (0, 1], S is connected. Therefore
S (called the topologist’s sine curve) is also connected. S = S∪ { (0,y)
| y ∈ [−1, 1]}. Claim : S is not path connected.

Proof. Suppose not and let f : [a, c] → S be a path from (0, 0) to a point
of S. Since f is continuous and {0} x [−1, 1] is closed, f −1 ({0} x [−1, 1])
is closed and hence has a largest element b. Then f : [b, c] → S is a path
that maps b into {0} x [−1, 1] and other points of [b, c] to points of S. For
convenience replace [b, c] by [0, 1] and let f (t) = (x(t), y(t)). Then x(0) = 0
1
and when t > 0 we have x(t) > 0 and y(t) = sin x(t) . Construct a sequence
1
{tn } as follows: For n ∈ N let u be 0 < u < x( n ) s.t. sin( u1 ) =(−1)n . By IVT,
since x(t) is continuous, there’s tn with 0 < tn < n1 s.t. x(tn ) =u. Thus the
sequence {tn } is a sequence of points converging to 0. But y(t)= sin( x(t1n ) )
= sin( u1 ) = (−1)n which is divergent. This contradicts to the fact that f is
continuous.

8.2.2 Connected Components

Definition 8.2.7. Given X, define an equivalance relation on X by setting
x ∼ y if there is a connected subspace of X containing both x and y. The
equivalence classes are called the components of X.

Recall that the union of two connected subsets, having a point in common,
must also be connected.

Theorem 8.2.8. The components of X are connected disjoint subspaces of

X whose union is X, such that each nonempty connected subspace of X
intersects only one of them.

44
Definition 8.2.9. Define an equivalence relation ∼ on X by x ∼ y if and
only if there is a path in X from x to y. The equivalence classes are called
path components.

A path from x to y was a continuous function f : [a, b] → X such that

f (a) = x and f (b) = y. For convenience let’s take [a, b] = [0, 1] (any two
closed interval in R are homeomorphic). Let’s see that ∼ is an equivalence
relation:

i) (Reflexive) f : [0, 1] → {x} defined as f (t) = x the constant map is a

path from x to x i.e. x ∼ x

ii) (Symmetric) If f : [0, 1] → X is a path from x to y, then f (1 − t) is a

path from y to x ⇒ ”x ∼ y ⇒ y ∼ x”

iii) (Transitive) Let f : [0, 1] → X and g : [0, 1] → X be paths from x to y

and from y to z, respectively.
Define (
f (2t) 0 ≤ t ≤ 12
h(t) =
g(2t − 1) 12 ≤ t ≤ 1
which is continuous by pasting lemma ( f and g coincide on the inter-
section). Hence h is a path from x to z ⇒ x ∼ z

Example 8.2.10. The topologist’s sine curve has one component but two
path components.

Example 8.2.11. Q has components consisting of single points.

45
Chapter 9

Compact Spaces

Compactness is not as natural as connectedness but it is an important prop-

erty of some Euclidian subspaces. The proofs of many important results in
analysis require compactness.

Definition 9.0.1. A collection A of subsets of a space X is a cover to X

A is equal to X.
(or covering), if the union of elements of S
i.e. A = {Aα | α ∈ Λ, Aα ⊂ X} and X = α∈Λ Aα

If elements of A are open in X, then A is an open covering of X.

Definition 9.0.2. If a subcollection B of a covering A is a cover itself, then

B is said to be a subcover of A.

Definition 9.0.3. A space X is said to be compact if every open covering

has a finite subcover.

Remark 9.0.4. Negating the above definition we get: ”X is not compact if

it has an open cover without a finite subcover”.

Example 9.0.5. R is not compact.

Consider A = {(n, n + 2) | n ∈ Z} which is an open cover for R. But

no finite subcollection of A can cover R.

Example 9.0.6. If X is a finite set, then any open cover will be finite. Hence
its subcovers are also finite. So X must be compact.

Example 9.0.7. X = {0} ∈ { n1 | n ∈ N} is compact.

Proof. Suppose A is an open cover for X and let U ∈ A be any open element
of A containing 0.

46
All but finitely many points of X are in U . If U1 , U2 , ..., Uk are elements of
A containing these points that are not in U , then B = {U, U1 , U2 , U3 , ..., Uk }
is a finite subcover.

Example 9.0.8. If K = { n1 | n ∈ N} then K is not compact.

Proof. For each n ∈ N let n = 12 ( n1 − n+1

1 1
) = 2n(n+1) then Un = ( n1 − n , n1 +
n ) ∩ K = { n1 } is an open set in the subspace topology of K that contains
only n1 . The open cover A = {U1 , U2 , ...} has no subcover.

Example 9.0.9. The interval (0, 1] is not compact. A = {( n1 , 1] | n ∈ N} is

an open cover without a finite subcover.

Example 9.0.10. (0, 1) is not compact.

Example 9.0.11. [0, 1] is compact.

Lemma 9.0.12. A subspace Y of X is compact iff every covering of Y by

sets open in X has a finite subcovering of Y .

Proof. (⇒) Let Y be compact (i.e. every open (in Y ) cover of Y has a finite
subcover) and A = {Aα | α ∈ j} be a covering of Y by sets open in X.
Then {Aα ∩ Y | α ∈ j} is an open cover of Y by sets open in Y .
Since Y is compact, this open cover must have a finite subcover:
{Aα1 ∩ Y, Aα2 ∩ Y, ..., Aαn ∩ Y }.
Thus, {Aα1 , Aα2 , ..., Aαn } is a subcover of A for Y

(⇐) Let A = {Aα } be a covering of Y by sets open in Y . For each α,

choose a set Bα open in X s.t. Aα = Bα ∩ Y .
The collection B = {Bα } is a covering of Y by sets open in X. By hypothe-
sis some fine subcollection {Bα1 , ..., Bαn } covers Y . Then {Aα1 , ..., Aα1 } is a
subcover of A for Y .

Theorem 9.0.13. Every closed subspace of a compact space is compact.

Proof. Let Y be a closed subspace of the compact X. Given a covering A of

Y by sets open in x (By above lemma we need to find a finite subcollection
of A covering Y ) the union B = A ∩ X − Y is an open cover for the compact
space X. If this subcollection contains X-Y discard X-Y , otherwise leave it
as it is.

Theorem 9.0.14. Continuous image of a compact space is compact.

47
Proof. Let f:X→ Y be continuous and X be compact. Let A be a covering
of the set f (X) by sets open in Y . Then {f −1 (A) | A ∈ A} is a collection
of sets covering X. These are open in X because f is continuous. Hence
finitely many of them , say f −1 (A1 ), . . . ,f −1 (An ) cover X. Then A1 , . . .
,An cover f (X).
Theorem 9.0.15. Let X be a simply ordered set having the least upper bound
property. In the order topology, each closesd interval in X is compact .
Theorem 9.0.16. Every closed interval in R is compact.
Theorem 9.0.17. (Heine-Borel Theorem)A subspace of Rn is compact
if and only if it is closed and bounded in the euclidean metric or the square
metric.
Example 9.0.18. K = { n1 | n ∈ N} is not compact because it is not closed.
0 is its limit point but 0 ∈
/ K.
Example 9.0.19. S n−1 and Dn = B n in Rn are compact because they are
closed and bounded.
Example 9.0.20. Rn is not bounded hence Rn is not compact.
Example 9.0.21. Rn =S n−1 -{point} is not compact.
Example 9.0.22. {(x, x1 ) | 0 < x ≤ 1} ⊂ R2 is closed but not bounded.
Therefore it cannot be compact.
Example 9.0.23. {(x, y) | y = sin x1 , x ∈ (0, 1]} is bounded but not closed
⇒ not compact.
Theorem 9.0.24. (Extreme Value Theorem) Let f : X → Y be con-
tinuous where Y is an ordered set in the ordered topology. If X is compact,
then there exists points c and d in X such that f (c) ≤ f (x) ≤ f (d) for every
x ∈ X.
In Calculus courses we have seen this theorem with X as a closed interval
and Y = R.
Proof. Since f is continuous and X is compact, the set A = f (X) is compact.
We must show that A has a largest element M and a smallest element m.
Suppose A has no largest element. Then the collection {(−∞, a) | a ∈ A}
forms an open covering of A . Since A is compact this cover must have a
finite subcover:
B = {(−∞, a1 ), (−∞, a2 ), ...(−∞, an )}. If ai is the largest of the elements
a1 , ..., an , then ai does not belong to these sets. This contradicts to the fact
that B is a cover for A. With a similar argument we can show the existence
of a smallest element of A.

48
Theorem 9.0.25. (Tychonoff theorem)) An arbitrary product of compact
spaces is compact in the product topology.

Thus, a product space is compact if and only if all factor spaces are
compact. Note that the theorem is true even for uncountable product of
spaces.

Example 9.0.26. i. The n-dimensional torus T n = (S 1 )n = S 1 ×· · ·×S 1

is compact since S 1 is compact.

ii. [0, 1] × [0, 1) is not compact since [0, 1) is not closed and hence not
compact.

iii. [0, 1]ω is compact.

49
Chapter 10

Separation Axioms

Consider the sequence {an | an = a f or all n ∈ N} = {a, a, a, ...} of the

space X = {a, b}. If X is endowed with the trivial topology τ = {φ, X},
then this sequence converges to a, and also to b (lim an = x ∈ X iff for every
neighborhood U of x there’s N ∈ N such that an ∈ U for n ≥ N ).
In Calculus courses, we used to have convergent sequences with unique
limits or ,if existed, we expect lim f (x) to be a unique number.The above
x→x0
example suggests that the trivial topology does not have enough number of
open sets to separate two points, even on a small set like X = {a, b}.
Definition 10.0.1. A topological space X is a T0 -space if whenever x and
y are distinct points in X, then there is an open set containing one and not
the other.
Example 10.0.2. Trivial topology (τ = {φ, X}) on any set with more than
one point is not T0 .
Example 10.0.3. Subspaces and products of T0 spaces are T0
Definition 10.0.4. A topological space X is a T1 -space if, whenever x and y
are distinct points in X, then there is a neighborhood of each not containing
the other.
Clearly, every T1 -space is T0 .
Example 10.0.5. τ = {φ, {a}, {a, b}} is T0 but not T1 on X = {a, b}
Theorem 10.0.6. The following are equivalent, for a topological space X:

a) X is T1 ,

b) One-point sets (singletons) are closed in X,

50
 c) Each subset of X is the intersection of the open sets containing it.

Proof. (a⇒b): If X is T1 and x ∈ X, S

then each y 6= x has a neighborhood Uy
disjoint from {x}. Then X − {x} = Uy is open and hence {x} is closed.
y6=x
(b⇒c): Let A ⊂ X be a subset. For each x ∈ X − A the set S
X - {x}
is open by the hypothesis, and A ⊂ X − {x}. Then: A = X − {x} =
T x ∈A
/
(X − {x}) (De Morgan).
x∈X−A
(c⇒a):TBy c), the set {x} is also intersection of open sets containing x:
i.e {x} = Uα where x ∈ Uα for all α. Thus, if y 6= x then for some α0 we
must have y ∈
/ Uα0 .

The real importance of T1 -spaces lies in the observation above: one-point

sets are closed in T1 -spaces.

Theorem 10.0.7. Let A be a subset of T1 -space X. Then x ∈ X is a limit

point of A if and only if every neighborhood of x contains infinitely many
points of A.

Proof. (⇐): Definition of limit requires that every neighborhood of x should

intersect A in some point other than x itself. Since neighborhoods of x
intersect A in infinitely many points, it must intersect in some point other
than x.
(⇒): Suppose U is a neighborhood of x that intersects A in finitely many
points and let {x1 , ..., xm } = U ∩(A − {x}). This set is closed since X is T1 .
Therefore X −{x1 , ..., xm } is open. Setting V = U ∩(X −x1 , ..., xm ) we obtain
a neighborhood of x which does not intersect A − {x}. This contradicts to
the fact that x is a limit point of A.

10.1 Hausdorff Spaces

A space X is T2 (or Hausdorff) if, whenever x and y are distinct points of
X, there are disjoint open sets U and V in X with x ∈ U and y ∈ V .
In order to have unique limits for convergent sequences in X, the topology
on X must be Hausdorff.

Example 10.1.1. Let X be an infinite set with the cofinite topology. (i.e. a
subset A ⊂ X is open if and only if X-A is finite) Then, the closed sets are
X and finite sets, in particular one-point sets are closed. Therefore X is a
T1 -space. Non-empty open sets cannot be disjoint, hence X is not T2 .

51
Example 10.1.2. Every metric space is Hausdorff. If x and y are distinct,
then let ε = d(x, y) > 0 be distance between them. The ε-balls Bd (x, 2ε ) = U
and Bd (y, 2ε ) = V are open disjoint sets containing x and y, respectively.

Theorem 10.1.3. If X is a Hausdorff space, then a sequence of points of X

converges to at most one point in X.

Proof. Let {xn } be a sequence converging to x ∈ X. If y 6= x, let U and

V be disjoint neighborhoods of x and y respectively. Since U contains all
but finitely many terms of {xn }, the set V cannot. Therefore {xn } cannot
converge to y.

Theorem 10.1.4. Every simply ordered set is a Hausdorff space in order

topology. Product of Hausdorff spaces is a Hausdorff space. Subspace of a
Hausdorff space is Hausdorff.

10.1.1 Properties of Hausdorff Spaces

Theorem 10.1.5. Every compact subspace of Hausdorff space is closed.

Proof. Let Y be a compact subspace of the T2 -space X. Let x0 ∈ X − Y .

We need to show that x0 has a neighborhood U ⊂ X − Y (or U ∩Y = ∅) so
that X − Y is open. Since X is Hausdorff, for each y ∈ Y there are disjoint
neighborhoods Uy and Vy of x0 and y respectively. Then A = {Vy |y ∈ Y }
is an open cover of Y . Since Y is compact A must have a finite subcover
{Vy1 , ..., Vyn }. Setting V = Vy1 ∪ ... ∪ Vyn and U = Uy1 ∪ ... ∪ Uyn we obtain
disjoint open sets U and V where Y ⊂ V and x0 ∈ U .

Theorem 10.1.6. Continuous image of a compact space is compact.

Proof. Let f : X → Y be continuous, X be compact and A be an open cover

for f (X). Then the collection {f −1 (Aα )|Aα ∈ A} is an open cover for X,
since f is continuous. Hence finitely many of them {f −1 (A1 )},...,{f −1 (An )}
cover X. This implies A1 , ..., An cover f (X).

Theorem 10.1.7. Let f : X → Y be 1 − 1, onto and continuous. If X is

compact and Y is Hausdorff, then f is a homeomorphism.

Proof. We need to show that f −1 is also continuous. This is equivalent to

f being a closed function i.e. f (A) must be closed for every closed subset
A ⊂ X. Closed subsets of compact spaces are compact. Hence A is compact.
Since f is continuous, f (A) is also compact. By the above theorem, f (A)
must be closed.

52
Chapter 11

Countability Properties

A space X is said to have a countable basis at x, if there is a countable

collection B of neighborhoods of x such that each neighborhood of x contains
at least one of the elements of B. A space that has a countable basis at each
of its points is said to be first-countable.

Example 11.0.1. Every metrizable space is first-countable. Let x be a point

of the metric space (X, d) and set B = {Bd (x, n1 |n ∈ N} which is a countable
collection of concentric ε-balls. If U ⊂ X is any neighborhood of x then there
is ε > 0 s.t. Bd (x, ε) ⊂ U . For any natural number N ∈ N s.t. N > 1ε we
have Bd (x, N1 ) ⊂ Bd (x, ε) ⊂ U .

Definition 11.0.2. The collection B in the above definition is also called a

local basis at x. Therefore a first-countable space must have a local basis at
every point in it.

Theorem 11.0.3. Let X be a topological space;

(a) If there is a sequence of points of A ⊂ X converging to x ∈ X, then

x ∈ A; the converse holds if X is first-countable.

(b) Let f : X → Y be continuous and {xn } converges to x. Then {f (xn )}

converges to f (x). The converse holds if X is first-countable.

Definition 11.0.4. If a space X has a countable basis for its topology, then
X is said to be second-countable. Clearly: second countable ⇒ first countable.
(Countable basis B for X can be taken as a local basis at any x ∈ X)

Example 11.0.5. B = {(a, b)|a, b ∈ Q} is a countable basis for the real line
R. Thus R is second-countable.

53
Theorem 11.0.6. Subspaces and countable products of first-countable (re-
spectively second-countable) spaces are also first-countable (respectively second-
countable).
Definition 11.0.7. A subset A of a space X is said to be dense in X if
A = X.
Example 11.0.8. Q is dense in R i.e. Q = R.
Theorem 11.0.9. Let X be a second-countable space.
(a) Every open cover of X has a countable subcover.
(b) There is a countable dense subset of X.
Proof. Let {Bn } be a countable basis for X.
(a) Let A be an open cover of X. For n ∈ N, if possible, choose An ∈ A
s.t. Bn ⊂ An . The collection A0 = {An |n ∈ j ⊂ N} is countable and is
a subcover of A : Let x ∈ X and choose A ∈ A containing x. Since A
is open, there is Bn s.t. x ∈ Bn ⊂ A. This means n ∈ j because it is
possible for Bn to find an element of A (A in this case) containing Bn .
Thus An is defined and hence x ∈ Bn ⊂ An ⇒ A0 covers X.
(b) From every nonempty basis element Bn choose a point xn . Let D be
the set of such points. Then D is dense in X. If x ∈ X then every
basis element containing x intersects D ⇒ x ∈ D.

Definition 11.0.10. A space for which every open cover has a countable
subcover is called Lindelöf space.
Definition 11.0.11. A space having a countable dense subset is said to be
separable.
The above theorem says that a 2nd countable space is both Lindelöf and
separable.
Example 11.0.12. Rl is 1st countable, Lindelöf and separable but not 2nd
countable:
Recall that sets of the form [a, b) constitute a basis for Rl .
1st countable : Given x ∈ Rl , βx = {[x, x+(1/n)] | n ∈ N} is a countable
basis at x.
2nd countable : For any basis β for Rl , choose an element of B such that
x ∈ βx ⊂ [x, x + 1) for every x ∈ Rl . If x 6= y, then βx 6= βy because
x = inf βx and y = inf βy . Therefore β must be uncountable.
Lindelöf : See Munkres.

54
Remark 11.0.13. Product of Lindelöf Spaces need not be Lindelöf.

Remark 11.0.14. Subspaces of Lindelöf Spaces need not be Lindelöf.

11.1 Uncountability
Recall that an infinite set X is countable if there is a surjection:
f : N → X or there is an injection g : X → N
The result below uses topological properties of R to show no such f exists.

Definition 11.1.1. If X is a space, a point x ∈ X is an isolated point of X,

if the one-point set {x} is open in X.

Theorem 11.1.2. Let X be a nonempty compact Hausdorff space. If X has

no isolated points, then X is uncountable.

Corollary 11.1.3. Every closed interval in R is uncountable.

55
Chapter 12

Regular and Normal Spaces

Let X be a topological space.

Definition 12.0.1. X is a regular space if, for each pair consisting of a point
x and a closed set B disjoint from x, there exist disjoint open sets containing
x and B, respectively.

Definition 12.0.2. X is a normal space if, for each pair A, B of disjoint

closed sets of X, there are disjoint open sets containing A and B.

Remark 12.0.3. If X is a T1 -space so that one-point sets are closed, then

i. it is a T3 space if it is also regular, and

ii. it is a T4 space if it is also normal.

T4 ⇒ T3 ⇒ T2 ⇒ T1 ⇒ T0

56
Theorem 12.0.4. a) X is regular if and only if for any x ∈ X and a
neighborhood U of x there is a neighborhood V of x such that V ⊂ U

b) X is normal if and only if for any closed set A and an open set U
containing A there is an open set V such that A ⊂ V ⊂ V ⊂ U

Proof. (a) (⇒) Let X be regular and suppose that the point x and the
neighborhood U of x are given. Let B = X − U which is a closed set not
containing x. Then there are disjoint open sets V and W containing x and
B respectively.

claim: V and B are disjoint.

Proof. A point y ∈ B can not be a limit point of V , since W is a neighborhood

of y disjoint from V . Therefore, V ⊂ U .
(⇐) Let x ∈ X and B be a closed set not containing x. Set U = X − B
which is a neighborhood of x. Then there is a neighborhood V of x such that
V ⊂ U . The open sets V and X - V are disjoint open sets containing x and
B respectively. Thus X is regular.

Theorem 12.0.5. Subspaces and products of Hausdorff spaces are Hausdorff.

Theorem 12.0.6. Subspaces and products of regular spaces are regular.

Remark 12.0.7. The above theorem does not hold for normal spaces.

Example 12.0.8. The space RK is Hausdorff but not regular. For K =

{(1/n)|n ∈ N} the sets of the form (a, b) − K with open intervals (a, b) form
a basis for RK . Since RK is finer than Rst obviously RK is T2 .

Claim : RK is not regular.

Proof. The set K is closed and 0 6∈ K. We want to show that we cannot sep-
arate 0 and K. Suppose the open sets U and V contain 0 and K respectively.
Basis elements containing 0 and lying in U must be of the form (a, b) − K,
otherwise they intersect K. For large enough n, let (1/n) ∈ (a, b). Since V
is a neighborhood of (1/n), a basis element containing (1/n) and lying in
V must be of the form (c, d). Choose z such that z > max{c, (1/(n + 1))}.
Then z belongs to both U and V , so they are not disjoint.

57
Example 12.0.9. Rl is normal.

Let A and B be disjoint closed sets in Rl . For each point a ∈ A choose

a basis element [a, xa ) not intersecting B, and for each b ∈ B choose [b, xb )
not intersecting
S A. Then the S open sets
U = a∈A [a, xa ) and U = b∈B [b, xb )
are disjoint open sets containing A and B, respectively.
Example 12.0.10. R2l is not normal.

12.1 Normal Spaces

Theorem 12.1.1. Every regular space with a countable basis is normal.
Theorem 12.1.2. Every metrizable space is normal
Proof. Let (X, d) be a metric spaces. Let A and B be disjoint closed subsets
of X. For each a ∈ A choose a so that Bd (a, a ) ∩ B = ∅. Similarly for each
b ∈ B, choose b so that Bd (b, b ) ∩ A = ∅

B a, 2a and U = b∈B B b, 2b

S
S

U= a∈A

Then U and V are open sets containing A and B, respectively. They are
also disjoint. If z ∈ U ∩ V , then z ∈ Bd (a, 2a ) ∩ Bd (b, 2b ) for some a ∈ A
and b ∈ B. By the triangle inequality d (a, b) < a + 2
b
. If a ≤ b , then
d (a, b) < b which means a ∈ Bd (b, b ) If b ≤ a , then d (a, b) < a implying
that b ∈ Bd (a, a ) Both cases are impossible
Theorem 12.1.3. Every compact Hausdorff space is normal
Proof. Let X be a compact Hausdorff space.
Claim: X is regular.

Proof: Let B ⊂ X be a closed set and x ∈ X with x ∈

/ B. Then B is compact.
Since X is Hausdorff, fore every b ∈ B choose open sets b ∈ Ub and a ∈ Vb

58
s.t. Ub ∩ Vb = ∅. Then {Ub |b ∈ B} is an open cover for B, which should have
a finite subcover, say {Ub1 , Ub2 ...Ubn }. If we let
Sn Tn
U= i=1 Ubi and V = i=1 Vbi ,

then we find disjoint open sets containing B and x respectively. 

Claim: X is normal.

Proof: The proof is essentially the same as above. Let A and B disjoint
closed sets of X. For each point a ∈ A, choose disjoint open sets Ua and Vb
containing a and B, respectively (We can do this, since X is regular). Since
A is compact its open cover {Ua } has a finite subcover {Ua1 , Ua2 ...Uan }. Then,
Sn Tn
U= i=1 Uai and V = i=1 Vai

are disjoint open sets containing A and B respectively 

Theorem 12.1.4. Every well-ordered set in the order topology is normal

(In fact, every order topology is normal)

12.1.1 Urysohn’s Lemma

This is a deep result used in proving a number of important theorems. Its
proof involves an original idea which is beyond the scope of this course (See
Munkres).

Theorem 12.1.5. (Urysohn’s Lemma) Let X be a normal spaces. Let A and

B be disjoint closed subsets of X. Let [a, b] be a closed interval in the real
line. Then there exist a continuous map f : X → [a, b] such that
f (A) = a and f (B) = b.

As a consequence we have the Urysohn metrization theorem.

Theorem 12.1.6. Every regular space X with a countable basis is metrizable.

59
Bibliography

[1] J. Munkres, Topology, 2nd Edition, Pearson, (2014).

[2] S. Willard, General Topology, Addison-Wesley, (1970).

60