A considerable portion of Northern Eurasia, and particularly its continental shelf, was glaciated by inland ice during late Weichsel time. This was first inferred from such evidence as glacial striae, submarine troughs, sea-bed diamictons, boulder trains on adjacent land, and patterns of glacioisostatic crustal movements. Subsequently, the inference was confirmed by data on the occurrence and geographic position of late Weichselian end moraines and proglacial lacustrine deposits.

The south-facing outer moraines in the northeastern Russian Plain, northern West Siberia, and on Taimyr Peninsula are underlain by sediments containing wood and peat, the radiocarbon dating of which yielded ages of 22,000 to 45,000 yr B.P. The youngest late-glacial moraines are of Holocene age: the double Markhida moraine in the lower Pechora River basin, presumably associated with “degradational” surges of the Barents Ice Dome, is underlain by sediments with wood and peat dated at 9000 to 9900 yr B.P.: this suggests that deglaciation of the Arctic continental shelf of Eurasia was not completed until after 9000 yr B.P.

The reconstructed ice-front lines lead to the conclusion that the late Weichselian ice sheet of Northern Eurasia (proposed name: the Eurasian Ice Sheet) extended without interruptions from southwestern Ireland to the northeastern end of Taimyr Peninsula, a distance of 6000 km: it covered an area of 8,370,000 km2, half of which lay on the present-day continental shelves and a quarter on lowlands that were depressed isostatically below sea level. Hence, the ice sheet was predominantly marine-based.

A contour map of the ice sheet based both on the dependence of the heights of ice domes upon their radii and on factual data concerning the impact of bedrock topography upon ice relief has been constructed. The major features of the ice sheet were the British, Scandinavian, Barents, and Kara Ice Domes that had altitudes of 1.9 to 3.3 km and were separated from one another by ice saddles about 1.5 km high. At the late Weichselian glacial maximum, all the main ice-dispersion centers were on continental shelves and coastal lowlands, whereas mountain centers, such as the Polar Urals and Byrranga Range, played only a local role.

The portions of the ice sheet that were grounded on continental shelves some 700 to 900 m below sea level were inherently unstable and could exist only in conjunction with confined and pinned floating ice shelves that covered the Arctic Ocean and the Greenland and Norwegian Seas.

The Eurasian Ice Sheet impounded the Severnaya Dvina, Mezen, Pechora, Ob, Irtysh, and Yneisei Rivers, and caused the formation of ice-dammed lakes on the northern Russian Plain and in West Siberia. Until about 13,500 yr B.P. the proglacial system of lakes and spillways had a radial pattern; it included large West Siberian lakes, the Caspian and Black Seas, and ended in the Mediterranian Sea. Later, the system became marginal and discharged proglacial water mainly into the Norwegian Sea.

The oxygen isotopic stage 5/4 boundary in deep-sea sediments marks a prominent interval of northern hemisphere ice-sheet growth that lasted about 10,000 yr. During much of this rapid ice growth, the North Atlantic Ocean from at least 40°N to 60°N maintained warm sea-surface temperatures, within 1° to 2°C of today's subpolar ocean. This oceanic warmth provided a local source of moisture for ice-sheet accretion on the adjacent continents. The unusually strong thermal gradient off the east coast of North America (an “interglacial” ocean alongside a “glacial” land mass) also should have directed low-pressure storms from warm southern latitudes north-ward toward the Laurentide Ice Sheet. In addition, minimal calving of ice into the North Atlantic occurred during most of the stage 5/4 transition, indicative of ice retention within the continents. Diminished summer and autumn insolation, a warm subpolar ocean, and minimal calving of ice are conducive to rapid and extensive episodes of northern hemisphere ice-sheet growth.

Analysis of some sedimentary series near Rome and comparison with other series in northern and southern Italy give evidence that Arctica islandica, considered a marker species for the beginning of the Pleistocene, appears during the Réunion paleomagnetic event (about 2 my ago). A period of erosion, called the Aullan erosional phase, is evident in the lower part of the late Villafranchian (= early Pleistocene, Olivola faunal unit).

Precipitation and river-flow data from peninsular Florida show a trend toward increased aridity beginning about 1960. The trend follows by 10–20 yr the onset of Northern Hemispheric cooling and indicates a possible present-day link between this cooling and the recent decrease in precipitation in peninsular Florida. This phenomenon tends to support the hypothesis of a return to cooler and drier conditions, perhaps, similar to those prevailing during the Little Ice Age.

Among numerous stream-valley terrace deposits of the Gulf and Atlantic coastal plains that occupy chronologically and spatially intermediate positions between the youngest coastwise (coast-parallel) Pleistocene surface and the present, active floodplain, those of the Tunica Hills seem to provide by far the best opportunity for radiometric dating. Earlier views on the age of the Little Bayou Sara creek alluvium, represented by a single, 8 to 10-m-thick unit, ranged between the last interglaciation and middle Holocene. Reexamination of these deposits in the Little Bayou Sara and adjacent valleys clearly suggests their Late Pleistocene (apparently Farmdalian Interstade) age. The majority of the 14 available dates from the Little Bayou Sara and Tunica Bayou valleys proved to be too young, due to postdepositinal contamination. Dates ranging between 33,720 and 25,965 yr B.P. came from samples thought to be uncontaminated. Plant and faunal elements with boreal affinities in the unique fossil assemblage appear to be relicts of a preceding, full-glacial period, as regarded by Brown (1938). The absence of colder climate taxa from the Wilcox Bluff flora on Bayou Sara is insufficient evidence for a suggested Sangamon Interglacial age of the flora, and the terrace stratigraphy holds no proof for that view either. Only a single, loess-mantled, constructional, Quaternary, valley-terrace surface is present in the area. A narrow, low, actively developing floodplain terrace along Little Bayou Sara, cut into the Pleistocene alluvial unit, is primarily erosional in origin and has no bearing on the age of that unit. The age of the Tunica Hills terrace unit may provide comparison for dating intermediate valley-terrace deposits in favorable coastal settings elsewhere.

The Winter Gulf site near North Collins, New York is a 0.8-m sequence of organic detritus dated at approximately 12,700 yr B.P. A 260-kg sample produced over 500 identified individuals representing five orders of insects. Three insect zones were recognized, the lowest representing an open mire with sparse numbers of carices and other sedges; aquatic beetles were rare, and no trees were in the vicinity of the sample site. The middle zone indicated an extensive marsh or moist meadow environment with spruce nearby. The upper zone represented a mixed riparian and spruce forest insect assemblage.

The results of the Winter Gulf insect analyses support published hypotheses that the site was probably a protected embayment of early lake Warren. Although the flora at the site has been previously interpreted as representing parkland tundra, temperature analyses of the fauna indicate that the thermal conditions were much warmer than parkland tundra regions of North America today. These results have important significance for the region immediately south of the Laurentide Ice Sheet during this period of the Late Wisconsinan.

Nonconnah Creek, located in the loess-mantled Blufflands along the eastern wall of the Lower Mississippi Alluvial Valley in Tennessee displays a sedimentary sequence representing the Altonian Substage through the Woodfordian Substage of the Wisconsinan Stage. The site has a biostratigraphic record for the Altonian and Farmdalian Substages that documents warm-temperate upland oak-pine forest, prairie, and bottomland forest. At 23,000 yr B.P., white spruce and larch migrated into the Nonconnah Creek watershed and along braided-stream surfaces in the Mississippi Valley as far as southeastern Louisiana. The pollen and plant-macrofossil record from Nonconnah Creek provides the first documentation of a full-glacial locality in eastern North America for beech, yellow poplar, oak, history, black walnut, and other mesic deciduous forest taxa. During the full and late glacial, the Mississippi Valley was a barrier to the migration of pine species, while the adjacent Blufflands provided a refuge for mesic deciduous forest taxa. Regional climatic amelioration, beginning about 16,500 yr B.P., is reflected by increases in pollen percentages of cooltemperate deciduous trees at Nonconnah Creek. The demise of spruce and jack pine occurred 12,500 yr B.P. between 34° and 37° N in eastern North America in response to postglacial warming.

Fossil diatom assemblages from a 12-m core from Kirchner Marsh were compared with modern surface assemblages from 159 Minnesota and Labrador lakes using cluster analysis. The deepest levels of the core (spruce pollen zone 13,000 to 10,200 yr B.P.) resemble modern diatom assemblages from deep oligotrophic lakes of northeastern Minnesota. Diatom assemblages of the pine pollen zone (about 10,200 to 9500 yr B.P.) have few modern analogs. In the oak zone (9500 yr B.P. to present) after a brief pulse of diatom species indicative of eutrophication, the assemblages are dominated by species characteristic of shallow lakes, suggesting a drop in the lake water level during the prairie period (5500 to 7500 yr B.P.). Macrofossil data of W. A. Watts and T. C. Winter (1966, Geological Society of America Bulletin 77, 1339–1360) show that this shift to shallow-water diatoms occurred when aquatic macrophytes appeared at the site in abundance.