Present theories of colloid science do not account for specific ion effects as exemplified by the Hofmeister effect, surface tension of electrolyte solution interfaces, binding to micelles, effective charge in double layer interactions, and attractive interactions in low Hamaker constant systems. It is argued that specificity emerges naturally and can be rationalized if dispersion interactions acting on ions are included in the theory. These are in principle accessible from bulk solution properties. Specific ion adsorption due to dispersion interactions can be dominant even at charged interfaces especially at high salt concentrations (similar to 0.1M). The effects can be qualitatively different at air-water and oil-water surfaces. That part of extended Lifshitz theory for low Hamaker constant systems, in which the forces are mainly due to temperature and salt dependent interaction, is re-examined. It is shown to be at the same level of approximation as, and precisely equivalent to, the Onsager limiting law for the interfacial tension change due to dissolved salt at a single interface, i.e., to linearization of the Poisson-Boltzmann distribution, and restriction to electrostatic potentials as the sole determinant of adsorption excesses. The usual description of interactions into separate electrostatic, and dispersion interaction, is invalid, even at the level of continuum (primitive model) theories.