Numerical solutions are examined for isolated, intense vortices as influenced by western bounding bottom tgpography through the use of a rigid-lid, two-layer primitive ~-plane numerical model. Systematic studies are made of the sense of rotation (cyclonic/anticyclonic), the consequence of varying the gradient of bottom slope, and the different vertical shear in a two layer ocean. In the basin with a bottom slope, the nearly barotropic anticyclonic vortex forms a modon-like vortex for ~<< ~ with fixed Ro2 < O(1) (where ~ is the ratio between the variation of the Coriolis parameter across the eddy to the Coriolis parameter in the center, 8S the topographic effect and, Roz the Rossby number in the lower layer) and its generation is due to a compound effect of the planetary beta, topographic beta, advection, and mirror image. The formation of the modon-like vortex and the propagation of the original vortex onto the bottom slope depends on the strength of slope gradient and the baroclinicity of the vortex. The nearly barotropic anticyclonic vortex evolves into the stronger upper ocean one with increasing ~: the gradient of the bottom slope becomes steeper. Then the original vortex lives longer because the barotropic component of the energy is converted to the baroclinic one and it moves toward southeast in forming a modon-like vortex in the lower layer. The evolution of a vortex in the model results are compared to observational results of a Kuroshio warm core ring (KWCR) obtained from hydrographic data (June, 1985) and from NOAA satellite infrared images (April, 1985 to July, 1985. It is shown that a KWCR (June, 1985) is influenced by the western continental slope/shelf of the East Japan.