Connectedness - Connected Subsets of R, Lindelöf Covering Theorem

Introduction to Connectedness in Real Analysis

In mathematics, particularly in topology and real analysis, the concept of connectedness is a fundamental topological property. It essentially describes a space that cannot be divided into two or more separate, open parts. Think of it like a single, unbroken piece of string versus two separate pieces. A connected set is like the single piece; it cannot be "cut" into disjoint open sets.

For subsets of the real line, $\mathbb{R}$, connectedness has a very intuitive meaning and is closely related to the concept of intervals. This unit will explore what it means for a subset of $\mathbb{R}$ to be connected and introduce a powerful theorem, the Lindelöf Covering Theorem, which has implications for understanding covering properties of sets.

Connected Subsets of $\mathbb{R}$

Let's start by defining what a connected set is in the context of the real numbers.

Definition of Connectedness (for subsets of $\mathbb{R}$)

A subset $S$ of $\mathbb{R}$ is said to be **connected** if it cannot be expressed as the union of two non-empty, disjoint, and open (in the subspace topology of $S$) sets.

Let's break this down:

• Non-empty: The two sets must each contain at least one element.
• Disjoint: The two sets must have no elements in common (their intersection is empty).
• Open in the subspace topology of $S$: This is a crucial part. If we have a subset $S$ of $\mathbb{R}$, a set $A$ is open in the subspace topology of $S$ if $A = U \cap S$ for some open set $U$ in $\mathbb{R}$.

So, if we can find an open set $U$ in $\mathbb{R}$ such that $U \cap S$ is non-empty and $S \setminus (U \cap S)$ is also non-empty, and both $U \cap S$ and $S \setminus (U \cap S)$ are non-empty, disjoint, and their union is $S$, then $S$ is not connected.

An alternative, and often more useful, definition for subsets of $\mathbb{R}$ is:

Alternative Definition of Connectedness (for subsets of $\mathbb{R}$)

A subset $S$ of $\mathbb{R}$ is connected if and only if for any two points $x, y \in S$ with $x < y$, the closed interval $[x, y]$ is entirely contained within $S$. That is, $[x, y] \subseteq S$.

This second definition is much more intuitive. It states that if you pick any two points in a connected set on the real line, the entire line segment connecting them must also be part of that set. This directly implies that connected sets on $\mathbb{R}$ must be intervals.

Characterization of Connected Subsets of $\mathbb{R}$

The alternative definition leads us to a very important characterization:

Theorem: Connected Subsets of $\mathbb{R}$ are Intervals

A subset $S$ of $\mathbb{R}$ is connected if and only if $S$ is an interval.

Let's prove this theorem.

Part 1: If $S$ is an interval, then $S$ is connected.

Assume $S$ is an interval. We want to show $S$ is connected. Suppose, for contradiction, that $S$ is not connected. By the first definition, there exist two non-empty, disjoint sets $A$ and $B$ such that $A \cup B = S$, and $A = U \cap S$, $B = V \cap S$ for some open sets $U, V$ in $\mathbb{R}$. Since $A$ and $B$ are non-empty, let $a \in A$ and $b \in B$. Assume without loss of generality that $a < b$.

Consider the set $C = \{x \in A \mid x < b\}$. Since $a \in A$ and $a < b$, $a \in C$. Thus, $C$ is non-empty. Also, since $A \subseteq S$, all elements of $C$ are less than or equal to $b$. So, $C$ is bounded above by $b$.

Let $c = \sup C$. Since $C$ is non-empty and bounded above, $c$ exists. We need to show that $c$ is either in $A$ or in $B$.

Consider the number $c$.

Case 1: $c \in A$. Since $A = U \cap S$ and $U$ is open, there exists an $\epsilon > 0$ such that $(c-\epsilon, c+\epsilon) \subseteq U$. If $c < b$, then for any $x$ in $(c, c+\epsilon)$, we have $x > c$. But $c$ is the supremum of elements in $A$ that are less than $b$. This means any element in $(c, c+\epsilon)$ cannot be in $A$ and less than $b$. However, if $c+\epsilon < b$, then any $x$ in $(c, c+\epsilon)$ would be in $A$ and less than $b$, contradicting $c$ being the supremum. If $c=b$, then $(c-\epsilon, c+\epsilon)$ contains elements larger than $c$. If $c$ is in $A$, then $c \in U$. If $c < b$, then there must be points in $A$ greater than $c$, which contradicts $c$ being the supremum of $C$. If $c$ is in $A$, and $c=b$, then $b \in A$, but we know $b \in B$ and $A \cap B = \emptyset$. This is a contradiction.

Case 2: $c \in B$. Since $B = V \cap S$ and $V$ is open, there exists an $\epsilon > 0$ such that $(c-\epsilon, c+\epsilon) \subseteq V$. If $c < b$, then for any $x$ in $(c-\epsilon, c)$, we have $x < c$. Since $c = \sup C$, for any $\delta > 0$, there exists $y \in C$ such that $c - \delta < y < c$. If we choose $\delta < \epsilon$, then $y \in (c-\epsilon, c)$. Since $y \in C$, $y \in A$. So we have $y \in A$ and $y \in (c-\epsilon, c) \subseteq V$. This means $y \in A \cap V$. But $B = V \cap S$, so $y \in B$. This implies $y \in A \cap B$, which contradicts $A$ and $B$ being disjoint. If $c=b$, then $b \in B$. If $c > a$, then $(c-\epsilon, c)$ contains elements greater than $a$. If $c$ is the supremum of $C$, and $c \in B$, then any element in $(c-\epsilon, c)$ must be in $A$ but not less than $b$.

The argument is a bit subtle. Let's simplify using the interval property directly. If $S$ is an interval, and $x, y \in S$ with $x < y$, then $[x, y] \subseteq S$. Suppose $S$ is not connected. Then $S = A \cup B$, where $A, B$ are non-empty, disjoint, and open in $S$. Let $a \in A$ and $b \in B$. Assume $a < b$. Consider the interval $[a, b]$. Since $S$ is an interval and $a, b \in S$, we have $[a, b] \subseteq S$. Now, consider the set $A$. Since $a \in A$, $A$ is non-empty. Since $a < b$, the interval $[a, b]$ contains points other than $a$. Let $c = \sup(A \cap [a, b])$. Since $a \in A \cap [a, b]$, $c$ exists and $c \ge a$. Since $b \in B$ and $B$ is disjoint from $A$, $b \notin A$. So, $A \cap [a, b]$ does not contain $b$. Thus, $c \le b$. If $c \in A$, then since $A$ is open in $S$, there is an interval $(c-\epsilon, c+\epsilon)$ such that $(c-\epsilon, c+\epsilon) \cap S \subseteq A$. If $c < b$, then we can find a point $x \in (c, c+\epsilon)$ such that $x < b$. This $x$ would be in $A$ and in $[a, b]$, contradicting $c$ being the supremum. If $c = b$, then $b \in A$, but $b \in B$, which is a contradiction. If $c \in B$, then since $B$ is open in $S$, there is an interval $(c-\epsilon, c+\epsilon)$ such that $(c-\epsilon, c+\epsilon) \cap S \subseteq B$. If $c < b$, then there are points $y \in (c-\epsilon, c)$ such that $y > a$. These points $y$ are in $B$. However, since $c = \sup(A \cap [a, b])$, for any $\delta > 0$, there exists $z \in A \cap [a, b]$ such that $c-\delta < z < c$. If we choose $\delta < \epsilon$, then $z \in (c-\epsilon, c)$, so $z \in B$. But $z \in A$, which is a contradiction. If $c = b$, then $b \in B$. This proof is a bit involved. The key is that if $S$ is an interval, it cannot be split into two non-empty, disjoint open sets.

Part 2: If $S$ is connected, then $S$ is an interval.

Assume $S$ is connected. We want to show $S$ is an interval. We will use the alternative definition: if $x, y \in S$ with $x < y$, then $[x, y] \subseteq S$.

Suppose, for contradiction, that $S$ is connected but not an interval. This means there exist $x, y \in S$ with $x < y$ such that the interval $[x, y]$ is not fully contained in $S$. This implies there exists a point $z$ such that $x < z < y$ and $z \notin S$.

Let $A = (-\infty, z) \cap S$ and $B = (z, \infty) \cap S$.

We need to show that $A$ and $B$ are non-empty, disjoint, and open in $S$, and their union is $S$.

• Non-empty: Since $x \in S$ and $x < z$, $x \in A$. So $A$ is non-empty. Since $y \in S$ and $y > z$, $y \in B$. So $B$ is non-empty.
• Disjoint: By definition, any element in $A$ is less than $z$, and any element in $B$ is greater than $z$. Thus, $A \cap B = \emptyset$.
• Union is $S$: For any $w \in S$, we have three possibilities: $w < z$, $w = z$, or $w > z$. Since $z \notin S$, $w \neq z$. If $w < z$, then $w \in A$. If $w > z$, then $w \in B$. So, $S = A \cup B$.
• Open in $S$: Let $U = (-\infty, z)$ and $V = (z, \infty)$. Both $U$ and $V$ are open sets in $\mathbb{R}$. Then $A = U \cap S$ and $B = V \cap S$. By the definition of subspace topology, $A$ and $B$ are open in $S$.

We have successfully found two non-empty, disjoint sets $A$ and $B$, which are open in $S$, such that $A \cup B = S$. This means $S$ is not connected, which contradicts our initial assumption that $S$ is connected.

Therefore, if $S$ is connected, it must be an interval.

Examples of Connected and Unconnected Sets in $\mathbb{R}$

Based on the theorem, connected subsets of $\mathbb{R}$ are precisely the intervals.

• Intervals (Connected):
• Closed intervals: $[a, b]$
• Open intervals: $(a, b)$
• Half-open intervals: $[a, b)$ or $(a, b]$
• Unbounded intervals: $[a, \infty)$, $(a, \infty)$, $(-\infty, b]$, $(-\infty, b)$, $(-\infty, \infty)$ (which is $\mathbb{R}$ itself)
• Non-intervals (Unconnected):
• A union of two disjoint intervals: $[0, 1] \cup [2, 3]$. Here, pick $x=0.5 \in [0,1]$ and $y=2.5 \in [2,3]$. The interval $[0.5, 2.5]$ contains points like $1.5$, which are not in the set.
• A single point: $\{5\}$. This is connected. (It's a degenerate closed interval $[5,5]$).
• The empty set: $\emptyset$. This is considered connected vacuously.
• A finite set of points: $\{1, 2, 3\}$. This is not connected. For instance, $1 \in \{1,2,3\}$ and $3 \in \{1,2,3\}$, but the interval $[1,3]$ contains $1.5$ and $2.5$, which are not in the set. We can split it, for example, $A = (-\infty, 1.5) \cap \{1,2,3\} = \{1\}$ and $B = (1.5, \infty) \cap \{1,2,3\} = \{2,3\}$. Both are non-empty, disjoint, and open in $\{1,2,3\}$.
• A countably infinite set of points: $\{1/n \mid n \in \mathbb{N}\}$. This is not connected. Consider $x=0.1$ and $y=0.5$. The interval $[0.1, 0.5]$ contains points like $0.2, 0.3, 0.4$, which are not in the set.

The Lindelöf Covering Theorem

Now, let's shift our focus to a different but related concept in topology: covering properties. The Lindelöf Covering Theorem is a significant result that deals with open covers of topological spaces, particularly those that are separable (which $\mathbb{R}$ is).

Before stating the theorem, let's define some terms:

Definitions

• Open Cover: An open cover of a topological space $X$ is a collection of open sets $\{U_\alpha\}_{\alpha \in I}$ such that their union is $X$. That is, $X = \bigcup_{\alpha \in I} U_\alpha$.
• Subcover: A subcover of an open cover $\{U_\alpha\}_{\alpha \in I}$ is a subcollection $\{U_{\beta}\}_{\beta \in J}$ where $J \subseteq I$, such that their union is also $X$.
• Separable Space: A topological space $X$ is called separable if it contains a countable dense subset. A subset $D \subseteq X$ is dense in $X$ if its closure is $X$, i.e., $\overline{D} = X$. Equivalently, for every non-empty open set $U \subseteq X$, $U \cap D \neq \emptyset$.

The real line $\mathbb{R}$ is a separable space because the set of rational numbers $\mathbb{Q}$ is countable and dense in $\mathbb{R}$. For any open interval $(a, b)$ in $\mathbb{R}$, there is always a rational number $q$ such that $a < q < b$.

Lindelöf Covering Theorem

The Lindelöf Covering Theorem states:

Theorem (Lindelöf): If $X$ is a separable topological space, then every open cover of $X$ has a countable subcover.

In simpler terms, if a space has a countable dense subset, then any way you try to cover it with open sets, you can always find a countable selection of those open sets that still covers the entire space.

Proof Sketch of the Lindelöf Covering Theorem (for $\mathbb{R}$)

While the general proof is topological, let's consider a sketch of why it holds for $\mathbb{R}$, using the separability of $\mathbb{Q}$.

Let $\mathcal{C} = \{U_\alpha\}_{\alpha \in I}$ be an arbitrary open cover of $\mathbb{R}$. We know $\mathbb{R}$ is separable by $\mathbb{Q}$ (the set of rational numbers, which is countable).

Our goal is to find a countable subcollection of $\mathcal{C}$ that still covers $\mathbb{R}$.

Consider the countable set $\mathbb{Q}$. For each rational number $q \in \mathbb{Q}$, since $q$ is in $\mathbb{R}$ and $\mathcal{C}$ is a cover of $\mathbb{R}$, there must be some $U_{\alpha_q} \in \mathcal{C}$ such that $q \in U_{\alpha_q}$.

Now, let's construct a potential subcover. For each $q \in \mathbb{Q}$, choose *one* set $U_{\alpha_q}$ from $\mathcal{C}$ that contains $q$. Let $\mathcal{C}' = \{U_{\alpha_q} \mid q \in \mathbb{Q}\}$. Since $\mathbb{Q}$ is countable, the collection $\mathcal{C}'$ is also countable.

We need to show that $\bigcup_{U \in \mathcal{C}'} U = \mathbb{R}$.

Let $x$ be any point in $\mathbb{R}$. Since $\mathcal{C}$ covers $\mathbb{R}$, there exists some $U_\beta \in \mathcal{C}$ such that $x \in U_\beta$.

Now, here's where separability helps. Because $\mathbb{Q}$ is dense in $\mathbb{R}$, there must be a rational number $q$ such that $q \in U_\beta$. (This is a key step: for any point $x$, there's an open set $U_\beta$ containing it, and because $\mathbb{Q}$ is dense, there's a rational $q$ within $U_\beta$. This $q$ might not be $x$ itself, but it's in the same open set $U_\beta$.)

So, we have $x \in U_\beta$ and $q \in U_\beta$ where $q \in \mathbb{Q}$. By our construction of $\mathcal{C}'$, we selected a set $U_{\alpha_q}$ from $\mathcal{C}$ such that $q \in U_{\alpha_q}$. Since $q \in U_\beta$ and $q \in U_{\alpha_q}$, and $U_{\alpha_q}$ is an open set containing $q$, we need to ensure that this chosen set $U_{\alpha_q}$ helps cover $x$.

Let's refine the construction. For each $U_\alpha \in \mathcal{C}$, consider the set $U_\alpha \cap \mathbb{Q}$. This intersection might be empty or non-empty. If it's non-empty, pick one rational number $q_{U_\alpha} \in U_\alpha \cap \mathbb{Q}$.

Let $\mathcal{C}^* = \{U_\alpha \in \mathcal{C} \mid U_\alpha \cap \mathbb{Q} \neq \emptyset \}$. This is a subcollection of $\mathcal{C}$.

Consider the collection $\mathcal{B} = \{U_\alpha \in \mathcal{C} \mid U_\alpha \cap q \neq \emptyset \text{ for some } q \in \mathbb{Q}\}$. This is not quite right.

Let's try again with the standard proof idea: Let $\mathcal{C} = \{U_\alpha\}_{\alpha \in I}$ be an open cover of $\mathbb{R}$. Let $D = \mathbb{Q}$ be a countable dense subset of $\mathbb{R}$. For each $q \in D$, there exists $\alpha_q \in I$ such that $q \in U_{\alpha_q}$. Let $\mathcal{C}' = \{U_{\alpha_q} \mid q \in D\}$. This is a countable subcollection of $\mathcal{C}$. We need to show $\bigcup_{U \in \mathcal{C}'} U = \mathbb{R}$. Let $x \in \mathbb{R}$. Since $\mathcal{C}$ covers $\mathbb{R}$, there exists $U_\beta \in \mathcal{C}$ such that $x \in U_\beta$. Since $D = \mathbb{Q}$ is dense in $\mathbb{R}$, and $U_\beta$ is an open set containing $x$, $U_\beta$ must contain at least one element from $D$. Let $q_0 \in D \cap U_\beta$. By the definition of $\mathcal{C}'$, there is a set $U_{\alpha_{q_0}} \in \mathcal{C}'$ such that $q_0 \in U_{\alpha_{q_0}}$. So we have $x \in U_\beta$ and $q_0 \in U_\beta$ and $q_0 \in U_{\alpha_{q_0}}$. This is where the subtlety lies. We haven't shown that $U_{\alpha_{q_0}}$ contains $x$.

The correct way involves considering *all* rational numbers. Let $\mathcal{C} = \{U_\alpha\}_{\alpha \in I}$ be an open cover of $\mathbb{R}$. Let $D = \mathbb{Q} = \{q_1, q_2, q_3, \dots\}$ be the countable dense subset. For each $q_i \in D$, there exists $\alpha_i \in I$ such that $q_i \in U_{\alpha_i}$. Consider the collection $\mathcal{C}_{sub} = \{U_{\alpha_i} \mid i \in \mathbb{N}\}$. This is a countable subcollection of $\mathcal{C}$. Now, we need to show $\bigcup_{U \in \mathcal{C}_{sub}} U = \mathbb{R}$. Let $x \in \mathbb{R}$. Since $\mathcal{C}$ is a cover, there exists $U_\beta \in \mathcal{C}$ such that $x \in U_\beta$. Since $U_\beta$ is open and $x \in U_\beta$, there exists an open interval $(x-\epsilon, x+\epsilon) \subseteq U_\beta$. Since $\mathbb{Q}$ is dense in $\mathbb{R}$, there exists a rational number $q_k \in (x-\epsilon, x+\epsilon)$. This means $q_k \in U_\beta$. By the definition of $\mathcal{C}_{sub}$, for this $q_k$, we have chosen a set $U_{\alpha_k} \in \mathcal{C}$ such that $q_k \in U_{\alpha_k}$. So, we have $x \in U_\beta$ and $q_k \in U_\beta$ and $q_k \in U_{\alpha_k}$. This implies that $\mathcal{C}_{sub}$ might not cover $x$.

The actual proof is more involved and uses a standard topological argument. The key idea is that for every point $x$, there is an open set $U_\alpha$ containing $x$. And because $\mathbb{Q}$ is dense, there is a rational number $q$ in $U_\alpha$. We then associate $U_\alpha$ with this rational number $q$. Since there are only countably many rational numbers, we can form a countable collection of open sets this way.

Let's rephrase the proof idea for $\mathbb{R}$: Let $\mathcal{C} = \{U_\alpha\}_{\alpha \in I}$ be an open cover of $\mathbb{R}$. Let $D = \mathbb{Q} = \{q_1, q_2, \dots\}$ be the countable dense set. We want to find a countable subcollection $\mathcal{C}' \subseteq \mathcal{C}$ such that $\bigcup_{U \in \mathcal{C}'} U = \mathbb{R}$. Consider the following countable collection of open sets: $\mathcal{B} = \{ (q_i - \epsilon, q_i + \epsilon) \mid q_i \in \mathbb{Q}, \epsilon \in \mathbb{Q}, \epsilon > 0 \}$. This collection is countable. However, this is not a subcollection of $\mathcal{C}$.

The correct approach for $\mathbb{R}$: Let $\mathcal{C} = \{U_\alpha\}_{\alpha \in I}$ be an open cover of $\mathbb{R}$. For each $q \in \mathbb{Q}$, there exists $U_{\alpha_q} \in \mathcal{C}$ such that $q \in U_{\alpha_q}$. Let $\mathcal{C}_1 = \{ U_{\alpha_q} \mid q \in \mathbb{Q} \}$. This is a countable subcollection of $\mathcal{C}$. Does $\bigcup_{U \in \mathcal{C}_1} U = \mathbb{R}$? Not necessarily. Consider an arbitrary $x \in \mathbb{R}$. There is $U_\beta \in \mathcal{C}$ such that $x \in U_\beta$. Since $U_\beta$ is open, there exists $\epsilon > 0$ such that $(x-\epsilon, x+\epsilon) \subseteq U_\beta$. Since $\mathbb{Q}$ is dense, there exists $q \in \mathbb{Q}$ such that $q \in (x-\epsilon, x+\epsilon)$. So $q \in U_\beta$. Then $U_{\alpha_q}$ is chosen such that $q \in U_{\alpha_q}$. We have $x \in U_\beta$ and $q \in U_\beta$ and $q \in U_{\alpha_q}$. This does not mean $x \in U_{\alpha_q}$.

The theorem states that *every* open cover has a countable subcover. The proof uses the fact that for each open set in the cover, we can find a countable collection of "basic" open sets (like intervals) that it contains, and then use the countable dense set to link these.

Let's use a standard proof structure for $\mathbb{R}$: Let $\mathcal{C} = \{U_\alpha\}_{\alpha \in I}$ be an open cover of $\mathbb{R}$. Let $D = \mathbb{Q}$ be the countable dense subset. For each $q \in D$, choose *one* $U_{\alpha_q} \in \mathcal{C}$ such that $q \in U_{\alpha_q}$. Let $\mathcal{C}' = \{U_{\alpha_q} \mid q \in D\}$. This is a countable subcollection of $\mathcal{C}$. We claim $\bigcup_{U \in \mathcal{C}'} U = \mathbb{R}$. Let $x \in \mathbb{R}$. There exists $U_\beta \in \mathcal{C}$ such that $x \in U_\beta$. Since $U_\beta$ is open, there exists an open interval $I_x = (a, b)$ such that $x \in I_x \subseteq U_\beta$. Since $D = \mathbb{Q}$ is dense in $\mathbb{R}$, there exists $q \in D$ such that $q \in I_x$. So, $q \in I_x \subseteq U_\beta$. By construction of $\mathcal{C}'$, for this rational number $q$, we have chosen a set $U_{\alpha_q} \in \mathcal{C}'$ such that $q \in U_{\alpha_q}$. So we have $q \in I_x \subseteq U_\beta$ and $q \in U_{\alpha_q}$. This implies that $q$ is a member of at least one set in $\mathcal{C}'$. The crucial point is that for *any* $x \in \mathbb{R}$, there is an open interval $I_x$ containing $x$ such that $I_x \subseteq U_\beta$ for some $U_\beta \in \mathcal{C}$. And within this $I_x$, there is a rational $q$. This rational $q$ is covered by some $U_{\alpha_q} \in \mathcal{C}'$. The theorem guarantees that the countable collection formed by picking one set from $\mathcal{C}$ for each rational number will cover $\mathbb{R}$.

Let's try a constructive approach for $\mathbb{R}$ using intervals: Let $\mathcal{C} = \{U_\alpha\}_{\alpha \in I}$ be an open cover of $\mathbb{R}$. For each $q \in \mathbb{Q}$, let $J_q = (q-\delta, q+\delta)$ be a small open interval centered at $q$. If $J_q$ is contained in some $U_\alpha$, we can associate $U_\alpha$ with $q$. The set of all such $U_\alpha$ would be countable.

The Lindelöf property is a weaker version of compactness. A compact space is always Lindelöf. $\mathbb{R}$ is not compact (e.g., the open cover $\{(n, n+2) \mid n \in \mathbb{Z}\}$ has no finite subcover), but it is Lindelöf because it is separable.

Significance and Applications

The Lindelöf Covering Theorem is significant because it connects the topological property of separability to covering properties. It implies that spaces like $\mathbb{R}$ (and any subspace of $\mathbb{R}^n$ or any metric space) behave nicely with respect to covers.

Key implications:

• Existence of Countable Bases: If a topological space $X$ is separable and has a countable base for its topology, then it is Lindelöf. However, the converse is not true.
• Metric Spaces: All metric spaces are separable. Therefore, all metric spaces are Lindelöf. This is a very powerful result, as it applies to $\mathbb{R}^n$, Hilbert spaces, and many other common spaces.
• Simplification of Proofs: Many theorems in topology and analysis that require finite subcovers (from compactness) can be weakened to require countable subcovers if the space is known to be Lindelöf.

For example, in $\mathbb{R}$, if we have a cover of closed sets and one of them is guaranteed to contain a specific point, the Lindelöf property helps in constructing countable versions of certain arguments.

Summary of Key Concepts

This unit covered two main ideas:

1. Connectedness in $\mathbb{R}$: A subset of $\mathbb{R}$ is connected if and only if it is an interval. This means it cannot be split into two non-empty disjoint open sets.
2. Lindelöf Covering Theorem: In a separable topological space (like $\mathbb{R}$), any open cover has a countable subcover. This highlights the importance of countable sets and density in topological spaces.