According to a theorem in my book, it follows that

"If the series $\sum_{k=1}^{\infty}a_k$ converges $\Longrightarrow \lim_{k\rightarrow\infty}a_k=0$, but the reverse implication is not true."

This clearly means that I can't use this theorem to conclude whether a series converges.


  1. What use is this theorem for then?
  2. Is this statement true: $$\lim_{k\rightarrow\infty}a_k=\pm\infty\Longrightarrow \sum_{k=1}^{\infty}a_k \ \text{diverges} \ ?$$
  3. What if $k$ starts from $0$, will above theorems still hold?
  • $\begingroup$ It's a divergence criterion $\lnot(a_k \to 0) \implies \sum a_k$ diverges. $\endgroup$ Dec 30 '17 at 14:38
  • $\begingroup$ Note that in $p$-adic Analysis, a series is convergent if and only if its general term tends to $0$. $\endgroup$
    – Bernard
    Dec 30 '17 at 14:42
  • $\begingroup$ Yes, but I can't compute the limit of the general term to conclude if the series is convergent, according to that theorem above? $\endgroup$
    – Parseval
    Dec 30 '17 at 14:49

1.The theorem is mainly used for establish the divergence of a series. (Contrapositive to the original).

2.Your statement is true. Actually, whenever the limit is not going to 0, you can conclude the series is divergent.

3.$k=0$ is fine. Actually, you can trim away the first finite terms in a series and its divergence could still be established with this theorem.

(Some more thoughts for students studying real analysis: We can trim off the first finite terms of a sequence due to this theorem:
Let $X=\{X_n:n\in\mathbb N\}$ be a sequence of real numbers and let $m\in \mathbb N$. Then the m-tail $X_m=\{X_{n+m}:n\in\mathbb N\}$ of X converges if and only if X converges. In this case, $\lim_{x \to \infty} X_m=\lim_{x \to \infty} X$ )

  • $\begingroup$ One more question, what would happen if $\lim_{k\rightarrow\infty}a_k=m$? The number $m$ is just an arbitrary real constant. $\endgroup$
    – Parseval
    Dec 30 '17 at 15:01
  • 1
    $\begingroup$ As long as $m\neq 0$ we conclude that the series is divergent. $\endgroup$
    – Macrophage
    Dec 30 '17 at 15:04

As for every implication, for negation we have that:

$$\lim_{k\rightarrow\infty}a_k\neq 0\Longrightarrow \sum_{k=1}^{\infty}a_k \ \text{does not converges}$$

It could diverge or oscillate.

This is used to prove easily that the series does not converges.

The series behaviuor is independent by a finite number of initial terms.


In a nutshell:

$\lim _{k \rightarrow \infty }a_k = 0$ is

a necessary but not sufficient condition

for the conververgence of the series

$\sum_{k=0}^{\infty} a_k.$

Necessary :

Convergence of series implies

$\lim_{k \rightarrow \infty} a_k =0.$

The geometric series with terms: $a_k =(1/2)^k$ converges.

The criterion is not sufficient:

For the terms of the harmonic series: $a_k = 1/k $, we have

$\lim _{k \rightarrow \infty}a_k= 0.$

However: The series $\sum_{k=1}^{\infty} 1/k$ is not convergent.


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.