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

I would like to show that

$$\sum \frac{1}{nn^{1/n}}$$ diverges, and I'm quite certain I will have to use the comparison test. I don't immediately see how that would be useful, though.

share|cite|improve this question
Do you know how $n^{1/n}$ behaves as $n$ gets big? – Jonas Meyer Sep 13 '12 at 2:06
Well.... actually this is an answer that I DO understand very well (didn't see this comment before). As n gets big that goes to 1, so essentially this summation "converges" to the harmonic series for large n consequently it diverges. – Squirtle Sep 13 '12 at 2:36
dustanalysis: Once you realize that, you can go directly to the limit comparison test: which is what DonAntonio alludes to in his second method. But Brian indicates how you can use that limit together with the direct comparison test, by comparing to $\frac{1}{cn}$ instead of $\frac{1}{n}$, with $c>1$. – Jonas Meyer Sep 13 '12 at 2:42
up vote 5 down vote accepted

According to Cauchy's Condensation Test , our series converges iff the following series converges:



$$2^{n/2^n}=e^{\frac{n}{2^n}\log 2}\xrightarrow [n\to\infty]{}1\neq 0$$

thus our series diverges.

Added: Another easy test: choose $\,\{b_n:=\frac{1}{n}\}\,$ ,then:

$$\frac{\frac{1}{n\,n^{1/n}}}{b_n}=\frac{1}{\sqrt[n] n}\xrightarrow [n\to\infty]{}1\neq 0$$

Thus, our series converges iff the series $\,\sum b_n\,$ does...which it doesn't as it is the harmonic series.

share|cite|improve this answer
Don't you have to plug in 2^n everywhere then? In particular when taking the (1/n) power you didn't replace that with (1/2^n). Also, I'm not certain which technique you used under the "added" section. I appreciate your help, btw.... but could you add a little more explanation. – Squirtle Sep 13 '12 at 2:31
About your first question: yes, but you're completely right and I oversaw that power. I already fixed it in my post, thanks. About the second proof: it is a standard test for positive series and you can find it here – DonAntonio Sep 13 '12 at 2:49
@DonAntonio.... I had a blonde moment though, because it turns out that its obviously true that what you now have is "bigger" than what you had. (I.e. for all x>1, n a positive integer,we have x^(1/2^n) < x^(1/n) therefore, (1/x^(1/2^n)) > (1/x^(1/n)). In other words, the old way I actually think is "better" because the solution looks so clean, thanks for the insightful typo. – Squirtle Sep 13 '12 at 3:54
Yes @dustanalysis, I know, yet it should be what it must be, and the factor $\,2^n\,$ must appear in that exponent. Good you found it helpful – DonAntonio Sep 13 '12 at 10:17

HINT: Suppose that you can find a positive number $c$ and a positive integer $n_0$ such that $n^{1/n}<a$ for all $n\ge n_0$; then you’d have $$\frac1{nn^{1/n}}\ge\frac1{cn}$$ for all $n\ge n_0$. What do you know about $$\lim_{n\to\infty}n^{1/n}\;?$$

share|cite|improve this answer

$\displaystyle \sum \frac{2^n}{2^n(2^n)^{1/2^n}}>\sum \frac{2^n}{2^n(2^n)^{1/n}}=\sum\frac{1}{2}=∞$

share|cite|improve this answer
After applying Cauchy's Condensation Test – Squirtle Sep 13 '12 at 4:01

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.