In Part I we focussed on a convergent representation of the gravitational potential generated by topographic masses on top of the equipotential surface at Mean Sea Level, the geoid, and by those masses which compensate topography. Topographic masses have also been condensated, namely represented by a ''single layer'', Part II extends the computation of the gravitational field of topographic-isostatic masses by a detailed analysis of its force field in terms of vector-spherical harmonic functions. In addition, the discontinuous mass-condensated topographic gravitational force vector (''head force'') is given, Once we identify the Moho discontinuity as one interface of isostatically compensated topographical masses, we have computed the topographic potential and the gravitational potential which is generated by isostatically compensated masses at Mean Sea Level, the geoid, and illustrated by various figures of geoidal undulations. In comparison to a data oriented global geoid computation of J. Engels (1991) the conclusion can be made that the assumption of a constant crustal mass density, the basic condition for isostatic modeling, does not apply. Instead density variations in the crust, e.g. between oceanic and continental crust densities, have to be introduced in order to match the global ''real'' geoid and its topographic-isostatic model. The performed analysis documents that the standard isostatic models based upon a constant crustal density are unreal.