Am Thu, 31. March 2005 um 10:02:40 +0200 schrieb Dr. med. Roland Linder:
> My hypothesis: Starting with one-digit primes, you can use a Mersenne prime as
> Mersenne exponent to yield a further Mersenne prime.

    This sequence C_i = 2^C_{i-1} - 1 with C_0 = 2 is known as Catalan-
Mersenne sequence.  I find these numbers very interesting, too, but
little seems to be known about them.  The best pointers I have found are

http://mathworld.wolfram.com/Catalan-MersenneNumber.html
(plus the references stated there) and
http://www.utm.edu/research/primes/mersenne/#c .

C0...C4 are prime (2, 3, 7, 127 and 170141183460469231731687303715884105727),
but the status of all the others is unknown.

    One thing is for sure: Either all C_i are prime, or only the first
few of them (C_0...C_j for some j).  That's because if some C_i is
composite, C_{i+1} = 2^C_i - 1 can't be a (Mersenne) prime and so on.

    It is believed by "Guy's strong law of small numbers" that only the
first few (only C_0...C_4?) are prime, but who knows?  I think this must
be solved by some wise number theorist rather than sheer computing
power.  In any case I don't know how it could be done to run the Lucas-
Lehmer test in parallel, if possible at all.  But maybe there is a way
to simplify the Lucas-Lehmer test just for this special case, or
something like that?

    By the way, this sequence is a subset of the so called Double
Mersenne numbers MM_p = 2^(2^p-1)-1 with 2^p-1 prime.  The first four of
them are prime, the next four are composite, and the status of the rest
is unknown.

    And I think I've seen some prove that the decimal representation of
all C_i with i >= 2 ends with "7", that of all C_i with i >= 3 ends with
"27", and so on, but I don't remember where.

    - Stephan Niemz.

Attachment: signature.asc
Description: Digital signature

_______________________________________________
Prime mailing list
[email protected]
http://hogranch.com/mailman/listinfo/prime

Reply via email to