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

Since I'm studying real analysis using this book by myself, I'm not sure whether or not my method to prove convergence of sequence is right. I'm working on the above question's (d), and my solution was kind of different from the solution provided by the following website. The solution in the website is much more beautiful and more concise than mine, but I believe my method also works.

3.14 If $\{s_n\}$ is a complex sequence, define its arithmetic means $\sigma_n$ by

$$\sigma_n = \dfrac{s_0+s_1+...+s_n}{n+1} (n=0,1,2,...).$$

(d) Put $a_n=s_n-s_{n-1}$, for $n\geq 1$. Show that

$$s_n-\sigma_n=\dfrac{1}{n+1}\sum_{k=1}^n k a_n$...$(1)$$

Assume that lim$(n a_n)=0$ and that ${\sigma_n}$ converges. Prove that $\{s_n\}$ converges.

My solution:

Since $\sigma_n$ converges, it is a Cauchy sequence. So,

$\forall\epsilon,\forall\delta>0, \exists N\in\mathbb{N}$ s.t.$\forall n, \forall m> N , |n a_n-0|<\epsilon$ and $|\sigma_n-\sigma_m|<\delta$

Let $M=\max\{k |a_k|:1\leq k\leq m\}$

From eq. $(1)$,

\begin{align} |s_n-s_m|&=|(\sigma_n-\sigma_m)+(\dfrac{1}{n+1}\sum_{k=1}^{n}k a_k-\dfrac{1}{m+1}\sum_{k=1}^{m}k a_n)|\\&\leq|\sigma_n-\sigma_m|+\dfrac{|(m+1)(a_1+2a_2+...+n a_n)-(n+1)(a_1+2a_2+...m a_m)|}{(n+1)(m+1)}\\&=|\sigma_n-\sigma_m|+\dfrac{|(m+1)((m+1)a_{m+1}+(m+2)a_{m+2}+...+n a_n)+(m-n)(a_1+2a_2+...m a_m)|}{(n+1)(m+1)}\\&<\delta+\dfrac{(m+1)(n-m)\epsilon+(m-n)mM}{(n+1)(m+1)}\\&=\delta+\dfrac{(n-m)\epsilon}{n+1}+\dfrac{(m-n)mM}{(n+1)(m+1)}\end{align}

As $\epsilon, \delta \to 0, n \to \infty$ and $m \to n$, the right side converges to $0$; therefore, the Cauchy sequence $\{s_n\}$ converges.

share|cite|improve this question
Am I allowed to assume that M such that $M=max\{k|a_k|:1≤k≤m\}$ exists? – Math.StackExchange May 20 '14 at 0:53
up vote 4 down vote accepted

You're overcomplicating things. Suppose that $\sigma_n\to s$. I claim $s_n\to s$.

Cesaro's theorem says that if $x_n\to x$, then $\widehat{x_n}\to x$ where $\widehat{x_n}=\displaystyle \frac{1}{n+1}\sum_{k=0}^nx_k$.

Let $x_k=ka_k$. Then $ka_k\to 0$; so $\widehat{x_k}=s_k-\sigma_k\to 0$. Since $\sigma_k\to s$, $s_n\to s$.

Thus, all you have to do is prove Cesaro's theorem. You can assume that $x_n\to 0$ (why?), and then it is all pretty easy.

share|cite|improve this answer
Thanks for your suggestion. Although I didn't know this theorem before, surprisingly, I unconsciously proved a part of the theorem in the question (a). I'm really glad to know a simpler way to solve the question. – Math.StackExchange May 20 '14 at 1:07

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.