I have been exploring the fascinating world of prime numbers, particularly Mersenne Primes, and have noticed an interesting pattern. It seems to me that $2^n - 1$ is prime as long as $n$ itself is a Mersenne prime.

I already know that $2^n - 1$ is not necessarily prime just because $n$ is a normal prime number, but I have not found a counterexample that to the claim that $2^n - 1$ is a prime whenever $n$ is a Mersenne prime.

Is there a proof that this claim is false?

  • $\begingroup$ True for the first four Mersenne primes, and false for the fifth one already - see here. $\endgroup$ – Dietrich Burde Nov 6 '17 at 16:23
  • $\begingroup$ I know some facts don't related directly with your question. Maybe you know those, I cited these because I think that are important. You can see from this The Prime Page, by Professor Chris Caldwell the graph for Mersenne primes in section 6 Where is the next larger Mersenne prime?. On the other hand I like very much also several conjectures due to Professor Farideh Firoozbakht, you can read these in The Prime Puzzles & Problems Connection, by Carlos Rivera. I say for example the Puzzle 434 and Puzzle 517. $\endgroup$ – user243301 Nov 6 '17 at 21:33

$2^{13}-1 = 8191$ is a Mersenne prime, but $2^{8191}-1$ is composite.

  • $\begingroup$ Excellent, I knew that logically there had to be a counterexample, I just couldn't find one on the internet. If there is an article that goes into depth on this, or where you got this example, please link to it! $\endgroup$ – Kernel Stearns Nov 6 '17 at 16:23
  • $\begingroup$ @KernelStearns But this is on the internet; the above list says that $2^{8191}-1$ is not a (Mersenne) prime. $\endgroup$ – Dietrich Burde Nov 6 '17 at 16:25
  • $\begingroup$ @DietrichBurde I see now the small type at the bottom that implies every Mersenne prime between 0 and M37,156,667 is on the list. However, this post was meant to provide a quick answer to other people who are new to primes and have the same question. $\endgroup$ – Kernel Stearns Nov 6 '17 at 16:28
  • $\begingroup$ That's fine. I only wanted to argue about your statement "I just couldn't find one on the internet." Here the homepage on Mersenne "www.mersenne.org" immediately suggest itself. $\endgroup$ – Dietrich Burde Nov 6 '17 at 16:33
  • $\begingroup$ In the past I tried to combine some problems as the mentioned in the top, Puzzle 434 and Puzzle 517 by Firoozbakht from The Prime Puzzles & Problems Connection, or Puzzle 774 by Noahiro Nomoto with the equation for integers $n>1$ involving the sum of remainders function (see the article [1] or [2]), but I didn't any remarkable deduction. Maybe you or @DietrichBurde can do experiments in your home about it. [1] T. Cross, A note on almost perfect numbers, Mathematics Magazine, 47 (1974), 230–231. [2] Z. Spivey, The Humble Sum of Remainders Function, Mathematics Magazine, 78, No. 4 (2005). $\endgroup$ – user243301 Nov 6 '17 at 22:02

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.