I haven't read the paper, but in other 'fun with primes' exercises, 1 is often excluded.