MathOverflow is a question and answer site for professional mathematicians. 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

Let $C$ be an infinite, separable and connected metric space. If $C$ becomes totally disconnected when one of its points $p\in C$ is removed, does every closed ball of $C$ with positive radius and center $p$ always contain an infinite connected subset?

share|cite|improve this question
I have a question: if C becomes totally disconnected when the point p is removed, does it necessarily mean that every connected component of C has to include p? Thx for clarifying. – Carlo Von Schnitzel Aug 18 '10 at 23:15
@alephomega: I believe the answer is yes. For if $B\subset C$ and contains more then one point, but $p\notin B$, then when removing $p$ we still have $B\subset C\setminus\{p\}$, and $B$ is still connected, which is a contradiction to the totally disconnectedness of $C\setminus \{p\}$. – Dror Atariah Aug 19 '10 at 8:07
up vote 5 down vote accepted

Here I construct an example which proves the answer is NO.

Take the KK fan:

enter image description here

Remove its dispersion point at the top. Now you have a Cantor set of lines of rationals/irrationals that cannot be separated horizontally. Stretch this into a "Cantor-like tube" and weave it closer and closer to a point $p$ in the plane while shrinking its diameter and making sure that every loop goes a distance of $7$ away from $p$.

enter image description here

Remove $p$ and you have a hereditarily disconnected space ($\simeq$ KK fan minus its vertex). If $A$ is nonempty and clopen in $X$ then $A$ must snake around the tube forever, so it limits to $p$ and thus $p\in A$. Therefore $X$ is connected. The ball of radius $1$ around $p$ is hereditarily disconnected.

EDIT: I am basically taking the space which consists of the curve below, and the origin $p=(0,0)$ (so $\{0\}\times (0,1]$ is not included). The difference is that instead of weaving a line, I am weaving this "Cantor tube" while shrinking its diameter. In the first case if I remove $p$ then I get something $\simeq [0,\infty)$, whereas in the second case I get something $\simeq$ my Cantor tube.

enter image description here

share|cite|improve this answer
Be patient. As for me I like it very much but have not digested it yet. – მამუკა ჯიბლაძე Apr 30 at 19:56
Why removing $p$ gives $\simeq$ KK minus vertex? – მამუკა ჯიბლაძე Apr 30 at 20:01
Aren't you creating new converging sequences by weaving closer and closer to $p$? The points involved will become more and more closer to each other... – მამუკა ჯიბლაძე Apr 30 at 20:08
The only convergent sequences will converge to $p$. The tube does not limit to any point of itself. Maybe it is not clear in the picture, but the loops will stop before they get to, say, the 9:00 position. – Forever Mozart Apr 30 at 20:11
But there are infinitely many of them, and their points which were apart in KK get closer and closer to each other – მამუკა ჯიბლაძე Apr 30 at 20:30

I don't know if this answers your question precisely but here's an interesting example. First, let's start with the following:

The Knaster-Kuratowski Fan

Let $K$ be the Knaster-Kuratowski fan, also called Cantor's teepee. The space $K$ is defined as follows. Let $C$ be the Cantor set. Let $Q \subset C$ be the set of endpoints of the deleted middle-third intervals. Let $P=C \setminus Q$. We also let $p=(1/2,1/2) \in \mathbb R^2$. Now for each $x \in C$, we let $L_x$ be the line joining $p$ and $x$. Now our space $K$ is the union over all $x \in C$ of the sets

  • $\{ (x,y) \in L_x : y\in \mathbb Q \}$, if $x\in Q$, and
  • $\{ (x,y) \in L_x : y\notin \mathbb Q \}$, if $x \in P$.

The space $K$ is connected but $K\setminus \{p\}$ is totally disconnected. Also $K$ is punctiform, that is it contains no compact connected T2 subspace.

A related property of $K$

Let $f$ be a continuous function from $K$ to $K$. Let $U$ be some closed connected open set about a point $x \neq p$. Then there exists a closed connected open set $V$ about $p$, which can be written down as the set of all $(x,y)$ with $1/2 - \epsilon < y < 1/2$ such that $f(V) \subset U$ since $f$ is continuous. Then $V$ is necessarily homeomorphic to the space $K$.

share|cite|improve this answer
The section "Answer, a related property" is still hard for me to parse. Perhaps simple sentences (in the grammar sense) plus additional equivalent formal logical statements with quantifiers would help me. – Włodzimierz Holsztyński Apr 30 at 21:02
@WłodzimierzHolsztyński Last quote - it is Schnitzel, not Schnitze. Are you trying to notify the author of the answer? Are you sure the @ works in such cases too? – მამუკა ჯიბლაძე Apr 30 at 21:48
მამუკა ჯიბლაძე, I apologize for my unfortunate typo. I've fixed it, and I gave up on "@". Thank you for pointing to my oversight. – Włodzimierz Holsztyński May 1 at 0:04

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.