Can osmotic pressure be negative?

We have established cases (Murad, S., Powles, J. G. and Holtz, B., 1995, Molec. Phys., 86, 1437) in which the osmotic pressure, as conventionally defined, is negative even though the conventional criteria for osmotic equilibrium are satisfied, admittedly, mainly but not exclusively, for some rather extreme non-ideal mixtures and for computer 'experiments' and theory only. The present closer investigation of these particular cases shows that although these mixtures are stable in isolation they split up when in osmotic contact with the pure solvent into two mixture phases such that the total osmotic system has a lower Helmholtz free energy. In the examples here one of these mixtures is a very dilute solution so the osmotic pressure, which previously was negative, is very small indeed but is positive. The other mixture is concentrated. The possibility of negative osmotic pressure, if only in metastable systems, is of both scientific and technological interest. The osmotic three-phase system extends the interest in osmosis to more complex and possibly more useful situations. It also demands new theories and simulation techniques some of which are provided.