There has recently been much discussion regarding entanglement transformati
ons in terms of local filtering operations and whether the optimal entangle
ment for an arbitrary two-qubit state could be realized. We introduce an ex
perimentally realizable scheme for manipulating the entanglement of an arbi
trary state of two polarization-entangled qubits. This scheme is then used
to provide some perspective to the mathematical concepts inherent in this f
ield with respect to a laboratory environment. Specifically, we look at how
to extract enhanced entanglement from systems with a fixed rank, and, in t
he case where the rank of the density operator for the state can be reduced
, show how the state can be made arbitrarily close to a maximally entangled
pure state. In this context we also discuss bounds on entanglement in mixe
d states.