Thank you Hans Havermann. That frequency of the digits 1-9 looks a little like Zipf's Law or Benford's Law. So OEIS does not yet have the sequnce where a(n) is the smallest integer whose n-th power begins with n identical digits, base 10?