A theory of liquids, to be valid, must not disagree with any pertinent experimental fact such as: (1) The entropy of vaporization of all simple liquids to the same concentration of vapour is the same. It is larger for associated liquids, hence simple liquids are in a state of maximum disorder. They have no structure. (2) Simple liquids flow freely when expanded by only a few percent over their intrinsic volume. There is no room for “holes”, or “lattices”, or jumps longer than a small fraction of a molecular diameter. Self-diffusion occurs because thermal motion prevents any molecule from remaining in the same spot. (3) A parameter has been discovered and evaluated for molecules that are not hard spheres. It is the molal volume Vt at which soft molecules begin to be sufficiently separated between collisions to acquire fractions of their momentum in free space. At volumes smaller than Vt they are in fields of force not appropriately described by pair-potentials. The role of temperature is only to determine volume. Values of Vt depend upon the capacity of a molecular species to absorb collision momentum by bending, vibrating or rotating. The kinetic theory equation for the viscosity of dilute gases becomes applicable to polyatomic vapours when σ2 is replaced by Vt2/3.