by Douglas E. Cox
This article is a revised version of one originally published in the Creation Research Society Quarterly, 14:149-155, December, 1977, and reprinted in: Corliss, William R. 1980. Unknown Earth: a Handbook of Geological Enigmas, The Sourcebook Project, Glen Arm, Md. p. 5-15.


Three kinds of column-like formation are discussed. There are columns of unconsolidated material, often differing from their surroundings, embedded in sand or drift. There are pillars of consolidated material, often sandstone, embedded in rock. And there are free-standing solid pillars. It is proposed that all of these kinds of pillars, and also potholes, are related; they can be ascribed to a process of rock disintegration or crystallization following upon the release of pressure as the land was raised out of the water at the end of the Noachian Flood. It is hard to believe that sedimentary processes could have caused these formations; hence the cross stratification, often found with them, can not be an effect of sedimentation.


Strange polystrate pillars occur in sandstone and in drift in many places. The writer has examined pillars in gravel and sand in the region of Waterloo County and other places in Southern Ontario. Pillars in the Potsdam sandstone, near Kingston, Ontario, were investigated by the writer and H. L. Armstrong.

These pillars, also called "pipes", or "pots", transect the pattern of cross stratification. They are anomalous and difficult to explain in terms of sedimentary deposition.

Polystrate fossils such as trees, vertically embedded in sediments, have been cited by some Creationists as evidence that rock strata accumulated rapidly. [1] This explanation cannot account for the unconsolidated pillars; but they, too, may have been formed rapidly.

Processes forming vertical pillars in cross stratified sandstone and drift are difficult to find in nature today, but past causes may have been quite different from those existing now. One such cause is a possible disintegration process, due to rapid release of former high pressure. Uplift of the continents and erosion of overburden by currents of the retreating flood waters in certain areas would be accompanied by a decrease of pressure on rocks.

In this article, geologic evidence is presented in support of the idea that the pattern of cross stratification in drift gravels may be an effect of a disintegration process, and that similar patterns of cross strata in some sandstones and conglomerates are the effects of a crystallization process, instead of a sedimentary phenomenon. [2] Certain features of cross stratified formations support this interpretation; for example, the configuration of the strata around pebbles generally does not exhibit stoss and lee effects as would be expected in an environment of rapid current flow. As well, the cross strata is typically smooth, even in conglomerates and gravel, and it is not "bumpy" in the vicinity of pebbles; this bumpiness would be expected if the pattern were a sedimentary one, due to the ridges and streamling effects on the lee side of embedded pebbles.

The presence of vertical polystrate pillars in cross stratified sandstone and drift is another sort of evidence that cross stratification is not a sedimentary phenomenon, but may result from a pressure-related disintegration process. Such a process can account for similarities between the pillars in sandstone and those in the drift. Also, from this point of view, a relationship between pillars and the contents of potholes would be expected.

Drift Pillars

A few years ago writer observed a group of pillars in cross stratified drift gravel as they were exposed in the sides of a gravel pit at Blair, a few miles south of Kitchener, Ontario. See Figure 1. The gravel pit was operated by Forwell Ltd., of Kitchener. In the course of excavations over a few years, removal of the coarse gravel revealed many pillars.

Figure 1

They were distinguished from the enclosing gravel in various ways. Many were stained dark brown or black, and appeared as finger-like extensions of the soil profile into the gravel below. Some of these contained a light coloured interior, thus exhibiting a concentric structure. The pillars were cylindrical in shape, with tapering or rounded bottoms. All were in the uppermost part of the drift, beginning within or just below the soil. The pillars extended downwards to varying depths. At this site, pillars one to two feet in diameter, and up to 10 feet in vertical extent were common. Don Metzloff, foreman at the site, also observed the pillars in the course of excavations. He reported that the pillars were mostly clustered together in one section of the pit. The largest pillars reached 20 feet vertically and were 3 to 4 feet in diameter. These were described by Metzloff as consisting of "dirty coloured, fine sand, just like ground."

Some of the pillars consisted of sand and clay which transected coarse gravel, and others were composed of gravel, although different in appearance from that of the enclosing drift. More clay was present in some pillars, which made them more compact and resistant to weathering than the enclosing gravel.

Concentric structures were present in many pillars. In one pillar exposed in cross section, there was distinct colour banding, and an outer lining of pebbles formed the perimeter of the pillar. Another pillar was lined by sand and clay, which formed a smooth cylindrical margin about the pillar. The pattern of cross stratification in the gravel near some pillars was bent downwards.

Several features of the pillars resemble features of other structures described in the geologic literature. A report by Conant et al. described "pots" in gravels in Maryland and Virginia, which were somewhat different in shape to the pillars in the Blair gravel pit, but otherwise seemed quite similar. [3] The pots they described were more spherical or bulbous in shape than the pillars in the Blair gravel pit. The pots also exhibited concentric structure, and the stratification of the surrounding gravel was bent downwards around some of the pots.

The pots were about 7 ft in depth and about the same width, though some were much larger. All were at the top of the gravel. One pot was wrapped in a layer of white clay 1-2 inches thick. Describing the contents of the pots, the authors say:

The filling of the pots is chiefly a clayey silt containing a lew percent to perhaps 40 percent ol admixed sand and gravel. The silt is generally medium gray and mostly structureless, but in some pots it is faintly or distinctly stratified parallel with the margins. In some pots, flat pebbles also tend to be aligned parallel with the margins. The uppermost 1 ft. of the filling is commonly more gravelly. [4]
"Till clumps" similar in shape to the pots in Maryland and Virginia were reported by Mather et al. from Cape Cod, Massachusetts. [5]

The mechanism favoured by Conant et al. for the formation of the pots involved seasonal frost action during a Glacial Period; but such a mechanism would seem unworkable in the case of the pillars at the Blair gravel pit which are much narrower and deeper than the Maryland pots. Yet there are similarities which suggest a common orgin for all these peculiar structures.

A Pillar with Cross Stratified Contents

Some pillars in the drift seem clearly to indicate that the pattern of cross stratification cannot be of sedimentary origin The drift in Southern Ontario is explained by geologists in terms of the Glacial Theory either as till, thought to be a direct deposit ot the ice-sheets, or outwash, which is considered to be glacial debris which has been transported in streams flowing from the melting ice-sheets, and deposited in rapid currents. Till is unstratified while outwash exhibits the pattern of cross stratification.

The portion of a drift pillar illustrated in Figure 2 was found in a gravel pit near Campbellville, Ontario. Its contents, on the left, were internally cross stratified. This pillar was enclosed by fine, cross stratified sand, and consisted of coarse sand and pebbles, with nearly horizontal strata. No pebbles were present in the enclosing sand. Only a part of this pillar was intact, about 5 feet below the original level of the gravel pit. The diameter of the pillar was about 1 ft.

Figure 2 [Click on picture for a larger image, about 80 kbytes.]

The perimeter of this pillar was well defined, and consisted of a thin film of clay enclosing the structure. It is hard to imagine how such a structure could have formed in an environment of rapidly flowing currents, as assumed in the glacial explanation for cross stratified drift.

In particular, the cross stratification of the contents of the pillar could hardly have been caused by currents within the structure. Since the pillar was unconsolidated, and was enclosed by unconsolidated sand, it would likely have been washed away by currents of outwash streams. Another, non-sedimentary explanation is needed.

Some geologists have proposed rather special circumstances to explain pillars in drift. A pillar of sand, with concentric layers, was found in coarse and medium Quaternary sand at St. Jerome, north of Montreal, Quebec, by Dionne. [6] The height of the pillar was 152.5 cm, and diameter was 34 cm at the top and 24 cm at the base. The sand comprising the pillar was fine at the center, and the pillar transected cross stratified beds. It was proposed that a whirlpool eroded a deep cylindrical hole in the unconsolidated sand, and refilled the hole immediately afterwards, forming a pillar.

While some such mechanisms may seem plausible for individual structures, this could hardly apply to a large group of pillars in drift. The pillars appear to occur in clusters in certain areas, with distributions similar to potholes. The tops of several pillars, resembling "hoodoos" in drift were exposed along a roadside as loose sand was eroded by wind during construction of Balmoral Road, Cambridge, Ontario, as shown in Figure 3.

Figure 3

 It is hard to understand how these structures could have been formed in unconsolidated sand in any sedimentary environment. However, these features can be explained in terms of an in situ disintegration process. As more observations of these peculiar structures are reported the improbability of isolated polygenetic causes for the pillars is increased.

Can Freezing Explain Pillars?

Explanations for pillars in drift that involve frost effects fail to account for many similarities between the drift pillars and those in sandstone, described below.

As mentioned above, the deep, narrow shape of many pillars does not seem to support a theory of frost action.

Further evidence against freezing as a mechanism for the formation of the drift pillars is the difference in the composition of some of the pillars compared with the surrounding drift material. This was evident in the examples shown in Figure 2 described above, where there is coarse sand and pebbles in the structure and finer material surrounding it. A pillar consisting of coarse sand and pebbles, that occurs in fine sand, could hardly have been derived from the enclosing sand by any uniformitarian mechanism such as frost action. The presence of such a pillar in sand is also anomalous if the sand is explained in terms of glaciofluvial deposition.

Formation of the Pillars

The idea of a non-uniformitarian process of rock disintegration, causing patterns of cross stratification, would lead to a simple explanation for the pillars.

Many Creationists think a major part of the phanerozoic sedimentary rock record (i.e., rocks which contain fossils), that occur in the earth's crust, was formed during the Noachian Deluge. Uplift of the continents at the end of the flood would generate currents that could cause considerable erosion of sediment; the removal of overburden, tectonic movements, and jointing, may have released pressure on newly formed rocks, providing an environment for the process of rock disintegration.

Disintegration would likely be limited to consolidated rocks. Sediments formed in the Deluge were probably compacted as the continents were elevated and the flood waters retreated towards the ocean basins. Lithification occurred as diffused cementing agents crystallized. As vertical pressure on the topmost rocks decreased, due to lower depths of water and erosion of overlying sediments, the horizontal stress tended to became the maximum stress. Some component minerals in the sediment mixture separated out, as the sediments were lithified; water tended to be expelled from the rock. Near the surface of the rock, crystallization of rock forming minerals, and expansion of the occluded water and volatiles may have generated successive thin layers of sand. Concretions developed in rocks, becoming pebbles and boulders as the disintegration process altered the rock about them into sand and clay. It is proposed that the pattern of cross stratification in the sand of the drift was formed by a mechanism of disintegraton such as this.

In many places the disintegration began in a small area of the rock surface, and penetrated vertically downwards, resulting in a cavity filled with the disintegration product. In effect, a pothole was formed.

Where the surrounding rock was also subsequently disintegrated, this pothole (with its contents) became a pillar within the cross stratified sand and gravel. This mechanism for the formation of the pillars in the drift implies a relationship exists between the pillars and the contents of potholes, before excavation.

Concentric structure in many pillars may be explained by lateral enlargement of the pillars during the pothole stage. Disintegration of the rock walls would account for linings of pebbles, orientation of the pebbles parallel to the margins, and lining of the margins by clay which has been reported in some pillars.

Discolouration of the contents of the pillars, which also produces a concentric effect, probably occurred during the pothole stage. Migration of volatiles, calcite, iron oxide, dissolved silica and other minerals from the rock enclosing the pillar towards the low-pressure surface would cause concentric banding. Precipitation of these minerals at the margin of the pillars occurred due to lower pressure.

Down warping or bending of the pattern of cross stratification of the gravel around some pillars indicates the pillars were present when the pattern of cross stratification was formed. This bending may indicate more rapid disintegration of the rock occurred near some pillars.

Pillars Compared with Potholes

Conventional geology says potholes were formed by erosional processes such as the action of streams; but it is difficult to explain all the features of potholes in this way. In my opinion, potholes in the rocks of a stream bed are merely expsoed by existing processes, not formed by them. [7] The potholes which are exposed by streams, or other agents such as wave action along the shores of lakes or coasts, may have been present in rocks prior to the initiation of present conditions. Before exposure, potholes are usually filled with drift, consisting of sand, clay, and gravel. In some rocks potholes may contain sandstone.

The mechanism outlined above for the formation of drift pillars also explains potholes, and a resemblance between the pillars and the contents of potholes would support the proposed theory. The pillars in drift resemble potholes in shape, many are bulbous like potholes, and the range of sizes is similar. Like potholes, the pillars are rounded at the base, and penetrate to varying depths. In distribution patterns, the pillars resemble potholes, which may occur individually or in clusters.

The characteristics of the contents of potholes are rarely given much attention in the geologic literature, but some large potholes in Norway were excavated in 1874, and described by Brogger and Reusch. [8]

The largest kettle excavated was near Bakhelagel, and the work occupied 3 men for 50 days. Careful records were kept of the position of the boulders found in the kettle. The depth was found to be 33.5 ft, or 44 ft if measured from the highest side.

The larger rocks in the kettle formed layers at various levels, and the authors wrote, "The contents of this kettle, therefore, plainly showed a sort of stratification." [9]

Deep within the pothole two large boulders were found, which appeared to have a smaller hole begun in them. The small hole was at the centre of the kettle.

In the theory of disintegration causing potholes, the smaller hole may represent part of an initial pothole, which widened and deepened by further disintegration of the enclosing rock.

Indications of a concentric structure in the contents of a pothole were found by the writer in a large filled pothole at Rockwood, Ontario. Here the larger rocks seemed to be concentrated around the walls. This may also indicate widening of the pothole by disintegration.

Further studies of pothole contents are needed to confirm the presence of a concentric arrangement of the contents, and other points of resemblance between the pillars in drift and the contents of potholes.

The "Park of Pillars"

From the point of view of uniformitarianism, a relationship is not readily perceived between the drift and cross stratified sandstone. But the theory of disintegration would require such a relationship.

Pillars or pipes are present in many sandstones, which show similarities to pillars in the drift. A well known example of pillars occurs in an outcrop of Potsdam sandstone on the property of Bill Hughes, R.R. 6 Kingston, Ontario. See Figure 4.

Figure 4

Several vertical pillars are visible, the largest measuring 14 ft in diameter. Some exposed in the sides of a cliff, evidently tlie site of an old quarry, extend 20 ft vertically. Concentric bands occur within many of the pillars, and also in tlie rock enclosing them. This feature led some to conclude the pillars were "fossil trees", and sections of the pillars were exhibited as such in shop windows in Kingston in 1888. [10]

Hughes operated a quarry near the site of the pillars, and he told me a pothole had been found during the excavations. It was filled with sand and small pebbles.

One of the explanations which has been offered for the pillars is that they are potholes filled with alluvial sediments. [11] It seems clear that they are not fossil trees, but, being polystrate, they may show as well as polystrate trees that the rock around them was formed quickly.

The pillars are described in an article by Hawley and Hart, [12] and a thorough description is not attempted here. These authors proposed that the pillars were formed hy springs rising through the strata while it was still unconsolidated, and referred to the pillars as "quicksands enclosed by concretions."

Similar pillars have been reported from other exposures of the Posdam formation. Some occur across the Rideau Canal opposite the Hughes farm, others are present at Morton, Ontario, and at Redwood New York. '

The rock in which the pillars occur is a reddish sandstone with cross stratification. Perhaps the sandstone formed in the subsurface by crystallization of hydrous amorphous silica. Hematite coatings on quartz sand grains and within flaws in the grains can be explained by exsolution during the crystallization process. This in situ crystallization process formed cross stratified sand, which was then recemented by deposition of silica and iron oxide in pore spaces.

The pillars may have been formed in a manner similar to that described for the formation of pillars in drift. Vertical pillars were formed by penetration of the crystallization surface in localized areas, and subsequently a larger area of the surrounding rock was also altered. Concentric structures were formed in the pillars due to decreased pressure upon disintegration, and deposition of hematite at successive surfaces of the rock during disintegration.

Structures similar to those in the Potsdam sandstone near Kingston have been reported from cross bedded siltstone and conglomerate in the Bush Creek region of Eagle County, Colorado. [13] These were up to 4 ft in length and 8 inches across. Some of the pillars interlocked, and some completely enclosed smaller ones. Concentric bands surrounded the pillars.

Another report described pillar-like structures up to 200 feet in height. The[14]se occur in the Laguna area, New Mexico, and are referred to by Schlee as "sandstone pipes." The pipes range from a few inches up to 150 feet in diameter. The pipes are grouped in clusters, and many have a concentric internal structure. Cross bedding was present within the pipes and in the enclosing sandstone.

Other similar structures have been reported in the St. Peter Sandstone of Arkansas, and in northern Arizona and Colorado. [15]

Dietrich described cylindrical pillars in the Potsdam sandstone at Redwood, New York, and noted that similar structures also occur at East Anglesey, England, near Brussels, Illinois: along the coast of Syria and Palestine: at Barnstaple Bay, Devonshire, England: and at Canadon Hondo, Chubut, Argentina. [16]

In some regions the disintegration of the rock enclosing pillars may have been more complete than the disintegration which formed the pillar itself. In such cases a free-standing pillar may be formed by erosion of the loose sand.

Free-Standing Pillars

Free-standing pillars of sandstone were described by Simpson at Canadon Hondo, Central Patagonia, and associated with the Potsdam pillars near Kingston. [17]

These pillars were exposed by erosion of the soft sandstone beds in which they were embedded. Most of the pillars were from 1 to 2 ft in diameter, and from 3 to 10 leet in height.

Many of the features of the pillars mentioned by Simpson have their counterpart in typical potholes. Some twin pillars were observed, some contained horizontal flutings on the sides, some tended to bulge in the middle. The bases of some weathered pillars are undercut.

The pillars seem to be the antithesis of potholes, which may be explained in terms of the disintegration theory of cross stratification. The pillars originally formed by rock disintegration in small areas, and subsequently the enclosing rock also disintegrated more completely. The material in the illars was indistinguishable from the matrix except for a slightly coarser sand and firmer cementation.

The pillars described by Simpson seem to be sandstone counterparts of pillars in drift observed by the writer.

Islands, composed of sandstone, such as those in the Wisconsin River at Wisconsin Dells, may really be large pillars. One of these island pillars is known as the "Inkstand. " Many potholes occur in crevices along the Wisconsin River.

The pillars are cross stratified, and these may have been formed by more complete disintegration of the rock enclosing pillars. Sand enclosing the pillars would have been washed away by the Wisconsin River, leaving the pillars as islands.

Predictions Based on the New Theory

Four kinds of phenomena can be related by the theories of in situ rock disintegration forming the pattern of cross stratification in drift gravel, and of sandstone formation by a crystallization process in successive layers. These are potholes, pillars in drift, sandstone pipes, and free-standing pillars. All of these show similarities, which suggest a similar cause.

One of the ways in which theories can be tested is by means of predictions. The following two predictions can be made on the basis of the theory outlined in this article:

  1. The contents of potholes should reveal concentric features, or internal cross stratification, or other features similar to those of pillars in sandstone and in drift.
  2. Pillars similar to those which occur in drift may also be expected in cave fill. Potholes are common in caves, and the fill of caves has been interpreted in terms of the disintegration similar to that proposed for the drift. [18]
Both these predictions can be easily tested, and future observations may either support or refute the present theory. Another prediction, based on the uniformitarian premise, is possible, to test its validity: potholes ought to occur in rock strata at all levels, as fossils of bygone times. Where are all the examples of these fossil potholes?


  1. Rupke, N.A. 1970. Prolegomena to a study of cataclysmal sedimentation. in W.E. Lammerts, ed. Why Not Creation? Presbyterian and Reformed Publishing Co., Nutley, N.J., pp. 141-147.
  2. Cox, D.E. 1975. The formation of cross stratification: a new explanation. Creation Research Society Quarterly, 12(3):166-173.
  3. Conant, L.C., R.F. Black, and J. W. Hosterman, 1976. Sediment-filled pots in upland gravels of Maryland and Virginia, Journal of Research, U.S. Geological Survey 4(3):353-358.
  4. Ibid., p. 354.
  5. Mather, K.F., R.P. Goldthwait, and L.R. Thiesmeyer, 1942. Pleistocene geology of western Cape Cod, Massachusetts: Bulletin of the Geological Society of America. 53(8):1127-1174.
  6. Dionne, J-C., and C. Laverdiere, 1972, Structure cylindrique verticale dans un depot meuble Quaternaire, au nord de Montreal, Quebec. Canadian Journal of Earth Sciences, 9(5):528-543.
  7. Cox, D.E., 1975, On the interpretation of potholes, Creation Research Society Quarterly, 12(1):25-31.
  8. Brogger, W.C. and Reusch H.H. 1874. Giant's kettles at Christiania Quarterly Journal of the Geological Society, 30, 750-771.
  9. Ibid., p. 760
  10. Weston, T.C., 1899. Reminiscences among the Rocks, Warwick Bros and Rutter, Toronto. p. 278.
  11. Baker, M.B., 1916. The geology of Kingston and vicinity, Ontario Bureau of Mines Annual Report, vol 25, pt. 3. pp. 19-20.
  12. Hawley, J.E. and Hart, R.C., 1934. Cylindrical structures in sandstone. Bulletin of the Geological Society of America 45(6):1017-1034.
  13. Gabelman, J.W. 1955. Cylindrical structures in Permian (?) siltstone, Eagle County, Colorado. Journal of Geology, 63(3):214-227.
  14. Schlee, J.S., 1963. Sandstone pipes of the Laguna area, New Mexico. Journal of Sedimentary Petrology, 33(1): 112-123.
  15. Ibid. , p. 115.
  16. Dietrich, R.V., 1953. Conical and Cylindrical structures in the Potsdam sandstone, Redwood, New York. New York State Museum Circular 34, pp. 8-9.
  17. Simpson, G.G., 1935. Cylindrical structures in sandstone, (discussion). Bulletin of the Geological Society of America, vol 46, pp. 2011-2014.
  18. Cox, D.E. 1976. Cave formation by rock disintegration. Creation Research Society Quarterly 13(3):155-161.

Return to main page