The graphics on recursion in HOP book help a lot to visualize how recursion works in the code. Take a binary tree as a matter of example for representing the recursive solution to the problem. In a recursive procedure the flow goes from the most left-side branch to the most left-side leave all the way down. In the iterative procedure, the flow goes top-down, horizontal layer of nodes after horizontal layer of nodes
When you combine both procedures (recursion + iteration), Recursion takes priority over iteration. That means if recursion takes places when the loop variable is set to, for instance 1, at the first level of your tree, then iteration is "interrupted", recursion goes all the way down till all loops of the nested branched nodes are resumed, then the flow returns back from bottom to top (through the parent nodes) till reaches the first level again and goes on subsequently with the next value of the loop variable (set from 1 to 2) and so on and so forth till the first loop of the first level is accomplished.
in reply to Mark Jason Dominus And Me - The Partition Problem