The starting point is given by Fermat's little theorem. It seems almost magical to discover the presence of (a) among the factors of 2^a-2 only if (a) is prime. Unfortunately it should also be noted in 2^a-2 the presence of many prime factors lower than (a); this can ruin the magic when (a) is a number made up of some of the factors that make up 2^a-2. I imagined that (2^a-2)/a can be considered as the comparison between two sequences with different growth rates. This revision is due to two reasons. The first is to modify the for() function I use below, to make it more efficient. The second is to point out that I have found two new sequences which from my verifications seem to work perfectly; I describe them in the first article of the list that I report at the end of this one.