Harold N. Fisk: a luminescent man

The publication of the “Geological Investigation of the Alluvial Valley of the Lower Mississippi River” (Fisk, 1944) was the greatest achievement of Harold N. Fisk's remarkable career. Its creative rigor and vision marked a great leap forward in understanding the alluvial and sedimentological processes of the Mississippi Valley and the fundamental value of these insights to river engineering strategies and techniques. Fisk's monumental effort shares many common elements with great strides in other disciplines made during this era. The period from the 1920s to the 1940s was marked by breakthroughs in fields such as hydraulics, medicine, chemistry, and small-particle physics that featured grand vistas of discovery. Fisk's triumph, as in other cases, can be attributed to factors that supplemented and complemented his personal scientific brilliance. He displayed less tangible elements of personality, shrewdness, and leadership that enabled him to attract interest and support for his undertakings, build a highly devoted team of associates, maintain the confidence of sponsors, and “market” his discoveries through clever media relations. This remarkable and in some respects enigmatic man was able to make the most of his opportunities in much the same way as his stellar contemporaries. Doggedness, vision, insight, personal qualities, and the availability of Federal needs and generous funding were all synergetic, crucial elements in the preparation and acceptance of the “Fisk Report.”     

Loess stratigraphy of the Lower Mississippi Valley

Loesses of the Lower Mississippi Valley (LMV) are world-famous. Sir Charles Lyell (1847), Hilgard (1860), Stafford (1869), Call (1891) and Mabry (1898), thought the LMV loess was a single water deposit although “double submergence” was noted by Call (1891) and Salisbury (1891). Shimek (1902) and Emerson (1918) recognized LMV loess as a wind deposit which came from the valley. Although wind-deposited loess gained wide acceptance, Russell (1944a) published his controversial theory of “loessification” which entailed weathering of backswamp deposits, downslope movement and recharge by carbonates to form loess. Wascher et al. (1947) identified three LMV loesses, mapped distributions and strongly supported eolian deposition. Leighton and Willman (1950), identified four loesses and supported eolian deposition as did Krinitzsky and Turnbull (1967) and Snowden and Priddy (1968), but Krinitzky and Turnbull questioned the deepest loess. Daniels and Young (1968) and Touchet and Daniels (1970) studied the distribution of loesses in south-central Louisiana. West et al. (1980) and Rutledge et al. (1985) studied the source areas and wind directions which deposited the loesses on and adjoining Crowley's Ridge. B.J. Miller and co-workers (Miller et al., 1985, 1986, Miller and Alford, 1985) proposed that the Loveland Silt was Early Wisconsin rather than Illinoian age and advanced the name Sicily Island loess. They proposed the underlying loess was Illinoian and advanced the name Crowley's Ridge. We termed the loesses, from the surface downward, Peoria Loess, Roxana Silt, Loveland/Sicily Island loess, Crowley's Ridge Loess and Marianna loess. Researchers agree that the surfical Peoria Loess is Late Wisconsin and the Roxana Silt is Late to Middle Wisconsin, but little agreement exists on the age of the older loesses. Pye and Johnson (1988) proposed Early Wisconsin for the Loveland/Sicily Island. McKay and Follmer (1985) suggested this loess correlated with a loess under Illinoian till. Clark et al. (1989) agreed on Crowley's Ridge, but suggested the Loveland/Sicily Island loess on Sicily Island was older. Mirecki and Miller (1994) and Millard and Maat (1994) suggested an Illinoian age for the Loveland/Sicily Island loess. Miller and co-workers suggested, as did Pye and Johnson (1988), an Illinoian age for the Crowley's Ridge loess. McKay and Follmer (1985) suggested it correlated with a loess under “Kansan” till. Stratigraphy indicates the Marianna is the older of the five loesses.

Researchers identified loess on both the east and west side of the LMV as well as on higher terraces within the valley. Many researchers assumed unaltered loesses were commonly yellowish brown, and silts or silt loams (West et al., 1980; Miller et al., 1986). The nonclay fraction of unweathered LMV loesses was dominated by quartz followed * Corresponding author. by carbonates, mainly dolomites, followed by feldspars, and micas. Clays were dominated by montmorillonite followed by micaceous minerals, kaolinite and vermiculite (Miller et al., 1986). Soils in the Crowley's Ridge loess are most developed, followed by the soils in the Loveland/Sicily Island which are more developed than the modern soils in the Peoria Loess. Soils in the Roxana and Marianna loesses are least developed and the Farmdale Soil of the Roxana is the weaker of the two (Miller et al., 1986). There is certainly overlapping range in the degree of soil development in the various loesses.     

Mississippi Valley regional source of loess on the southern Green Bay Lobe land surface, Wisconsin

Loess has been recognized on the glacial land surface of the Green Bay Lobe for over 100 yr, but no systematic explanation of the source of the loess has been advanced. Intriguingly, the loess on the Green Bay Lobe land surface is thicker than predicted by regional thinning trends from the Mississippi Valley and is geographically separated from much loess of southwest Wisconsin by a sandy region devoid of loess. Mapping based on soil survey interpretation indicates that the loess occurs above an escarpment marking the eastern end of the sandy loess-free region. Particle size fining trends demonstrate that the loess was transported by northwesterly winds. Clay mineralogy of the Green Bay Lobe loess is distinctly different than glaciogenic sediments and matches loess of the Mississippi Valley, indicating a regional source and long distance transport of the loess. We propose the loess was transported from the regional source along a surface of transport produced by migration of eolian sand through low-relief landscapes including the glacial Lake Wisconsin basin. Eolian sand migration caused repeated entrainment of dust leading to east-southeastward transport. The loess accumulated above an escarpment that limited sand mobility and re-entrainment of loess beyond this topographic barrier.     

Quaternary paleoecology of the Lower Mississippi Valley

In 1938, Clair A. Brown published his classic paleobotanical discoveries from the Tunica Hills of southeastern Louisiana, indicating ice-age plant migrations of more than 1100 km. Brown collected fossils of both boreal trees such as white spruce (Picea glauca) and southern coastal plain plants from deposits mapped as the Port Hickey (Prairie) river terrace by Harold N. Fisk. Subsequent revisions of terrace mapping, radiocarbon dating, and paleoecological analysis reconciled Brown's conceptual and stratigraphic “mixing” of these two ecologically incompatible fossil plant groups. An older Terrace 2 (of Sangamonian to Altonian age) contains the warm-temperate assemblage. A younger Terrace 1 (of Farmdalian, Woodfordian, and Holocene age) includes full-glacial and late-glacial remains of both boreal and cool-temperate plants; and a warm-temperate suite of plants dates from the Holocene interglacial. New plant fossil localities with radiocarbon chronologies are now available from within the Lower Mississippi Valley of Missouri and Arkansas as well as from the adjacent Ozark Plateaus, the Interior Low Plateaus of Kentucky and Tennessee, and the bordering Blufflands of Tennessee, Mississippi, and Louisiana. These studies demonstrate that glacial and interglacial patterns of vegetation have been influenced by regional changes in climate, glacial runoff, and regime of the Mississippi River.     

AMS Radiocarbon Age of the Upper Mississippi Valley Roxana Silt

Accelerator mass spectrometer radiocarbon ages of the Roxana Silt (loess) along the Upper Mississippi Valley of Wisconsin and Minnesota indicate that loess sedimentation of the Roxana Silt occurred between about 55,000 and 27,000 ¹⁴ C yr B.P. However, due to local environmental controls, the basal age at any given site may range from 55,000 to 35,000 ¹⁴C yr B.P. The radiocarbon ages presented here are in agreement with previous radiocarbon ages for the Roxana Silt in its type area of west-central Illinois, but indicate that long-term sedimentation rates along the bluffline of the Upper Mississippi Valley were very slow (4-8 cm/1000 yr) compared to long-term sedimentation rates along the bluffline of the type area (40-70 cm/1000 yr). Comparison of radiocarbon ages for midcontinent middle Wisconsinan loess deposits indicates that sedimentation along the Mississippi River valley may have preceded loess sedimentation along the Missouri River valley by as much as 20,000 yr or that basal ages for middle Wisconsinan loess along the Missouri Valley are erroneously young. The bracketing ages for the Upper Mississippi Valley Roxana Silt indicate that the Mississippi River valley was receiving outwash sedimentation between 55,000 and 27,000 ¹⁴C yr B.P.     

A late quaternary record of eolian silt deposition in a maar lake, St. Michael Island, western Alaska

Recent stratigraphic studies in central Alaska have yielded the unexpected finding that there is little evidence for full-glacial (late Wisconsin) loess deposition. Because the loess record of western Alaska is poorly exposed and not well known, we analyzed a core from Zagoskin Lake, a maar lake on St. Michael Island, to determine if a full-glacial eolian record could be found in that region. Particle size and geochemical data indicate that the mineral fraction of the lake sediments is not derived from the local basalt and is probably eolian. Silt deposition took place from at least the latter part of the mid-Wisconsin interstadial period through the Holocene, based on radiocarbon dating. Based on the locations of likely loess sources, eolian silt in western Alaska was probably deflated by northeasterly winds from glaciofluvial sediments. If last-glacial winds that deposited loess were indeed from the northeast, this reconstruction is in conflict with a model-derived reconstruction of paleowinds in Alaska. Mass accumulation rates in Zagoskin Lake were higher during the Pleistocene than during the Holocene. In addition, more eolian sediment is recorded in the lake sediments than as loess on the adjacent landscape. The thinner loess record on land may be due to the sparse, herb tundra vegetation that dominated the landscape in full-glacial time. Herb tundra would have been an inefficient loess trap compared to forest or even shrub tundra due to its low roughness height. The lack of abundant, full-glacial, eolian silt deposition in the loess stratigraphic record of central Alaska may be due, therefore, to a mimimal ability of the landscape to trap loess, rather than a lack of available eolian sediment.     

Pleistocene stratigraphy in the southern Lower Mississippi Valley

Scientific inquiry into Pleistocene stratigraphy of the Lower Mississippi Valley (LMV) dates to early writings of European naturalists in the late 19th century. By the early 20th century, landscape evolution concepts, stratigraphic models, and regional syntheses had developed for most areas. The 1944 monograph of H.N. Fisk marks the advent of a predictive stratigraphic and landscape evolution model that links form and process to a predominantly glacioeustatic mechanism. The Fiskian model gained widespread acceptance, and decades passed before significant alternate models began to emerge. Revised stratigraphic and geomorphic concepts are presently developing from newly acquired environmental and engineering data. Present scenarios classify Pleistocene outcrop areas into erosional and constructional landscapes, and veneers of eolian, colluvial, fluvial, coastal, and marine origin can drape both types of surfaces.

The southern LMV and adjacent Gulf Coastal Plain (GCP) experienced significant landscape change during the Pleistocene. Late Tertiary (Pliocene?) to Early Pleistocene deposition of the Upland Complex was by streams with a high sand and gravel load relative to its mud load. The regional drainage network and fluvial system behavior was probably significantly different from the modern. Braided stream alluvial fan complexes received sediment from highland source areas adjacent to the LMV and the glaciated mid-continent. It is plausible that part of Upland Complex deposition predates initial glacial advances.

From Early to Middle Pleistocene, an erosional landscape formed during a dissection period that chiefly postdates soil formation on stable landscape positions of the Upland Complex. Slope evolution truncated a regionally extensive geosol in multiple phases, and parts of the erosion surface complex are graded to the oldest preserved constructional alluvial plains in present valleys. Toe and foot slope positions of the erosion surface complex and its correlative alluvial plains are presently delineated as the Intermediate Complex. Constructional landscapes formed at this time are sparsely preserved; Fisk's Montgomery Terrace in the Lower Red River Valley (LRRV) is the best preserved example. Influences on the development of erosion surfaces in the LMV are not well understood; however interactions of relative sea level fall, climate change, and epirogenic crustal movement are plausible factors.

From the latter part of Middle Pleistocene to the Holocene, there was widespread evolution of modern constructional landscapes. Constructional alluviation preserved lithofacies of mixed load, laterally accreting, meandering streams that developed over large areas of the southern LMV to form parts of the Prairie Complex. Lateral planation in valleys and stable rates of upland sediment generation were dominant processes during Prairie Complex deposition.

Pleistocene stratigraphic examples considered important by Fisk are still considered relevant to modern stratigraphic investigators. Presently, Pleistocene units of the southern LMV, the adjacent LRRV, and central GCP can be correlated only by relative stratigraphic relationships. Refined chronostratigraphic and paleoenvironmental models for these areas would help improve the understanding of the geomorphic influences on Quaternary landscape evolution in the region.     

Rapid and cyclic aeolian deposition during the Last Glacial in European loess: a high-resolution record from Nussloch, Germany

This paper reports the results of an investigation of the Weichselian Upper Pleniglacial loess sequences of Nussloch (Rhine Valley, Germany) based on stratigraphy, palaeopedology, sedimentology, palynology, malacology and geochemistry (δ¹³C), supported by radiocarbon, TL and OSL dating. Grain-size and magnetic susceptibility records are taken at 5 cm intervals from the Upper Pleniglacial (UPG) loess. The data indicate cyclic variations in loess deposition between ca 34 and 17 ka, when the sedimentation rate is especially high (1.0–1.2 m per ka for more than 10 m). The grain-size index (GSI: ratio of coarse silt versus fine silt and clay) shows variations, which are assumed to be an indirect measurement of wind intensity. The sedimentation rate, interpreted from the profiles, indicates high values in loess (Loess events LE-1 to LE-7) and low or negligible values in tundra gley horizons G1 to G8. OSL ages from the loess and ¹⁴C dates from organic matter in the loess show that loess deposition was rapid but was interrupted by shorter periods of reduced aeolian sedimentation. Comparison between the data from Nussloch and other European sequences demonstrates a progressive coarsening of the loess deposits between ca 30 and 22 ka. This coarsening trend ends with a short but major grain-size decrease and is followed by an increase to a new maximum at 20 ± 2 ka (“W” shape). Correlation between the loess GSI and the Greenland ice-core dust records, suggests a global connection between North Atlantic and Western European global atmospheric circulation and wind regimes. In addition, the typical Upper Pleniglacial loess deposition begins at ca 30–31 ka, close to Heinrich event (HE) 3, and the main period of loess sedimentation at about 25 ± 2 ka is coeval to HE 2. Correlation of magnetic susceptibility and grain-size records shows that the periods, characterised by high GSI, coincide with an increase in the amount of ferromagnetic minerals reworked from the Rhine alluvial plain. They suggest enhancement in the frequency of the storms from N–NW. These results are integrated within a palaeogeographical model of dust transport and deposition in Western Europe for the Weichselian Upper Pleniglacial (or Late Pleniglacial).     

A Mid–Late Quaternary loess–paleosol record in Simmons Farm in southern Illinois,USA

In unglaciated areas of the Mississippi Valley region, the typical full loess–paleosol succession contains the Modern Soil developed in Peoria Silt, weakly developed Farmdale Geosol developed in Roxana Silt, Sangamon Geosol developed in Loveland Silt, and Yarmouth Geosol developed in Crowley's Ridge Silt. Although a fifth loess called the Marianna Silt is reported at one area, the paleosol that separates the Crowley Ridge and Marianna Silts is not well defined. Previous thermoluminescence (TL) and optical stimulated luminescence (OSL) age chronology has suggested multiple phases of Sangamon Geosol developed in Loveland Silt, but clear morphological evidence of polygenetic Sangamon Geosol profiles have not been found. Recently, a thick loess–paleosol sequence has been studied in the middle Mississippi Valley in unglaciated southern Illinois, USA. Soil morphology and analytical results revealed five loesses and associated paleosol units. Two Sangamon Bt horizons were found separated by a thick ACtk horizon, interpreted to indicate two phases of Sangamon Geosol development. This well-preserved loess–paleosol succession provides one of the most complete mid–late Quaternary loess records in the middle Mississippi Valley to date, and is important for studying the stratigraphic framework and paleoclimate and environment changes.     

Sand dunes as a major proximal dust source for late Pleistocene loess in the Negev Desert, Israel

Grain size analyses of three hilltop, primary eolian loess sequences in the Negev desert, southern Israel, show a bimodal grain-size distribution at 50-60 μm and 3-8 μm. Using analyses of mineralogy and OSL ages we demonstrate that the coarse mode is composed mostly of quartz grains and its relative magnitude increases regionally with time, suggesting an enhancement of a time-transgressive proximal dust source compared to a distal, Saharan fine-grain dust. The only proximal dust source for large amount of coarse silt quartz grains is the sands that advanced into Sinai and the Negev concurrently with the loess accretion during the late Pleistocene as a result of the exposure of the Mediterranean shelf. We therefore propose that the coarse silt quartz grains were formed through eolian abrasion within the margins of an advancing sand sea. This relationship between desert sand seas as a source for proximal coarse dust and desert margin loess deposits can be applicable to other worldwide deserts such as Northern Africa, China and Australia.     

Loess sedimentation in Tibet: provenance,processes, and link with Quaternary glaciations

Well-preserved loess deposits are found on the foothills of mountains along the middle reaches of the Yarlung Zangbo River in southern Tibet. Optically stimulated luminescence (OSL) dating is used to determine loess ages by applying the single-aliquot regeneration technique. Geochemical, mineralogical, and granulometric measurements were carried out to allow a comparison between loess from Tibet and the Chinese Loess Plateau. Our results demonstrate that (i) the loess deposits have a basal age of 13–11 ka, suggesting they accumulated after the last deglaciation, (ii) loess in southern Tibet has a “glacial” origin, resulting from eolian sorting of glaciofluvial outwash deposits from braided river channels or alluvial fans by local near-surface winds, and (iii) the present loess in the interior of Tibet has accumulated since the last deglaciation when increased monsoonal circulation provided an increased vegetation cover that was sufficient for trapping eolian silt. The lack of full-glacial loess is either due to minimal vegetation cover or possibly due to the erosion of loess as glaciofluvial outwash during the beginning of each interglacial. Such processes would have been repeated during each glacial–interglacial cycle of the Quaternary.     

Source Regions and Formation of the Loess Sediments on the High Mountain Regions of Northwestern China

Loess on the high mountains of northwestern China is largely controlled by the topographic features, distributed mainly on the northern slopes of the Tianshan and Kunlun Mountains as well as in the Ili Valley (between the northern and central Tianshan ranges). Loess distributed in different regions has different ages, with the oldest (early Pleistocene) loess occurring in the Ili Valley. Geochemical, mineralogical and granulometric analyses demonstrate that the loess in the three different regions has different provenance. The loess on the northern Tianshan and Kunlun Mountains is mainly derived from the Junggar and Taklimakan Deserts, respectively, whereas the loess in the Ili Valley is derived mainly from the Sary–Ishikotrau Desert located in the Republic of Kazakhstan. However, these deserts serve mostly as holding areas for the silts rather than original source regions. The production of the silt-sized particles is mainly associated with glacial abrasion and tectonic-induced rock denudation. Thus, the loess in the studied regions can be described as “mountainous” loess rather than simply “glacial” loess or “desert” loess. It is unlikely that the mechanisms of loess formation are the same in different regions, given their dependence on specific geomorphological conditions and associated processes.     

The contributions of H.N. Fisk to engineering geology in the Lower Mississippi Valley

Fisk's work provided a three-dimensional pattern of occurrence and composition of sediments in the alluvial valley and deltaic plain of the lower Mississippi River. He also interpreted the processes by which the sediments were deposited and the relevance of tectonism in their history. These data provide site characterizations that are basic for all major categories of engineering. His work has found applications in better methods for control of the river, stabilizing its banks, locating sources of aggregate for concrete, management of groundwater, maintenance of wetlands, and generally for more reliable, timely and economical evaluations in selecting optimum site locations and determining foundation conditions. Since then, work in the Lower Mississippi Valley (LMV) has expanded enormously and continues to be greatly indebted to formative interpretations by Fisk.     

山东青州地区黄土的粒度组成及成因分析

在鲁中山地北侧的山间谷地及山麓地带广泛发育厚度不一的黄土堆积。对该区青州傅家庄黄土剖面进行了系统的粒度分析, 并与黄土高原的第四纪黄土-古土壤、北京现代降尘、剖面附近的河流相样品进行了对比。结果表明, 青州黄土的粒度分布特征与北京现代降尘、黄土高原黄土非常相似, 与河流相样品则有很大的不同; 在整个序列中, 青州黄土粒度变化与黄土高原典型的风成黄土、古土壤一致, 黄土层颗粒粗, 古土壤层颗粒细。上述结果从沉积学的角度为青州黄土风成成因提供了证据。青州黄土的粒度组成特征及前人的研究表明, 青州黄土主要来源于沉积区以北的黄泛平原和莱州弯等地出露的海相地层, 其次是高空气流携带的西北内陆的远缘粉尘。      

A continuous record of Holocene eolian activity in West Greenland

Eolian landforms are widespread alongside proglacial valley-sandurs in West Greenland and comprise low-relief sand sheets, climbing dunes, and upland loess. Sedimentary facies mainly reflect distance to outwash-source zones and the influence of vegetation cover. The sediments show stratification types typical for poorly to moderately vegetated sand-sheets, alternately laminated silt/peat sequences, and unstratified loess. Twenty-five accelerator mass spectrometry ¹⁴C dates provide the basis for the chronostratigraphy of the inland eolian deposits. ¹⁴C dates from interstratified sand-sheets suggest that the bulk of eolian sands were deposited prior to 3400 cal yr B.P. and after 550 cal yr B.P. This two-phase formation for the inland dunes most likely reflects local changes in proglacial floodplain development and meltwater rerouting associated with a significant recession of the Greenland ice sheet during the mid Holocene climate optimum. Subsequent floodplain regeneration and renewed sand-sheet formation after 550 cal yr B.P. followed when the ice margin readvanced to its present position. In contrast, atmospheric deposition of regionally derived silt in upland peat mires has been continuous since at least 4750 cal yr B.P. Silt influx data demonstrate a strongly episodic history of the intensity of eolian activity over the past five millennia, which tentatively reflects alternating periods of (winter) aridity associated with the variable incursion of maritime air masses over the interior ice-free areas of West Greenland.     

Late Pleistocene Environments of the Central Ukraine

The Vyazivok loess sequence from the Dnieper Plain, Ukraine, documents regional environmental changes during the late Pleistocene and Holocene. Pedological and palynological analyses and low-field magnetic susceptibility document changes from dense temperate forest during the last interglacial maximum to open, harsh, loess–steppe during the latest Pleistocene. The Vyazivok section overlies hillwash derived from a lower Pleistocene terrace and consists of two stratified soil complexes (Kaydaky and Pryluky; marine isotope stage [MIS] 5 equivalent) separated by a layer of eolian dust (Tyasmyn silt). The lower soils in both complexes formed within forest. These soils are overlain by the Uday (MIS4) and Bug (MIS2) loess units, which are separated by boreal soils of the Vytachiv (MIS3) complex. The coldest conditions within the record occurred in the youngest loess. Holocene soils cap the Bug loess. The Vyazivok section shows remarkable similarities with other classical loess sequences in western Europe, the Czech Republic, and Austria. The Kaydaky, Pryluky, and Vytachiv deposits, correlate with the PKIII, PKII, and PKI soil complexes, respectively, of the Czech Republic. The Tyasmyn and Prylyky silt layers correspond to marker horizons from central Europe.     

Geoarchaeology of the Cerro La China locality (Buenos Aires,Argentina): Site 2 and site 3

Cerro La China, within the Tandilia Range, has a long archaeological record in which three different human occupations were identified. The geoarchaeological approach allowed us to correlate the archaeosedimentary columns at Site 2 and Site 3 and place the local stratigraphy in a regional context. The archaeologically rich stratigraphic sequence is composed of three episodes of loess deposition. The first bears a 10,600 yr B.P. Paleoindian occupation with Fishtail or Fell's Cave Stemmed points. It is followed by a soil-forming interval which ended about 5000 yr B.P. with an erosive episode. The second event of loess deposition continued through an unknown span of time until a new soil-forming interval started; it includes triangular stemless medium size projectile points corresponding to a different human group. By the sixteenth century another erosive event interrupted the soil-forming interval. After this, the third short-term eolian episode occurred and bears the most recent archaeological occupation marked by small triangular stemless points.     

“In-situ” theories of loess formation and the significance of the calcium-carbonate content of loess

It is generally accepted that loess deposits form by action and yet the concept of “loessification” is still encountered, particularly in some European literature. Loessification often seems to be invoked when the deposit being investigated is rich in calcium carbonate because the use of the concept depends upon an eclectic definition of loess. Loessification is irrelevant if loess is defined in terms of the fine quartz material, but may be meaninful if the loess is defined in terms of calcium carbonate. The Russell theory of in-situ formation can possibly be reconciled with deposition by eolian action and production of quartz material by glacial grinding but the Berg theory cannot, and appears to be based on false premises. The “soil science” approach to the problem of loess formation may be responsible for the overvaluing of the significance of the carbonate content; the “geological” approach, with some emphasis on transportation problems, leads logically to the concept of eolian deposition but has, so far, failed to focus attention on the problem of the formation of the actual loess material. The fine quartz material is mostly formed by glacial grinding; there appear to be no other natural forces powerful enough to produce appreciable quantities of loess-sized quartz particles. Deposition of the secondary carbonate is usually the last of the significant stages in the formation of the deposit; scanning-electron-microscope studies suggest that it exists as discrete encrustations (gnarls) rather than as continuous coatings on the quartz grains.