10

For a given $k$ define

$$s_k = 1 + \prod_{i=1}^k p_i$$ $$t_k = \text{NextPrime}(s_k)$$ $$v_k = t_k - s_k +1$$

Where $p_i$ is the $i$th prime number.

Conjecture: $v_k$ is prime

Example: $$k=3$$ $$s_3 = 1+(2)(3)(5)=31$$ $$t_3 = 37 $$ $$v_3 = 37-31 +1 = 7 \,\text{(prime)} $$

Tested numerically up to $k=400$ and it seems to hold.

StubbornAtom
  • 17,052
sheppa28
  • 929
  • can you please share your program? – Pruthviraj Aug 12 '19 at 19:29
  • 2
    Note $\ v_k = t_k - 2\cdot 3\cdots p_k,$ is coprime to $,2,3,\ldots, p_k\ \ $ – Bill Dubuque Aug 12 '19 at 19:36
  • As @AccidentalFourierTransform notes, this sequence has received some attention. For that matter, easy plausibility/probabilistic heuristic arguments around the Prime Number Theorem strongly suggest that it is asymptotically true. E.g., maybe RH implies it's asymptotically true. Surely less than RH would give something, etc. – paul garrett Aug 12 '19 at 20:09
  • one plus any prime give an even number which can be written as the sum of two primes smaller than the initial prime.moving the primes before the initial prime gives your relation. unless one proves opposite. – sirous Aug 12 '19 at 20:12
  • 1
    This is known as Fortune's conjecture – Sil Aug 12 '19 at 20:19
  • If next prime is within ${p_{k+1}}^2$ it's guaranteed as until then only distances would be forced to be 1 less than a prime. –  Aug 12 '19 at 22:18

1 Answers1

6

This is sequence A005235 on OEIS. It is conjectured to always be prime, although it is not known yet. Check the page for useful remarks and references.

AccidentalFourierTransform
  • 2,592