The succession of coral reefs in the Corallian Beds (Upper Jurassic) of southern England is briefly described, and they are shown to contain essentially the same molluscan fauna irrespective of their stratigraphical position. On the other hand all the reefs of one age pass laterally into shell-beds which are distinctive for the strata of that age and differ in composition from the shell-beds above and below, equivalent to coral reefs of later and earlier age. The same is found to hold for certain living and raised Pleistocene reefs and their contemporaneous shell-beds examined by the author near Qosseir on the Red Sea. The shell-beds are the normal shelly stratal sequence, and the differences in their constituent spp. might be attributed to evolution, were it not for the control supplied by the abnormal coral reef facies. The mollusks inhabiting the coral reefs, and thereby fixed geographically, show no corresponding evolutionary modifications. Hence the faunal changes in the vertical sequence of normal shell-bed strata are explained as due almost entirely to migration. If this conclusion is right, much current geological thought is without sound basis.