Let $a_n=\frac{n!}{n^n}$ be a sequence. Prove that $\limsup(a_n)=0$.

I know that the series $\sum^{\infty}_{n=1} \frac{n!}{n^n}$ converges, so the nth term must converge to 0. But I cannot use the convergence of series to prove this.

Basically what I have is: The sequence $a_n$ is decrescent and bounded by 0 $(a_n>0 \; \forall n)$, therefore is convergent and it's limit is the $\limsup(a_n)$.

So I must show that $\lim\frac{n!}{n^n}=0$

Let $\epsilon>0$, prove that exists $n_0 \in \mathbb{N}$ such that, if $n>n_0$

$$\left|\frac{n!}{n^n}-0\right|<\epsilon \Rightarrow -\epsilon<\frac{n!}{n^n}<\epsilon$$

I know that $\frac{n!}{n^n}>=-\epsilon$ because $\frac{n!}{n^n}>0$.

How can I prove that $\frac{n!}{n^n}<\epsilon$?

  • 1
    $\begingroup$ Bound it by something you know goes to zero. Here's an idea: you have $\prod_{i=1}^n \frac{i}{n}$, note that all the terms in this product are less than or equal to $1$ and $\lfloor n/2 \rfloor$ of them are less than or equal to $1/2$. Can you conclude? $\endgroup$ – Ian Oct 8 '16 at 21:53
  • $\begingroup$ How do you know the series converges? $\endgroup$ – πr8 Oct 8 '16 at 22:12

$$\dfrac{1}{n^n}\leq\dfrac{n!}{n^n}=\dfrac{n-1}{n}\dfrac{n-2}{n}\cdot\cdot\cdot\dfrac{1}{n}\leq1\cdot1\cdot 1\cdot\cdot\cdot\dfrac{1}{n}=\dfrac{1}{n}.$$

Does this help?

  • $\begingroup$ +1 for you. Using this you can say $\frac{n!}{n^n} \leq \frac{1}{n} < \epsilon$. So pic $n_0 = \frac{1}{\epsilon}$ $\endgroup$ – Ahmad Bazzi Oct 8 '16 at 22:20

Hint $$\frac{n!}{n^n}=\frac{1}{n}\cdot \frac{2}{n} \cdot \frac{2}{n} \cdot...\cdot \frac{n}{n} \leq \frac{1}{n} \cdot 1 \cdot 1 \cdot ... \cdot 1$$


Another way more to show: using the quotient test for positive series, we get


so the series $\;\sum\limits_{n=1}^\infty\frac{n^n}{n!}\;$ converges, and thus



Suppose we choose any positive, sufficiently small $^{\mathbf{\dagger}}$ epsilon. We have $\displaystyle \frac{n!}{n^n} < \varepsilon \iff n! < \varepsilon (n^n)$.

Now taking the log of both sides and applying relevant properties, the above is true $\iff \displaystyle \sum_{k=1}^n \ln(k) < \ln(\varepsilon) + n \ln(n)$.

Rearranging yields $\displaystyle n \ln(n) - \sum_{k=1}^n \ln(k) \ = \ \sum_{k=1}^n \ln(n/k) > -\ln( \varepsilon)$.

We'll have $\displaystyle \left\lceil -\ln( \varepsilon) \right\rceil = M$ for some $M \in \mathbb{R}^+$. The above inequality holds for all $n > e^M$, so we may conclude $\displaystyle \lim_{n \rightarrow \infty} \left( \frac{n!}{n^n} \right) \rightarrow 0$.

$^{\mathbf{\dagger}}$ Important only so we don't have to worry about the sign of $\ln(\varepsilon)$.


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.