By the recent models of change in sea level , glacio-isostasy and hydro-isostasy have been consideyed. This paper presents the concept of absolute isostasy and relative isostasy. The first represented thefinal result of isostasy,and the last represented the mid process of isostasy.
    Absolute isostasy is the relationship between the crust of the glacier and the crust of the ocean. Thesimple model was presented by A. L. Bloom(1967)，W. E. Farrell and J. A. Clark (1976 ) present that the gravitational attraction of an ice mass upon a nearby ocean tends to hold sea level high in the vicinity of the ice. This extra load near the ice may have a significant influence on postglacial isostatic adjustment. The result of my research is that the gravitational attraction of an ice mass and the extra load ofwater near the ice tends to make the core move. All of these can change the geoid and the sea level. It may be a bridge between the Clarks s isostasy model and Morner"s gravitation model.
    Suppose we extract a mass M,representing an ice sheet,from the ocean (it caused a uniform fall 130m in sea level),and locate M as a spherical cap of radius B (woodward，1888). The mass M of the spherical cap is different in graYitational attraction from the mass M of a point on the Earth"s surface.In the first case,the mass M can make the core move 1. 2km and it caused the Maximal differentia 3. 89m of the change in sea level.
    For the relative isostasy,the crust of continent near the ocean will join the glacio-isostatic adjustment. The pinnacle islands is not a dipsticks,because the change in sea level upon melting of the ice and filling of the ocean would not be uniform and the isostatic adjustment in the whole earth can not take place at the same time.