The primes $p$ are, of course, in one-to-one correspondence with the squares of primes $p^2$. But is there any interval $a < x < b$ possible where the primes thin out so much, that it contains more squares of primes than primes?

  • 2
    $\begingroup$ Trivially, yes. Any interval where $a = p^2-1$ and $b = p^2+1$, for some prime $p$. $\endgroup$
    – LarsH
    Jan 17 '16 at 3:36
  • 1
    $\begingroup$ @Ross: Maybe you can make it less easy by requiring that the interval contains $n$ primes with $n>>0$. $\endgroup$ Jan 17 '16 at 3:57
  • $\begingroup$ How embarrassing. Oh well. $\endgroup$ Jan 18 '16 at 13:31
  • $\begingroup$ @TitoPiezasIII: Having been spanked on the simpleminded phrasing of the question I'm hesitant to try again. $\endgroup$ Jan 18 '16 at 13:32
  • $\begingroup$ I'd be interested to know the motivation for this question. $\endgroup$
    – Josh Chen
    Jan 20 '16 at 10:38

The interval $23 \lt x \lt 29$ is one such

  • $\begingroup$ And $47<x<53$, etc. $\endgroup$ Jan 17 '16 at 3:52

I think the intent of your question was that the number of primes in the interval, call it $\pi(n)'$, is non-zero. If so, then the simplest case is,

$$p_n^2\leq x \leq p_{n+1}^2$$

for prime $p_i$. Your question then assumes the count as $\pi(n)'<2.$ However,

  1. A consequence of Legendre's conjecture is that the number of primes in that interval is at least 2.
  2. More strongly, if Brocard's conjecture is true then, for $n>1$, there are at least 4.

As the count $\pi(n)' =2, 5, 6, 15, 9, 22, 11, 27, 47,\dots$ (A050216) goes up fast, it is highly doubtful that Brocard's conjecture is false.

  • $\begingroup$ I believe Bertrand's postulate would also be relevant. $\endgroup$ Jan 20 '16 at 10:07
  • $\begingroup$ No, in fact when I asked my question, I was imagining a desert containing no primes, not even one, but where at least two prime squares fall. You're saying that Legendre's conjecture makes that impossible. $\endgroup$ Jan 20 '16 at 15:11
  • $\begingroup$ And (after reading it) I see Bertrand's postulate, which is proven, completely rules out my scenario. $\endgroup$ Jan 20 '16 at 15:15
  • $\begingroup$ @RossPresser: Yes, Legendre's (and Brocard's) conjecture make it impossible since they postulate at least $2$ and $4$ primes, respectively, in that interval. $\endgroup$ Jan 20 '16 at 15:46
  • $\begingroup$ Those two are conjectures, but Bertrand's postulate is proven. $\endgroup$ Jan 20 '16 at 15:58

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.