Pingos and pingo scars: Their characteristics, distribution, and utility in reconstructing former permafrost environments

Pingos and pingo scars: Their characteristics, distribution, and utility in reconstructing former permafrost environments

QUATERNARY Pingos RESEARCH and Pingo Utility Department 6, 37-53 Scars: in Reconstructing of Geology, (1976) Their Characteristics, Former ...

2MB Sizes 5 Downloads 32 Views




and Pingo

Utility Department

6, 37-53


in Reconstructing of Geology,






RONALD C. FLEMAL Northern Illinois University, Received April



Environments DeKalb,



24, 1975

Pingos are large frost mounds which develop in permafrost as the result of the segregation of massive ground-ice lenses. At least two genetic varieties of pingos, open- and closed-systems, form under differing conditions of climate, topography, and groundwater occurrence. Active pingos are known to occur in many high latitude regions. Pingo scars are the degeneration products of pingos. Ideally they are ramparted, circular depressions, although they may be expressed in a variety of divergent forms due to differing conditions of topography, substrate materials, degree of thermokarst overprint, and erosionalldepositional histories. Pingo scars occur in many modern permafrost regions. Presumed pingo scars have also been identified in many regions beyond the present permafrost limit and therefore may have utility in reconstructing former permafrost environments.

INTRODUCTION Pingos have been described from many modern permafrost regions. Studies of these have provided a fairly clear picture of their varieties and significant observations on their genesis and destruction. In the last two decades pingo study has taken on an added dimension in that a large number of features found in nonpermafrost regions have been ascribed to the former presence of pingos. Besides being of interest unto themselves, such former pingos offer evidence of past permafrost conditions.

used by the Eskimos of the Mackenzie Delta region to denote the many hundreds of conical hills of that area. It was introduced into scientific usage by Porsild (1938) to refer to these same hills. The term has subsequently been applied to similar hills in other regions as well as to scars marking the sites of former pingos. Among the descriptive criteria commonly employed to define a pingo are (1) intrapermafrost location; (2) hill-form, typically with a conical transverse section and circular or oval plan form; (3) interior core of massive ice; (4) existence for longer than a single season (i.e., “perennial”); and (5) comparatively large size TERMINOLOGY with respect to other varieties of frost Pingos are members of the larger class mounds. Pingos may differ according to of intrapermafrost features variously both genesis and state of decay, which known as ice mounds, ice-and-earth has led to a terminology reflective of mounds, or frost mounds. In addition to these differences. Review of the genesis pingos this class includes palsas and vari- of pingos are provided by Muller (1963) ous varieties of hummocks (turf, earth, and Washburn (1973). It is generally peat, bog, etc.). Pingos are the largest considered that there are two major and most spectacular members of this genetic varieties of pingos, those formed class. The smaller pingos are transitional from cryostatic pressure and those in size with p&as. formed by hydrostatic pressure. CryoThe term pingo is a word meaning statically generated pingos result from the encroachment of a freezing front on “small hill, ” “up there,” or “up above” 37 Copyright @ 1976 by the University of Washington All rights of reproduction in any form reserved.




subpermafrost saturated ground or on a talik. Cryostatic pressure causes the confined water to be ejected upward where it freezes to form the ping0 core. Pingos of this type have been termed “closedpingos. system” or “Mackenzie-type” Pingos generated from hydrostatic pressure result from the upward movement and near-surface freezing of water propelled by a hydrostatic head. Pingos of this type have been termed “open-system” or “East Greenland-type” pingos. Both theoretical calculations and actual measurements (Williams and von Everdigen, 1973) suggest that hydrostatic pressure alone is insufficient to form open-system pingos in the classically suggested manner. Accordingly, some workers have suggested that opensystem pingos result from a combined cryostatic and hydrostatic pressure system. Maarleveld (1965) recognized a third variety of pingo, the bugor. A bugor is formed when water is “sucked up by an ice lenticle in clay or silt” (Maarleveld, 1965, p. 19). The two terms “Bulganniakh” and “hydrolaccolith” are often used as synHowever, Maarleveld onyms for pingo. (1965) suggested that the two terms are more directly equatable to closed-system and open-system pingo, respectively. Miiller (1968) also suggested that hydrolaccolith should be applied to all varieties of ice intrusions, and not exclusively to pingos. Partially decayed pingos typically have forms resembling volcanic craters. Advance stages of decay may be marked by depressions which often have a central lake and an enclosing rampart. Various names have been applied to such late stage forms, including pingo scar, pingo remnant, pingo ruin, collapsed pingo, extinct pingo, relict pingo, fossil pingo, and Pleistocene pingo. Most of these terms are objectionable on one or more grounds. Of the available terms, “pingo

scar” seems the most appropriate. It is accordingly used herein to refer to any morphologic feature left as a residual upon the destruction of the ice core of a pingo. Pingo scars occur in modern permafrost regions and have been inferred to exist in modern nonpermafrost regions. Palsas have much in common with pingos; the two overlap in size, form, composition, and mechanisms of origin. They are difficult to distinguish in active varieties and the distinction is even more difficult in the scar forms. Accordingly, much that is said herein is applicable to palsas as well as to pingos. The term “palsa scar” will be employed as a parallel to ping0 scar. CHARACTERISTICS OF ACTIVE PINGOS Active pingos are known from Alaska, ~~aa~,~eit~r~~~,,-~~d, and ‘the WR: __They-&e most prevalent in the latitude range 65-75”N where the permafrost is either discontinuous or continuous but thin (Miiller, 1968). In the Asian steppes pingos occur abundantly as far south as 43”N, and in isolated occurrences to 36’N (Frenzel, 1959). No pingos are known from the south polar regions. Pingos range in height from a few meters to as much as 100 m (Jahn, Basal diameters range from 1968). several meters up to 1200 m (Holmes, Foster, and Hopkins, 1966). Most pingos are circular or oval in plan, but some ridge-form or esker-like varieties have been recognized (e.g., Porsild, 1938; Mackay, 1962; Pissart, 1967). Departure from circularity is most pronounced in pingos which occur on slopes. The long axes of such pingos may be aligned either parallel or perpendicular to the slope. The ice core of pingos is typically massive, with only a small amount of entrained debris (Mackey, 1973). In the varieties transitional with palsas, lenses and irregular masses of debris may be in-



terstratified with the ice. Ice cores may thin alluvium or bedrock (Miiller, 1963). attain thicknesses of 80 m or more and a It has been inferred that the slope of diameter almost as great as that of the some open-system pingos may be propingo itself (Miiller, 1968). The upper vided by a glacier front (Flemal, 1972). surface of the ice core mirrors the surface Closed-system pingos tend to occur in form of the pingo; such undersides as more or less flat terrain. They are most have been seen are typically planar common in deltaic or lacustrine terrain (Mackay, 1973). Ice cores are often ex- such as typifies the Mackenzie Delta. posed in tension fractures opened in the Brown and PCwe (1973) pointed out summits of pingos during the uplifting of that almost all of the open-system pingos the dome. in northwestern North America occur The largest known pingo field, that of outside areas glaciated within the last the Mackenzie Delta, contains approxi25,000 yr. Hughes (1969) concluded mately 1450 individuals (Mackay, 1973). that local topography, in part con&The field of central Alaska and the cen- tioned by the extent and relative ages of tral Yukon contains more than 700 in- former glaciations, is the major factor dividuals (Mackay and Black, 1973). It controlling Thus the this distribution. is possible that some of the fields of the development of open-system pingos in USSR contain even greater numbers. Re- other physiographic settings does not apgion&-densities are of the order of one pear to be precluded by occurrence and per 2-3 km2 (e.g., Mackay, 1962), al- timing of prior glaciations. Distribution though local densities may be very much of closed-system pingos has not been obhigher. Open-system pingos in particular served to bear any relationship to glacial tend to overlap one another (e.g., Miiller, borders. 1963; Holmes, Hopkins, and Foster, The time required for the full develop1968). ment of a ping0 may range from as little Distribution of the two major genetic as a few minutes (Strugov, 1955) to as types of pingos bears a strong relationmuch as 1000 yr (Mackay and Black, ship to the extent of permafrost developThe majority of active pingos 1973). ment. Most closed-system pingos occur and pingos which retain an appreciable in areas of continuous permafrost; openamount of their original ice core had system pingos are essentially confined to their origins within Holocene time (Pew& discontinuous permafrost. Therefore, if 1973). Many are less than 1000 yr old the exceptions in the distribution are ex- and some less than 25 yr old (Mackay, cluded, a general association of pingo 1973). type with mean annual air temperature CHARACTERISTICS OF can be made: - 6” C or lower for closedPING0 SCARS system pingos and a range of about - 1 to Pingo scars have been described from a - 6°C for open-system pingos (Washburn, large number of localities and environ1973). ments, including both modern permafrost Modern open-system pingos typically occur near the base of slopes. Those of regions and regions where permafrost no longer exists. central Alaska and the central Yukon Modern permafrost regions. The least occur dominantly on south- or southeastproblematical of pingo scars are those facing slopes of alluvium-filled valleys, found in modern permafrost regions, parprobably in response to variations in ticularly when found in juxtaposition groundwater conditions (Brown and with pingos in various stages of generaPew& 1973). Some of the open-system tion or decav. These pingos and pingo pingos of East Greenland occur in either _ I




scars indicate that formation of a pingo scar need not involve complete degradation of the permafrost. Rather, pingo scars may form as the natural product of pingo degeneration independent of the history of the permafrost. Complete decay sequences have been described by many authors (e.g., Porsild, 1938; Mtiller, 1963; Holmes, Hopkins, and Foster, 1968; O’Brien, 1971; Mackay, 1973). Similar decay sequences have been described for palsas (e.g., Svensson, 1969; Friedman et al., 1971). Freshly raised pingos degenerate as a result of breaching of the ice core, most D often at the summit, either by exposure in tension fractures (Fig. 1A) or erosion of overburden. Exposure of the ice core allows seasonal melting and sublimation of the core. Intermediate forms often E have a crater-like shape (Fig. 1B) which may enclose a small lake. With progressively greater melting of the core, relief is gradually lowered until eventually relief FIG. 1. Possible stages in the development of is inverted and the original mound has a pingo scar. Explanation in text. become a depression (Fig. 1C). At this by stage the pingo has become a pingo scar. of the rampart may be accentuated frost and ice heaving (Svensson, 1969) Pingo scars mirror the outline of the original pingo; that is, they tend to be or decreased as the result of erosion. Ramparts may also be destroyed by circular to oval. The central depression itself often contains a lake, pond, or bog, slumping into the depression, particularly those portions of ramparts which lie upwhich is the site of secondary deposition slope from the depression (Miiller, 1963). (Fig. lC, D). This deposition may result Breached rims of the latter type apparfrom the accumulation of organic debris, ently require that the depression lake be wind-blown debris, debris washed in from adjacent higher ground, or debris welled fed by springs, a condition common in open-system pingos (Miiller, 1963). up from below by rising groundwater. The latter phenomena is known to conSeveral generations of pingos may detribute to the sedimentary fills o velop at or near a single site. Thus, earlier forms may be disrupted or even +ygg~~s~.t~~~y; destroyed by later forms. Repeated associated with the waters of many open- generation is more common with opensystem ping0 scars. system pingos than with closed-system A rampart of debris, formed in part pingos. from material eroded from the original Very old pingo scars, particularly in slopes of the pingo (Fig. lB, C) and in areas where active pingos are absent, may part as the result of the original up- be confused with thaw lakes (Holmes, thrusting of the dome (Fig. lA), may Hopkins, and Foster, 1968). Thus, pingo encircle the central depression. The relief scars may be more common in the mod-


em permafrost environments than is reflected in the literature. A critical gap in knowledge of the history of pingo scars exists between those stages observed in modem permafrost regions and forms found in former permafrost regions. That gap concerns the effects imposed by degradation of the permafrost adjacent to the pingo scar (Mackay, 1973). This massive thermokarst overprint can be presumed to have been imposed on all pingo scars which occur in modern nonpermafrost regions. Its effects, however, can only be postulated. One postulate is that there will be accentuation of relief. This stems from the observation that permafrost is either absent or thinner under periglacial lakes of any sort (e.g., Washburn, 1973). Destruction of the permafrost in areas adjacent to a pingo scar will therefore tend to lower those areas more than the scar itself will be lowered. With minimal differential lowering the overall form of the scar may not be appreciably altered (Fig. ID). However, if the thermokarst lowering is great, the pingo scar may actually be inverted from a depression to a rise or hillock (Fig. 1E). This phenomenon may explain the problem (Mackay and Black, 1973) of apparent excess of material in some ping0 scars. Nor~q,&a~&Finland. A series of geometrically regular lakes described from northern Norway (Svensson, 1962, 1964, 1969,197O) and northern Sweden (Rapp and Rudberg, 1960) were considered by Svensson to represent scars of either pingos of palsas. Similar features described from northern Finland were considered to be pingo scars by Seppala (1972). All are circular to oval and are enclosed by a low, uniform wall. A shallow annular depression, often waterfilled, typically occurs outside the wall and accentuates the overall relief. Average diameters are of the order of a few tens of meters, although the largest have diameters greater than 100 m. Wall


heights range from less than 1 m to in excess of 8 m. The lakes always occur in groups or clusters, typically on flat or slightly sloping ground. They are most prevalent over glaciofluvial deposits of high moisture content, but also occur in thin drift over bedrock. The lake fills consist of stratified silts and peat. The floors of the lakes commonly lie above the level of the adjacent ground. The lakes occur in areas which are now essentially permafrost-free. However, Svensson (1969) noted that they form a continuum with decayed but still active frost mounds which occur in nearby permafrost areas. Rapp and Rudberg noted that boulders within the lake walls are oriented with their long axes tangential to the walls, suggestive of a radially outward dilation mechanism of formation. They noted the similarity in form and size of the Swedish lakes to small pingo-crater lakes in East Greenland, but otherwise left open the question of origin. Denmarh. Small ponds in southeastern Sjaelland whose basins are developed in moraines of the last glaciation were considered by Cailleux (1957) to have formed as the result of the melting of the ice from postglacial pingos. The previous interpretation was that the lakes were kettles. The absence of encircling ramparts or walls has caused the pingo scar origin to be questioned (e.g., (Maarleveld, 1965). Germany. Presumed pingo scars have been identified in several localities from Germany by Frenzel (1959), Muckenhausen (1960), Kaiser (1960), Picard Wiegand (1965), Schneider (1961), (1970), and Bruning (1972). These are particularly noteworthy for the broad range of their substrates, including crystalline rocks, sandstone, and a variety of Quatemary deposits, and their topographic environments, including both plains and steep valley sides. However,




Netherlands. Numerous pingo scars Wiegand (1965) noted that in all cases have been identified from the northern they occur in areas underlain by waterpart of the Netherlands (Maarleveld and saturated materials. Most of the forms are similar: regular, van den Toorn, 1955; Nossin, 1961; round, closed depressions, usually sur- Polack, 1964; Ploeger and Groenman-van rounded by a rampart. Diameters range Watteringe, 1964; Maarleveld, 1965). from 6 to 250 m. Fills are usually ap- These are round depressions or lakes preciable. Pollen in the basal sections of which typically have an encircling rampart. Diameters range upward of 500 m. the fills indicates a late Wtirm initiation The scars are most prevalent in areas of of deposition and hence a probable Wtirm surficial till underlain by sand. The scar age for the origin of the depressions. Most of the depressions lie outside the depressions bottom in sand and are parlimits of Wiirm glaciation and therefore tially filled with peat, gyttja, and sand. Fill thicknesses are of the order of 5 m. can not have been formed by any mechanism of Wiirm glaciers. In some cases Although the scars all occur in areas last upturning of bedrock strata can be seen glaciated during the Riss, pollen data from the basal sections of the fill indicate beneath the depression ramparts. initiation of deposition during the latter Muckenhausen (1960) described several Wiirm. Nossin indicated that the pollen unusual pingo scars from the Eifel region. data suggests two periods of formation, These consist of shallow ramparted furone pre-Allerod and one immediately rows which terminate in the downslope preceding the Preboreal. The ramparts direction in a ramparted circular depresare composed of redeposited till suggession. They are developed in a clayey tive of slumpage away from a formerly solifluction layer. Muckenhausen invisioned the forms as resulting from the high central area. Belgium and Luxemburg. Pingo scars downslope movement of subsurface water. Freezing of the water at the toe of have been described from Belgium and the slope formed an initial ice block, adjacent parts of Luxemburg by Pissart which then grew in the upslope direction (1956, 1963, 1965), Slotboom (1963), as additional ice was added to the block Mullenders and Haesendonck (1963), from that side. Paepe (1968), and Mullenders and GulPoland. .Presumed pingo scars have lentops (1969). Most occur in the high b%en identified by Dylik (1963a, 196313, plateaus of east Belgium where they 1965) and Dylikowa (1964). Discussion number about 1000. Two common of these features is also given by Dylik forms occur, closed circular or oval de(1964, 1973), Gawlik (1970), and Jahn pressions, and open elongate depressions. (1970). The common form is that of a Both forms are typically encircled by regular rimmed depression. Substrata raised ridges or ramparts. Closed forms may be deformed upward concentrically range in diameter to 150 m; open forms with the depression rim. Dylik (1965) range to 800 m. The depressions contain believed the pingo scars at Jozefow de- fills of stratified clay, gyttja, and peat. veloped during the maximum of the last Pollen analysis indicated that initiation glaciation. Gawlik considered the pre- of deposition occurred from the Wtirm cursors of some of the forms to have maximum (Pissart, in Washburn, 1973) been bugors rather than pingos sense to the Late Dryas (Mullenders and Gulstricto. Jahn discounted the pingo origin lentops, 1969). The typical substrate is of all but the ramparted depressions. poorly drained clay. Rim compositions Nonramparted depressions he considered are heterogeneous mixtures of loess, clay, soil, and bedrock. to be preserved thermokarst depressions.


Mullenders and Gullentops (1969) tentatively identified a single large, rimmed depression (900 by 1600 m) located in Low Belgium as a pingo scar. It is developed in Eocene sands and overlain by loess. Pollen analysis indicates its origin in the Older Dryas. France. Possible pingo scars have been identified from four areas, the regions of Paris (Cailleux, 1956, 1960, 1961; Pissart, 1958,1960,1968), Bordeaux (Boye, 1957, 1958), Brittany (Riviere and Vernhet, 1962), and the Vosges (Wiegand, 1965). With the exception of the latter, all lie beyond the limits of glaciation. The most common form is a circular, water-filled depression lacking an encircling rampart. In the Paris region there are more than 4500 such features. They are restricted to areas underlain by carbonate rocks, and hence some may be sinkholes (Pissart, 1958). Prince (1961) and Gulley (1962) also considered the possibility of human origins of the depressions. Pissart (1958, 1960) and Cailleux (1960, 1961) argued against human origins and in favor of a ground ice origin, either in the form of thaw lakes or pingo scars. England. Sparks, Williams, and Bell (1972) ascribed a large number of depressions located in East Anglia to a frost mound origin. Sparks and West (1972) also regarded these depressions as having been derived from some type of frost mound, but not necessarily any type known from present permafrost regions. Most of the depressions ‘are circular or oval in plan, contain a marsh or pond, and are surrounded by ramparts. In areas of cultivation the ramparts are lower than elsewhere, implying partial destruction by agricultural practices. Depression densities locally reach lOO/km’ and average 30-50/km2. In areas of high density the forms mutually interfere. Most have major diameters ranging from 10 to 120 m, although composite forms


are larger. The depression floors, which are underlain by peat and muck, lie less than 3 m below the ramparts. Internal stratigraphy suggests that the depressions are confined to poorly drained, almost drift-free, low-lying areas, often at the base of chalk cliffs. Although no deformation of the underlying bedrock was noted, the rims contain rubble of the underlying chalk. Prince (1961, 1962, 1964) described circular depressions from Norfolk. Although he considered the possibility of their periglacial origin, he favored an anthropogenic origin. Watson (1971) considered ten elongate ponds on the Isle of Man to be pingo scars. The ponds are located on an alluvial fan and are oriented with their long dimension perpendicular to the fan slope. The largest measures 40 by 140 m. The ponds lack ramparts, but Watson attributed this to destruction as a result of extensive cultivation. Watson considered the ponds to have direct parallels in the slope-elongate open-system pingos of Prince Patrick Island described by Pissart (1967). Wales. Several groups of presumed pingo scars have been described from central and southwestern Wales (Pissart, 1963; Trotman, 1963; Watson, 1971, 1972; Watson and Watson, 1972). All are similar in being ramparts enclosing peat-filled basins. Both circular and elongate forms are present; elongate forms are typically open in the upslope direction and pass into ridges on steeper slopes. Rampart heights are as great as 6.5 m and maximum diameters extend to 120 m for circular forms and 250 m for elongate forms. All occur in valley bottoms or lower valley sides in host material of alluvium and bedrock. The bedrock is locally warped upward toward the depressions. Watson and Watson (1972) reported fills within the depressions exceeding 11 m and consisting of peat overlying



clay-silt lake sediments. Basal material has been dated to Pollen Zone II/IV, suggesting a late glacial origin for the depressions. The total number of pingo scars is difficult to determine due to interference of individual forms, but they definitely number in the several hundreds. The major clusters occur at the foot of northfacing slopes. Crowding of forms is most pronounced on the steeper slopes. Ireland. Mitchell (1971) reported that pingo scars have wide distribution beyond the limit of the last glaciation in the south of Ireland. Detailed investigation of a cluster of 200 of these structures located within a 5.75 km2 area indicated that the scars were formed at the end of the last cold stage, in Zone III (Mitchell, 1973). The structures are of the typical ramparted form, and are circular to elongate in plan. Rim diameters range from 30 to 75 m. Canada. Outside the modern permafrost region possible pingo scars have been identified from the Sturgeon Lake (Henderson, 1959) and Fort St. John (Mathews, 1963) areas of British Columbia and from southern Alberta and Saskatchewan (Bik, 1967, 1968, 1969). They are known locally under the term “prairie mounds.” The mounds of the Fort St. John area number several thousand, with individuals ranging in average diameter from 100 to 300 ft (30-90 m) and average height from 5 to 20 ft (1.5-6 m). Extreme members reach 1500 ft (450 m) in diameter and 40 ft (12 m) in height. The prevalent form is a circular to oval miniature plateau, although a shallow summit crater is present in some of the mounds. Mounds may be clustered or separated by intervening flats. The mounds overlie and incorporate postglacial Iacustrine sediments (Peace Lake), indicating an origin following glacial retreat. The lake sediments are locally folded, suggestive of cryoturbation activity. A cap of post-


Peace Lake sediment occurs on several of the mounds, but is absent from the intermound tracts. Mathews considered the simple pingo scar origin, but favored a mechanism wherein water-saturated soil, rather than water alone, was mobilized beneath permafrost toward weak spots. Eruption of the soil then formed the mounds. The prairie mounds of the Sturgeon Lake area are morphologically similar to those of the Fort St. John area. However, the majority have a swampy central depression and are composed of till rather than deformed lacustrine sediments. The Alberta and Saskatchewan field contains many hundreds of thousands of prairie mounds. They are oval in plan, usually less than 600 ft (183 m) in diameter, and less than 30 ft (9 m) in height. The center of the mounds contains always one and frequently several depressions. The highest part of the mounds is a ring wall or rim around the central depression. The ring wall is always breached in at least one spot. The mounds occur commonly in fields, less commonly in groups of two to five, and rarely singlely . Earlier workers (e.g., Gravenor, 1955; Gravenor and Kupsch, 1959; Stalker, 1960) considered the mounds to have originated in contact with glacial ice. However, Bik (1969) argued that the breached rims and the occurrence of mounds over both glacial and proglacial deposits was inconsistent with any ice-contact mechanism of origin. He favored a periglacial mechanism similar to that offered by Mathews (1963). United States. In addition to the many pingo scars of Alaska, presumed pingo scars have been described from Illinois (Flemal, Hesler, and Hinkley, 1970; Flemal, 1972; Flemal, Hinkley, and Hesler, 1973) and a single presumed scar has been described from Indiana (Wayne, 1967). Fairbridge (1968a) has also sug-


gested that some depressions from New York and New Jersey (Wolfe, 1953) may be pingo scars. The Illinois scars number over 500 in an area of approximately 300 km2. They are confined to a low-relief till plain underlain by thick glacial deposits. The scars range from circular to elliptical in plan and from ring-walls to flat-topped plateaus in transverse section. The average diameter is several tens of meters, although several with diameters greater than 400 m and one elongate form with a major diameter of 1200 m have been noted. Rim heights range from less than 1 to more than 5 m. Clustering is common and both overlapping and superimposed forms occur. Annular depressions ring many of the scars. The scars consist largely of stratified lacustrine sediments, with maximum thicknesses in excess of 7 m. The sediments are mineralogically similar to the underlying till and outwash. A Wisconsinan age for the scars is based on the stratigraphic bracketing of the lacustrine sediments between an overlying cover of Wisconsinan loess and an underlying sequence of Wisconsinan till and glaciofluvial deposits. Summary The most characteristic form of a pingo scar is a regular circular to oval depression surrounded by a rampart and having horizontal dimensions in the range of lo-150 m. However, considerable variation from the ideal may occur. The regular outline may be disturbed by interfering successive generations of pingos. Elongation may occur to the point of producing ridge and furrow forms, particularly on slopes. The depression may be filled, or the adjacent ground may have undergone differential thermokarst lowering, such that the depression floor lies above adjacent areas and the form is plateau-like. Ramparts need not be obvious nor necessarily present due to ei-


ther deposition in the scar depression or erosion of the rampart. On sloping ground the rampart is typically absent in the upslope direction. Large scars may have diameters of several hundreds of meters, ranging perhaps to more than 1 km. Annular depressions may occur outside the rampart of the scar. Most reported pingo scars occur in fields of several hundreds to thousands of individuals. Within the fields they most commonly occur in clusters where densities may approach 100/km2. Solitary pingo scars have been reported. Ping0 scars may occur in any type of substrate. However, they are most prevalent in areas where the substrate is wet and particularly at places of emerging ground water. Plains, valley bottoms, and lower valley sides are the most common topographic environments. In some cases upward eruption can be demonstrated by the presence of upturned strata within the ramparts. Age of formation can be demonstrated by dating of the basal deposits within the depression or by stratigraphic bracketing. The most logical type of parent pingo for the majority of pingo scars is the open-system pingo. This conclusion is borne out by the observation that the majority of ping0 scars occur in overlapping and interfering relationships, suggesting multiple generation of pingos at a single locality. This is a common characteristic of open-system pingos (e.g., Miiller, 1963) but not of closed-system pingos. Since open-system pingos require less rigorous permafrost conditions than do closed-system pingos, this is also in accord with the assumption that permafrost was not as extensively developed along the Pleistocene ice margins as it is in some polar regions today. The degree to which p&as or other varieties of frost mounds, rather than pingos sense &i&o, may be the ancestral forms to presumed pingo scars is uncertain. The work of Svensson (1969) sug-



gests that palsa degeneration leads to forms very similar to those of pingo degeneration. It is possible that many of the smaller ping0 scars are actually palsa scars. This distinction is not critical for use of the scar forms as simple permafrost indicators, since both forms so indicate. However, it may be critical if attempts are made to determine temperature regimes of the former permafrost since these would not necessarily be identical. ALTERNATIVE MECHANISMS OF FORMATION Since the identification of a pingo scar is indisputable evidence for the former existence of permafrost, the use of pingo scars as permafrost indicators is predicated on the ability to correctly identify them. This may not be a straightforward matter, since there are several other known or presumed mechanisms which can produce forms resembling pingo scars. These may be placed into five catagories: I. Glacial processes A. Erosional forms B. Depositional forms 1. Ice-press forms 2. Ice-disintegration forms 3. Burial/melting forms II. Thermokarst processes A. Disintegration of segregated ice B. Disintegration of dispersed ice C. Eruptive forms III. Karst processes IV. Eolian processes V. Anthropogenic processes Glacial processes. Erosion of preexisting material by a glacier or its meltwaters is not known to have produced anything resembling the large fields of ping0 scars. However, milling by subglacial streams may be able to produce small numbers of depressions or ramparted forms under suitable conditions.


The crescent ridges of Lundquist (1969), a streamlined form transitional between drumlins and Rogen moraine, are also forms with some resemblence to pingo scars. These are insufficiently well known to further assess the similarity. Glacial depositional processes include several mechanisms capable of producing features resembling pingo scars. One possibility is the pressing or squeezing of sub-ice material into nearby tunnels, holes, or crevasses. Among features thought to have been formed in this way are the plains plateaus of Stalker (1960). In the ideal form the plains plateau is a circular or polygonal, flat-topped mound consisting of a till rim and a central fill of water-laid deposits. Some forms have little or no central fill, and hence are expressed as ramparted depressions. Diameters range from less than 15 m to more than 185 m and heights are typically 1.5-9 m. Their most common occurrence is in ground moraine regions behind hummocky end moraines, where they typically occur in groups of several hundred. In each of these characteristics, with the possible exception of association with hummocky moraine and the till rim, they resemble described pingo scars. The moraine plateaus of Stalker (1960) generally resemble his plains plateaus, with the exception that they are typically less numerous, larger (often in excess of 1 mile in diam), less regular in outline, and occur in morainal areas. A second possibility is differential deposition on a disintegrating, stagnant glacier. Among the many forms thought to be produced in this environment, that which most closely resembles the pingo scar is the ice-contact ring of Parizek (1969). This feature has also been variously known as a closed or circular disintegration ridge, doughnut, rimmed kettle, and prairie mound. Its origin has been discussed by Gravenor (1955), Gravenor and Kupsch (1959), Clayton (1967), and Parizek (1969), among oth-


ers. The general form is that of a circular ridge. The ridge may be locally breached and the interior may be filled such as to produce high-center forms. Dimensions are of the order of 15-150 m in diameter and 2-6 m in height. Most are composed of till with pockets of washed drift. The ice-walled lake plain of Clayton (1967) and Clayton and Cherry (1967) is another ice-disintegration form which may resemble a pingo scar. Cited examples of ice-walled plains are typically much larger and less regular than presumed pingo scars. Nevertheless, there is a possible overlapping in morphologic characteristics. Flemal, Hinkley, and Hesler (1973) reviewed the comparative characteristics of the two forms. Burial/melting processes, particularly the formation of kettles, can produce depressional forms of the type of general interest. These will not normally be ramparted, but can be confused with Cailleux nonramparted pingo scars. (1957) in particular has questioned whether many presumed kettles are not actually ping0 scars. Thermokarst processes. Thermokarst processes might produce circular forms in several ways. These include differential melting of massive segregated ice or dispersed ice and eruptive processes. The first class includes pingos and other varieties of frost mounds, as well as ice wedges. The distinction between scars of pingos and other frost mounds is not simple, but is also probably not critical for most purposes. Pingo scars would normally tend to be larger than the scars of other frost mounds. However, Salmi (1970) described one palsa with a long dimension of 85 m and palsas with diameters of 40 m are not uncommon (e.g., Lindqvist and Mattsson, 1965). These lie well within the size range of pingo scars. Disintegration of ice wedges offers at least two possibilities of interest. The first is formation of low-center polygon ponds, with inversion of relief upon


melting of the ice wedges. This mechanism might produce either ramparted forms, with the rampart composed of excessive material accumulated at the pond margins, or plateau forms if the ponds received uniform sedimentation. Such features are likely to be of lesser diameter and more regularly arrayed than most presumed pingo scars. Henderson (1959) offered a second alternative in which the initial form is a high-center polygon. According to the theory, pressure of freezing causes the development of central ice lenses as well as local upheaval and movement of watersaturated material toward the polygon centers. Subsequent melting of both the ice wedges and the ice lenses produces a ring-form. Henderson allowed that the polygons would have to be unusually large to produce features of the dimensions typical of ping0 scars. Differential melting of dispersed ground ice may produce depressions of the down-wearing variety (Czudek and Demak, 1970). These span a broad range of shapes and sizes, including some of pingo scar outlines and dimensions. They often also contain lakes or are the sites of subaerial deposition, In the modern permafrost environment some regular thaw lakes are not readily distinguishable from old pingo scars (Holmes, Hopkins, and Foster, 1968). The effects produced on a thaw depression by complete disintegration of the permafrost is subject to the same uncertainty as that concerning the passage of pingo scars through a total permafrostdisintegration stage. Jahn (1970) considered that thaw hollows will preserve their depression form. However, it may be that the ultimate disposition of many thaw depressions is inversion of relief to ring-wall or even plateau forms. This follows if the original ground ice is uniformly distributed and the thaw depressions receive any sedimentation while active. All areas will thus eventually




hams (1973). In most cases anthroposubside an equal amount and the deposits genie structures are less regular and of of the depression will stand in relief against the adjacent terrain. In the case lesser density than presumed pingo scars. However, some activities such as marl where deposition was predominantly pitting (Prince, 1964) may produce both against the depression margins a ring-wall regular and high density structures. might result; if deposition was uniform over the depression floor a plateau form Summary might result. Mathews (1963) and Bik (1969) sugMany processes other than pingo degested that the prairie mounds of western generation can produce regular ground Canada were formed under permafrost forms of the type under discussion. No conditions by an eruptive mechanism in- single criterion allows these to be disvolving flow of supersaturated till into tinguished from pingo scars. Rather, areas of low stress. This mechanism is suites of criterion must be employed and therefore like the ice-press mechanism, these are different and not always certain but involves ground ice rather than gla- for each comparison. cial ice stresses. Other eruptive processes As a rule, a periglacial-thermokarst of thermokarst terrains, such as suffosion origin as opposed to glacial origin can be bursts (Paterson, 1940) and formation of assumed if it can be established that the tundra craters (Jahn, 1948) could lead to structure in question originated either forms resembling pingo scars. As Fair- outside the limits of glaciation or over bridge (1968a) pointed out, little attenpro- or postglacial deposits. An exception has been paid to these phenomena tion exists for kettle holes. The case can and hence it is difficult to assess further be strengthened if there is independent the forms they might produce. proof of the existence of permafrost in Karst processes. Formation of karst the region at the time the structure origsinkholes must be considered as an alterinated. Absence of till in the structure, native to pingo scar development in areas except from the rampart, is also strongly underlain by surficial or near-surface sol- suggestive of a periglacial origin. uable rock. Sparks, Williams, and Bell It is worth noting that there is a major (1972) discussed some of the varieties of imbalance in the literature on the quessolution forms which resemble presumed tion of glacial versus periglacial origin of ping0 scars. circular forms. European literature very Eolian processes. Eolian processes may strongly favors the periglacial origin be responsible for some circular mounds, whereas North American literature is such as the prairie mounds of Arkansas dominated by various glacial mecha(Quinn, 1968). On the basis of loessal nisms. It may well be that this is refleccomposition these should not be easily tive of differences in processes in the two confused with pingo scars. regions. But it may also be reflective of An thropogenic processes. Anthropopreconditioning to one or the other genie structures of many origins may re- model, and therefore suggests that some semble pingo scars. These include bomb reevaluation in both regions is warranted. craters, ancient silage pits, well borings, Differentiation of pingo scars from cisterns, cellars, and a variety of quarries other varieties of thermokarst features is and barrow pits. Characteristics of these most difficult. Form and size are not structures as compared to pingo scars definitive, although a pingo scar origin is favored for regular, ramparted forms in have been reviewed by Pissart (1958), Prince (1961,1962, 1964), Troll (1962), the dimension range of a few tens of Gulley (1962), Wiegand (1965), and Wil- meters to a few hundreds of meters. Per-


haps the best criteria are upwarping of strata beneath a rampart or fabrics otherwise indicative of an initial upwarping, and overlapping and super-positioning of individuals indicative of multiple generation of forms. Location on valley sides is also suggestive, as is disposition of the substrate-sediment contact below the level of the adjacent terrain. Possible modes of origin other than glacial or periglacial must also be considered. CONCLUSIONS 1. Pingo scars are indisputable evidence of former or present permafrost. 2. Pingo scars occur interspersed with active pingos in broad areas of the modem permafrost regions. They have also been tentatively identified in permafrost areas where active pingos no longer occur and in many areas of former permafrost. 3. Identification of pingo scars in former permafrost regions is a relatively recent development in the field of periglacial studies. Within the last 20 yr many such pingo scars have been recognized, most abundantly in northwest Europe but also in parts of North America and northern Asia. It is likely that further investigation will reveal many more occurrences. 4. Pingo scars occur in a morphologic spectrum which ranges from simple closed depressions through ramparted depressions to miniature plateau forms. The detailed morphology is a function of many variables, including ground slope, types and characteristics of substrate and surficial materials, type of parent ping0 or other frost mound, degree of thermokarst overprint, and postformation history of erosion and deposition. 5. Ideal pingo scars are highly symmetrical in plan (circular to oval) and have horizontal dimensions in the range of about lo-200 m. Variant forms occur due to multiple generation and interference of individual forms as well as orig-


inal nonsymmetry of the parent pingos. Unusually large ping0 scars may range upward of 1 km in diameter. 6. Pingo scars typically contain an infilling of water-laid deposits. These result from the presence of a lake or bog occupying the pingo scar center. Common materials are silt, clay, marl, gyttja, and peat. Study of these sediments can provide important information on the time of formation and history of the scars. 7. Ping0 scars may occur in any type of material. They are probably best developed in water-saturated and loosely consolidated surficial materials, but may occur in bedrock. 8. Open-system pingos appear to be the ancestral form of the majority of described pingo scars. 9. Of the pingo scars in former permafrost regions, the majority appear to have formed during the last stage of the last glaciation or during the last deglaciation. No pingo scars of pre-Wisconsinan time have yet been identified. 10. Distinction of pingo scars from forms of other possible origins requires both detailed knowledge of the presumed pingo scars as well as further exploration of the morphologic expression to be expected from forms produced by other processes. Pingo scars are likely to be most difficult to distinguish from scars of other thermokarst features. 11. Identification of pingo scars offers much promise for delimiting former permafrost areas and conditions. Further investigation of both pingos and pingo scars is therefore deemed to be highly desirable. ACKNOWLEDGMENTS The author wishes to express his appreciation to Sharon Ervin for assistance with the bibliographic search, to Karin Klingelhofer, Hans Knoll, and Stanley Frost for assistance with translations, and to Robert Morris, Sidney White, Troy P&w&, and James Benedict for offering critical readings of the manuscript.



Bik, M. J. J. (1967). On the periglacial origin of prairie mounds. In “Glacial Geology of the Missouri Coteau” (L. Clayton and T. F. Freers, Eds.), pp. 83-94. North Dakota Geological Survey Miscellaneous Series 30. Bik, M. J. J. (1968). Morphoclimatic ohservations on prairie mounds. Zeitschrift fiir Geomorphologie 12,409-469. Bik, M. J. J. (1969). The origin and age of the prairie mounds of southern Alberta, Canada. Biuletyn Peryglacjalny 19, 85-130. BoyB, M. (1957). Clots, langiies et lagunes de la Lande girondine. Acaddmie des Sciences, Comptes Rendus Hebdomadaires des Skances (Paris) 244,1058-1060. Boy& M. (1958). Laguny Land6w. Biuletyn Peryglacjalny 6, 29-56. Brown, R. J. E., and PQwB, T. L. (1973). Distribution of permafrost in North America and its relationship to the environment: A review, 1963-1973. In “Permafrost: The North American Contribution to the Second International Conference,” pp. 71-100. National Academy of Sciences, Washington, D.C. Bruning, H. (1972). Das Rhein-Main-Gebiet in den quartaereiszeitlichen Periglazialbereichen. Oberrheinischer Geologischer Verein, Jahresberichte und Mitteilungen (Stuttgart) 54, 79-100. Cailleux, A. (1956). Mares, mardelles et pingos. Acadimie des Sciences, Comptes Rendus Hebdomadaires des Skances (Paris) 242,19121914. Cailleux, A. (1957). Les mares du sud-est de Sjaelland (Danemark). Acade’mie des Sciences, Comptes Rendus Hebdomadaires des Siances (Paris) 245, 1074-1076. Cailleux, A. (1960). Sur les mares et lacs ronds des plaines aujourd’hui temperees. Revue de Geomorpholgie Dynamique 11, 28-29. Cailleux, A. (1961). Mares et lacs ronds et loupes de glace du sol. Biuletyn Peryglacjalny 10,35-41. Clayton, L. (1967). Stagnant-glacier features of the Missouri Coteau in North Dakota. In “Glacial Geology of the Missouri Coteau” (L. Clayton and T. F. Freers, Eds.), pp. 25-46. North Dakota Geological Survey Miscellaneous Series 30. Clayton, L., and Cherry, J. A. (1967). Pleistocene superglacial and ice-walled lakes of westcentral North America. In “Glacial Geology of the Missouri Coteau” (L. Clayton and T. F. Freers, Eds.), pp. 47-52. North Dakota Geological Survey Miscellaneous Series 30. Czudek, T., and Demek, J. (1970). Thermokarst in Siberia and its influence on the de-

C. FLEMAL velopment of lowland relief. Quaternary Research 1, 103-120. Dylik, J. (1963a). Traces of thermokarst in the Pleistocene sediments of Poland. Bull. Sot. Sci. et Lettres Lodz 14, 16 pp. Dylik, J. (196313). Nowe problemy wiecznej zmarzliny plejstocenskiej. Acta Geographica Lodziensia 17, 93 pp. Dylik, J. (1964). Periglacial investigations and their significance for paleogeography. Polish Academy of Sciences, Review (Warsaw) 9, 14-32. Dylik, J. (1965). L’Btude de la dynamique d’Qvolution des dhpressions fermees a Jozefow aus environs de Lodz. Revue Geomorphologie Dynamique 15,158-171. Dylik, J. (1973). Znaczenie baden wiecznej zmarzliny dla pelniejszego poznania peryglacjalnej morfogenezy w Polsce plejstocenskiej. Czasopismo Geograficze 44, 207-215. Dylikowa, A. (1964). Etat des recherches periglaciaires en Pologne. Biuletyn Peryglacjalny 14, 41-60. Fairbridge, R. W. (1968a). Glacial geology: Periglacial and global effects. In “The Encyclopedia of Geomorphology” (R. W. Fairbridge, Ed.), pp. 440-444. Reinhold, New York. Fairbridge, R. W. (196813). Suffosion and tundra craters. In “The Encyclopedia of Geomorphology” (R. W. Fairbridge, Ed.), pp. 1099-1100. Reinhold, New York. Flemal, R. C. (1972). Ice injection origin of the DeKalb mounds, north-central Illinois, U.S.A. In “Proceedings, 24th International Geological Congress, Montreal,” Sect. 12, pp. 130-135. Flemal, R. C., Hesler, J. L., and Hinkley, K. C. (1970). The DeKalb mounds: Possible remnants of pingos. In “Thirty-fourth Annual Tri-State Field Conference Guidebook” (I. E. Odom and M. P. Weiss, Eds.), pp. 65-72. Northern Illinois University, DeKalb. Flemal, R. C., Hinkley, K. C., and Hesler, J. L. (1973). DeKalb mounds: A possible Pleistocene (Woodfordian) pingo field in northcentral Illinois. In “The Wisconsinan Stage” (R. F. Black, R. P. Goldthwait, and H. B. Willman, Eds.), pp. 229-250. Geological Society of America Memoir 136. Frenzel, B. (1959). Die Vegetationsund Landschaftszonen Nord-Eurasiens wghrend der letzten Eiszeit und wghrend der postglazialen Warmeziet-I. Teil: Allgemeine Grundlagen. Abhandlungen der Akademie der Wissenschaften und der Literature in Mainz, Mathematisch-Naturwissenschafliche Klasse 13,937-1099.



Friedman, J. D., Johansson, C. E., Oskarsson, Environment Past and Present” (T. L. Pew& N., Svensson, H., Thorarinsson, S., and WilEd.), pp. 203-215. McGill-Queens University liams, R. S. (1971). Observations on Icelandic Press, Montreal. polygon surfaces and palsa areas. Photo interLundqvist, J. (1969). Problems of the so-called pretation and field studies. Geografiska Rogen moraine. Sveriges Geologiska UnderAnnaler 53-A, 115-145. siikning, Arsbok (Stockholm) 64, l-32. Gawlik, H. (1970). Rola procesow perglacja/ Maarleveld, G. C. (1965). Frost mounds, a nych w rozwoju rzeiby kotliny Szczercowsummary of the literature of the past decade. skiej. Acta Geographica Lodziensia 24,165Mededelingen van de Geologische Stichting, 179. Nieuwe Serie 17,7-20. Gravenor, C. P. (1955). The origin and signifMaarleveld, G. C., and van den Toorn, J. C. icance of prairie mounds. American Journal Pseudo-solle in Noord-Nederland. (1955). of Science 253, 475-481. Tijdschrift Koninklijk Nederland. AardrijkGravenor, C. P., and Kupsch, W. 0. (1959). sundig Genoot 72,344-360. Ice-disintegration features in western Canada. Mackay, J. R. (1962). Pingos of the Pleistocene Journal of Geology 67,46-64. Mackenzie Delta area. Geographical Bulletin Gulley, J. L. M. (1962). Le marnage dans le l&20-63. bassin parisien. Revue Geomorphologie DyMackay, J. R. (1973). The growth of pingos, namique 13, 26-30. western arctic coast, Canada. Canadian JourHenderson, E. P. (1959). “Surficial geology of nal of Earth Sciences 10, 979-1004. the Sturgeon Lake Map Area.” Canada GeoMackay, J. R., and Black, R. F. (1973). Origin, composition, and structure of perennially frological Survey Memoir 303, 180 pp. zen ground and ground ice: A review. In Holmes, G. W., Foster, H. L., and Hopkins, D. M. (1966). “Permafrost: The North American ContribuDistribution and age of pingos tion to the Second International Conference,” of interior Alaska. In “Proceedings of Interpp. 185-192. National Academy of Sciences, national Conference on Permafrost, Lafayette, Washington, D.C. Indiana, 1963,” pp. 88-93. NAS-NRC PubMathews, W. H. (1963). “Quaternary Stratig lication No. 1287. raphy and Geomorphology of the Fort St, Holmes, G. E., Hopkins, D. M., and Foster, John Area, Northeastern British Columbia.” “Pingos in Central Alaska.” H. L. (1968). British Columbia Department of Mines and United States Geological Survey Bulletin Petroleum Resources, Victoria. 1241-H, 40 pp. Mitchell, G. F. (1971). Fossil pingos in the Hughes, 0. L. (1969). “Distribution of Opensouth of Ireland. Nature (London) 230, System Pingos in Central Yukon Territory with Respect to Glacial Limits.” Canada Geo43-44. Mitchell, G. F. (1973). Fossil pingos in logical Survey Paper 69-34, 8 pp. Jahn, A. (1948). Research on the structure and Camaross Townland, Co. Wexford. Irish Royal Academy Proceedings Sect. B 73, 269temperature of the soils in the western Greenland. Academic Polonaise des Sciences et 283. Milckenhausen, E. (1960). Eine besondere Art Lettres (Krakow) Bull. Ser. A, 1940-1946, 50-59. von Pingos am Hohen Venn/Eifel. Eiszeitalter Jahn, A. (1968). Patterned ground. In “The und Gegenwart 11, 5-11. Encyclopedia of Geomorphology” (R. W. Mullenders, W., and Gullentops, F. (1969). Fairbridge, Ed.), pp. 814-817. Reinhold, The age of the pingos of Belgium. In “The New York. Periglacial Environment Past and Present” Jahn, A. (1970). “Zagadnienia Strefy Per(T. L. P&we, Ed.), pp. 321-335. McGillglacjalnej.” Panstwowe Wydawnictwo NauQueen’s University Press, Montreal. kowe, Warsza. Mullenders, W., and Haesendonck, F. (1963). Kaiser, K. (1960). Klimazeugen des periglazialen Note preliminaire sur la palynologie des pinDauerfrostbodens in Mittel- und Westeuropa. gos de Plateau des Tailles (Belgique). ZeitEin Beitrag zur Rekonstruktion des Klimas schrift fiir Geomorphologie 7,165-168. Observations on pingos der Glaziale des quartiiren Eiszeitalters. EisMiiller, F. (1963). (Beobachtungen iiber Pingos). Canada Nazeitalter und Gegenwart LL, 121-141. tional Research Council Technical Translation Lindquist, S., and Mattsson, J. 0. (1955). Studies on the thermal structure of a pals. 1073,117 pp. Miiller, F. (1968). Pingos modern. In “The Lund Studies in Geography, Series A, 34, Encyclopedia of Geomorphology” (R. W. 38-39. Reinhold, Fairbridge, Ed.), pp. 845-847. Lundqvist, J. (1969). Earth and ice mounds: A terminological discussion. In “The Periglacial New York.



Nossin, J. J. (1961). De Reliefontwikkeling in Zuidoostelijk Friesland. In “Boven-Boorngebied.” Publication 178 of the Frisian Academy, Drachten, Netherlands. O’Brien, Robert (1971). “Observations on Pingos and Permafrost Hydrology in Schuchert Dal, N. E. Greenland.” Meddelelser om Gronland, Bd. 195, Nr. 1, 20 pp. Paepe, R, (1968). The geometry of geomorphological forms as related to Late Pleistocene stratigraphy. Societe Beige de Geologie, de Paleontologie et d ‘Hydrologie Bulletin 77, 22-38. Parizek, R. R. (1969). Glacial ice-contact rings and ridges. In “United States Contributions to Quaternary Research” (S. A. Schumm and W. C. Bradley, Eds.), pp. 49-102. Geological Society of America Special Paper 123. Patterson, T. T. (1940). The effects of frost action and solifluxion around Baffin Bay and in the Cambridge district. Quarterly Journal of the Geological Society of London 96, 99-130. Pew&, T. L. (1973). Ice wedge casts and past permafrost distribution in North America. Geoform 15,15-26. Reste von Pingos bei Picard, K. (1961). Husum/Nordsee. Naturwissenschaftlicher verein fiir Schleswig-Holstein, Kiel 32, 72-77. Pissart, A. (1956). L’orgine periglaciaire des viviers des Hautes Fagnes. Annales de la Socikti Giologique de Belgique 79, 119-131. Pissart, A. (1958). Les depressions fermees dans al region parisienne. Le problem de leur origine. Revue Geomorphologie Dynamique 9,73-83. Pissart, A. (1960). Les depression fermees de la region parisienne. Les difficult& d’admettre une origine humaine. Revue de Geomorphologie Dynamique 11, 12. Pissart, A. (1963). Les traces de ‘pingos’ du Pays de Galles (Grande-Bretagne) et du plateau des Hautes Fagnes (Belgique). Zeitschrift fiir Geomorphologie 7,147-165. Les pingos des Hautes Pissart, A. (1965). Fagnes: Les problemes de leur genese. Annales de la Societi Giologique de Belgique 88,277-289. Pissart, A. (1967). Les pingos de 1’Ile Prince Patrick (76ON-120°W). Geographical Bulletin 9,189-217. Pissart, A. (1968). Pingos, Pleistocene. In “The Encyclopedia of Geomorphology” (R. W. Fairbridge, Ed.), pp. 847-848. Reinhold, New York. Ploeger, P. L., and Groeman-van Watteringe, W. (1964). Late glacial pingo and valley development in the Boorne region near Wijnjeterp,

C. FLEMAL Province of Friesland, Netherlands. Biuletyn Peryglacjalny 13,199-233. Polak, B. (1963). The sub-soil of the Blecke Meere, compared to the fluvioglacial deposit of Speulde; a contribution to the problem of thermophilous and exotic pollen in late-glacial sediments. Netherlands, Geologische Stitchting, Mededelingen, Niewe Serie 16, 39-47. Porsild, A. E. (1938). Earth mounds in unglaciated arctic northwestern America. Geographical Reviews 28, 46-58. Prince, H. C. (1961). Some reflections on the origin of hollows in Norfolk compared with those in the Paris regions. Reuue Geomorphologie Dynamique 12, 110-117. Prince, H. C. (1962). Pits and ponds in Norfolk. Erdhunde 16,10-31. Prince, H. C. (1964). The origins of pits and depressions in Norfolk. Geography 49,15-32. Quinn, J. H. (1968). Prairie mounds. In “The Encyclopedia of Geomorphology” (R. W. Reinhold, Fairbridge, Ed.), pp. 888-889. New York. Rapp, A., and Rudberg, S. (1960). Recent periglacial phenomena in Sweden. Biuletyn Peryglacjalny 8, 143-154. Riviere, A., and Vernhet, S. (1962). Accidents periglaciaires dans la region de Penestin (Morbihan). Acadimie des Sciences, Comptes Rendus Hebdomadaires des Skances (Paris) 255,744-746. Salmi, M. (1970). Investigations on palsas in In “Ecology of the SubFinnish Lapland. arctic Regions, Proceedings of the Helsinki Symposium,” pp. 143-153. UNESCO, New York. Schneider, F. K. (1970). Fossile Pingos in der Niederrheinischen Bucht. Fortschritte in der Geologie uon Rheinland und Westfalen 17, 333-342. “Pingo-like Remnants in Seppala, M. (1972). the Peltojaarvi Area of Finnish Lapland. Turku, Universitet, Institutem Geographicurn, Publication 55, 8 pp. Slotboom, R. T. (1963). Comparative geomorphological and palynological investigation of the pingos (viviers) in the Hautes Fagnes (Belgium) and the mardellen in the Gutland (Luxemburg). Zeitschrift fur Geomorphologie 7,193-231. Sparks, B. W., and West, R. G. (1972). “The Ice Age in Britain.” Metheun, London. Sparks, B. W., Williams, R. B. G., and Bell, F. G. (1972). Presumed ground-ice depressions in East Anglia. Proceedings of the Royal Society of London, Series A, 327,329-343. Stalker, A. Mac. S. (1960). Ice-pressed drift forms and associated deposits in Alberta. Canada Geological Survey Bulletin 57,38 pp.

PINGOS Strugov, A. S. (1955). Vzryv girdrolakkolita. Priroda 44, 117. Svensson, H. (1962). Observations on palsas. Photographic interpretation and field studies in North Norwegian frost ground areas. Norsk Geogmfish Tidsskrift 18, 212-227. Svensson, H. (1964). Traces of pingo-like frost mounds. Svensk Geografisk Arsbok (Lund) 40, 93-106. Svensson, H. (1969). A type of circular lake in northernmost Norway. Geografiska Annaler 51A, 1-12. Svensson, H. (1970). Frozen-ground morphology of northeasternmost Norway. In “Ecology of the Subarctic Regions, Proceedings of the Helsinki Symposium,” pp. 161-168. UNESCO, New York. GlaTroll, C. (1962). ‘Solle” and “mardelles,” cial and periglacial phenomena in continental Europe. Erdkunde 16,31-34. Trotman, D. M. (1963). Peat deposits within a ping0 near Llangurig, Wales. Zeitschrift fiir Geomorphologie 7,168-171. Washburn, A. L. (1973). “Periglacial Processes and Environments. ” Edward Arnold, London, 320 pp. Remains of pingos in Watson, E. (1971). Wales and the Isle of Man. Geological Journal 7,381-392.


Watson, E. (1972). Pingos of Cardiganshire and the latest ice limit. Nature (London) 236, 343-344. Watson, E. and Watson, S. (1972). Investigations of some pingo basins near Aberystwyth, Wales. In “Proceeding 24th International Geological Congress, Montreal,” Sect. 12, pp. 212-223. Wayne, W. J. (1967). Periglacial features and climatic gradient in Illinois, Indiana, and western Ohio, east-central United States. In “Quaternary Paleoecology” (E. J. Cushing and H. E. Wright, Jr., Eds.), pp. 393-414. Yale University Press, New Haven, Conn. Fossile Pingos in MitWiegand, G. (1965). teleuropa. Wunberger Geographische Arbeiten 16, 152 pp. Williams, J. R., and van Everdigen, R. 0. Groundwater investigations in per(1973). mafrost regions of North America: A review. In “Permafrost: the North American Contribution to the Second International Conference,” pp. 435-446. National Academy of Sciences, Washington, D.C. Williams, R. N. G. (1973). Frost and the works of man. Antiquity 47, 19-31. Wolfe, P. E. (1953). Periglacial frost-thaw basins in New Jersey. Journal of Geology 61, 133-141.