On the divisibility of numerator of B(2n) by primes
benoit
abcloitre at wanadoo.fr
Tue Feb 10 17:06:44 CET 2004
Following TD Noe. For a given prime p here is an essay to define
general rules for the sequence of n such that p divides the numerator
of B(2n). I considered whether p is irregular or not.
(i) if q is a regular prime :
For a regular odd prime q, the sequence of n such that q divides the
numerator of B(2n) "is" given by :
{q*k}_{k>=0 and k not congruent to (q-1)/2}
Should be a direct consequence of Von Staudt and Clausen.
Exemples :
(i.1) the sequence of n such that 5 divides the numerator of B(2n)
(divided by 5) "is" given by :
1,3,5,7,9,11,13,15,17,19,21,23,25,27,29,31,33,35,37,39,41,43,45,47,49,51
,53,55,57,....
and even numbers = multiple of (5-1)/2=2 are missing.
(i.2) the sequence of n such that 17 divide the numerator of B(2n)
(divided by 17) "is" given by :
1,2,3,4,5,6,7,9,10,11,12,13,14,15,17,...
and multiple of (17-1)/2=8 are missing.
(ii) if p is an irregular prime :
For a given irregular prime p let {x1,x2,...,xt} denotes the set of
distinct values 1<=x1<...<xt<(p-1)/2 such that p divides numerator of
B(2*xi) for i=1,...,t.
Then I suspect the sequence of n such that p divides the numerator of
B(2n) "to be" given by :
{p*k} U_{i=1}^{t} {(p-1)/2*k+xi} for k>=0.
Exemples :
(ii.1) the sequence of n such that 37 divides the numerator of B(2n) is
(here there is a single value x1=16) :
16,34,37,52,70,74,88,106,111,124,142,148,160,178,185,196,214,222,232,250
,259,268,286,
which "is" the sequence {37*k} U {18*k+16} k>=0 .
(ii.2) the sequence of n such that 157 divides the numerator of B(2n)
is (here there are 2 values x1=31 and x2=55) :
31,55,109,133,157,187,211,265,289,314,343,367,
which "is" the sequence {157*k} U {78*k+31} U {78*k+55} k>=0 .
If all of this is true : B(2n)/(2n)/(2n-1) has a different numerator
than B(2n)/(2n) means that at least one prime factor p of (2n-1)
divides B(2n).
Since (2n,2n-1)=1 clearly p can't be regular from (i). Otherwise
2n=2*p*k contradicting (2n,2n-1) =1.
Determining the sequence of n for which numerator of B(2n)/(2n)/(2n-1)
differs from numerator of B(2n)/(2n) should imply the computation of
{(xi)} for all irregular primes.
For info. I submitted this complement of "indexed" sequences for
irregular primes (as A073277) :
%S A000001
0,0,0,0,0,0,0,0,0,0,1,0,0,0,0,1,0,1,0,0,0,0,0,0,1,1,0,0,0,0,1,0,0,1,0,2,
0,0,0,0,0,0,0,0,0,0,0,0,0,1,0,0,0,1,1,0,1,0,0,1,1,1,1,0,0,0,0,1,0,2,0,0,
0,2,0,1,0,1,1,0,1,0,1,0,0,0,0,1,1,2,0,0,3,0,0,0
%N A000001 number of k, 1<=k<=(n-1)/2, such that n divides the
numerator of B(2k) where n runs through the odd primes.
%F A000001 a(n) = 0 if n is a regular prime ; a(n) > 0 if n is an
irregular prime
B.Cloitre
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: text/enriched
Size: 2876 bytes
Desc: not available
URL: <http://list.seqfan.eu/pipermail/seqfan/attachments/20040210/21e4525f/attachment.bin>
More information about the SeqFan
mailing list