INTRODUCTION

Linearly H-closed spaces are a concept within the realm of general topology,
which is a branch of mathematics concerned with studying the properties of spaces
and mappings that are preserved under continuous deformations.The notion of
H-closedness and linearly H-closed spaces specifically arise from investigating
certain closure properties of subsets within a topological space.
Linearly H-closed spaces extend this notion to subsets within a topological
space.Intuitively, linearly H-closed subsets have closure properties that resemble
those of H-closed spaces but within a smaller subset of the space.They ensure that
certain kinds of infinite intersections of closed sets within the subset do not vanish,
thus capturing a kind of compactness within that subset.
These spaces find applications in various areas of mathematics, including
functional analysis, topology, and set theory, where properties related to
compactness and closure are crucial for understanding the behavior of
mathematical objects and structures.The aim to provide a foundational
understanding of linearly H-closed spaces, setting the stage for a deeper
exploration of their intricate mathematical properties and their significance in
theoretical and applied contexts.

In Chapter-1: We discuss the preliminaries.

In Chapter-2: We see the basics of Linearly H-Closed spaces.
In Chapter-3: We see the linearly H-Closed spaces.
In Chapter-4: We see the real-life applications of linearly H-closed spaces in
topology.

1
 CHAPTER 1
PRELIMINARIES

Introduction
Now we can see about the preliminaries of H-closed spaces in topology.

Definition 1.1
Let X be a non empty set. A family τ of subsets of X is said to be
topology on X .If τ satisfy the following
(1) X ∈ τ
(2) ϕ ∈ τ
(3) A ∈ τ and B ∈ τ ⇒ A ∩ B ∈ τ and A ∪ B ∈ τ
The pair ( X, τ ) is called Topological Space.

Definition 1.2
A topological space ( X, τ ) is said to be Compact if every open cover of X
has a finite subcover.

Definition 1.3
A topological space X is said to be Hausdorff Space if for each points of
distinct points x and y in X there exists an open neighborhood U of x and an
open neighborhood of V of y such that U ∩ V = ϕ .
It is also known as τ2 space.

Definition 1.4
A hausdorff space is said to be H-Closed if it is closed in every hausdorff
space containing it as a subspace.
(i.e,) a Hausdorff topological space is H-closed if it is closed in every
Hausdorff topological space in which it is embedded.

2
Definition 1.5
Let ( X, τ1 )and ( X, τ2 ) be topological spaces.
The function f: X → Y is said to be Continuous if the inverse image of
every open set of Y is open in X .
If V ∈ τ Y ,then f −1 (V ) ∈ τ Y .

Definition 1.6
A topological space ( X, τ ) and a non empty subset S of X ,the Subspace
topology on S is defined by τs = {S ∩ U ; U ∈ τ } .
The subspace topology is given by ( S, τs ).

Definition 1.7
A topological space ( X, τ )is Countable if there is countable neighborhood
base at each of its points.

Definition 1.8
A topological space (X, τ ) is Feebly Compact if every locally finite cover
of non empty open sets is finite.

Definition 1.9
Let X and Y be topological spaces.
A function h : X → Y is called a Homeomorphism between X and Y if
(1) h is continuous on X to Y
(2) h−1 is continuous on Y to X .

Definition 1.10
Let X be a topological space. A Sequence in X is defined as a mapping
from the set of natural numbers N to X . The general term of a sequence is
denoted by xn and the sequence is denoted by (xn ) .

3
Definition 1.11
A Subcover of a cover of a set is a subset of the cover that also covers the
set.

Definition 1.12
A topological space (X, τ ) is said to be Pseudocompact if its image under
any continuous function to R is bounded.

Definition 1.13
A topological space is Maximal R if every continuous bijection from a
space (Y, V ) with property R to (X, τ ) is a homeomorphism.

Definition 1.14
A point x is called an Isolated Point of a subset S of a topological space
X if x is an element of S and there exists a neighborhood of x that does not
contain any other points of S .

Definition 1.15
A space( X, τ ) is said to be Linearly H-Closed if in every open chain
cover of the space there is a dense element.
Every linear order with no maximum element has a cofinal subset indexed
by a regular cardinal, we may assume that each open chain cover is of the form Uα
where k is a regular infinite cardinal and Uα ⊆ Uβ if α < β .

Definition 1.16
An Accumulation Point or a limit point of a set S in a topological space
( X, τ ) is a point x that can be ”approximated” by points of S in the sense that
every neighborhood of x with respect to the topology on X also contains a point
of S other than x itself.

4
Definition 1.17
A Complete Accumulation Point of a set M in a topological space X
is a point x ∈ X such that the intersection of M with any neighborhood of x has
the same cardinality as the entire set M .

Definition 1.18
The Euclidean Space of n dimensions En , is defined as the topology
generated by the basis (Rn , d) ,where Rn is the set and d is the Euclidean metric
d(X, Y ) .

Definition 1.19
If ( X, τ ) is a topological space, then the Semi-Regularization topology
τs on X of T is the coarser topology on X generated by the family of all open
domains of ( X, τ ).
( X, τ ) is called semiregular if τ = τs .
( X, τs ) is the semi-regularization space of a topological space ( X, τ ).

Definition 1.20
Let (X, τ ) be a topological space and F be a family of open subsets of X .
The family F is said to be Locally Countable if it satisfies the following
property: for every x ∈ X , there exists an open subset x ∈ U (x) ⊆ X such that
there are only countably many members in τ which intersect U (x) .

Definition 1.21
Two spaces X and Y are said to be Topologically Equivalent if there
exists a homeomorphism h : X → Y between them.

5
Definition 1.22
In mathematics, a Paracompact space is a topological space in which
every open cover has an open refinement that is locally finite.
Every compact space is paracompact.

Definition 1.23
A Cover of a set X is a family of subsets of X whose union is all of X .
If C = Uα : α ∈ A is an indexed family of subsets of Uα ∈ X then C is
S
a cover of X if X ⊆ α∈A Uα .
A cover of a topological space X is open if all its members are open sets.

Definition 1.24
In a topological space ( X, τ ), a Neighborhood of a point p is defined as
a set which contains the point p together with some open set that contains p .
N is a neighborhood of p if there exists U ∈ τ such that p ∈ U ⊂ N .

Definition 1.25
A topological space is termed a Tychonoff space if it is Hausdorff, and
given any point x ∈ X and closed subset A ⊆ X such that x ∈
/ A , there exists a
continuous map f : X → [0,1] such that f (x) = 0 and f (a) = 1 for all a ∈ A .

6
 CHAPTER-2
ON LINEARLY H-CLOSED SPACES

Now we can see about linearly H-closed spaces related theorems and examples.

Theorem 2.1
A regular closed subspace of a linearly H -closed space is linearly H -closed.
Proof:
Suppose that A ⊆ X is a regular closed subset and that U = {Uα : α < κ} is
an open chain cover of A .
Then for each α < κ there are open sets Vα in X such that Uα = A ∩ Vα and
then Vα ∪ (X\A) = Uα ∪ (X\A) and {Uα ∪ (X\A) : α < κ} is an open chain
cover of X .
Thus there is some β < κ such that Uβ ∪ (X\A) is dense in X .
Hence Uβ is dense in int(A) and it follows that cl (Uβ ) ⊇ cl(int(A)) = A .

Theorem 2.2
If D is dense in X and D is linearly H -closed, then X is linearly H -closed.
Proof:
Let A be a linearly H -closed subset of X .
Since D is dense in X , for any point x in X , there exists a sequence (dn )
in D converging to x .
Since A is linearly H -closed, any sequence in A converging in X must have
its limit in A .
Consider a sequence (an ) in A converging to x .
Since (dn ) converges to x , you can construct a new sequence (dnk ) by taking
elements from (dn ) such that dnk is within 1/k of x .
Show that (dnk ) is a sub sequence of (an ) by choosing indices appropriately.

7
 Since A is linearly H -closed, the limit of (dnk ) is in A .
Since limits are unique in a Hausdorff space, the limit of (dnk ) must also be x .
Therefore, x is in A .
Hence, A is closed in X .
This completes the proof that X is linearly H -closed.

Theorem 2.3
A continuous image of a linearly H-closed space is linearly H-closed.
Proof:
To prove that a continuous image of a linearly H -closed space is linearly
H -closed, we need to show that any sequence in the image space that converges
has its limit in the image space as well.
Let X be a linearly H -closed space, and let f : X → Y be a continuous
function where Y is another topological space.
Let (yn ) be a sequence in f (X) that converges to some y ∈ Y .
Since (yn ) is in the image of f , there exists a sequence (xn ) in X such that
f (xn ) = yn for all n .
Since X is linearly H -closed, any sequence in X converges to a limit in X .
Therefore, the sequence (xn ) converges to some x ∈ X .
Since f is continuous, f (xn ) converges to f (x) .
By the uniqueness of limits, y = f (x) , which means y is in the image of f .
Thus, any sequence in f (X) that converges has its limit in f (X) .
Therefore, f (X) is linearly H -closed.
This completes the proof.

Theorem 2.4
The property of being linearly H-closed is an inverse invariant of perfect open
maps.
Proof:

8
 Suppose that f : X → Y is a (surjective) open perfect map, where Y is
linearly H -closed and let F = {Uα : α < κ} be an open chain cover of X .
Since f is closed and surjective, it follows that Vα = Y \f [X\Uα ] is an open
subset of Y and we claim that G = {Vα : α < κ} is an open cover of Y .
To prove this claim, suppose that y ∈ Y ; the set f −1 [{y}] is a compact
subset of X which is covered by the family F and hence there is some γ < κ
such that f −1 [{y}] ⊆ Uγ .
But then, y ∈
/ f [X\Uγ ] and so y ∈ Vγ .
Since G is an open chain family and Y is linearly H -closed, it follows that for
some β < κ, Vβ is dense in Y and we then claim that Uβ is dense in X .
For if not, then, since f is an open map, f [X\ cl (Uβ )] is a non-empty open
subset of Y disjoint from Vβ , which is a contradiction.
As an immediate corollary, we have that the product of a compact space and a
linearly H -closed space is linearly H -closed.
It follows from a space which is feebly compact is linearly H -closed.
A topological space (X, τ ) is said to be densely countably compact , if it
possesses a dense subspace D with the property that every infinite subset (or
equivalently, every countably infinite subset) of D has an accumulation point in
X.
A feebly compact space with a dense set of isolated points is easily seen , while
a feebly compact space with a countable dense set of isolated points is ccc and
hence linearly H -closed; in particular a space is linearly H -closed.
The properties of being (respectively, countably compact) and being linearly
H -closed are not comparable however: A sub maximalization of [0, 1] is H -closed
, while ω1 is countably compact but not linearly H -closed.
T
A filter F on a space X is said to be free if {cl(F ) : F ∈ F} = ∅ .
A filter on a space X is a regular closed filter (respectively, an open filter) if it
has a base of regular closed sets (respectively, open sets).
The existence of a free regular closed filter with a nested base of regular closed

9
sets on a space X is easily seen to be equivalent to the existence of a free filter on
X with a nested base of open sets .
Furthermore, if a filter has a nested base, then it has a nested base of regular
cardinality.
Note that if U = {Uα : α < κ} is an open chain cover of a space X then so is
U ∗ = {int (cl (Uα )) : α < κ} .
Furthermore, some element of U is dense in X if and only if the
corresponding element of U ∗ is dense in X .

Theorem 2.5
A space X is linearly H -closed if and only if there is no free open filter on X
with a nested base.
Proof:
If there exists some free regular closed filter F on X with a nested base of
regular closed sets B = {Bα : α < κ} , then {X\Bα : α < κ} is an open chain
cover of X which clearly has no dense element. Conversely, if there is some open
chain cover {Uα : α < κ} no element of which is dense in X , then
{int (cl (Uα )) : α ∈ κ} is also an open chain cover of X with no dense element and
so {X\ int (cl (Uα )) : α ∈ κ} is a nested base for a free regular closed filter on X .
The result follows from the remarks preceding the theorem.
We note in passing that the non-existence of a free open filter on X with a
countable (nested) base is equivalent to the feeble compactness of X .

Example 2.6
Let κ and λ be infinite cardinals; a generalized (deleted) Tychonoff plank of
the form (κ+ 1) × (λ + 1)\{(κ, λ)} is linearly H -closed if and only if
cof(κ) ̸= cof(λ) .

Corollary 2.7
A feebly compact space is linearly H-closed.

10
Proof:
To prove that a feebly compact space is linearly h -closed, we need to show
that every closed linearly ordered subset in the space is itself closed.
Let X be a feebly compact space and A be a closed linearly ordered subset of
X.
We want to show that A is closed in X .
Since A is closed, its complement X − A is open.
We will show that X − A is open by showing that every point in X − A has
a neighborhood contained entirely in X − A .
Let x be a point in X − A .
Since A is closed, every point a ∈ A has a neighborhood Ua in X such that
Ua ∩ A = {a} .
Since X − A is open, x has a neighborhood Vx such that Vx ⊆ X − A .
Consider the collection of neighborhoods {Ua }a∈A .
Since A is linearly ordered, this collection is locally finite. By the feebly
compactness of X , there exists a finite subcover {Ua1 , Ua2 , . . . , Uan } of {Ua }a∈A .
Let U = ni=1 Uai . U is an open set in X .
S

Now, consider the intersection U ∩ Vx . Since U only contains points from A

and Vx only contains points from X − A , their intersection is empty.
Thus, for every point x ∈ X − A , there exists a neighborhood Vx such that
Vx ⊆ X − A , which implies that X − A is open.
Therefore, A is closed in X , and hence, X is linearly h -closed. This
completes the proof.

Corollary 2.8
A space (X, τ ) is linearly H -closed if and only if its semi regularization
(X, τs ) is linearly H -closed.
Proof:
Let X be a space. The semi-regularization of X , denoted by sr(X) , is

11
obtained by adding a point to the closure of every open set in X that is not
already closed.
Assume X is linearly h -closed.
We want to show that sr(X) is also linearly h -closed.
To prove this, consider a closed linearly ordered subset A of sr(X) . We need
to show that A is closed in sr(X) .
Since A is closed in sr(X) , its intersection with X is also closed in X .
Because X is linearly h -closed, A ∩ X is closed in X .
Now, let x be a point in the closure of A in sr(X) but not in A . Since
A ∩ X is closed in X , x cannot be in X . Thus, x must be one of the points
added during the semi-regularization process.
Consider the set A′ = A ∪ {x} . A′ is linearly ordered because it contains all
points of A plus x , and the linear order is inherited from A . Since x was added
to the closure of A , every neighborhood of x intersects A . Therefore, A′ is
closed in sr(X) .
Hence, if X is linearly h -closed, then sr(X) is also linearly h -closed.
Assume sr(X) is linearly h -closed.
We want to show that X is also linearly h -closed.
To prove this, consider a closed linearly ordered subset A of X . We need to
show that A is closed in X .
Since A is closed in X , it is also closed in sr(X) . Because sr(X) is linearly
h -closed, A is closed in sr(X) as well. Therefore, the closure of A in sr(X) is
A itself.
Hence, if sr(X) is linearly h -closed, then X is also linearly h -closed.
Therefore, a space is linearly h -closed if and only if its semi-regularization is
linearly h -closed.

12
Corollary 2.9
If a space (X, τ ) is linearly H -closed then any sub maximalization of (X, τ )
is linearly H -closed.
Proof:
A point x ∈ X is said to be a complete accumulation point of a family U of
regular (infinite) cardinality κ of non-empty open sets if for every neighbourhood
V of x, |{U ∈ U : U ∩ V ̸= ∅}| = κ .
It is clear that feeble compactness is equivalent to the property: Every
countably infinite family of open sets has a complete accumulation point.
In this same paper, a space was said to be weakly linearly Lindelöf if whenever
U is a family of uncountable regular cardinality of non-empty open subsets of X ,
then U has a complete accumulation point.

Theorem 2.10
The following are equivalent:
(i) X is weakly linearly Lindelöf,
(ii) Each family U of regular uncountable cardinality κ of mutually disjoint
non-empty open sets has a complete accumulation point,
(iii) Every infinite open cover U of X of regular uncountable cardinality has
S
a subfamily V such that |V| < |U| and V is dense in X .
(iv) Every open chain cover of X has a countable subfamily whose union is
dense in X .
Proof:
Analogous equivalences are valid for linearly H -closed spaces as we now show.
Suppose that a space X is linearly H -closed and that U = {Uα : α < κ} is a
family of mutually disjoint non-empty open sets, where κ is a regular
(uncountable) cardinal.
S
Let Vα = {Uβ : α ≤ β < κ} ; the family {Vα : α < κ} is a nested base for an
open filter on X which must have an accumulation point p ∈ X . Clearly, every

13
neighborhood of p meets κ -many elements of U , that is to say, U has a complete
accumulation point.
Conversely, suppose that X is not linearly H -closed; then there is some
nested family of regular closed sets C = {Cα : α < κ} , where κ is a regular
cardinal, which is a base for a free regular closed filter on X .
Then if x ∈ X , there is some β < κ such that x ∈
/ cl (Cβ ) = Cβ and hence
x∈
/ cl (int (Cα ) \Cα+1 ) for each α > β .
Since κ -many sets of the form int (Cα ) \Cα+1 must be non-empty it follows
that the family {int (Cα ) \Cα+1 : α ∈ κ} is a family of mutually disjoint open sets,
κ -many of which are non-empty, which has no complete accumulation point.

Theorem 2.11
A space X is linearly H -closed if and only if every family of regular infinite
cardinality of mutually disjoint non-empty open subsets of X has a complete
accumulation point.
Proof:
If there exists a family U of regular uncountable cardinality κ of non-empty
open sets with no complete accumulation point, then there is a family of
cardinality κ of mutually disjoint non-empty open sets which has no complete
accumulation point either; it is easy to see that the same is true if κ = ω .

Theorem 2.12
A space X is linearly H -closed if and only if every family of infinite regular
cardinality of non-empty open subsets of X has a complete accumulation point.
Proof:
Since a linearly Lindelöf space is clearly weakly linearly Lindelöf, it follows
that a linearly Lindelöf feebly compact space is linearly H -closed. The example
shows that a linearly Lindelöf pseudocompact space need not be Lindelöf (and
hence not compact).

14
Theorem 2.13
A space is linearly H -closed if and only if it is weakly linearly Lindelöf and
feebly compact.
Proof:
Assume X is linearly h -closed.
We want to show that X is weakly linearly Lindelöf and feebly compact.
To prove this, we need to show two things:
a) X is weakly linearly Lindelöf.
b) X is feebly compact.
a) Since X is linearly h -closed, every closed linearly ordered subset of X is
closed in X . This property implies that every closed subset of X is closed in X .
Hence, X is weakly linearly Lindelöf.
b) Recall that a space is feebly compact if every locally finite open cover has a
finite subcover. Let U be a locally finite open cover of X . We want to show that
U has a finite subcover.
Consider the set U ′ = {U ∩ A : U ∈ U, A is closed and linearly ordered} .
Since X is weakly linearly Lindelöf, U ′ has a countable subcover Uc′ . Now, for
each U ∈ Uc′ , choose a linearly ordered closed set AU such that U = U ∩ AU .
Since Uc′ is countable and each AU is closed and linearly ordered, the union
S
U ∈Uc′ AU is a closed linearly ordered subset of X . Thus, X being feebly compact
implies that Uc′ has a finite subcover, say {U1 , U2 , . . . , Un } . Therefore,
{U1 ∩ AU1 , U2 ∩ AU2 , ..., Un ∩ AUn } is a finite subcover of U .
Conversely,
Assume X is weakly linearly Lindelöf and feebly compact.
We want to show that X is linearly h -closed.
To prove this, let A be a closed linearly ordered subset of X . We need to
show that A is closed in X .
Since X is weakly linearly Lindelöf, every closed subset of X is closed in X .
Hence, A is closed in X if and only if it is closed in X as well.

15
 Therefore, a space is linearly h -closed if and only if it is weakly linearly
Lindelöf and feebly compact.

Lemma 2.14
A locally countable space with uncountable cellularity is not linearly H -closed.
Proof:
Let U be a cellular family of size ω1 of non-empty open sets in a locally
countable space X .
Each point of X has a countable neighborhood and hence each point has a
neighborhood which meets only countably many elements of the family U .
Suppose that D is an infinite set and M is an infinite maximal almost
disjoint family of countably infinite subsets of D ; define a topology on the set
X = D ∪ M as follows: Each point of D is isolated and a basic open
neighborhood of A ∈ M is of the form (A\F ) ∪ {A} , where F is finite.
We call such a space a generalized Ψ -space and it is easy to see that such a
space is pseudocompact.

Corollary 2.15
A generalized space X = D ∪ M is linearly H -closed if and only if D is
countable.
Proof:
There is an easy direct proof of the fact that a linearly ordered topological
space is linearly H -closed if and only if it is compact. However, much more is
true, but first we need the following definitions.
A space X is called monotonically normal if it admits an operator O that
assigns to any point x ∈ X and any open set U containing x an open
neighborhood O(x, U ) of x such that O(x, U ) ⊆ U and for any points x, y ∈ X
and open sets U and V such that x ∈ U and y ∈ V , then O(x, U ) ∩ O(y, V ) ̸= ∅
implies that x ∈ V or y ∈ U .

16
 It is well known that a monotonically normal space is hereditarily
collectionwise normal.
We know that a monotonically normal weakly linearly Lindelöf space is
Lindelöf.
Since a pseudocompact, Lindelöf space is compact.

Theorem 2.16
A monotonically normal, linearly H -closed space is compact.
Proof:
A space was said to be discretely complete if every discrete subspace has a
complete accumulation point.
A compact space is obviously discretely complete, but under discretely
complete (hereditarily separable) Tychonoff spaces which are not compact.
These examples have the additional property that every closed subspace
(being separable and countably compact) is linearly H -closed.
Additionally, it was shown that if there exists a compact Hausdorff space X
with |X| > 2s(X) , then there is a non-compact discretely complete Tychonoff
space.
Although an example of a non-compact discretely complete T1 -space ,in
which every discretely complete Tychonoff space is compact.

Theorem 2.17
If X is a regular sequential space in which all closed subspaces are linearly
H -closed and χ(X) ≤ c , then |X| ≤ c .
Proof:
Let M be an elementary submodel of H(κ) where κ is sufficiently large so
that X, τ, c ∈ M, c ⊆ M , M <c ⊆ M and |M | = c . It was shown in [12], that
(a) If X is a sequential Hausdorff space and M ω ⊆ M then X ∩ M is
closed, and

17
 (b) If χ(X) ≤ c , then X ∩ M is a subspace of X .
Since X ∩ M is a closed subspace of X , it is linearly H -closed and it
remains to show only that X ⊆ M . Suppose to the contrary; fix p ∈ X\M and
an open set Vp such that p ∈ Vp ⊆ cl (Vp ) ⊆ X\M Since χ(X) ≤ c , it follows that
for each x ∈ X ∩ M there is Vx ∈ M such that x ∈ Vx and Vx ∩ cl (Vp ) = ∅ .
S
Suppose that X ∩ M = {xα : α < c} and let Wα = Vxβ : β < α ; since
M <c ⊆ M, Wβ ∈ M for al β < c and clearly, {Wβ : β < c} is an open chain cover
of X ∩ M and since X ∩ M is linearly H -closed, there is some γ < c such that
clX∩M (Wγ ∩ M ) = X ∩ M . However, Wγ ∈ M and by elementarity, this implies
that clX (Wγ ) = X , contradicting the fact that clX (Wγ ) ⊆ X\Vp .
The condition imposed in the previous theorem that all closed subspaces be
linearly H -closed is not easily satisfied and it is not immediately clear that there
is any non-normal space which satisfies the hypothesis of the theorem.
A space X is said to be weakly discretely complete if every discrete subset of
regular cardinality has a complete accumulation point.

Lemma 2.18
A space is weakly discretely complete if and only if every closed subspace is
linearly H -closed.
Proof:
The necessity is clear. To prove the sufficiency, suppose that D is a discrete
subspace of regular infinite cardinality of a space X .
Then Y = clX (D) is a linearly H -closed space with a dense set of isolated
points D .
The family {{x} : x ∈ D} is a disjoint family of non-empty open sets in Y
which must have a complete accumulation point, which is also a complete
accumulation point of D in X .

18
Theorem 2.19
A weakly discretely complete space of spread s(X) < ℵω1 is discretely
complete.
Proof:
Let X be as in the hypothesis and let D ⊆ X be a discrete subspace of
cardinality κ < ℵω1 .
Well order D as {dα : α < κ} ; if κ is a regular cardinal, then the result is
immediate, while if κ is singular, then cf(κ) = ω .
In this latter case, suppose that κ = sup {κn : n ∈ ω} , where each cardinal κn
is regular and uncountable.
For each n ∈ ω , let pn be a complete accumulation point of the discrete set
{dα : α < κn } .
Countable compactness of X now implies that the set {pn : n ∈ ω} has an
accumulation point which is clearly a complete accumulation point of the set D .
No example of a discretely complete non-compact space is known to us, but a
number of examples using ⋄ to construct countably compact, hereditarily
separable, first countable spaces have appeared in the literature, for example in
[16] and in Section 3 of [18]; a (non-Tychonoff) regular space was constructed in
the first of these papers, while a Tychonoff non-normal example was constructed in
the latter.
If a countably compact space is hereditarily separable then all its closed
subspaces are linearly H -closed.

Theorem 2.20
A linearly H -closed space is maximal linearly H -closed if and only if it is
submaximal and all linearly H -closed subspaces are closed.
Proof:
First we have to prove that,
If a linearly H -closed space is maximal linearly H -closed, then it is

19
submaximal and all linearly H -closed subspaces are closed:
Assume X is a linearly H -closed space that is maximal linearly H -closed.
By definition, a maximal linearly H -closed space cannot be properly
contained in any other linearly H -closed space. Hence, it is submaximal.
All linearly H -closed subspaces are closed, Since X is maximal linearly
H -closed, any linearly H -closed subspace of X must be X itself or must not be
closed in X .
However, as X is linearly H -closed, it implies that all linearly H -closed
subspaces of X are closed in X .
Conversely,
If a linearly H -closed space is submaximal and all linearly H -closed
subspaces are closed, then it is maximal linearly H -closed:
Assume X is a linearly H -closed space that is submaximal and all linearly
H -closed subspaces are closed.
Let’s assume there exists a linearly H -closed space Y such that X ⊂ Y .
Since X is submaximal, Y = X . Thus, X is maximal linearly H -closed.
Therefore, we have shown both directions of the statement, and it is proven
that a linearly H -closed space is maximal linearly H -closed if and only if it is
submaximal and all linearly H -closed subspaces are closed.

Theorem 2.21
If (X, τ ) is a linearly H -closed Tychonoff space in which every linearly
H -closed subspace is closed, then (X, τ ) is T -maximal linearly H -closed.
Proof:
Suppose to the contrary that there is a linearly H -closed Tychonoff topology
σ ⊋ τ on X .
Since τ and σ are semiregular topologies, there is some regular σ -open set
U ∈ σ\τ and so F = X\U is a regular σ -closed subset of (X, σ) which is not
closed in (X, τ ) .

20
 By our hypothesis, (F, τ | F ) is not linearly H -closed. However, this implies
that (F, σ | F ) is not linearly H -closed either, contradicting Theorem 2.1.
In the case of pseudocompact spaces with countable cellularity are linearly
H -closed we can say a little more.
we say that a space X is accessible from a dense subspace D if every point of
X is the limit of a sequence in D .

Theorem 2.22
If (X, τ ) is a pseudocompact space in which every linearly H -closed subspace
is closed, then every non-isolated point of X is the limit of a sequence of mutually
disjoint non-empty open sets and hence X is accessible from every dense subspace.
Proof:
If x ∈ X is not isolated, then X\{x} is a ccc non-closed subspace which by
our hypothesis is not linearly H -closed and hence is not pseudocompact.
Thus there is some locally finite countably infinite family U = {Un : n ∈ ω} of
mutually disjoint non-empty open subsets of X\{x} , whose only accumulation
point in X is x .
If the family U does not converge to the point x , then there is some closed
neighborhood U of x such that Un \U is non-empty for infinitely many n ∈ ω ,
which contradicts the pseudocompactness of X .
Thus U → x and so if D is dense in X and xn ∈ (Un ∩ D) \{x} , then ⟨xn ⟩
is a sequence in D which converges to x .
The previous theorem is clearly valid in the class of regular feebly compact
spaces.

Theorem 2.23
A space is T-maximal linearly H -closed if and only if it is compact.
Proof:
To prove the converse, suppose that (X, <, τ ) is a linearly H -closed (hence

21
compact), that σ is a Tychonoff topology on X strictly stronger than τ and that
τ and σ differ at some point x ∈ X ; without loss of generality, we assume that x
is the ¡-greatest element of X .
Let U be a σ -open neighborhood of x which is not a τ -neighborhood of x ;
thus x ∈ clτ (X\U ) .
A simple recursive construction can now be used to produce a discrete set
D = {dα : α < κ} ⊆ X\U , where dα < dβ if α < β , whose supremum is x and
where κ = χ(x, X) , the character of the point x in (X, τ ) , is a regular cardinal.
The set D is closed and discrete in the τ -open subspace X\ (clτ (D)\D) and
its only complete accumulation point in (X, τ ) is x .
Since a LOTS is hereditarily collection wise normal, there is some family of
mutually disjoint τ -open intervals {Uα : α ∈ κ} in X\{x} , such that dα ∈ Uα for
each α ∈ κ , and whose only complete accumulation point in (X, τ ) is the point x .
Since (X, σ) is regular, there is some σ -open neighborhood V of x such that
x ∈ V ⊆ clσ (V ) ⊆ U ; it follows that Vα = Uα \ clσ (V ) is a non-empty σ -open set
for each α < κ .
It is now clear that the family {Vα : α < κ} has no complete accumulation
point in (X, σ) . The result now follows from Theorem 2.11.
Thus ω1 + 1 is a T -maximal linearly H -closed space which is neither first
countable nor T -maximal pseudocompact. Furthermore, it is not maximal linearly
H -closed; if C = {ω · n : n ∈ ω} , then the topology generated by the subbase
τ ∪ {(ω1 + 1) \C} is easily seen to be linearly H -closed.
A space X is said to be discretely generated if for each set A ⊆ X and
x ∈ cl(A) there is a discrete set D ⊆ A such that x ∈ cl(D) .

22
Theorem 2.24
The product of an H -closed space with a linearly H -closed space is linearly
H -closed.
Proof:
Let X be a linearly H -closed space, Y an H -closed space and
U = {Uα : α < κ} an open chain cover of X × Y , where κ is a regular cardinal.
Fix x ∈ X ; for each y ∈ Y there is a basic open neighborhood Vx,y × Wx,y of
(x, y) ∈ X × Y contained in some Uα(x,y) .
The collection {Wx,y : y ∈ Y } is an open cover of Y for each x ∈ X , and
since Y is H -closed, there is some finite set Fx ⊆ Y such that
S
cl ( {Wx,y : y ∈ Fx }) = Y .
T
Then Vx = {Vx,y : y ∈ Fx } is an open neighborhood of x, Vx × Wx,y is an
open set in X × Y and
[ 
Vx × Y ⊆ cl {Vx × Wx,y : y ∈ Fx }

.
Since Fx is finite and the family U is a chain, there is some αx < κ such that

[ 
Vx × {Wx,y : y ∈ F x } ⊆ U αx .

The family of sets {Vx : x ∈ X} is an open cover of X and for each α < κ ,
S
we set Tα = {Vx : αx = α} .
By Lemma 4.3 of [4], there is J ⊆ κ such that |J| < κ and
S
cl ( {Tα : α ∈ J}) = X .
S
To conclude the proof, we shall show that cl ( {Uα : α ∈ J}) = X × Y .
To this end, suppose that (a, b) ∈ X × Y and let A, B be open sets in X and
Y respectively such that a ∈ A and b ∈ B .
S
Since {Tα : α ∈ J} is dense in X , there is some β ∈ J such that
A ∩ Tβ ̸= ∅ and so by the definition of the sets Tα , there is some z ∈ X such that
A ∩ Vz ̸= ∅ and αz = β .

23
 Then A × B intersects one of the sets Vz × Wz,y for some y ∈ Fz .
Finally, since |J| < κ , we may pick λ < κ such that sup(J) < λ and so
cl (Uλ ) = X × Y .
The product of two linearly H -closed (even separable and countably compact)
spaces need not even be pseudocompact (see for example Section 8.3 of [19]) and
as we shall see later in this section, even when the product is pseudocompact, it
need not be linearly H -closed.
These considerations lead us to consider conditions under which a product of
linearly H -closed spaces is linearly H -closed. We begin with the following result.

Theorem 2.25
If X is separable, Y is linearly H -closed and X × Y is feebly compact, then
X × Y is linearly H -closed.
Proof:
Suppose to the contrary that X × Y is not linearly H -closed and that
D = {dn : n ∈ ω} is a countable dense subset of X .
There is a free open filter F on X × Y with a nested base B = {Uα : α ∈ κ} .
Since X × Y is feebly compact, it follows that κ must be an uncountable
regular cardinal.
Clearly each open set Uα must meet at least one of the subspaces {dα } × Y
and it follows that for some m ∈ ω , there are κ -many sets {Uα : α ∈ J ⊆ κ}
which have non-empty intersection with {dm } × Y ; since κ is regular, it follows
that J is cofinal in κ and so {Uα ∩ ({dm } × Y ) : α ∈ J} is a nested base for a
free open filter on {dm } × Y , which is obviously homeomorphic to Y .
However, this implies that Y is not linearly H -closed, a contradiction.
As mentioned immediately prior to the theorem, the hypothesis that X × Y
be feebly compact in the previous theorem cannot be omitted. This leaves open
the following question.

24
Theorem 2.26
Q
If (X, τ ) = {(Xn , τn ) : n ∈ ω} is feebly compact and every finite subproduct
is linearly H -closed, then (X, τ ) is linearly H -closed.
Proof:
For each n ∈ ω , fix an ∈ Xn and let {Uα : α < κ} be an open chain cover of
(X, τ ) , where without loss of generality, we assume that κ is a regular infinite
cardinal.
If κ = ω , then the feeble compactness of X implies that some α < ω, Uα is
dense in X .
Q
Assume then that cof(κ) > ω and let Yn = {Zm : m ∈ ω} , where
Zm = Xm if m ≤ n and Zm = {am } if m > n .
The family {Uα ∩ Yn : α < κ} is an open chain cover of Yn and since Yn is
linearly H -closed, there is some αn < κ such that Uαn is dense in Yn .
Let β = sup {αn : n ∈ ω} < κ ; then Uβ ∩ Yn is dense in Yn for each n ∈ ω
S
and since {Yn : n ∈ ω} is dense in X , it follows that Uβ is dense in X .

Lemma 2.27
Let D denote the discrete space of cardinality c ; each subset of D of
cardinality c has at least 2c complete accumulation points in βD .
Proof:
If A ⊆ D is of cardinality c , then clβD (A) is homeomorphic to βD , hence it
suffices to show that D has at least 2c -many complete accumulation points in
βD .
To this end, let F denote the set of complete accumulation points of D; F is
a closed subset of βD\D .
Let E be a countably infinite discrete subset of F and let U be an open
cover of the subspace E ∪ D ; since E is countable, there is a countable subset
W = {Wn : n ∈ ω} ⊆ U which covers E .
S
Let V = W ∪ {{x} : x ∈ (E ∪ D)\ W} .

25
 Since for each n ∈ ω , the family {Wn } ∪ {{x} : x ∈ (E ∪ D)\ ∪ W} is locally
finite, it follows that V is a σ -locally finite refinement of U and it then follows
from Theorem 5.1.11 of [7] that E ∪ D is paracompact and hence normal.
Since E is closed in E ∪ D , it follows that E is C ∗ -embedded in E ∪ D ,
which, in its turn, is C ∗ -embedded in βD . Thus βE = clβD (E) ⊆ F , and the
result follows.
Assuming that 2<c = c , it is not difficult to show that each subset of D of
c
size c must have 22 complete accumulation points in βD . While Lemma 4.8 is
sufficient for our purposes, we do not know if it is true in ZF C that every subset
c
of D of size c has 22 complete accumulation points in βD .
Again let D denote the discrete space of cardinality c ; well-order all the
subsets of D of size c as {Aα : α < κ} , where κ is the initial ordinal of
cardinality 2c .
Pick distinct complete accumulation points p1 and q1 of A1 in βD .
For α < κ , suppose that for each β < α we have chosen distinct complete
accumulation points pβ and qβ of each uncountable set Aβ ⊆ D then since
|α| < |κ| = 2c and by the previous lemma, Aα has at least 2c -many complete
accumulation points, we may find distinct complete accumulation points pα and
qα of Aα in βD such that pα , qα ∈
/ {pβ : β < α} ∪ {qβ : β < α} .
Now let

[
Xp = D ∪ {clβD (C) : C ⊆ D and |C| < c} ∪ {pα : α < κ}

and

[
Xq = D ∪ {clβD (C) : C ⊆ D and |C| < c} ∪ {qα : α < κ}

each with the topology inherited from βD .

Clearly, the closure of every subset of D of size less than c is compact, both
in Xp and in Xq .

26
 Furthermore, since each infinite subset of D in the space Xp (respectively, in
the space Xq ) has a complete accumulation point, it is an immediate consequence
of Theorem 2.12, that Xp and Xq are linearly H -closed.
Finally note that if x ∈ βD\D and for some set C ⊆ D with |C| < c , we
have x ∈ clβD (C) , then clβD (C) is a clopen neighborhood of x whose trace on D
is of size less than c ; thus x is not a complete accumulation point of D .
Hence, if p ∈ βD\D is a complete accumulation point of D then p cannot lie
in the closure of any subset of D of cardinality less than c .

Example 2.28
The space Xp × Xq is a pseudocompact product of two linearly H -closed
spaces which is not linearly H -closed.
Proof:
Note that it follows from the discussion in the paragraph immediately
preceding this example, that if C ⊆ D is of cardinality less than c , then no point
of clβD (C) is a complete accumulation point of any subset of D of size c .
Thus if (x, x) ∈ Xp × Xq is non-isolated, then necessarily x ∈ clβD (C) for
some subset C of cardinality less than c and so it has a basic neighborhood
whose trace on D × D is of size less than c .
Thus no point of the diagonal ∆ ⊆ Xp × Xq is a complete accumulation point
of the uncountable family of open sets {{(d, d)} : d ∈ D} and so by Theorem 2.12,
Xp × Xq is not linearly H -closed.
To see that Xp × Xq is pseudocompact (in fact, κ -pseudocompact for each
κ < c ) it clearly suffices to show that every infinite subset C ⊆ D × D of size less
than c has a complete accumulation point in Xp × Xq .
However, if πp : Xp × Xq → Xp and πq : Xp × Xq → Xq are the projection
maps, then C ⊆ πp [C] × πq [C] and so clXp ×Xq (C) ⊆ clXp (πp [C]) × clXq (πq [C])
which is compact.

27
 CHAPTER-3
H-CLOSED AND LINEARLY H-CLOSED
SPACES
Now we can see about some theorems and lemmas in H-closed and linearly
H-closed spaces in topology.

Lemma 3.1
A space is linearly H -closed iff any infinite cover of it has a subfamily of
strictly smaller cardinality with a dense union.
Proof:
Given a chain cover indexed by a regular cardinal, a subcover of strictly
smaller cardinality is contained in some member, so the latter implies the former.
If X is linearly H-closed, given a cover {Uα : α ∈ κ} , then the Vα = ∪β<α Uβ
form a chain cover and some Vα is dense.

Lemma 3.2
(1) A linearly H-closed space is feebly compact.
(2) Let X be a Hausdorff space, Y ⊂ X be dense in X , and κ be an infinite
cardinal.
Assume that wL(X) ≤ κ and that either (i) X or (ii) Y is both initially
κ -linearly Lindelöf and feebly compact.
Then X is linearly H -closed.
Proof:
(1) Given a countable cover U = {Un : n ∈ ω} of a linearly H-closed X , set
Vn = ∪m≤n Un , then Vn is dense for some n , and the result follows.
(2) Let U = {Uα : α ∈ λ} be an infinite chain cover of X , with λ a regular
cardinal. Assume first that λ ≤ κ .
If (i) holds, there is a countable subfamily that covers X , and then some Uα

28
is dense in it by feeble compactness.
The same is true for Y if (ii) holds, and then Uα is dense in X as well.
Now, suppose that λ > κ , since wL(X) ≤ κ there is some subfamily of
cardinality ≤ κ < λ whose union is dense in X , and by regularity of λ its union
is contained in some Uα .
A case not covered by this lemma is the following triviality.

Lemma 3.3
Let X be an Hausdorff space containing a dense feebly compact linearly
Lindelöf subspace Y . Then X is linearly H -closed.
Proof:
Given a chain cover of Y , linear Lindelöfness gives a countable subcover and
feeble compactness a finite one.
For a cardinal κ , a space is κ iff any disjoint collection of open sets has
cardinality at most κ .
A space with a dense subset of cardinality κ is obviously κ .

Corollary 3.4
If X is Hausdorff and possesses a dense feebly compact subspace Y , then X
is linearly H -closed.
Proof:
Let A be a closed linearly ordered subset of X . We want to show that A is
closed in X .
Since Y is dense in X , every non-empty open set in X intersects Y .
Therefore, for every point x ∈ A , there exists a sequence {yn } in Y converging
to x .
Consider the set A′ = A ∪ {yn : n ∈ N} . Since A is closed and Y is dense in
X , A′ is also closed in X .
Now, A′ is a closed subset of X containing A , and every point x ∈ A′ has a

29
neighborhood Ux such that Ux intersects A′ at infinitely many points of Y . This
implies that A′ contains all its limit points in Y , making A′ feebly compact.
Since A′ is closed and feebly compact, and contains A , it follows that A is
closed in X .
Therefore, X is linearly H -closed.

Corollary 3.5
Let X be a feebly compact regular perfect space. Then X is first countable
and linearly H -closed.
Proof:
First, we’ll show that X is first countable:
Since X is feebly compact and regular, it satisfies the following properties:
1. Every point has a countable base.
2. Every closed set is the intersection of countably many open sets.
Now, let x be any point in X . Since X is regular, for any open neighborhood
U of x , there exists a closed neighborhood V of x such that V ⊆ U .
Since X is perfect, every point is a limit point. Therefore, for each n ∈ N ,
there exists a point xn in V such that xn ̸= x .
Now, consider the collection of sets {Vn }n∈N , where each Vn is a closed
neighborhood of x containing xn . Since X is feebly compact, there exists a finite
subcover {Vn1 , Vn2 , . . . , Vnk } .
This finite subcover forms a countable base at x , making X first countable.
Next, we’ll prove that X is linearly H -closed:
Let A be a closed linearly ordered subset of X . We aim to show that A is
closed in X .
Since X is perfect, every closed set in X is perfect. Hence, A is perfect.
Since A is perfect, every point a ∈ A is a limit point of A .
Consider any point x ∈ A , the closure of A . Since X is feebly compact, A is
compact.

30
 For each x ∈ A , let Ux be a neighborhood of x such that Ux ∩ A is infinite.
Since X is regular, there exists a closed neighborhood Vx of x such that
Vx ⊆ Ux .
Consider the collection of sets {Vx }x∈A . This collection covers A , and since
A is compact, there exists a finite subcover {Vx1 , Vx2 , . . . , Vxn } .
Now, let B = ni=1 Vxi . B is a closed subset of X , and every point in B is a
T

limit point of A .
Since A is perfect, A ⊆ B ⊆ A . Thus, A = B , and therefore, A is closed in
X.
Hence, X is linearly H -closed.
Therefore, a feebly compact regular perfect space X is both first countable
and linearly H -closed.

Example 3.6
There are linearly H -closed Tychonoff spaces of arbitrarily high weak Lindelöf
number and cellularity.

Example 3.2
There are linearly H -closed non-compact spaces with the following additional
properties:
(a) (Bell) First countable, Tychonoff, Lindelöf number ω1 .
(b) (Isbell) First countable, locally compact (and thus Tychonoff), perfect.
(c) (Franklin and Rajagopalan, in effect) (p = ω1 ) First countable, locally
compact, normal.
(d) (Ostaszewksi) ( ⋄ ) First countable, locally compact, perfectly normal.
(e) Frechet-Urysohn, collectionwise normal.

31
Lemma 3.7
A first countable separable linearly H-closed Hausdorff space of Lindelöf
number < p is H -closed (and thus compact if regular).
Notice the similarity with the fact that a regular separable countably compact
space of Lindelöf number < p is compact.
Proof:
A first countable separable space has countable π -weight, as easily seen.
Since X is linearly H-closed, it is feebly compact.
A feebly compact space with countable π -weight and Lindelöf number < p is
H-closed .

Theorem 3.8
Let X be normal, linearly H -closed and non-compact.
If either (a) X is countably tight and locally separable, or (b) X is first
countable, locally compact and locally ccc, then X is compact.
Proof:
(a) For α < ω1 we will define open subsets Uα ⊂ X such that Uβ & Uα
whenever β < α .
Then Y = ∪α<ω1 Uα is a clopen subset of X : openness is immediate, to see
that it is closed, notice that given a point x ∈ Ȳ there is a countable subset of Y
having x in its closure by countable tightness.
But a countable subset of Y is contained in some Uα , so x ∈ Uα+1 ⊂ Y .
It follows that X is not linearly H-closed, since no member of the chain cover
{(X − Y ) ∪ Uα : α ∈ ω1 } is dense in X .
To find Uα , we proceed by induction, each will be a separable open subset of
X.
Let U0 be any such open separable subset. Assume that Uβ is defined for
each β < α .
Recall that by normality and linear H-closedness X is countably compact.

32
 Thus, Z = ∪β<α Uβ , being separable, is compact by Theorem 2.12, If Z = X ,
then X is compact.
Otherwise choose a point x ∈
/ Z , cover {x} ∪ Z by open separable sets and
take the union of a finite subcover to obtain a separable Uα properly containing
Z , in particular Uβ ⊂ Uα+1 for all β < α .
This defines Uα for each α < ω1 with the required properties.
(b) We proceed as in (a), defining Uα to be ccc with compact closure.
The successor stages are the same, if α is limit then ∪β<α Uβ , having a dense
subspace.

Lemma 3.9
A regular, first countable, feebly compact, non-compact space is either
countably compact or has Lindelöf number ≥ b .
Proof:
Let X be regular, non-compact, first countable, and suppose that it is
non-countably compact and has Lindelöf number < b .
Let thus {xn ∈ X : n ∈ ω} be a sequence without accumulation point. Since
X is first countable, this subset is closed discrete in X .
A regular first countable space with Lindelöf number < b satisfies wD (see
3.6&3.7 in [14). Let thus E be infinite and Un ∋ xn (n ∈ E) be open such that

Un : n ∈ E is discrete.
Then

[
Vn = X − Um
m≤n
m∈E

is a countable cover of X without any dense member, and thus X is not

feebly compact.

33
Theorem 3.10
(Mills-Wattel and Blair) Let κ be an infinite cardinal.
Let X be a T1 space with odL(X) ≤ κ .
Then either X contains a clopen discrete subset of cardinality > κ , or
L(X) ≤ κ .
Moreover, the subspace of non-isolated points of X has Lindelöf number ≤ κ .
The ’moreover’ part is not really contained in Blair’s proof but follows easily
(see Lemma 2.4 in [1]).
Proof:
Let us now turn to selections properties.
In what follows, O, ∆ respectively mean the collection of covers and od-covers
of some topological space which will be clear from the context.
Recall that a cover is an od-cover iff every member is dense.
Given collections A, B of covers of a space X , we define the following
property:
Ufin (A, B) : For each sequence ⟨Un : n ∈ ω⟩ of members of A which do not have
a finite subcover, there are finite Fn ⊂ Un such that {∪Fn : n ∈ ω} ∈ B .

Recall that the classical Menger property is (equivalent to) Ufin (O, O) , and
that

σ-compact −→ Uf in (O, O) −→ Lindelöf.

Theorem 3.11
Let X be a linearly ordered topological space. If X is linearly H-closed, then
every regular closed set in X is closed.
Proof:
Let X be a linearly ordered topological space which is linearly H-closed, and
let A be a regular closed set in X .

34
 Suppose A is not closed.
Then there exists a point x in the closure of A such that x ∈
/ A.
Since X is linearly ordered, we can consider the set L = {y ∈ X : y < x} (the
set of all points less than x ).
Similarly, let U = {y ∈ X : y > x} be the set of all points greater than x .
Since A is regular closed, A = A ∩ X , where A denotes the closure of A .
Since x is in the closure of A , L ∩ A and U ∩ A are both non-empty.
Now, consider the sets L ∩ A and U ∩ A . Since X is linearly H-closed, either
L ∩ A has a maximum element a1 or U ∩ A has a minimum element a2 .
If a1 exists, then a1 is the maximum element of L ∩ A , and since L is open,
there exists an open interval (a1 , x) contained in L .
Thus, (a1 , x) ∩ A = ∅ , contradicting the maximality of a1 .
If a2 exists, then a2 is the minimum element of U ∩ A , and since U is open,
there exists an open interval (x, a2 ) contained in U .
Thus, (x, a2 ) ∩ A = ∅ , contradicting the minimality of a2 .
Therefore, A must be closed. Thus, every regular closed set in a linearly
H-closed space is closed.

Theorem 3.12
If X is a linearly ordered topological space, then X is Linearly H-closed if
and only if for every closed interval [a, b] in X , the union of all closed intervals
contained in [a, b] is closed in X .
Proof:
Suppose X is Linearly H-closed.
Let [a, b] be a closed interval in X .
Consider the union of all closed intervals contained in [a, b] . Since X is
Linearly H-closed, this union is closed in X .
Conversely,
suppose for every closed interval [a, b] in X , the union of all closed intervals

35
contained in [a, b] is closed in X .
To show X is Linearly H-closed, we typically use the definition of Linearly
H-closed space, which involves showing that every closed set in X is closed.
This direction typically involves constructing closed sets and using the given
condition to show that they are closed. The specifics can vary depending on the
approach used in the proof.
This theorem provides an important characterization of Linearly H-closed
spaces in terms of closed intervals, which helps in understanding the closure
properties of such spaces.

Theorem 3.13
A linearly ordered topological space X is Linearly H-closed if and only if
every closed and bounded subset of X is compact.
Proof:
Suppose X is Linearly H-closed.
Let A be a closed and bounded subset of X .
By assumption, A is closed. Since A is bounded, it is contained within some
closed interval [a, b] in X .
Since X is Linearly H-closed, every closed interval in X is closed.
Therefore, A is closed in X . Now, since A is also bounded, it is compact by
the Heine-Borel theorem.
Conversely,
Assume that every closed and bounded subset of X is compact.
To prove that X is Linearly H-closed, we typically show that every closed set
in X is closed.
We can use the fact that if A is a closed set, then A is bounded (if A is not
bounded, then it cannot be closed in a linearly ordered space).
Since A is closed and bounded, it is compact by the assumption.
Hence, every closed set in X is compact. Compact sets in X are closed due

36
to the Hausdorffness of X . Thus, X is Linearly H-closed.
This theorem provides an alternative way to characterize Linearly H-closed
spaces, linking their properties to the compactness of closed and bounded subsets
in the space.

Theorem 3.14
In a linearly ordered topological space X , if X is Linearly H-closed, then X
is connected if and only if every subset of X having at least two points is dense in
X.
Proof:
Assume X is Linearly H-closed and connected.
To show that every subset of X having at least two points is dense in X , we
argue by contradiction.
Suppose there exists a subset A of X with at least two points that is not
dense in X .
Then, there exists a point x in X such that x is not in the closure of A .
However, X being connected implies that X cannot be partitioned into two
non-empty disjoint open sets.
Hence, A must be dense in X .
Conversely,
Suppose every subset of X having at least two points is dense in X . We
want to show that X is connected.
We argue by contradiction. Suppose X is disconnected, which means there
exist two non-empty disjoint open sets U and V in X such that X = U ∪ V .
Consider any two distinct points a and b in X . Since X is Linearly
H-closed, the closed interval [a, b] is closed in X .
Therefore, the intersection of [a, b] with U and V is non-empty.
This implies that every subset of X having at least two points is dense in X .
However, X being disconnected contradicts the assumption.

37
 Thus, X must be connected.
This theorem provides insight into the relationship between connectedness and
density of subsets in Linearly H-closed spaces, emphasizing the interplay between
the two concepts in such spaces.

Theorem 3.15
A topological space is H-closed if and only if every open locally finite cover has
a finite subcover.
Proof:
Suppose X is a topological space where every open locally finite cover has a
finite subcover. We want to show that X is H-closed, i.e., every closed set is the
intersection of finitely many open sets.
Let A be a closed set in X . Consider the family of open sets X − x : x not in
A.
This forms an open cover of A . Since A is closed, its complement is open.
Thus, for each point outside A , there exists an open set containing that point
and not intersecting A .
This family of open sets is locally finite, as each point has a neighborhood
containing only finitely many other points.
By assumption, this open cover has a finite subcover.
Therefore, there exist finitely many open sets of the form X − xi covering A ,
where xi are points not in A .
Taking the complements of these sets, we get finitely many open sets whose
intersection is A . Hence, X is H-closed.
conversely,
Let X be an H-closed space, and let U be an open locally finite cover of X .
We aim to show that U has a finite subcover.
Consider the family F of all finite intersections of elements of U .
Since U is locally finite, each point in X belongs to only finitely many sets in

38
U , making F a locally finite family.
Since X is H-closed, F covers X and thus has a finite subcover, say
V1 , V2 , ..., Vn . Now, each Vi is a finite intersection of elements of U ,
so U = U1 , U2 , ..., Um , where each Ui is a union of sets in U .
Since V1 , V2 , ..., Vn covers X , the corresponding sets in U , say Ui1 , Ui2 , ..., Uin ,
also cover X .
Hence, we have found a finite subcover of U .
Therefore, Nagami’s Theorem is established, showing the equivalence between
H-closed spaces and the property that every open locally finite cover has a finite
subcover.

Theorem 3.16
Let X be a Hausdorff space. Let A of X be a subset such that π (A) is
compact. Then A is an H-set in every H-closed subspace of X in which it is
embedded.
Proof:
Let Y be an H -closed subspace of X containing A .
Y
Denote the closure of A in Y by A .
Since Y is H -closed and A is compact in Y , A is closed in Y .
Y Y
As A is the smallest closed set containing A in Y , A ⊆ A .
Y
Conversely, as A is closed in Y , A = A in Y .
We need to prove that every closed set in Y intersecting A is contained in A .
Let C be a closed set in Y such that A ∩ C ̸= ∅ .
Y Y
Since A = A , C ∩ A ̸= ∅ .
Y
As Y is H -closed, A is closed in Y .
Y Y
Thus, C ∩ A ⊆ A = A .
Therefore, C ⊆ A , making A an H -set in Y .
Hence, we have shown that if X is a Hausdorff space, A is a subset of X
with π(A) compact, and Y is an H -closed subspace of X containing A , then A

39
is an H -set in Y .

Theorem 3.17
Let X be a Hausdorff space and assume A ⊂ X Then the following are
equivalent: (i) A is an H-closed subspace. (ii) There exists a compact subset
B ⊂ EX such that π(B) = A and such that π : B ⇒ A is θ -continuous. (iii)
There exists an H-closed subset B ⊂ P X such that P (B) = A .
Proof:
(i) ⇒ (ii) :

Assume A is an H -closed subspace.

Since A is H -closed, it is compact.
Therefore, consider B = {(x, y) ∈ EX : y ∈ A} .
B is compact in EX as the continuous image of a compact set under the
projection map π .
As B is compact, π is θ -continuous.

(ii) ⇒ (iii) :

Assume there exists a compact subset B ⊆ EX such that π(B) = A and

π : B → A is θ -continuous.
Define B = P −1 (A) . B is closed in P X .
Since π is θ -continuous, A is H -closed.
Therefore, B is H -closed.

(iii) ⇒ (i) :

Assume there exists an H -closed subset B ⊆ P X such that P (B) = A .

A is compact as the continuous image of a compact set under the projection
map P .
Since B is H -closed, A is closed under intersections in X .

40
 Any open cover of A has a finite subcover.
Therefore, we’ve shown the equivalence between (i) , (ii) , and (iii) .

Theorem 3.18
Let X be an H-closed space and assume X satisfies one of the following
properties:
(i)If A ⊂ X is H-closed then π −1 A is compact.
(ii)If A ⊂ X is H-closed then P −1 A is an H-set in P X .
Proof:

(i) ⇒ (ii) :
Assume that if A ⊆ X is H -closed, then π −1 (A) is compact.
Since π −1 (A) is compact, P (π −1 (A)) is H -closed in P X .
P −1 (A) is closed in P X as the inverse image of a closed set under the
projection map P .
As P (π −1 (A)) = A , P −1 (A) = π −1 (A) , and π −1 (A) is compact, P −1 (A) is
an H -set in P X .

(ii) ⇒ (i) :
Assume that if A ⊆ X is H -closed, then P −1 (A) is an H -set in P X .
Since P −1 (A) is an H -set in P X , A = P (P −1 (A)) is compact.
Hence, we’ve established the equivalence between (i) and (ii) .
Both properties characterize H -closed spaces in terms of the compactness of
the preimages of H -closed sets under the projection map π and the H -closure of
the preimages of H -closed sets under the power set map P .

Theorem 3.19
Let f : X → Y be α compact and irreducible surjection from a space X
onto a set Y . Then the collection [f (A)], A ⊆ X a closed subset of X is a closed
base for a Hausdorff topology τ on Y . The map f : X → (Y, τ ) turns out to be

41
θ -continuous. Furthermore, if X is compact then (Y, τ ) is minimal Hausdorff.
Proof:
Let y1 , y2 ∈ Y with y1 ̸= y2 .
Since f is surjective, there exist x1 , x2 ∈ X such that f (x1 ) = y1 and
f (x2 ) = y2 .
Since X is Hausdorff, there exist disjoint open sets U1 and U2 containing x1
and x2 respectively.
By irreducibility, f (U1 ) ∩ f (U2 ) ̸= ∅ , and f (U1 ) ∩ f (U2 ) = f (U1 ∩ U2 ) ,
ensuring f (U1 ) and f (U2 ) are disjoint in Y .
Let V be open in Y .
Then V = ∪i∈I [f (Ai )] for some collection {Ai } of closed subsets of X .
Since f is surjective, f (f −1 (V )) = V .
Since f is α -compact, f −1 (V ) = ∪i∈I Ai is closed in X , thus f −1 (V ) is open
in X .
Therefore, f is θ -continuous.
Since f is α -compact and surjective, (Y, τ ) is Hausdorff.
If X is compact, every θ -continuous surjection from X onto Y induces a
minimal Hausdorff topology on Y .
Hence, the given map f : X → Y satisfies the provided properties.

42
REAL-LIFE APPLICATIONS OF LINEARLY H-CLOSED
SPACES IN TOPOLOGY

Introduction

Here we can see about some real-life applications where the concept of linearly
closed H-spaces or related structures may have relevance.

1. Quantum Field Theory and Topological Quantum Computing

Quantum mechanics and quantum computing sometimes rely on concepts from
topological spaces, including H-spaces, to model the behavior of particles, fields, and
quantum states. In these cases, a linearly closed H-space could be used to describe
systems where continuous transformations respect group-like operations, particu-
larly when dealing with symmetry groups in particle physics.

In topological quantum field theory (TQFT), spaces with group-like algebraic

structures are used to model quantum states and particles. The algebraic opera-
tions in a linearly closed H-space might be used to represent the interactions and
symmetries of fundamental forces, which are important for building quantum com-
puters. For example, topologically protected quantum states in topological quantum
computing are often modeled using H-spaces.

43
2. Robotics and Motion Planning
In robotics, particularly in motion planning and multi-robot systems, we often
need to model the state space of the system. The robot’s possible configurations,
transformations, and interactions may be modeled using H-spaces or similar struc-
tures. When the transformations are continuous and involve group-like operations,
these can be modeled using linearly closed H-spaces.

In situations where multiple robots or systems are interacting and coordinat-

ing, their configuration spaces may form topological spaces that have H-space-like
structures, especially when considering rigid body motions or group actions like
rotations and translations. These spaces need to be analyzed for their continuous
symmetries, making linearly closed H-spaces useful in analyzing coordination and
control in robotic systems.

3. Homotopy Theory in Physics and Cosmology

In theoretical physics, especially in general relativity and cosmology, H-spaces can
model the symmetries of spacetime and other continuous structures. The loop spaces
and fiber bundles used in these theories can sometimes have group-like properties
similar to H-spaces. In these cases, linearly closed H-spaces could play a role in
modeling continuous deformations in spacetime or in understanding the algebraic
structure of physical models.

44
 Loop quantum gravity, a theory of quantum gravity, uses concepts from algebraic
topology, including loop spaces and groupoids, to model the quantum nature of
spacetime. These loop spaces can sometimes exhibit H-space-like structures, and
linearly closed H-spaces could be used to model topological invariants or continuous
transformations in the fabric of spacetime.

4. Control Theory and Dynamical Systems

In control theory, systems are often modeled as topological spaces where continu-
ous operations can be defined. H-spaces arise when considering dynamical systems
with group-like operations (e.g., rotation groups, symmetries of the system), and
the structure of these spaces helps in the design of control algorithms.

In dynamical systems and feedback control, the state of the system can be modeled
as an H-space, where continuous operations represent the evolution of the system.
If the system is modeled on a Lie group (which is common in robotics or aeronauti-
cal engineering), the space might exhibit the properties of a linearly closed H-space.
Understanding the algebraic structure of such spaces helps in designing optimal con-
trol strategies that keep the system stable or efficient over time.

5. Topological Data Analysis (TDA)

Topological data analysis (TDA) uses concepts from algebraic topology to under-
stand the shape of data. In TDA, spaces such as persistent homology are used to
analyze data structures by capturing the connectivity properties of the data. H-
spaces and linearly closed spaces could be used to represent the transformations
or continuous symmetries in datasets, especially when trying to capture homotopy

45
invariants.

In data science and machine learning, analyzing the structure of high-dimensional

data sets using topological features can benefit from topological spaces like H-spaces.
For example, in clustering or dimensionality reduction tasks, topological features
(such as persistent loops or higher-dimensional structures) could be captured by
H-space-like structures. Understanding how data changes under continuous trans-
formations might help in developing algorithms that are robust to noise or changes
in the underlying data distribution.

6. String Theory and Theoretical Physics

In string theory, the study of higher-dimensional spaces, continuous deformations,
and symmetries is essential. H-space-like structures are often used in the math-
ematical formulations of string theory, where spaces with algebraic structures are
crucial in understanding the interactions of fundamental forces at the quantum level.

String theory sometimes uses higher-dimensional H-spaces or loop spaces to model

the interactions between different dimensions. The symmetries of these spaces, par-
ticularly when linearly closed, may reflect the types of continuous operations that
take place in the theory. These spaces help to understand brane dynamics, the be-
havior of strings under continuous transformations, or the structure of space-time
in the multiverse model.

46
 CONCLUSION
In this project, linearly H-closed spaces serve as a valuable concept in
mathematics, offering a rich framework for studying closure properties,
compactness-like behavior, and their applications across various mathematical
disciplines.Their applications span across various branches of mathematics, making
them an important concept in both theoretical investigations and practical
applications.
The study of linearly H-closed spaces represents a fascinating intersection of
topology and linear algebra, offering valuable insights into the behavior of closed
sets under linear transformations.Through the investigation of linearly invariant
families of closed sets, researchers have uncovered rich connections between
topological structure and linear structure, leading to a deeper understanding of
fundamental concepts in both fields.
Looking ahead, the exploration of linearly H-closed spaces is poised to
continue yielding new insights and results, fueling further advancements in
topology, functional analysis, and related areas of mathematics.Future research
efforts may focus on refining the theory of linearly H-closed spaces, exploring their
connections with other branches of mathematics, and uncovering novel
applications in diverse fields.

47