A.M. Karpenko, 2022

Institute of Geological sciences of the National Academy of Sciences of Ukraine, Kyiv, Ukraine

Weak links in the existing models of Kaniv dislocations formation are shown. A new concept of a complex of factors, conditions and processes of paragenesis of buried overdeepened depressions and glaciodislocations of the Kaniv type in continental glaciation areas has been proposed and substantiated. The theory of glacioisostasis is involved to the construction of the glaciotectonic model, in particular, the theory’s statement about occurrence of a temporary compensatory glacioisostatic elevation of relief in the form of a roll before the glacier front, subparallel to its edge. In the core of the roll rocks, previously buried at a considerable depth, will be elevated above the local erosion base. When the glacier’s movement slows or stops, its melt water flowing out from the periglacial lake through the low section of the roll crest, as the roll rises, will develop an antecedent valley across the roll - a breakthrough with steep slopes, cutting through the rocks raised from the depths. As the glacier accelerates and reaches the roll, this breakthrough gives the glacier tongue the shape of a wedge (in the horizontal projection). With further increase in pressure of the ice mass glacier wedge with its lateral surface tears off the rock blocks from the slopes of the breakthrough and moves them forward and to the side. Subsequently the compensatory roll is sinking (is smoothed out), and the erosion-exaration breakthrough turns out below the level of the local erosion basis, is buried and transforms into an overdeepened depression.

Key words: glaciodislocations; overdeepened depression; glacioisostasis; compensation roll; erosion; breakthrough; glacial wedge; exaration.

Introduction. The Kaniv mountains are a unique natural object with a mysterious origin, which has been discussed by geologists, geographers, and naturalists for more than a hundred years. These mountains are composed of crumpled, deformed layers of rocks that lie at depths of up to a hundred metres around them, and in the mountains are raised more than three hundred metres upwards. More than a hundred scientific papers have been devoted to these dislocations and the mechanism of their formation, and dozens of versions of how they were formed have been proposed by researchers.

Almost from the very beginning of the stage of detailed geological study of the Kaniv dislocations in the first half of the last century, their genetic connection with another outstanding geological object discovered nearby - a depression buried under thick Quaternary sediments in the bed of Paleogene and Mesozoic rocks, later named by G.I. Horetskyi [1970] Shevchenko Scour of glacial ploughing and glacial erosion - became clear.

In our opinion, the key to reconstructing the mechanism of formation of the Kaniv dislocations can be found in an adequate explanation of the genesis of the so-called ”overdeepened valleys” (overdeepened depressions) of the Shevchenko type.

Buried overdeepened valleys of the Russian Plain, concepts of their formation. The widespread occurrence of deep scouring in the glacial region of the European part of the USSR was noted by B.L. Lychkov [1942]. He considered the fragments of buried valleys discovered here to be traces of an ancient (pre-Quaternary) buried river network and even tried to reconstruct it. Initially, other researchers believed the same, supplementing the picture of the distribution of fragments of buried overdeepened valleys with new factual data, for example, E.V. Rukhyna [1957]. This point of view prevailed for a long time. But eventually, based on the data of detailed drilling coverage of the areas of development of such valleys, G.I. Horetskyi [1967, 1973] showed one of the most important features of their structure - they turned out to be elongated but closed.

This is the first argument for our version of the mechanism of their formation.

Other arguments are as follows.

Deep valley-like depressions in the relief of the Quaternary sediment bed are common also in watersheds and watershed slopes, where their depth reaches 300 m; the location of valley-like depressions in the plan view in most cases does not correspond to the modern pattern of the hydrographic network [Horetskyi, 1967].

In many cases, the bottom of these valleys is located tens or even hundreds of metres below the modern sea level.

G.I. Horetskyi [1967] noted a peculiarity of some of these depressions: the presence of narrow canyon-like, as he calls them, ”gullies of glacial runoff and erosion” at their base. They are several tens of metres deep below the moraine’s sole and are filled mainly with sands, sometimes with gravel and pebbles.

The cross-sectional profile of the valley-like depressions is V- and U-shaped (”gully” and ”trough” types of glacial depressions respectively - [Horetskyi, 1980]); the slopes are steep, sometimes steep-to, and the width of the valleys is small (E.V. Rukhyna [1957] in the Moscow area indicates a width of 1 to 2.5 km).

E.V. Rukhyna [1957] notes the typical properties of the Quaternary sediments that bury the valleys. Glacial formations account for the largest share in their composition. The lower moraine (there are up to three horizons) usually lies on bedrock, and there is no thick complex of alluvial deposits beneath it (gravel and pebble underlying sands of low thickness are found). The absence of alluvial deposits at the bottom of overdeepened valleys is also confirmed by G.I. Horetskyi [1967]. He also draws attention to the almost universal accompaniment of deep valley-like depressions by glaciodislocations, by inclusion in the moraine of numerous detached masses, ranging in size from small lumps to large blocks with a diameter of several tens of metres [Horetskyi, 1968, 1972].

The features pointed out by E.V. Rukhyna indicate the formation of valleys as a result of vigorous rapid incision caused, as she believes, by uplift of the territory; active erosion continued until the very beginning of glaciation. After that the territory was levelled.

Based on extensive factual material, G.I. Horetskyi [1967] denies the purely erosive origin of the overdeepened valleys, especially since they turned out to be closed depressions rather than buried river valleys, and proposes a version of their formation by the joint action of glacial exaration and glacial erosion.

The presence on modern watersheds and valley slopes of numerous and huge erratic masses and dislocated massifs involving those ancient rocks which lie at considerable depths in the glaciation zone (and which, however, are reached by nearby overdeepened depressions) indicates that masses of these rocks were torn away near the bottom of the depressions and brought to the surface by some powerful force.

The history of the discovery and study of the Kaniv dislocations and the Shevchenko Basin shows that initially researchers quite logically linked such grandiose phenomena to tectonic processes. However, the participation of glacial formations in the structure of the sedimentary strata, that fills the overdeepened depressions, gave rise to the version of another group of researchers - about the purely glacial origin of the overdeepened depression (Shevchenko), and therefore, the related with it Kaniv dislocations. What is obvious in this version is the exaggeration of the bulldozing (ploughing) capabilities of the glacier, attributing to the plastic-viscous movement of the glacier on flat terrain a dubious ability to have a powerful vertical component: at first wedge in into the bedrock to a considerable depth, and then emerge again to the surface, and even with huge masses of rock removed from depths of a hundred metres.  (A.V. Matoshko and Yu.G. Chuhunnyi [1993] estimate the normal depth of exaration of the Dnieper glacier to be a maximum of several meters, and the incredible depth of exaration in the Shevchenko Basin they call apotheous, without explaining the reasons for this anomaly). That is why one more the concept - of the two-factor origin of the Kaniv and other dislocations appeared. D.K. Bilenko [1939] mentioned the position of V.M. Chyrvynskyi on these issues, who simply and accurately stated that for glaciated dislocations to occur, one must assume the existence of an uplift within the moraine terrace before the glacier arrived; otherwise, the appearance on the flat terrain of such isolated hills through dislocating by glacier becomes unclear.

However, after further research failed to confirm the existence of tectonic structures in the Kaniv area, only the glacial component of the hypothesis remained valid and the hypothesis’ persuasiveness was reduced. In addition, M.H. Kostianoi [1962] carried out engineering and geological calculations of the tangential forces of the glacier in the Kaniv area. "The results of calculations show that the hypothesis of glaciodislocations in an elementary form does not find justification" - came to the conclusion of M.H. Kostianoi [1962, p. 81]. Feeling this flaw of the ”glacial” remnant of the hypothesis of the formation of a paragenetic pair "depression-dislocations", researchers began to add to it versions of various processes:  washing out of rock from under the glacier by subglacial waters; freezing of substrate blocks to the glacier sole and their subsequent movement to the surface together with the glacier; extrusion of incompetent substrate material into various cavities in the glacier body; squeezing of rocks, that have acquired a plastic consistency, from under the glacier edge under the weight of the glacier, etc.

For example, the authors of the most recent special study of the Kaniv dislocations opted for a model of their formation that assumes glacial pressure on the substrate to be so great that it caused ”active progressive movement of the rocks of the glacial bed under the influence of glacial loads” [Lavrushyn and Chuhunnyi, 1982, p. 92]. (Here Y.A. Lavrushyn and Yu.G. Chuhunnyi [1982] resorted to the model proposed by F. Wahnschaffe [1882] 100 years before - the circle is closed). But along the entire glacier path, at least along most of it, the geological conditions of the glacier bed often change only slightly, so this mechanism of glaciodislocations formation would have resulted in their almost continuous presence along most of its path; however, instead, we have a very limited distribution their in some relatively small segments of this long line - and this selectivity, having no explanation, makes the Wahnschaffe-Lavrushyn-Chuhunnyi version unconvincing.

Also and other additions did not help to create a coherent and convincing model of the paragenesis of such objects as the Kaniv dislocations and the Shevchenko overdeepened depression. And the same weakest link of all concepts is due to the search for a version of the mechanism within the framework of an unchanged postulate: a huge mass of rocks (up to 50 km³ in volume - [Matoshko and Chuhunnyi, 1993]), which was clamped on all sides at a depth of more than a hundred metres, was torn off and raised to a height of about two and a half hundred metres. One of the researchers wrote in this regard: ”It is known that the specific gravity of ice is less than 1 kG/dm³, while the specific gravity of sedimentary rocks is on average 2.5 kG/dm³. Thus, in order to compensate for the 270 metre rise in rock, the ice cover would have to be 675 metres high, and in a small area. But even if this condition was met, it is impossible to imagine how the glacier, whose bed was at levels close to 100 m above sea level, in a strip up to 10 km wide, plunged and tore a piece of rock from a depth of 130 m. It is unclear what forces must have acted to break and move the huge mass. And these two stages require maximum effort, since not only the force of inertia but also the force of adhesion is at work here” [Pazynych, 2007, p. 196].

In his major work [1970], G.I. Horetskyi in the most detailed way considers the history of research, structure, history of formation and development of the Shevchenko basin of ”glacial ploughing and erosion” but avoids considering the mechanism of formation of the Kaniv dislocations. However, due to the genetic relationship between the Kaniv dislocations and the Shevchenko basin, there should be a single convincing version of the mechanism of their joint formation. G.I. Horetskyi apparently was sensing that his hypothesis of glacial ploughing and glacial erosion was not such.

Thus, the main ways of glaciodislocations formation are now considered to be: disruption and movement of large blocks of rocks under the pressure of an advancing glacier (”bulldozer” effect, or pushing); squeezing of pliable rocks from under the glacier edge - squeezing; injection of loose plastic material into cavities in the ice thickness [Ananiev et al., 1992].

Compensation roll as a key node in the concept of the paragenesis of glaciodislocations and overdeepened depression. Researchers in vain have completely rejected the tectonic component of the hypothesis of the complex origin of the Kaniv dislocations and the Shevchenko overdeepened depression.

None of them, from the beginning of the research to the present day, has been able to assume that the tectonic uplift in the Kaniv area could have been temporary and not left any distinct traces in the earth’s crust. For example, S.L. Breslav wrote [1971, p. 636]: ”At the same time, it is difficult to explain such a significant deepening by water erosion alone. In this case, it is necessary to assume either an extremely sharp decrease in the overall erosion base or unusually large amplitudes of tectonic uplifts, which for some reason were subsequently replaced by equally powerful movements of the opposite sign. Both assumptions do not currently have any reliable justification”.

The latter is not true. At that time the theory of glacioisostasy already existed, which describes tectonic movements just of this nature.

The idea was proposed by T.F. Jamieson [1865] in the second half of the 19th century, and in the first third of the 20th century it was already widely accepted, and it is somewhat surprising that even in later times, domestic researchers of the Quaternary continental glaciation of the East European plain did not widely apply its provisions (only V.V. Riznichenko in his works of the early 1930s pointed to the possibility of oscillatory movements of the crust associated with glacial loading and unloading; some attempts to apply the theory of glacioisostasy to explain some features of the relief structure of glacial and extraglacial areas can be found in the works of B.L. Lychkov [1927, 1928, 1931, 1942, 1944]).

One of these provisions is the phenomenon of the earth’s surface uplift zone (compensation roll) along the leading edge of the glacioisostatic deflection area. As early as 1874, N.S. Shaler [1874] suggested the existence of such a phenomenon as a consequence of squeezing out from under the glacier of plastic substance in the earth’s shell, which lies below the crust. Later, this concept was developed by F. Nansen [1922, 1928] (who noted that after deglaciation the roll-like uplift should sink) and R.A. Daly [1934]. To a certain extent, the theoretical position on the compensatory glacioisostatic roll was developed in articles by B.L. Lychkov and in monographs [Nazarov, 1971; Levkov, 1980].

After a number of investigators in 60-80s of the last century managed to establish geological and geomorphological evidence of existence of compensating rolls in glacial regions of the North America (Anderson R.C.; Bloom A.L.; [Newman et al, 1971]; Walcott R.I.), northern-western Europe ([Boillot, 1964]; Emery K.O., Aubrey D.G.; Newman W.S. et al.; Mary G.; Mörner N.-A. et al.) and the European part of Russia (Bylynskyi E.N.), E.N. Bylynskyi stated in his monograph [1996, p. 138]: ”In accordance with modern concepts, during ice accumulation, large deflections occurred in the ice sheet bed under the influence of increased loads. These deflections spread for several hundred kilometres along the periphery of the glaciers within the platforms and in the periglacial zone, causing here the formation of a flexure sloping towards the ice sheet. In the distal direction from the glacier, it ended with a roll-like uplift (advanced roll). The amplitude of the uplift seems to have varied from place to place, depending on the structural features of the lithosphere, but according to many researchers, it could have been 50 – 200 metres”. The author substantiates also by the results of his own 30 years research that ”With each major glaciation, under the influence of elastic stresses and as a result of the substance movement asthenosphere from the area of formation of the glacier during its growth and in the opposite direction with his degradation, on the periphery of the ice sheets were raised and disappeared roll-shaped elevations up to 200 m. Together with flexure-like tilts of the lithosphere they had a significant impact on sedimentation and formation of the relief of the periglacial zone several hundred kilometres wide” [Bylynskyi, 1996, p. 166].

In the new XXI century, researchers (for example: J. Wallinga; F.S. Busschers et al. [2007]; A. Panin et al. [2015]; N.N. Nazarov, S.V. Kopytov [2020]; A.O. Utkina [2020]) often use the concept of the formation of a temporary roll-like uplift in front of the glacier - a glacial forebulge - in reconstructing the history of development and explaining the structure of river valleys in the periglacial zone.

Already in the last century, the development of mathematical models of the phenomenon of glacio- and hydroisostasis was initiated and continued in the new one (for example, Peltier W.R. and colleagues). The modeling confirms the bending of the lithosphere under the weight of the covering glacier and the compensatory rise at a certain distance from it.

Geophysicists joined the development of the glacioisostasy theory shortly after the idea emerged. One of the first was N.S. Shaler, who in 1874, analysing glacioisostatic phenomena, predicted the existence of the asthenosphere, explaining by the properties of this earth layer also and the formation of a temporary compensatory roll around the ice sheet. Among the geophysical studies of glacioisostasy in the next XX century, we can mention the works of B. Gutenberg (first half of the century), R.I. Walcott (70s).

E.V. Artiushkov [1967, p. 14] attributed a significant role to the phenomenon of compensation roll formation: ”When a glacier advances, the asthenosphere material squeezed out from under it leads to the formation of a gentle rise 100-200 m high outside the glacier when an average glacier thickness of 2-3 km. This rise on the flat territory actually forms a new watershed and causes flooding of large areas between the glacier, the rise and the uplands located perpendicular to the glacier edge”. Other geophysics S.A. Ushakov and M.S. Krass [1969] estimate the magnitude of such uplifts to be much smaller: a few, at most, the first tens of metres. However, these authors also admit that in the presence of areas of hardening and/or increased asthenosphere viscosity crustal uplifts of hundreds of metres are possible between each of them and the glacier edge.

N.F. Balukhovskyi [1968, 1973] came close to the idea of the formation of a temporary uplift in front of the glacier. However, in his version, the periglacial gentle roll (40-50 m high, 5 to 20 km wide and several tens of kilometres long) was formed by plastic Jurassic and Triassic clays extruded from under the glacier (in fact, this construction by N.F. Balukhovskyi is "squeezing"). In addition to the fact that the formation of the roll in this way is somewhat problematic, Balukhovskyi’s description of the further development of the situation is, in our opinion, not convincing. To explain the occurrence of dislocations, the author uses seismic waves, which, in his opinion, were caused by shocks emanating from the glacier and activated the plastic Jurassic clays of the periglacial roll, and these clays introduced in the cover stratum, what resulted in the formation of a system of folds-upthrows. No explanation is given as to why the glacier did not push back (did not ”roll”) the roll further forward or did not flow over it, but, instead, began to cut away it, and under the glacier at the same time ”a depression of detachment up to 150-160 m deep was formed and gradually expanded”.

The role of glacial lakes. Geological evidence for the extensive development of lake basins in front of the glacier edge is often described in the literature (J. Mangerud et al.; A. Panin et al.; D.D. Kvasov; E.N. Bylynskyi; N.N. Nazarov and S.V. Kopytov). The formation of glacial lakes reconstructed on the Russian Plain was explained by A.S. Lavrov and D.D. Kvasov [1963, 1975], and in Western Siberia - by I.A. Volkov et al. and S.A. Arkhipov et al. by the damming of river valleys by the glacier. In the becoming of the glacioisostasy theory another condition for the formation of periglacial lake basins was stated: in the band of undeglacier glacioisostatic depressions that extended beyond the glacier along its edge [Bylynskyi, 1975, 1996]. This method of lake formation was greatly facilitated by the uplift of the compensation berm [Artiushkov, 1967], which could provide a great accumulation of glacier meltwater with a sizable excess of its level over the local erosion base, because such lake basins received high banks: the glacier slope and the opposite slope - of the compensation roll.

The factor of breakthrough valleys. The phenomenon of the formation of drainage channels (glacial lake spillways), that discharge water from overflowing glacial lake basins, has long been established, and researchers are discovering new and new forms of this type of relief (e.g., J. Mangerud et al.; A. Panin et al.; [Nazarov and Kopytov, 2020]).

Water discharge usually starts through a saddle in a watershed with a neighbouring river basin and becomes catastrophic. Spillways often are large in size and have steeply slopes, and the largest of them (kulli´) are canyons that reach a depth of about a hundred metres, cutting through and through the watershed massiv.

From the descriptions of buried overdeepened depressions, summarised in the literature (see above), it is clear that they are morphologically very similar to spillways. This circumstance, as well as the distribution of buried overdeepened depressions exclusively within or close to the area that in past was covered by glaciers, leave no doubt about their genetic correspondence to spillways. The difference from the spillways lies in the burial of the depressions by Quaternary sediments, the location of their bottoms well below the local erosion bases, and the arcuate concavity of this buried bottom.

Glacier bulldozing effect. On flat terrain, there are no traces of significant exaration (in the sense of denudation of the glacier bed) activity of a glacier, for example, the Dnipro glaciation [Matoshko and Chuhunnyi, 1993]. This is explained by the visco-plastic nature of ice flow under the influence of its own weight [Kapytsa, 1958]. However, at certain stages of research on the mechanical impact of glaciers on the substrate, some researchers considered this impact to be significant. ”Based on the relationship between detached masses and erosion glacial forms, a number of researchers do not exclude that in some areas the glacier acted on the bed like a giant plough: it plowed out the troughs, deformed the sediments in their sides, and threw material onto the slopes.” [Levkov, 1980, pp. 27-28]. And it would seem that such objects as the Kaniv dislocations can confirm this point of view. However, the fact that a long range of versions of their formation mechanism have been proposed indicates that this old concept - the plow out of huge masses of rock by a glacier from depths of hundreds of metres (see above) - is not convincing to many researchers.

In this regard, it is worth mentioning that as early as 1880 H. Credner [1880], using examples from northwestern Saxony, concluded that dislocations in boulder clay layers were caused by lateral (tangential) glacial pressure, and that the formation of dislocations was facilitated by the presence of surface irregularities in the substrate.

Indeed, if the ice flows under the influence of its own weight, the ”pivotal” level of this flow must lie in the lowest layers of the continental glacier thickness, where the vertical weight vector of a huge ice thickness is transformed into a horizontal one. Therefore, in the glacial tongue slopes (both frontal and lateral), the lower ice layer permanently push forward, playing the role of a ”knife” that can ”cut away” uneven terrain, especially with a steep slope that is towards to the glacier movement. Glacier movement does not stop if, after the ”knife” is extended, the entire ice column immediately moves over it - this maintains pressure on the lower layer. This also implies that the slopes of a glacier advancing across flat terrain must be very steep. So, indeed, the effect of a ”mould blade of bulldozer” many hundreds of metres high is created, which can undercut and displace elevated landforms.

But even if such a mechanism does work, it does not at all provide the possibility of lifting to the surface of rocks lying at great depths under the glacier.

Discussion. And here we need to return to the question of the mechanism of formation of the overdeepened depressions.

The flow of viscous matter in the upper mantle, caused by the glacial load, being inert in nature, obviously lagged behind the advance of the glacier edge - but this advance could not but be uneven. It is quite logical that during the phases of the glacier suspension, the wave of subcortical stresses could have time to "roll out" from under the edge of the glacier in the form of a compensation roll. At the same time, the waters of the near-glacial lake could overflow the bowl of the lake basin and to found in the compensatory roll the lowered areas of its ridge, through which they overflowing to its other side. The slope of this flow was significant. Intense erosion formed a narrow, deep spillway in the roll (an analogy is the breach of a pond dam), at the bottom of which alluvium was not deposited, or was deposited already only at the stage of levelling by erosion the longitudinal profile of the spillway (see above description of the properties of buried overdeepened valleys). This process could accompany the growth of the roll from its very beginning, in particular along the valley of the already existing runoff in the direction from the glacier - an antecedent erosion segment could gradually form within the roll.

And this spillway, even if its bottom remained slightly higher than the general surface level outside the roll, in roll would cut through rocks that lie at a considerable depth outside it.

Within a short (on geological scales) time of 10³-10* years (Artiushkov, 1979) the restoration of isostatic equilibrium (relaxation) take place. And the further fate of the spillway is to be lowered (together with the compensatory roll, which isostatic sinking, including under the pressure of ice masses if the glacier continued actively to advance), and later, in the process of glacier melting, to be filled with glacial formations, then lake and other sediments and become a buried depression of an elongated closed shape. The longitudinal profile of the bottom of an erosion form in a roll, close to a straight line, becomes concave after it sinking - this arcuate shape is noted by researchers when studying the morphology of buried overdeepened depressions.

In our opinion, this is how buried valley depressions were formed in the Quaternary glacial zone of the Russian plain.

If the glacier accelerates its advance at the mature stage of formation of a spillway in the compensatory roll, then, having reached the roll, the visco-plastic ice flow will find a breakthrough in it, in which will inevitably form an ice wedge. The concept of an ice wedge is found in the works of researchers. However, they used it rather to give the text an emotional colouring: ”This is the moving huge mass that formed a living and plastic wedge driven into the Dnieper valley” [Soboliev, 1926, p. 205]; ”The Dnieper glacier invaded the Roslavl-Seshchyn combe in a powerful wedge, crushed the underlying interglacial deposits (especially clays), deformed the combe sides, captured many detached masses and deposited in the depression moraine deposits up to 25-35 m thick, and in places (Seshchynskaya station) up to 60-65 m thick” [Horetskyi, 1967, p.22]; the wedging role of the glacial tongue in the formation of the Seshchinskaya glaciodislocations somewhat more definite was stated by D.I. Pohuliaiev [1956, 1968, 1972]. We give this notion a key meaning in our concept, and we mean the wedge-shaped shape of the glacial tongue not in vertical section but in plan, because if the ice wedge is directed downwards with its ”tip” (wedge-shaped in vertical section), the lateral surface of such a wedge does not exert a direct upward squeezing force on the contact rocks but only a very indirect one through down and side pressure on these rocks. And in the case of pressure on the rocks of the erosion scarp of the spillway in the compensation roll, it is enough for the ”horizontal” ice wedge only to move the substrate rocks, that have already been raised and even exposed by erosion in the precipice of the spillway. The onslaught of ice only needs to overcome the forces of adhesion and internal friction of the rocks, and throw them forward, sideways and downward (to the distal foot of the roll), where the pressure vector of the lateral sub-vertical surface of the ”horizontal” ice wedge in the deep, steeply sloped compensation roll scour is directed.

This situation greatly enhances the dislocation capabilities of the glacier. By wedging the scour further, as if opening the gates wider for the main ice masses, the ice wedge moves the rocks, that component the scour sides, forming dislocations. The glacier’s task of destroying the sides of the scour was further simplified by the fact that, bending into the shape of a roll, the rock layers lost some of their strength (are decompacted and cracked); in addition, clay rocks could acquire a more plastic consistency.

This process continued until the main masses of ice arrived, when they began to overflow through the compensation roll.

The idea of the horizontal wedging effect has obviously occurred to researchers many times before. However, within the framework of the same approach (mobilisation of large masses of rock deep in the sedimentary series), it gave little for insight into dislocations such as those in Kaniv.

By the end of the last century, discussions about the mechanism of formation of the Kaniv dislocations freeze. The scientific thought stopped at the glacioextrusive concept [Lavrushyn and Chuhunnyi, 1982] - squeezing . There are no grounds to deny the involvement of glaciodiapirism and glacioextrusions in the formation of the Kaniv dislocations, but it seems that these processes played a far from major role. Researchers note the manifestations of diapirism in other similar depressions to the Shevchenko but do not assign it a leading role in the formation of the corresponding forms.

The authors of the most recent thorough study of the Kaniv dislocations stated: ”Thus, when creating the concept of the origin of the Kaniv dislocations, one has to take into account many factors, not just one of them. Apparently, the underestimation of this circumstance can be explained the current lively discussion about the origin of the Kaniv dislocations and the obvious inconclusiveness of both glaciotectonic and purely tectonic constructions” [Lavrushyn and Chuhunnyi, 1982, p. 84].

Fully supporting this conclusion of these and other researchers, we argue from the standpoint of a systemic paleogeographic approach that the complex multifaceted phenomenon of the formation of two genetically related objects (the overdeepened valley and glaciotectonic dislocations adjacent to it) requires a favourable combination of a number of factors and of specific phases of their dynamics.

To begin with, we point out the need for a geological structure that is favourable in this area for the formation of a sufficiently high compensation roll.

Further, much depends on the duration of the glacier’s suspension phase: it must be optimal, i.e. sufficient not only to form a compensation roll but also to create a deep and sufficiently wide spillway in it, otherwise the glacier will simply ”flow” over the roll; and if the glacier is stopped for too long, the roll will begin to flatten out, the bottom of the spillway will sink downwards, glacioaluvium will appear on him, and a flowing lake basin will eventually form, which will be filled with sediments. In the latter casen an overdeepened depression will form (the ”gully” type of glacial depressions - [Horetskyi, 1980]) but large-scale dislocations are unlikely.

If the gully does not cut through the roll to its base, there will be no alluvium in it, and moraine will lie down on the substrate rocks at the bottom of the prepared by the glacier exaration gully (”trough” type of glacial depressions - [Horetskyi, 1980]).

These simplified reconstructions of ours present only a few of the many possible combinations of the features and dynamics of the main factors and conditions of the phenomena discussed in this article. That is, the variability of the specific consequences of the described processes is quite wide. Therefore, not many overdeepened depressions have been found in the wide areas was covered by Quaternary glaciers, probably their number is in the tens; and even fewer large glaciodislocations are found near them.

An interesting question is the size of the compensation roll.

E.V. Rukhyna [1957] suggests that the formation of buried overdeepened valleys of the glacial zone required before glaciation the uplift of the territory by several hundred metres. Considering the same issues in major monograph [Markov et al., 1965], K.K. Markov agrees that the pre-glacial surface of the Russian plain exceeded its present-day surface by about 400 m. These values, in fact, can be considered to characterise the range of heights of compensation rolls, the formation of which was accompanied by vigorous cutting through by glacial meltwater. The estimate of 400 m is probably too high, as it was given by the authors who considered the overdeepened valleys to be part of the pre-glacial river network and, therefore, linked this value with the general erosion base, while only the local base should be taken into account. Therefore, when assessing possible roll heights, in each case it is need to start from the depth of the buried depression: the height of the roll above the surrounding surface should have been no less than the depth of the gully that cut through it. Whether this depth corresponds to the maximum height the roll reached remains unknown.

The literature contains empirical data on the height of the compensation roll. For example, G. Boillot [1964] calculated the height of the roll in water area of the English Channel in the Late Valdaian time - it reached at least 70 m. Studies by W.S. Newman et al. [1971] on the US Atlantic shelf showed that the height of the roll in the Late Wisconsinan could exceed 80 m.

The length of the buried trough also indicates the width of the roll. This estimate is very approximate, as several unknowns factors are involved: whether the erosion of the roll occurred perpendicular to the roll axis or at an acute angle (up to subparallelism to the glacier front); at which dynamic phase of roll formation the erosion cutting stopped; and as the roll moves, the length of the gully may exceed roll width.

Conclusions. Attraction of the glacioisostatic theory principle on the possibility of formation of a temporary compensation roll in front of the glacier allows us to construct a concept of the paragenesis of glaciodislocations such as the Kaniv and the overdeepened depressions, that is free from weak link in the previously proposed reconstructions: problematic of detachion at great depths and the high uplift of significant masses of rock by the glacier. It is quite logical that only a tectonic factor can cause such an uplift.

Another condition for the formation of both the overdeepened depressions and glaciodislocations is the development of a breakthrough across the roll during the slowdown (suspension) of glacial advance. Finally, the third condition is the timely resumption of glacier movement, which must have time to wedge into the breakthrough before the compensatory roll is sinking (flattened).

As the pressure of the ice mass continues to increase, the ice wedge by lateral surface destroys the sides of the gulley, detached blocks of rocks from them and pushed these blocks forward-side-down, thus forming the glaciodislocations. Subsequently the compensatory roll was sinking (smooth down) and the erosion-exaration scour, that formed in it, had found itself below the level of the local erosion base, was buried and became an overdeepened depression.

The uniqueness of the pair of geological monuments of the modern nature of the Dnipro region - the Kaniv mountains and the Shevchenko buried overdeepened valley - is undoubtedly due to a very favourable combination of the many circumstances and factors, described above, and of their dynamics during one of the episodes of glacial transgression.

We should also note that the palaeogeographic reality of the formation of a temporary glacioisostatic compensation roll in front of the glacier is strongly being confirmed by the eloquent fact of the existence of overdeepened depressions. After all, the only realistically possible agent for creating these landforms, which were subsequently buried, is erosion, and erosion can act only above its local base. Even if we assume that the buried depression was created as a tunnel (by subglacial pressure flows), the explanation of the next, main stage of dislocation formation - the uplift of huge masses of bedrock from a considerable depth - will remain very problematic without the factor of compensation roll.