FW: a special 3_F_2

["Daniel Lozier" <lozier@nist.gov>]

-----Original Message-----
From: opsftalk@nist.gov [mailto:opsftalk@nist.gov]On Behalf Of R.
Vidunas
Sent: Friday, March 30, 2001 9:11 AM
To: dlozier@nist.gov
Subject: Re: a special 3_F_2 (fwd)


Sender: "R. Vidunas" <vidunas@science.uva.nl>
Subject: Re: a special 3_F_2 (fwd)


It is not true that
3_F_2(-b2,b1+1,b1+1;b1+1-n,b1+2+n;1) <> 0
if b1,b2,n are positive integers such that b1<>b2 and b1>n.

For example, for b2=1 the hypergeometric sum is:
(b1+1-n-n^2)/(b1+2+n)/(b1+1-n),
so if we take b1=n^2+n-1 and n>=2 we get a contrexample.

R.V.