5
$\begingroup$

Is there any proof of the inequality $p_n<n^2$ (for all sufficiently large $n$) by elementary means and without using Prime Number Theorem?

I searched in google but in vain. The results that I have found (and from which the inequality can be proved) are mostly results that arises as a consequence of PNT.

So, is there any such proof? If so, then please include a link of the paper in your answer (or comment).

$\endgroup$
  • 2
    $\begingroup$ You can prove by elementary means $p_n=O(n\log n)$. $\endgroup$ – karvens Oct 21 '14 at 5:09
  • $\begingroup$ The statement is not true for $p_1=2$, so you need an exception for that case. $\endgroup$ – Esteban Crespi Oct 21 '14 at 5:50
  • $\begingroup$ @karvens: But how does it conclude that $p_n<n^2$, can you elaborate? $\endgroup$ – user 170039 Oct 21 '14 at 6:10
  • 1
    $\begingroup$ If $p_n=O(n\log n)$, we have $p_n<Cn\log n<n^2$ for all sufficiently large $n$. $\endgroup$ – karvens Oct 21 '14 at 6:14
  • 1
    $\begingroup$ @user170039 Chebyshev already proved something of the strength $p_n \le (1.2 + o(1)) n \log n$; this was nearly 50 years before PNT was established. The proof that karvens refers to is probably the simpler one due to Erdős. $\endgroup$ – Erick Wong Oct 21 '14 at 8:52
2
$\begingroup$

As said in the comments, Chebyshev already proved by elementary means that $p_n<C n \log(n)$ for all $n\ge n_0$. As there was explicitly asked for a reference, I include one (of many): Theorem $5.1$ here.

$\endgroup$

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.