Two basic approaches to the solution of surface problems exist in both theoretical and experimental research on metals. In the first approach, the bulk properties of metals and changes caused by surface processes are studied and the interpretation is based on the nearly-free-electron model of the metal. This approach was successful in the explanation of bulk properties. However, because of the complexity and asymmetry of the system (crystal and gas molecule), one is usually forced to use oversimplified models, which sometimes fail in the elucidation of all the observed effects. Another approach is based on the assumption that the surface interaction (resulting in the formation of a quasichemical compound) concerns a limited number of participants (as in a chemical reaction). In the simplest case: gas particle and a surface atom. New experimental evidence (field-ion-emission microscopy, field-electron-emission microscopy, ion-neutralization spectroscopy) has been found supporting the idea that the properties of a surface atom in the moment of the interaction can be roughly approximated by the properties of an individual atom. The type of approach which should be used in a particular case depends both on the experimental method and on the problems in which we are interested.