The Collatz mapping implies a partition of the natural numbers that is instrumental in solving the conjecture. But can we prove the partition?