The "excess entropy" around nonequilibrium steady states, (deltaS)(ss), is not a Liapunov function.

In a response to my recent paper [Fox, R. F. (1979) Proc. Natl. Acad. Sci. USA 76, 2114-2117], Nicolis and Prigogine [Nicolis, G. & Prigogine, I. (1979) Proc. Natl. Acad. Sci. USA 76, 6060-6061] reasserted that the "excess entropy" around nonequilibrium steady states, (delta(2)S)(ss), is a Liapunov function. A simple, explicit counterexample which invalidates this claim is presented. The existence of an alternative theory possessing a proper Liapunov function for steady states is reviewed.