Recent geophysical studies of glacial-isostatic deformation overlook the same fact that was overlooked in earlier studies: the loads of ice that were applied to continents during glacial ages were not loads added to the Earth's crust, but were loads transferred from the oceanic 70 percent to the glaciated 5 percent of the crust by the hydrologic cycle. A realistic model of glacial isostasy must be represented by a balance, in which glaciated areas totaling about 5 percent of the Earth's surface have loads of 140 to 170 bars added or removed on a time scale of 104 years, while synchronously the oceanic 70 percent of the Earth's surface has a load of 10 to 12 bars or more removed or added on a similar time scale. The subcrustal mass transfer involved in such a balance is not well represented by harmonic equations, for the water and ice loads are not symmetrically disposed on the Earth's surface.
The suitability of the proposed balance model depends on whether the ocean floor will respond isostatically to a load of as little as 10 bars. Evidence from Lake Mead, Arizona, and Lake Bonneville, Utah, suggest that the continental crust, at least, does deform under a regional load of 10 bars or less.
Pleistocene marine shorelines offer a means of testing the balance model of isostasy. If the ocean floor deforms under water loads, the amount of postglacial submergence of a coast should be in part a function of the regional proximity of deep ocean water. Coasts with nearby ocean water more than 100 m deep had the load of water from the postglacial rise of sea level added early and close; coasts bordering shallow seas had the load added late and generally far offshore. An averaging technique to show the regional water load on the Atlantic coast of the northeastern United States provides a basis for comparing the submergence histories of several localities with the average water depth offshore. In general, the amount of submergence is proportional to the proximity of deep water.
Oceanic islands should record different Pleistocene shoreline levels than continental coasts. Local, detailed, late Pleistocene histories of a variety of coasts will provide a test of isostasy better than the familiar test of postglacial uplift.