Hello SeqFans, It appears that (3n^2 - 5)*ceiling(n/sqrt(3))^2 > n^4 for all n > 1. Can someone prove this are cite something? Thanks. Clark Kimberling