A counterexample to "adaptive partitioned random search to global optimization"

The optimality theorem presented in the above paper by Tang (see ibid., vol.39, no.11, p.2235-44, 1994) is stated to be incorrect. A counterexample is presented.