%I #15 Jan 08 2013 13:44:16
%S 4,30,217,263,367,429,462,590,650,738,3385,3644,4522,4612,5949,14357,
%T 31545,40933,49414,104071,118505,149689,157680,165326,325852,415069,
%U 491237,566214,597311,733588,1319945,1736516,2850174,2857960,3183065
%N Values of k that show the slow decrease in the larger values of the Andrica function Af(k) = sqrt(p(k+1)) - sqrt(p(k)), where p(k) denotes the k-th prime.
%C a(n) are values of k such that Af(k) > Af(m) for all m > k. This sequence relies on a heuristic calculation and there is no proof that it is correct.
%D R. K. Guy, "Unsolved Problems in Number Theory", Springer-Verlag 1994, A8, p. 21.
%D P. Ribenboim, "The Little Book of Big Primes", Springer-Verlag 1991, p. 143.
%H H. J. Smith, <a href="/A084976/b084976.txt">Table of n, a(n) for n=1..128</a>
%H H. J. Smith, <a href="http://harry-j-smith-memorial.com/PrimeSR/">Andrica's Conjecture</a>
%H Eric Weisstein's World of Mathematics, <a href="http://mathworld.wolfram.com/AndricasConjecture.html">Andrica's conjecture.</a>
%e a(3)=217 because p(217)=1327, p(218)=1361 and Af(217) =sqrt(1361) - sqrt(1327) = 0.463722... is larger than any value of Af(m)for m>217.
%Y Cf. A078693, A079098, A079296, A084974, A084975, A084977.
%K nonn
%O 1,1
%A _Harry J. Smith_, Jun 16 2003