Alexander, Your's and Paolo's are interesting formulas: thanks to both of you. Maybe one exists, after all, that expresses a(n) in terms of n alone (which I supposed to be unlikely). Ed Jeffery