Mathematics Stack Exchange is a question and answer site for people studying math at any level and professionals in related fields. 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

A problem from "Problems in Real Analysis - Advanced Calculus on the Real Axis":

Let $p$ be a nonnegative real number. Study the convergence of the sequence $(x_n)_{n\ge1}$ defined by $$x_n=\left(1^{1^p}\cdot2^{2^p}\cdots(n+1)^{(n+1)^p}\right)^{1/(n+1)^{p+1}}-\left(1^{1^p}\cdot2^{2^p}\cdots n^{n^p}\right)^{1/n^{p+1}}.$$

If $p=0$, it is already proved in the book that $x_n$ converges to $\dfrac1e$. So just consider the case that $p>0$. If the sequence converges, I can prove by Stolz-Cesàro Theorem that the sequence must converge to $0$. But I don't know how to determine which $p$ gives convergent sequences.

Please don't give complete solution, just helpful hints. Thanks in advance.

share|cite|improve this question

HINT : Let $ a_n = \left( 1^{1^p} \cdot 2^{2^p} \cdots n^{n^p} \right)^{1/n^{p+1}}$ so that $x_n = a_{n+1}- a_n $. Now study the asymptotics of $\log a_n$ using the Euler-Maclaurin formula.

share|cite|improve this answer
I didn't know what Euler-Maclaurin formula is. I already googled it, but I don't have a slightest clue of how to apply it here. Can you give more elementary hint? – dkdsj93 Oct 5 '11 at 2:09
@3423847238 I'm sorry, I don't have more a more elementary method. Some of these types of questions are designed to be solved by special tricks and clever repeated applications of some theorems like Caesaro-Stoltz, but a general powerful method is Euler-Maclaurin. You should learn about it. The way to apply it here is simple: $\log a_n$ is a sum whose value we can estimate very well if we take enough terms of the Euler-Maclaurin estimate. You can then use that estimate to estimate $a_{n+1} - a_n$. Sorry about this late reply by the way. – Ragib Zaman Oct 9 '11 at 21:09

Use $\sum_{k=1}^n k^p \log(k) = \frac{\mathrm{d}}{\mathrm{d} p} H_n^{(-p)}$, as well as series expansion for harmonic numbers $H_n^{(-p)}$ of order $-p$, using $H_n^{(-p)} = \zeta(-p) - \zeta(-p, n+1)$.

I am getting zero value for the limit, as you claimed.

share|cite|improve this answer
Somehow I get the feeling that calling in the Hurwitz zeta function here is mosquito-nuking... – J. M. Oct 3 '11 at 16:14
This seems to be completely out of my reach.. – dkdsj93 Oct 5 '11 at 2:10

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.