Extinctions are continuous throughout the Phanerozoic. Although a single natural phenomenon brought about by environmental change at any scale, they may be conveniently divided into two classes: background extinction and mass extinction (Raup and Sepkoski, 1982). The Frasnian-Famennian extinction is an example of mass extinction, and a large body oceanic impact has been suggested as a triggering mechanism (McLaren, 1970). Organisms most affected were shallow water benthos, especially corals, stromatoporoids, and brachiopods living on tropical to subtropical continental areas extending from western Canada eastward around the world to western Australia. Plotting taxa on charts does not necessarily give the scale of the event which involved the virtual total destruction of the biomass. Famennian faunas are very different and corals remain rare in numbers until well into the Carboniferous. The horizon of the event appears to lie most probably within the Uppermost P. gigas or Lower Pa. triangularis conodont subzone, but it might be as early as Upper gigas or as late as the base of the Middle triangularis subzones. Its duration could be less than a single subzone—0.5 to 1.0 m.y. or less. In discussing immediate causes all are agreed that muddy, cold, or deoxygenated water would be fatal to the type of organism that disappeared. Suggested mechanisms fall into four categories: 1) sea level changes, either regression or transgression; 2) climatic change including cold water and atmospheric variation; 3) accident, by common occurrence of more than one unrelated mechanism, or an apparent extinction caused by plotting taxa against time interval; 4) astronomical event, a variety of possibilities from impact or near miss by meteorite or comet and others. No one mechanism can be identified as being the most probable, and an oceanic impact must remain a serious contender.

A subsurface investigation of the coastal areas in the Apalachicola delta region on the northwest Florida coast indicates that the Pleistocene sediments, which thicken to the southwest, have been deposited on an uneven Miocene surface of variable age. The thickest Pleistocene section, west of the mouth of the Apalachicola River, consists of two superimposed sequences of terrigenous clastic sediments. Each sequence grades upward from coarse to fine, and represents two major Late Pleistocene sea-level fluctuations. Radiocarbon dates and stratigraphic relationships indicate that the upper of these sequences probably represents a mid-Wisconsin transgression of the sea, and the lower represents a Sangamon transgression. The thickest section of Recent sediments lies in the old incised valley of the Apalachicola River which was cut during the last lowering of sea level. These deltaic, prodeltaic, and bay sediments represent the gradual filling of the drowned river valley, which was inundated by the Recent rise in sea level. The bays, barrier islands, and spits are the only other areas of any significant Recent sedimentation with the exception of one offshore basin to the southwest of the present river mouth. Radiocarbon dates, stratigraphic relationships, and environmental interpretations suggest that there was a relatively high stand of the sea that was very near present sea level sometime between 24,000 and 40,000 years B.P. Geomorphologic evidence in the area indicates that this high stand may have been slightly higher than present sea level and probably corresponds to the Silver Bluff shoreline of Florida and Georgia. No evidence has been found for a Recent higher sea-level stand, and radiocarbon dates and sediments indicate that sea level was approximately 10 to 15 feet below its present level sometime between 4000 and 4500 years ago. A rather rapid rise probably occurred after this time to a position slightly below present sea level, from which it has gradually risen to its present position.

Sea level is used as datum in geodesy and geology. Contemporary sea level is changed locally by climatic influences and universally especially by variation in glacier volume. These eustatic changes should now be ascertainable with sufficient accuracy by averaging world mareographs. At present a rise is going on. This should be allowed for in the interpretation of data from each individual station. There then remain climatic influences, to be evaluated meteorologically; local settling and crustal movements, to be distinguished by precise levelling. Sea level as datum for the geological past is involved in many problems. Nick just below sea level in the hypsographic curve: this is evidently due to the action of external processes. Cyclothems: it is not yet known whether these are of eustatic nature. Guyots: the flat tops are probably sinking 20 meters per million years. Glacial low levels: roughly 100 meters minus. The river bed on the Sunda shelf has not been warped since the peak of the last ice age. Raised beaches: reef terraces prove the recent rise of island arcs facing deep-sea trenches and the intermittent nature of the movement, each jump adding about 1 milligal or less to the local value of gravity. There is a deplorable lack of data on tilting, warping, and dating of these terraces, which could teach so much of crustal movements. Daly terraces: slightly raised young terraces are of common occurrence, and Daly ascribed them to eustatism. The amount and dating of the supposed movement is still debated. In some areas, however, there is a singular absence of evidence on a recent sinking of sea level. Isostatic recoil. The post-Glacial updoming of the Canadian and Fennoscandian shields is recorded in countless raised beaches. Field work appears to point toward the establishment of hinge lines, with updoming on the inside and permanent stability on the outside. The hinge line migrated from south to north in several jumps. Inter-Glacial terraces have been widely studied outside glaciated areas, but not all can agree to their eustatic nature. The origin of the older terraces that lie far above sea level for an ice-free world is a vexing problem.