Recent measurements show that the free surfaces of liquid metals and alloys are always layered, regardless of composition and surface tension; a result supported by three decades of simulations and theory. Recent theoretical work claims, however, that at low enough temperatures the free surfaces of all liquids should become layered, unless preempted by bulk freezing. Using x-ray reflectivity and diffuse scattering measurements we show that there is no observable surface-induced layering in water at $T=298\mathrm{}\mathrm{K},$ thus highlighting a fundamental difference between dielectric and metallic liquids. The implications of this result for the question in the title are discussed.

• Received 30 December 2003

DOI:https://doi.org/10.1103/PhysRevB.69.245423