Mathematics Stack Exchange is a question and answer site for people studying math at any level and professionals in related fields. It's 100% free, no registration required.

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

So I'm looking at this proof, which is presented as a problem in Gamelin & Greene but I'm having some trouble understanding it.

Part (a), (b) and (c) are straightforward enough. However I'm having trouble following his reasoning on (d). I understood how everything is defined, I just don't see how the Baire Category Theorem applies, which says that the intersection of an infinite sequence of dense subsets of a complete metric space is dense. Could someone please explain it further?

share|cite|improve this question
Can I ask what is the $J_\delta(t_m)$ in d)? – Hui Yu Oct 9 '12 at 15:53
You believe that the Baire Category Theorem applies to $\mathbb{R}$, right? And that a set dense in the Sorgenfrey topology is also dense in the regular topology? – rschwieb Oct 9 '12 at 16:33

$\newcommand{\cl}{\operatorname{cl}}\newcommand{\int}{\operatorname{int}}$We have an open set $V$ in $\Bbb R_S^2$ that contains $F$, where $F=\{\langle x,-x\rangle:x\in T\}$, and $T$ is $\Bbb R\setminus S$ for some countable dense subset (not ‘sequence’: van Gool’s terminology is wrong) of $\Bbb R$ with respect to the usual Euclidean topology. For any $x\in\Bbb R$ and $\epsilon>0$ we define $J_\epsilon(x)=[x,x+\epsilon)\times[-x,-x+\epsilon)$, a basic open nbhd of $\langle x,-x\rangle$ in $\Bbb R_S^2$.

For $n\in\Bbb Z^+$ let $T_n=\{t\in T:J_{1/n}(t)\subseteq V\}$; I claim that $\bigcup_{n\in\Bbb Z^+}T_n=T$.

Proof. Since $V$ is an open nbhd of $F$, for each $t\in T$ there is an open nbhd of $\langle t,-t\rangle$ contained in $V$. Thus, for each $t\in T$ there is some $\epsilon(t)>0$ such that $\langle t,-t\rangle\in J_{\epsilon(t)}(t)\subseteq V$. Then for any positive integer $n>\frac1{\epsilon(t)}$ we have $\frac1n<\epsilon(t)$ and therefore $\langle t,-t\rangle\in J_{1/n}(t)\subseteq J_{\epsilon(t)}(t)\subseteq V$, i.e., $t\in T_n$. $\dashv$

Suppose that $\int\cl T_n=\varnothing$ for each $n\in\Bbb Z^+$, where the interior and closure are taken with respect to the usual Euclidean topology on $\Bbb R$. For $n\in\Bbb Z^+$ let $G_n=\Bbb R\setminus\cl T_n$; then each $G_n$ is a dense open subset of $\Bbb R$. For each $s\in S$ let $G_s=\Bbb R\setminus\{s\}$; $G_s$ is clearly also a dense open subset of $\Bbb R$. Let $$G=\bigcap_{n\in\Bbb Z^+}G_n\cap\bigcap_{s\in S}G_s\;;$$ $S$ is countable, so by the Baire category theorem $G$ is dense in $\Bbb R$. In particular, $G\ne\varnothing$, so

$$\begin{align*}\Bbb R&=S\cup T\subseteq\bigcup_{s\in S}\{s\}\cup\bigcup_{n\in\Bbb Z^+}\cl T_n\\ &=\bigcup_{s\in S}(\Bbb R\setminus G_s)\cup\bigcup_{n\in\Bbb Z^+}(\Bbb R\setminus G_n)\\ &=\Bbb R\setminus\left(\bigcap_{s\in S}G_s\cap\bigcap_{n\in\Bbb Z^+}G_n\right)\\ &=\Bbb R\setminus G\\ &\ne\Bbb R\;, \end{align*}$$

which is absurd. Thus, there must be an $n\in\Bbb Z^+$ such that $\int\cl T_n\ne\varnothing$.

Note that this argument works generally, not just in this setting: a space $X$ has the property that every intersection of countably many dense open sets is non-empty iff it is not the union of countably many nowhere dense closed sets.

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.