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

Is this space $F[R]$ with the Pixley-Roy topology paracompact?

In general, when the space $F[X]$ is paracompact for general topological space?

Definition of Pixley-Roy topology: Basic neighborhoods of $F\in F[X]$ are the sets $$[F,V]=\{H\in F[X]; F\subseteq H\subseteq V\}$$ for open sets $V\supseteq F$, see e.g. here.

share|cite|improve this question
up vote 2 down vote accepted

The paracompactness of Pixley-Roy Hyperspaces seems to have been solved by Przymusiński (Normality and paracompactness of Pixley-Roy hyperspaces, Fund. Math. 113, pp.201-219). Therein he gives the following theorem:

Theorem: The following are equivalent for a Pixley-Roy hyperspace $\mathcal{F}[X]$:

  1. $\mathcal{F}[X]$ is paracompact;
  2. $\mathcal{F}[X]^n$ is paracompact for all $n \in \mathbb{N}$;
  3. $\mathcal{F}[X]$ is collectionwise Hausdorff;
  4. There is a neighbourhood assignment $F \mapsto V_F$ for the finite nonempty subsets of $X$ such that given finite, nonempty $F ,H \subseteq X$ the inclusions $F \subseteq V_H$ and $H \subseteq V_F$ imply $F \cap H \neq \emptyset$.

It follows from this that $\mathcal{F}[\mathbb{R}]$ is not paracompact. Suppose that $V \mapsto V_F$ is any neighbourhood assignment for the finite nonempty $F \subseteq \mathbb{R}$. Note that given any $x \in \mathbb{R}$ we may assume without loss of generality that $V_{\{x\}}$ is of the form $( x - \varepsilon_x , x + \varepsilon_x )$ for some $\varepsilon_x > 0$. From now on, I will write $V_x$ for $V_{\{x\}}$. The non-paracompactness of $\mathcal{F}[\mathbb{R}]$ is then an immediate consequence of the following claim:

Claim: There are distinct $x, y \in \mathbb{R}$ such that $y \in V_{x}$ and $x \in V_{y}$.

Proof. We construct sequences $\{ x_i \}_{i \in \mathbb{N}}$ and $\{ \delta_i \}_{i \in \mathbb{N}}$ so that

  • $x_0 = 0$;
  • $\delta_i = \varepsilon_{x_i}$;
  • $x_{i+1} = x_i + (-1)^i \frac{\delta_i}{2}$

Note that for $j > i$ we clearly have that $x_j \in V_{x_i}$. If for any $j > i$ we also have $x_i \in V_{x_j}$ we are done. So assume that this never happens.

Given any $i \in \mathbb{N}$, we have $x_i \notin V_{x_{i+1}}$. It thus follows that $$| x_{i+1} - x_i | > \varepsilon_{x_{i+1}} = \delta_{i+1} = 2 | x_{i+2} - x_{i+1} |,$$ and so the sequence $\{ x_i \}_{i \in \mathbb{N}}$ converges to some $y \in \mathbb{R}$.

It can be shown that $y \in V_{x_i}$ for all $i \in \mathbb{N}$. (If $i$ is even, then for all $j > i+1$ we have $x_i < x_j < x_{i+1}$ and therefore $x_i \leq y \leq x_{i+1}$; similarly for $i$ odd.) As $x_i \rightarrow y$, then $| y - x_i | < \varepsilon_y$ for all large enough $i$, and therefore $x_i \in V_y$ for all such $i$. $\Box$

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.