|
|
A109927
|
|
First primes p connected to two primes either by 2p+1 or 2p-1 upward [downward (p-1)/2 or (p+1)/2].
|
|
3
|
|
|
3, 5, 11, 23, 37, 83, 157, 179, 359, 661, 719, 877, 997, 1019, 1237, 1439, 1657, 2039, 2063, 2137, 2459, 2557, 2819, 2903, 2963, 3023, 3061, 3623, 3779, 3803, 3863, 4177, 4261, 4357, 4621, 4919, 5399, 5581, 5639, 6037, 6121, 6217, 6361, 6899, 6983, 7079
(list;
graph;
refs;
listen;
history;
text;
internal format)
|
|
|
OFFSET
|
1,1
|
|
COMMENTS
|
These primes may be part of Cunningham chains longer than three terms. It seems the two operators are never mixed, except for 3, 5 and 7: -for 3, we have: 2 through <2p-1> -> 3 through <2p+1> -> 7 -for 5: 3 <2p-1> -> 5 <2p+1> -> 11 -for 7: 3 <2p+1> -> 7 <2p-1> -> 13
For p > 7, such a mixed chain with p in the middle is impossible because the number 3 would be a nontrivial factor of either the smallest or the largest term. - Jeppe Stig Nielsen, May 05 2019
|
|
LINKS
|
|
|
EXAMPLE
|
a(3)=11 is here because 5->11->23 through <2p+1>;
a(4)=23 because 11->23->47 through <2p+1>;
a(5)=37 because 19->37->73 through <2p-1>.
|
|
PROG
|
Terms computed by Gilles Sadowski.
(PARI) forprime(p=3, 10^6, if(p%3==2, isprime((p-1)/2)&&isprime(2*p+1), isprime((p+1)/2)&&isprime(2*p-1))&&print1(p, ", ")) \\ Jeppe Stig Nielsen, May 05 2019
|
|
CROSSREFS
|
|
|
KEYWORD
|
easy,nonn
|
|
AUTHOR
|
|
|
STATUS
|
approved
|
|
|
|