“Early Wisconsin” is a time term introduced by T. C. Chamberlin more than a century ago. Right or wrong, it has been applied to almost 100 units of drift in North America. These embrace any unit too young to bear a true weathering gumbotil or deep pebble rotting (pre-Wisconsin), and yet too weathered and covered with moderately thick loess to be fresh “classical” Wisconsin (now Wisconsinan). Several such drifts, like the long-argued Iowan, have been denied a separate place in time by showing that they are younger soil on older eroded drift like “Kansan.” The revolution in dating generated by radiocarbon measurements of organic matter, from 1950 onward, eliminated a few other “early Wisconsin” cases, including those of Chamberlin. However, from 1950 to 1970 many >35,000 or >40,000 yr B.P. radiocarbon ages reinforced an early Wisconsin possibility. In Ohio we found six kinds of evidence for early Wisconsin glaciation, but only three proved to be good. The greatest confirmation during these two decades came from sea-floor coring; sea-surface temperatures and oxygen isotopes showed one cool long Wisconsinan stage 75 to 18 ka. By 1970 more precise radiocarbon dating confirmed some actual glacial sections on land, as well, and the records of the majority agreed there were climaxes in extent of ice at 70 ka and 20 ka.

The exposed and abandoned Teays Valley has been recognized and studied in south-central Ohio for a century and a half. By 1900, the upper reaches had been traced by bedrock strath up present deeper drainages through West Virginia into Virginia. By 1945, the 1.5-mi-wide main valley had been traced downstream by water/oil wells, under several glacial drifts, northward and westward from Chillicothe, Ohio, to Indiana. The average gradient in this reach is northwest 0.9 ft/mi. Horberg and others carried this valley westward to the Mahomet buried drainage of Illinois. Discontinuities such as in Madison County, Ohio, have been explored by geophysical profiles and test wells and show a marked narrowing into a canyon cut through resistant dolomites. As late as the 1970s, the largest tributary, which drains most of eastern Kentucky northward and on either side of Cincinnati, has been added. More and more basal floodplain sands/gravels and northward-inclined crossbeds are found in all these meandering valleys beyond the glacial limit. Near the limit these deposits are covered with lacustrine clay rhythmites, which grade to silty and then into sandy outwash containing glacially derived northern clasts. Clearly, glaciation dammed the valley system westward out of Ohio. This blockage was early Pleistocene, certainly pre-Illinoian, because clays are magnetically reversed. Other blockages took place southwestward from Cincinnati at a later date. The exposed Minford Clays in southcentral Ohio and western West Virginia must have filled to near the present 900-ft contour, because the former, dendritic Teays drainage is criss-crossed by an aimless, superimposed drainage that postdates the Deep Stage.

The Old Kentucky River system was a major contributor to the Teays River, draining southwestern Ohio and much of eastern Kentucky. The trunk river flowed northward from southeastern Kentucky throughout Frankfort and Carrollton, and then past Cincinnati and Dayton, joining the Teays River near Springfield, Ohio. North of the glacial boundary, which lies along the modern Ohio River, the course of the Old Kentucky River has been modified, and is today largely buried by drift. Although dissection is extensive to the south, there are many remnants of this entrenched and broadly meandering Teays-age valley system and of its sub-upland predecessors. These valleys contain areas of upward-fining, deeply weathered gravel, composed mainly of rounded quartz, chert, and silicified limestone pebbles derived from the headwaters of the system. Modern rivers have been entrenched 30 to 100 m below the Old Kentucky River valley and its main tributaries, the Old Licking and South Fork. The Old Kentucky River system was severed from the Teays when glaciation dammed its downstream reaches, forcing a reversal in flow direction between its junction with the Teays in west-central Ohio and Carrollton, Kentucky, and causing westward overflow into the Old Ohio River system. Piracy by the Old Ohio may also have contributed to the integration of the Old Kentucky and Old Ohio River basins. Ponded sediment is present in some of the now-abandoned valley remnants east of Cincinnati. As a result of glacial damming, the headwaters of the Teays River in southeastern Ohio and West Virginia overflowed westward across the Manchester divide into the Old Kentucky River drainage basin. All of these events led to establishment of the modern Ohio River system.

Location The Mount Washington-Crawford Notch area (Fig. 1) in the White Mountains of New Hampshire can be reached using U.S. 302 and New Hampshire 16. A network of hiking trailsprovides access to the off-road sites shown on Figures 2 and 3.

To the late Lewis M. Cline goes the credit for the initial impetus for this symposium volume. As Technical Chairman of the 1970 Annual Meeting of The Geological Society of America, he conferred with Black on possible topics for symposia at the Milwaukee meeting, and specifically questioned whether something on the Pleistocene would be timely. This led to the sanction by the Geomorphology Group of a symposium on the Pleistocene and to the appointment of the undersigned as a committee to carry out the wishes of the group. The committee selected the Wisconsinan Stage as a representative and amenable topic. Because of timing and the availability of contributors, it was not possible for the committee to obtain anywhere near the number of papers desired to provide representative coverage of the Wisconsinan Stage in the Midwest and other areas of the world. This volume includes all but one of the six papers actually presented at the symposium in Milwaukee. Some of the nine additional papers were presented in part at the regular sessions of the Annual Meeting, and others are more recent offerings. The committee recognizes both regional and topical gaps in this coverage of the Wisconsinan Stage. However, between papers on historical perspective for the Midwest and both polar regions, its scope is considered to be sufficiently broad to provide the reader with a reasonable concept of the development of thought on the Wisconsinan Stage, its present status, and to indicate some trends for the future. We wish to thank the individual contributors for their cooperation in the demanding and difficult task of bringing this volume to completion.

Outwash deposits in the Ohio Valley that were thought to be alluvium were illustrated by Volney in 1803. Drake (1815) proposed an iceberg origin for Wisconsinan erratics in Ohio. Hitchcock (1841a) in Massachusetts reviewed the glacial theory of Agassiz favorably, and in Ohio, St. John (1851) wrote in detail on the subject. Ice sheet origin of drift was widely and generally accepted by 1865. Multiple glacial advances, many of which later turned out to be Wisconsinan, had been noted by Lyell (1849), Whittlesey (1866), Worthen (1868), Orton (1870), Winchell (1873), and Newberry (1874). Chamberlin (1878) gave the first detailed description and analysis of the different ages of surface drifts in the Kettle Moraine region. After Chamberlin introduced the term Wisconsin (1894 Wisconsin (1895), Leverett (1899) divided the Wisconsinan into early, middle, and late, and Leighton (1931) assigned the Iowan to the Wisconsin and gave the names Tazewell, Cary, and Mankato to the early, middle, and late substages of Leverett. Rock- and time-stratigraphic terminology of the Lake Michigan Lobe was defined by Frye and Willman (1960), and the terms Altonian, Farmdalian, Woodfordian, and Twocreekan were introduced. Other geologists have recently used this classification to differentiate the drift of other lobes.

The age of the late Wisconsin maximum of the Laurentide ice sheet off the coast of New England and on Long Island, New York, is not closely designated. Radiocarbon and stratigraphic evidence from Martha’s Vineyard suggests that the glacier margin may have been close to its maximum position as late as 15,300 yrs ago; indirect evidence from Long Island infers that the ice sheet had reached a maximum and had begun to recede prior to 17,000 yrs ago. In any case, by at least 14,200 yrs ago the glacier margin had retreated from its maximum late Wisconsin position at the Ronkonkoma Moraine on Long Island, had constructed recessional frontal deposits, and had retreated north of Rogers Lake on the southern Connecticut coast. Subsequent readvances culminated near Middletown, Connecticut, some time after 15,000 yrs ago, and in Cambridge, Massachusetts, after 14,000 yrs ago. Whether these readvances were synchronous is unknown because of the absence of close limiting dates and because of the lack of evidence for readvance in the intervening area. The northwestward recession of the glacier margin from the present coast in eastern Maine was accompanied by a marine transgression and deposition of hundreds of submarine moraines between 13,500 and 12,500 yrs ago. This general recession was interrupted by a readvance which culminated at the Pineo Ridge Moraine approximately 12,700 yrs ago. Although it may have resulted from general climatic change, the Pineo Ridge readvance just as likely may have been caused by a vastly decreased calving rate associated with isostatic uplift and marine recession from coastal Maine. This is well documented as having occurred simultaneously with the Pineo Ridge readvance. Thereafter, the ice sheet thinned and separated over the highlands of northwestern Maine leaving residual ice to the southeast. Active ice, receding into the St. Lawrence Valley of southeastern Quebec deposited the Highland Front Moraine approximately 12,700 to 12,600 yrs ago. In summary, (1) a major amelioration of climate that began prior to 14,200 yrs ago resulted in very rapid dissipation of the ice sheet in New England at least by 12,500 yrs ago, with the exception of small glaciers that possibly persisted in the highlands; (2) no conclusive evidence has been recognized for any climatic reversals during the dissipation of the ice sheet in New England; and (3) although major events in New England compare with those of the Great Lakes region, no minor events have proven correlation with the possible exception of the Pineo Ridge and Port Huron readvances.

The Hudson-Champlain Valley is the only continuous lowland between the classic glacial areas of the Midwest and coastal New England, and presumably it contains the most complete Wisconsinan record east of the Erie-Ontario Lobe. A date of 26,800 yrs B.P. on intraglacial peat in New Jersey establishes a maximum age for the Woodfordian advance of the Hudson-Champlain Lobe. On western Long Island, deposition of the Ronkonkoma and Harbor Hill Moraines was followed by readvance and deposition of the Roslyn Till, and finally, by a stillstand on the north shore. Deglaciation from the Ronkonkoma Moraine began about 17,000 yrs B.P. In the Wallkill Valley, the southwestern physiographic continuation of the Hudson Valley, the terminal Woodfordian position is the Culvers Gap Moraine. Recessional positions are recorded at the Augusta, Sussex, Pellets Island, and Wallkill Moraines. The age of the Wallkill Moraine is established at 15,000 yrs B.P. The Woodfordian terminus of the Hudson-Champlain Lobe is traced northward from the Denville re-entrant in the Terminal Moraine in New Jersey, rather than westward, connecting the Ronkonko-ma and Culvers Gap Moraines.

The names of the Huron, Erie, and Ontario Lobes imply that glaciers followed these lake depressions, but the flow patterns of these lobes were complex and changed several times during the Wisconsin glaciation. The sublobes of the southwestern part of the so-called Erie Lobe were more often an extension of the ice coming down Huron Basin, and ice from both basins participated in these sublobes. A review of the studies of Wisconsin-age deposits in the area of the three lobes indicates the emphasis that has been placed on investigations of tills by multiple methods, on paleontological studies of the interstadial and late-glacial deposits, and on radiocarbon dating. A threefold time-stratigraphic division of the Wisconsin Stage in this area is based upon synchronous fluctuations by several glacial lobes, and upon climatologic inferences from paleontologic studies. Early and late Wisconsin experienced maximum glacial advances; middle Wisconsin was dominated by interstadial retreats. The first Wisconsin glacial advance reached into the St. Lawrence Lowland only, and was followed by a glacial retreat during the St. Pierre Interstade about 65,000 radiocarbon yrs B.P. The second major glacial advance (by several lobes) went farther, but did not reach as far south as the late Wisconsin glaciation in Indiana and Ohio. In Pennsylvania and New York it was more extensive than the late Wisconsin, if the Olean Drift is indeed of early Wisconsin age. The source of ice was centered in the eastern Canadian Laurentide area. Mid-Wisconsin glacier margins retreated several times far into the Huron and Ontario Basins, or even north of them. Three main retreats were probably interrupted by two readvances which reached into the Erie Basin but not south of the Lake Erie watershed. Mid-Wisconsin time began more than 50,000 yrs B.P. and ended about 23,000 yrs B.P. During late Wisconsin time the major glacial advance in the western part of the region investigated reached its farthest extent south in at least three pulses: 21,000, 19,500, and 18,000 yrs ago. The oscillating retreats were interrupted by three documented readvances: about 17,000, 15,000, and 13,000 yrs B.P. The source of ice was primarily in the western Laurentide center. No evidence has been found here for the Valders readvance which took place in the Lake Michigan Lobe.

The type section for the Erie Interstade shows a well-developed beach between two layers of offshore sands. The section is underlain by a major late Wisconsin till, the Catfish Creek Till, and is overlain by the Port Stanley Till. The position of the buried beach 3 to 4 m above present Lake Erie level indicates that the interstadial lake, here named Lake Leverett, was lower than previously estimated and that it drained eastward, probably via the Mohawk Lowland, during an ice recession into the Ontario Basin. The Erie Interstade correlates well with a world-wide amelioration of the climate at about 15,500 yrs B.P. that separates two glacial maxima of the late Wisconsin-Weichselian, the older 20,000 to 17,000 yrs B.P. and the younger 14,800 to 14,400 yrs B.P.

The climatic history of the Illinois region during Wisconsinan time is interpreted from the character and extent of the glacial deposits and buried soils of the Lake Michigan Lobe. The Illinoian-Sangamonian history of the lobe is briefly discussed. The history is complex; it includes three major glacial episodes during the Illinoian, and a succession of glacial advances and retreats during the Wisconsinan. The major interglacial stages were times of near climatic equilibrium, and the soil profiles indicate temperatures were higher than those of the present. The glacial stages, on the other hand, were characterized by sharply fluctuating climates, with episodes of glaciation alternating with minor intervals of stability, soil formation, and temperatures approaching those of the major interglacial stages.

The Superior Lobe of the Wisconsin glaciation was initially localized by the deep lowlands of the Lake Superior Basin, cut in relatively nonresistant late Precambrian red sandstone. It advanced southwest out of this lowland and crossed a low divide leading to the Minneapolis Lowland, which is underlain by Precambrian and Cambrian sandstones. The conspicuous drumlins of central Minnesota, and the rugged St. Croix Moraine that borders the drumlins on the west and loops across the Minneapolis Lowland, delimit the major stillstand of the Superior Lobe. Discovery of red drift with diagnostic rock types from the Lake Superior Basin (agate, amygdaloidal basalt, red and purple felsite, red sandstone) in southwestern Minnesota indicates that the Superior Lobe once extended farther southwest down the Minneapolis Lowland to the Minnesota River valley and beyond during a pre-Wisconsin or early Wisconsin glaciation. As the Superior Lobe wasted from the St. Croix Moraine, a series of sharp subparallel tunnel valleys were cut into the drumlin plain and even into the underlying bedrock by subglacial streams driven to high velocity by the hydrostatic pressure resulting from the load of many hundreds of meters of active ice. Subsequent thinning and stagnation of the Superior Lobe opened the tunnel valleys to atmospheric pressure and converted the subglacial streams from major erosional streams to small depositional streams, which formed discontinuous eskers along many of the tunnel valleys. After distant retreat the Superior Lobe readvanced three times out of its basin, twice after proglacial lakes had produced a supply of red clay to be overridden. These latter two readvances may represent surges of the ice lobe resulting from the buildup of basal meltwater behind the frozen toe of the ice lobe. The Des Moines Lobe, originating in the Red River Valley of Manitoba and western Minnesota, moved southeastward down the Minnesota River valley and thence northeastward up the Minneapolis Lowland, overriding a segment of the St. Croix Moraine and extending across the state to Wisconsin in the form of the Grantsburg Sublobe. Its drift is characteristically gray to yellowish-brown and highly calcareous. The ice incorporated masses of Superior Lobe drift as it overrode the St. Croix Moraine, stringing them out to produce a complex of foliated red and gray drift. The Grantsburg Sublobe blocked the Mississippi River and other drainage from central Minnesota to form glacial Lake Grantsburg about 16,000 yrs ago. By that time the Superior Lobe had withdrawn completely from central Minnesota, for it supplied meltwater (and red clay) only on the east, down the St. Croix River and its upper tributaries. Meanwhile, the main Des Moines Lobe, which thus far supplied ice only to the Grantsburg Sublobe, thickened sufficiently to spill southward out of the Minnesota Valley across a low divide into Iowa. This produced the lobe that reached Des Moines 14,000 yrs ago. Beheading of the Grantsburg Sublobe in this manner caused the stagnation of the latter, which then wasted to form the Anoka Sandplain in its stead. The St. Louis Sublobe protruded from the Des Moines Lobe in northwestern Minnesota at a later date (about 12,000 yrs ago). Its meltwater flowed down the St. Louis River toward Lake Superior, but it was diverted southward into the St. Croix drainage by the still-existing Superior Lobe. Final wastage of the entire Des Moines Lobe produced glacial Lake Agassiz in northwestern Minnesota and adjacent North Dakota and Manitoba.

Lennoxville Till of the Lac-Mégantic region is homogeneous with respect to fabric, texture, clay-mineral composition, and color. The homogeneity resulted from erosion and transportation by a glacier that moved southeastward over bedrock that had little petrologic variation. Lennox-ville Till is, however, also characterized by southeast-trending bands with high concentrations of certain rocks, minerals, and chemical components derived from scattered igneous bodies. Chromium, nickel, and magnetite are dispersed in ribbonlike bands at least 50 km southeast of their principal source areas in the ultrabasic-basic rock complex at Thetford Mines. Concentrations of surface erratics derived from these same sources have dispersal patterns similar to those of Cr, Ni, and magnetite. Plagioclase grains and granodioritic erratics are dispersed in bands southeast of the granodiorite stock of the Little Megantic Mountains. Granodioritic erratics that occur on the surface are thought to have been let down onto lodgment facies of Lennoxville Till during melting of the Lennoxville glacier, because they mantle the surface of its grano-diorite-poor clay-till facies. By extension, most of the boulder mantle that characteristically rests on Lennoxville lodgment facies may be an ablation deposit with dispersal characteristics locally reflecting glacier deflections and lobations that occurred during deglaciation. Dispersal data strongly confirm ice-flow patterns inferred independently from striation, fabric, and geomorphic data. Most till components discussed show evidence of long-distance transport, except where topographic prominences have blocked or deflected ice-transported sediment. No evidence was found to support the concept of glacier flow into Quebec from late-glacial highland centers of outflow located south or east of the area. Texture of till has an important influence on observed concentrations of trace elements. Zr is concentrated preferentially in the silt fraction of silt + clay; Ti, Cu, V, Zn, Pb, Cr, and Ni are concentrated preferentially in the clay fraction. Concentrations of Cr, Ni, and Ti are thought to be texture-independent in samples located down-ice from igneous sources rich in these components.

Geochemical investigations were performed on 109 samples of late Wisconsin glacial till from southern Ontario. The –0.037 mm size fraction was analyzed for Cu, Zn, and Cr by atomic absorption spectrophotometry and for Ni by colorimetry. Means and standard deviations (in ppm) for 109 samples (group 1) are: Cu, χ̄ = 23, σ = 6; Zn, χ̄ = 62, σ= 14; Cr, χ̄ = 51, σ = 15; and Ni, χ̄ = 28, σ= 8. The statistics were also computed for the following groups of samples: (2) oxidized till (67 samples); (3) unoxidized till (42 samples); (4) Erie Lobe till (22 samples); (5) Ontario Lobe till (13 samples); (6) Huron Lobe till (24 samples); and (7) Georgian Bay Lobe till (11 samples). Cr and Ni are slightly higher in group 3 than in group 2. Results of t and F tests performed on groups 4, 5, 6, and 7 suggest that the samples are derived from two populations; one characterized by the Erie Lobe samples (population E) and the other by samples from the Georgian Bay and Huron Lobes (population N). Population E has a higher content of all four elements than population N. Results of a discriminant analysis suggest that the Ontario Lobe samples are derived from both populations. Three samples from the northwestern end of the Ontario Lobe show definite affinities with population N, whereas the others belong to population E. This difference suggests that the glacier which deposited the former till may have incorporated drift material from a northern source. This preliminary investigation suggests that the application of multivariate statistical techniques to geochemical analysis of selected fractions of glacial deposits could provide another basis for differentiation of glacial tills.

More than 500 circular to elliptical mounds occur in the late Pleistocene (Woodfordian) deposits of north-central Illinois. The mounds rise from 1 to 5 m above the general ground level and are either flat-topped or have slightly depressed centers. They range in diameter from 30 m to approximately 1 km; the smaller mounds are most abundant. The mounds consist of a core of lacustrine silt and clay surrounded by a sandy rim. The lacustrine sediments overlie Woodfordian till and outwash, and are in turn overlain by Woodfordian loess. The morphologic and stratigraphic characteristics of the mounds suggest that they are deposits formed within the lakes of pingo craters. These characteristics include (1) a high degree of symmetry, (2) overlapping and superpositional relationships, (3) occurrence over both till and outwash, (4) surrounding annular depression, (5) confinement to a low-relief inter-morainic area, (6) mineralogical identity with the underlying materials, and (7) the association of the mounds with indicators of permafrost. The pingo lake hypothesis is also consistent with implied groundwater conditions and ice margin locations during the Woodfordian.

Wide stream-cut trenches, now filled with lakes, swamps, and underfit streams, transect the drumlin plain of the Superior Lobe in a subparallel pattern trending southwest, oblique to the modern regional slope and drainage. They are pictured as the products of high-velocity streams in subglacial tunnels, driven by the great hydrostatic pressure resulting from the thick mass of still-active ice. The water for such flow cannot have come from the glacier surface, because the cold upper part of the ice could not permit its penetration. It must have come rather from the base, through melting by the geothermal flux or by the frictional heat of basal ice flow. The collecting basin for such water may have extended to the center of the ice sheet in the Hudson Bay area, and the water may have been stored for thousands of years until it could break through the frozen toe of the Superior Lobe and "catastrophically" cut the tunnel valleys during its exit. As the ice thinned to stagnation and the hydrostatic head was lost, the subglacial streams changed their habit from erosional to depositional, forming small eskers along the trenches. After extensive wastage of the Superior Lobe at the end of the St. Croix phase, the ice readvanced to or slightly beyond the rim of the Lake Superior Basin at least three times. The later readvances, which involved the overriding of proglacial lake beds and the deposition of red clayey till, were not closely synchronized with advances of adjacent ice lobes, and the pollen sequence for northeastern Minnesota shows no pattern of climatic reversals that matches ice-lobe fluctuations. Accordingly, the hypothesis is presented that nonclimatic factors may have controlled the most recent minor advances of the Superior Lobe. Subglacial meltwater may have built up beneath the then-restricted Superior Lobe and behind the dam afforded by the cold ice of the thin toe until the reduction in basal friction permitted rapid glacial flow (a surge) through the dam.

Late Pleistocene-Holocene climatic fluctuations and environmental conditions in the Gulf of Mexico are reflected by changes in the vertical distribution of planktonic foraminifers in the bottom sediments. The events are related closely by radiocarbon dates to continental Wisconsinan glacial-interglacial substages. The almost inverse relationship between abundances of two species, Globorotalia menardii (warm) and Globorotalia inflata (cold), allows recognition of three major episodes of climatic cooling during the Wisconsinan. Moreover, minor fluctuations of climate are reflected in detail by the combined distributional patterns of warm- versus cold-water planktonic species. Paleotemperature curves quantitatively derived from the frequency ratio of warm- versus cold-water species from the Gulf of Mexico are strikingly similar to the oxygen-isotope curve of Emiliani (1966) from the Caribbean for about the last 75,000 yrs. Morphologic changes in the Globorotalia menardii group and the withdrawal of cold-water species, such as Globorotalia inflata, from the Gulf of Mexico characterize two climatically distinct assemblages. The older assemblage corresponds to the late Pleistocene and is characterized by Globorotalia menardii flexuosa (warm) and by Globorotalia inflata (cold). Withdrawal of Globorotalia inflata occurred between 4,000 and 11,000 yrs ago and corresponds closely to an incursion of abundant Globorotalia tumida, Globorotalia ungulata, and other warm-water species. This faunal boundary is interpreted to represent the transition from the last glacial to postglacial conditions in the Gulf of Mexico. Assuming a constant average rate of deposition for pelagic sediment, the age estimated for the boundary agrees very closely with that of the radiocarbon bracketed date of 7,000 yrs B.P. of Frye and others (1968) for the Wisconsinan-Holocene boundary. Paleontologie events and paleotemperature curves from the Gulf of Mexico correlate almost exactly with those from the Caribbean and adjacent Atlantic. Widely differing opinions expressed by several authors on Wisconsinan nomenclature in the marine section are based on correlations using different paleontological criteria and different geochemical dating methods. Carbon-14 determinations reported in this study indicate ages considerably younger than other published dates for what certainly appears to be a paleontologically equivalent unit. Climatic events recognized in the Gulf of Mexico can be correlated rather precisely on the basis of radiocarbon dates with the continental Wisconsinan glacial-interglacial substages and, therefore, development of an independent nomenclature for the marine section is neither necessary nor desirable.

The oxygen-isotope ratio in polar snow is determined mainly by the temperature of formation of the precipitating clouds. A continuous core 1,390 m long through the ice sheet at Camp Century, Greenland, reveals a climatic record, inferred from those ratios, spanning possibly the last 100,000 yrs. The depth-age relationship of the core is calculated from present ice-flow patterns and simple assumptions; the paleoclimatic data are interpreted from the analysis of oxygen-isotope-ratio measurements on nearly 7,000 individual samples cut from the core. The ice-core record reveals that the Wisconsin Stage started 73,000 yrs B.P. Many perturbations of the oxygen-isotope ratios are observed within the Wisconsin Stage that agree with climatic oscillations dated by radioactive methods. An 11 ‰ shift in the 0 isotope data shows that the Wisconsin Stage ended very rapidly, within a 2,500 yr interval, at about 13,000 yrs B.P. Spectral analyses of the data show oscillations with periods of 78, 181, 400, and 2,400 yrs.

Oxygen- and hydrogen-isotope analyses of ice cores from a hole drilled 2,164 m through the Antarctic Ice Sheet suggest that the Wisconsin cold interval began about 75,000 yrs B.P., reached its climax about 17,000 yrs B.P., and terminated about 11,000 yrs B.P.