On the Possible Existence of Closed Disentropic Systems

I construe the question "Is the existence of a closed physical system whose entropy is decreasing possible?" as equivalent to the question "Is it possible for an entropically increasing physical system to be coupled, as measurer to measured, with an entropically decreasing physical system such that the entropy of the resulting system decreases?". Norbert Wiener's suggestion that the answer is negative, implying that the 2nd Law of Thermodynamics is inviolable, is then shown to be incorrect by the construction of a counter-example achieved thru a slight modification of a machine first proposed and rejected by Leo Szilard. I then show, however, through a proof from n-dimensional geometry, that physical analogies to the counter-example are only possible on the microscopic level (that is, with systems involving few molecules). Though one can never envisage a macroscopic system having decreasing entropy, therefore, the possible existence of such systems on the microscopic level is mathematically affirmed.