Mathematics Stack Exchange is a question and answer site for people studying math at any level and professionals in related fields. Join them; it only takes a minute:

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

we say that a space $X$ has the finite derived set property (which we abbreviate as the $FDS$ property) if whenever $A$ is infinite, there is an infinite subset $B\subseteq A$ such that $B$ has only a finite number of accumulation points in $X$.

a topological space is called a $KC$ space if every compact set is closed.

Below Theorem comes from "THE FDS-PROPERTY AND SPACES IN WHICH COMPACT SETS ARE CLOSED " by O.T. Alas, M.G. Tkachenko, V.V. Tkachuk, and R.G. Wilson. I have questions from it.

Theorem $\bf1.7.$ A compact $KC$ space of cardinality less than $\mathfrak{c}$ has the $FDS$ property (and hence is sequentially compact).

Proof: Suppose that $(X,\tau)$ is a compact $KC$ space and $|X| < \mathfrak{c}$. We assume to the contrary that $X$ does not have the $FDS$ property and so there is some countably infinite subset $A \subseteq X$ such that every infinite subset of $A$ has infinitely many accumulation points in $X$. By Lemma $1.2$, without loss of generality, we can assume that $A$ is discrete and that $\operatorname{cl}_X(A) = X$. Let $x\in X \ A$; if every neighbourhood $U$ of $x$ is such that $A\setminus U$ is finite, then $A\cup\{x\}$ is compact, hence closed in $X$ and so $x$ is the unique accumulation point of $A$, a contradiction. Thus we can choose an open neighbourhood $V$ of $x$, such that $A \setminus V$ is infinite. The closed subspace $X \setminus V$ of $X$ is compact and $A$ is countable, and hence Lindelöf, and so $A\cup (X \setminus V )$ is Lindelöf. This in its turn implies that $A \cup (X \setminus V )$ is not countably compact, for otherwise it would be compact, but being a proper dense subspace of $X$, it is not closed in $X$, a contradiction. Thus there is a discrete set $D \subseteq A \cup (X \setminus V )$ which is closed in $A \cup (X \setminus V )$. However, since $X \setminus V$ is compact, only a finite number of points of $D$ lie in $X \setminus V$ and hence $D \cap V$ is infinite and all its accumulation points (by our hypothesis, an infinite number in $X$) must lie in $V$; that is to say, $\operatorname{cl}_X(D \cap V )\subseteq V$.

Thus we have constructed two infinite sets $D \cap V$ and $A \setminus V$ whose closures are disjoint in $X$. Since $D \cap V \subseteq A$, each of these sets has the property that every infinite subset has an infinite number of accumulation points and the above argument can be repeated using $D \cap V$ and $A \subseteq V$ in place of $A$. A standard binary tree argument can now be used to show that $|X| \ge\mathfrak{c}$.

my questions are:
(1) why he can say :

However, since $X \setminus V$ is compact, only a finite number of points of $D$ lie in $X \setminus V$ and hence $D \cap V$ is infinite and $\operatorname{cl}_X(D \cap V) \subseteq V$?

(2) how he reach to contradiction in the last line

"A standard binary tree argument can now be used to show that $|X|\ge\mathfrak{c}$.

please give me a hand.

share|cite|improve this question
You really ought to proofread when you cut and paste: the fi-ligature generally doesn’t copy, which is why this has nite instead of finite and innite instead of infinite. – Brian M. Scott Jul 31 '13 at 7:45

First question:

$X\setminus V$ is compact because it is a closed subset of the $KC$ space $X$. $D$ is discrete and closed in $A\cup(X\setminus V)$, so $D\cap(X\setminus V)$ is closed and discrete in $X\setminus V$. A compact set has no infinite closed discrete subset, so $D\setminus V=D\cap(X\setminus V)$ must be finite. But $D$ is infinite, so $D\cap V$ must also be infinite. Finally, $D$ is closed in $A\cup(X\setminus V)$, so $D$ has no accumulation points in $A\cup(X\setminus V)$ and therefore none in $X\setminus V$. Thus, all accumulation points of $D$ are in $V$, and it follows immediately that $\operatorname{cl}_X(D\cap V)\subseteq V$.

Second question:

In the first paragraph they construct the infinite sets $D\cap V$ and $A\setminus V$ whose closures in $X$ are disjoint; call these sets $A_0$ and $A_1$. They point out that $A_0$ and $A_1$ satisfy the same hypotheses as $A$, so the same construction can be applied to each of them. This produces infinite subsets $A_{00}$ and $A_{01}$ of $A_0$ and infinite subsets $A_{10}$ and $A_{11}$ of $A_1$ whose closures in $X$ are pairwise disjoint and which again satisfy the same hypotheses as $A$. Continuing recursively in this fashion we can construct sets $A_\sigma$ for each finite sequence $\sigma$ of zeroes and ones in such a way that for each such $\sigma$, $A_{\sigma^\frown0}$ and $A_{\sigma^\frown1}$ are infinite subsets of $A_\sigma$ whose closures in $X$ are disjoint. (Here $\sigma^\frown i$ is the sequence obtained by appending $i$ to the sequence $\sigma$. For instance, $01101^\frown0=011010$.)

Now let $\alpha$ be any infinite sequence of zeroes and ones. For each $n\in\omega$ let $\sigma_n$ be the finite sequence consisting of the first $n$ terms of $\alpha$. For each $n\in\omega$ the set $\operatorname{cl}_XA_{\sigma_n}$ is closed in $X$ and therefore compact. Let $$K_\alpha=\bigcap_{n\in\omega}\operatorname{cl}_XA_{\sigma_n}\;;$$ $\operatorname{cl}_XA_{\sigma_n}\supseteq\operatorname{cl}_XA_{\sigma_{n+1}}$ for each $n\in\omega$, so $K_\alpha\ne\varnothing$. You can easily check that if $\alpha$ and $\beta$ are distinct infinite sequences of zeroes and ones, then $K_\alpha\cap K_\beta=\varnothing$. Finally, there are $\mathfrak{c}$ infinite sequences of zeroes and ones, so $X$ has $\mathfrak{c}$ pairwise disjoint, non-empty subsets $K_\alpha$. Clearly this implies that $|X|\ge\mathfrak{c}$.

share|cite|improve this answer

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.