> ... should be corrected, I guess: > http://www.research.att.com/~njas/sequences/A007185 I think there is an invisible zero in front of a(4) that makes it subtly different from a(3).