I understand that similar questions to this have been asked here (1) and here (2) but I believe my question is different because with regards to (1) I am not asking about the set $X$ that contains the points of my sequence and with regard to (2) because I have the distinct definitions of limit and limit point in mind. That said:

My Question: My textbook, Fundamental Ideas of Analysis by Michael Reed, defines a limit point as:

Definition: Let $(a_n)$ be a sequence of real numbers. A number $d$ is called a limit point if, given $\epsilon > 0$ and an integer $N$, there exists $\color{red}{\mathrm{ an }} \ n \ge N$ so that $|a_n - d| \le \epsilon$.

The definition of a limit is as follows:

Definition: We say that a sequence $(a_n)$ converges to a limit $a$ if, for every $\epsilon \ge 0$, there is an integer $N(\epsilon)$ so that $|a_n - a| \le \epsilon \color{red}{\text{ for all }} n \ge N$.

From these definitions, and given this question:

If a sequence converges it has exactly one limit point. Is the converse true? If not, provide a counterexample

Can I conclude that the converse is indeed true simply by appealing to the definitions, or do I need a tighter proof. And if I do, I'm a little stuck on how to refine this proof.

  • 1
    $\begingroup$ Could an unbounded sequence have exactly one limit point? $\endgroup$ – David Mitra Aug 15 '12 at 16:52
  • $\begingroup$ Does a sequence of this form "count"?: $(a_n) = 1$ for $n = 2k, k \in \mathbb N$, $(a_n) = n$ for $n = 2k - 1, k \in \mathbb N$? $\endgroup$ – Moderat Aug 15 '12 at 16:54
  • $\begingroup$ Yes, its only limit point is $1$ but it does not converge. $\endgroup$ – David Mitra Aug 15 '12 at 16:58
  • $\begingroup$ So that sequence I have constructed works as a counterexample then. Thanks @DavidMitra $\endgroup$ – Moderat Aug 15 '12 at 17:00
  • $\begingroup$ Hi, How do you prove that if a sequence converges it has a unique limit point? $\endgroup$ – router Jul 4 '18 at 7:04

The sequence $2,1/2,3,1/3,4,1/4 \cdots, n,1/n \cdots$ has the only limit point $0$, but not converges.


You could also pick the sequence

$ (1 + (-1)^n).n $

which doesn't converge but has 0 as limit point.

  • 2
    $\begingroup$ To be explicit: it is the sequence $0,4,0,8,0,12,0,\dots$ $\endgroup$ – Ben Millwood Aug 15 '12 at 17:22

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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