Prime-index-primes and twin primes
cino hilliard
hillcino368 at hotmail.com
Mon Oct 27 14:10:46 CET 2003
H fans,
Consecutive prime-index-primes (PIPS) ie., prime(prime(prime..(prime(x))..)
and
prime(prime(prime..(prime(x+1))..)
contain twin prime pairs.
But prime(prime(x)) and prime(prime(prime(x))) (PIPS order 1 and 2) do not
for some x, x+1.
Now prime(prime(prime(prime(x)))) or PIPS of order 3 contain a twinprime
pair for all x < 33176.
This is the limit for my 2 gig ram and pari primelimit. I am guessing
consecutive PIPS order 4,5,..n will
always contain a twin prime pair.
It seems there should be a way to tie PIPS and twin primes together. If we
could prove PIPS of order
n must always contain a twin prime pair then we will have a proof of the
infinity of twin primes since
PIPS_n(x) are infinite in lieu of primes are infinite.
This Pari script is for PIPS of order 3. Print if no twin prime pair is
found.
piptwins3(m,n) =
{
for(x=m,n,
f=1;
p1 = prime(prime(prime(prime(x))));
p2 = prime(prime(prime(prime(x+1))));
for(y=p1,p2,
if(isprime(y) && y+2 <= p2 && isprime(y+2),f=0)
);
if(f,print1(x","))
)
}
Maybe some one has software to check this for > 33176 and for PIPS order
5,6,.. to create
new sequences disproving my conjecture?
I have been thinking about a link between PIPS and twin primes for over 30
years now and have only
been able to provide statistical evidence showing that the density of PIPS
is less than the density of
twin primes.
Any thoughts are appreciated.
Cino
_________________________________________________________________
Fretting that your Hotmail account may expire because you forgot to sign in
enough? Get Hotmail Extra Storage today!
http://join.msn.com/?PAGE=features/es
More information about the SeqFan
mailing list