Preliminary reports of sedimentation experiments held at Glen Rose, Texas, March 2007

Written by Ian Juby
Various footage taken during these experiments can be viewed in my Video Logs (VLOGs) on Youtube:

  In mid-march, 2007, M.E. Clark (Professor Emeritus, U of Illinois @ Urbana), Andrew Rodenbeck and myself performed a series of experiments over two weeks at Creation Evidence Museum in Glen Rose, Texas.  The museum grounds have a rotary flume which was constructed by M.E. and Dr. Henry Voss, and was transported to Glen Rose some years ago.  M.E. also brought down "Archimedes," a specially designed and constructed liquefaction tank which will be discussed later.  While we were there, we also constructed a linear flume, and had intentions to experiment with silica lithification processes, but ran out of time.

Many lessons were learned which altered my personal views on a number of things and have significance for the geology caused by the global flood of Noah.  Specifically, the rotary and linear flumes, and just about everything we did with water (including a simple garden hose) produced layers.  Probably the most dramatic results were the production of complex cross-bedding.  The process was remarkably easy and solidifies the arguments that crossbeds within the geologic record were indeed formed by a global flood, and not by desert dunes as some have argued.  Newts were also placed into the linear flume during runs and their behaviour also confirmed some hypotheses regarding the formation of the coconino fossil trackways that are so prolific throughout the Grand Canyon and area.
While it seemed everything we did led to sedimentary layers being formed, much like what is seen in road cuts, liquefaction was the ultimate destroyer of layers.  For myself, this was a fairly radical change in my thinking, as I had wanted for years to perform experiments in liquefaction, and the results were pretty much the exact opposite of what I expected.

The Rotary Flume:

Shown on the right is the rotary flume.  The operation is quite simple:  The outer, plexiglass wall and the inner, green wall form a tank roughly 12 feet in outer diameter and 8 feet in inner diameter.  The paddles are in the upright position in the photo, but spring-lock into a downward position during the runs (paddle at far left is in the "locked" position).  The tank is filled with water and sediments, and the paddles drag in the water.  The paddles are spun in a counter-clockwise direction, pushing the water in the tank around in the circle, which picks up and carries the sediments in suspension.  When the rotation is stopped, the now forward-moving water pushes the paddles out of the locked position, which then spring up out of the water to avoid the turbulence and drag of a stopped paddle in the now flowing water.
The sediments settle out of the water as the water slows down and eventually stops.

Click here to see a video of one of the run-ups.

The principle of the rotary flume is to produce an infinite flow or wave.  When we first arrived, this was the first time using the flume with the new, spring-loaded paddle mechanism.  We did not know what to expect entirely, but had some educated guesses.  Sand was hauled and cleaned, extremely fine dust was also obtained from a wash along the Paluxy River, and extremely fine, white, silica sand was bought from the local hardware store.

Upon filling the tank with water and pouring in sediments, we immediately saw what was to become the rule:  The sediments sorted themselves out in very clear layers.  This became so common that by the end of two weeks, we jokingly referred to Andrew's law as "It's difficult not to make layers," and Clark's law as "It's easy to make layers."  Later on, I proposed the "law" that liquefaction destroys layers, as much to my surprise as that was.

We ran up the flume in a series of tests with essentially the same sediments for the first couple of runs while varying the water depth.  Multiple layers of varying numbers were made throughout the flume, and numerous cuts made in specific locations (randomly selected at first, then simply copied in later runs), followed by a complete circumferential cut on all runs.  Posts on the outside frame were labeled by myself, and in hindsight I wish I had labeled them differently:  The first, double post for each section was labeled with a negative number; i.e., A-1.  Going counterclockwise, looking from the top, they then increased in sequence until the next double post marked the next section.  Thus, A-1 and A1 can easily be confused.  So, please be aware of this denotation throughout the rest of this report.

Because of Guy Berthault's previous research with flumes years ago, we half-expected to get three layers.  Instead we got everything from one uniform layer to seven layers.  Before the first run, Andrew correctly pointed out that the inner diameter of the flume would have slower-moving water than the outer diameter, and thus the sediments would settle on the inside first.  Not only was this true, but usually the sediments settled out without us seeing it at all, as the sediments would never reach the outside, plexiglass wall.
The differential water speeds also led to complex vortices and helix spirals in the water, which led to complex and confusing layering.  However, several principles were verified, namely the fact that layers are formed by flowing water - and quite easily.

Of especial interest was interbedding that was quite apparent, with three layers fingering in to one solid layer, then fingering to five layers.

Also of special interest was a small worm that accidentally got mixed in with the sediments.  Andrew happened to cut exactly the correct spot on one of his sectionings.  The worm was polystrate (yes, it cut through layers), and the top portion of it was bent over flat within a layer.  The reason this is of interest is because this is precisely how a fossilized worm was found in the overburden limestone removed from the Paluxy riverbed in 2003.  Also on display within the Royal Tyrell museum in Drumheller, Alberta, is a depiction of three polystrate worms found in the Burgess Shale of Canada.  The Paluxy is quite unique in that fossil worms (sometimes still with pigment) are plentiful, and I was quite happy to see the same effect in the Burgess shale.
The point here is that a sediment-laden water flow deposited a dead worm in the upright position, precisely the same way one was found in the Paluxy limestones, which also have plentiful indications of being deposited by a strong current.  (I apologize for the lousy photo - my macro mode got turned off without my realizing, and I weren't none too happy 'bout it neither!)

In the end, we saw pretty much every stratigraphical feature produced: Crossbedding, fingering, thinning and thickening of layers, interbedding, and scours.

The Linear Flume:
Due namely to time constraints, our linear flume was very simple.  It was a clear-walled (acrylic plexiglass), long box, measuring 6 inches wide, 1 foot tall and 8 feet long.  A steel trough, or funnel, was at one end to facilitate ease of loading sediments and water being poured in.  The other end was left open, emptying into a container which was merely to recycle the sediments while allowing the water to overflow the container.
A conventional cement mixer was used to keep the sediments homogenized, and a continuous stream of water was added to the mix during the runs.

The linear flume not only gave us plenty of radical lessons to ponder, but also enlightened us as to some of the complexities of layering within the rotary flume.  Specifically, we took the lessons learned from Berthault's experiments and not only found them to be true, but that they applied to a much broader scope of sedimentology than I personally thought - both in the field, and in the lab.  For example, it appears now that horizontal layers we see throughout the geological record (and which we produced in the flumes) may really just be extremely long-wave crossbeds.
Berthault's main point from his experiments is that sediments sort out by particle size, not density!  This certainly seemed true in all of our experiments.  While obviously density played a role, it was a minimal one which was usually so insignificant it could be safely ignored.

The reason is not so obvious at first.  I very much like the way Andrew explains the sediments being held in suspension:  He refers to the particles as "flying," which really is what they are doing.  They are flying in a very dense fluid - water. 
The density between two sediments may be as large as 0.1 g/cm3, for example - but when you are talking about two particles 10 microns in diameter, their difference in density is so small as to be extremely difficult to even measure.  However, the velocity of water needed to suspend and carry a particle 20 microns in diameter is significantly greater than that required to carry a 10 micron particle.
To bring this to layman's terms, envision a boulder made of quartz that's 30 centimeters in diameter, and a boulder of limestone that's 60 centimeters in diameter.  The quartz is considerably denser than the limestone, yet the larger rock is obviously much heavier than the smaller rock, and thus will require water moving at significantly higher velocity to pick it up and carry it.  If they were both the same size, the water speed required to pick up both rocks would be different, but the difference would be nowhere near as great as the difference between two boulders of differing sizes.

The unusual thing noted when observing settling sediments is the tendency to sort out into three layers: fine on the bottom, coarse in the middle, and fine on top.   Berthault's explanation seems to hold water:  The flow of water at the bottom of the tank (or river, or lake bed, or stream bed) is almost zero because the bottom of the tank is not moving with the water.  Friction causes a rolling of water along the bottom, thus there is a very thin layer of almost stationary water at the bottom of the tank.  We refer to this as the "boundary layer."
Because the larger grains require the fastest moving water to carry them, they wind up settling out of the flow first, as the flow slows down.  However, within this boundary layer, you get water velocities which may be slow enough for all grains to drop out of the flow.  The largest grain winds up settling first, and the gaps between it and the other largest grains are filled in with the finer grains - up to the top of the largest grain.  This makes the first, bottom layer that appears at first glance to be all fines.

As the water slows down, the large grains then drop out, largest to smallest, making a "pile" which grows horizontally.  Finally, the fines are the last to drop out because they require the least amount of water velocity, and thus they make up the final layer of fines on top.

I will continuously refer to these three-layer sequences as they continually cropped up, and are probably related in some way to cyclothems which are well known in the rock record.

Experiment #1:  rapid emptying of entire sedimentary batch.

For the first experiment, I was operating the mixer.  It was filled with our variety of sediments and topped off with water.  After a brief mixing run to homogenize the sediments, I simply poured out the entire contents rather rapidly.  Total contents was probably around 12 gallons worth of water and sediments, poured out in roughly five seconds.  I had built a hill in the middle of the flume, which was promptly wiped out by the flow and had little to know effect on the very evident layering:

The layering was very long and the layers thin.

Experiment #2: Slow, continuous pour

The second run was a continuous pour of the same contents, with continuous water flow.  The whole pour probably lasted roughly 8 minutes or so and also produced very distinct layering.

Experiment #3: Pulsed flow

M.E. and Dr. Voss produced a paper1 on the subject of tidal action during the flood of Noah for the 1991 ICC.  The scriptures are quite clear that it took 150 days for the floodwaters to rise above the highest mountains, and thus during this time you will have tidal action influencing the continually advancing floodwaters.  Every twelve hours would see a mini-tsunami encroach upon the land, each higher than the last one.

To simulate this, we pulsed the flow of sediment-laden waters.  This produced the most dramatic horizontal layering, with the number of three-layered sets corresponding the number of pulses, or waves, we sent through the flume.  This is probably related to the cyclothems we see within the rock record.  Note the repeating sequence of layers, from bottom to top: coarse, white, red;  coarse, white, red, etc....

Experiment #4: Uphill flow

This experiment led, serendipitously, to the most dramatic find of the two weeks.  We merely tilted the flume so that the water and sediments had to go uphill a mere 2 degrees.  This produced some rather dramatic crossbedding.

Allow me to introduce what a crossbed is.  This photograph is from the Navajo formation, taken within Zion National Park.  You'll notice thick layers on top of each other, and within those layers are angled layers.  These angled layers are called crossbeds, and the crossbeds are composed of three parts:  The topset (the top, swooping downward curve), the foreset (the face of the slope), and the bottomset (the curve leading from the slope, leveling out against the top of the last layer).

I had a personal goal to produce crossbedding while we were down there, so I was thrilled to say the least.  However, none of us were expecting the ease at which it was produced.  This one experiment led to an understanding of their genesis, and led to a series of experiments in the linear and rotary flumes.

The secret was standing water.  While Andrew and I were well aware that Berthault had produced crossbeds in the lab, we considered his method unrealistic in nature.  In Berthault's experiments, they had a horizontal, linear flume in which they had water and sediments flowing through.  He then dropped a door at the end of the flume, causing a backwash up the flume.  Neither Andrew nor I considered this realistic to nature, nor applicable to the global flood of Noah: What was this magical dam that suddenly appeared on land, blocking the floodwaters of a worldwide flood?

However, sediment-laden waters encroaching on land and encountering an uphill will pool standing water ahead of the sedimentary deposit it's producing.  It isn't the uphill that's the key, but merely standing water - which could be an inland lake, water coming from the other side of the continent during the flood, or pooled water from the last tidal wave flowing back out to sea.
The fast-flowing water is carrying sediments in suspension.  Once it hits the standing water, it suddenly drops speed dramatically - well below the velocity required to hold the sediments in suspension.  The sediments "drop like a rock" (pun intended), and make a steep slope much like a conveyor belt will as it drops sand in a pile.  In this case though, the conveyor belt moves along with the pile!
The sediments fill in the standing water area, moving the front edge of the standing water ever farther back and making an ever-longer platform for the fast water to ride on.  Thus, the crossbeds continually build into the standing water - sometimes at remarkable speeds.  Here is a video of them being produced.

This also has some interesting ramifications:  If the flow truly is going uphill, then the standing water and the incoming water have no place to go - thus, the crossbeds will thicken inland as the standing water deepens.

Back to the rotary flume:
At this point, Dr. Clark suggested tilting the rotary flume to acheive an uphill on one side.  The rotary flume is mounted on several jackscrews, so we applied roughly a 2 degree tilt.  We added extra water and ran it.  If there were crossbeds, they were formed from the center out, on an extending, radial arm.  However, this experiment demonstrated that it was not the uphill nature of the deposition that produced crossbeds, rather it was flowing water hitting standing water.  Because all of the water in the rotary flume travels together, there was essentially no standing water and only brief pulses of backflow.

The high point was at C-1, with the low point obviously being between E2 and E3.  Layers were produced, but I would say less that we had before - it seemed to make a mess more than orderly layers, but still produced them in line with Andrew's and Clark's laws.  Essentially no recognizable crossbeds were formed.  The following radial cut was made at E1:

More experiments in the linear flume:

We then proceeded with a couple of experiments relating to crossbedding. 

We first performed a run with a very aggressive introduction of sediments and water into a 1 degree uphill slope.  Andrew and M.E. were operating the equipment, and both Dr. Carl Baugh and myself witnessed very steep-sloped crossbedding being formed, but within a fairly thin bed (the reasons for this will be discussed later).  This is mentioned in passing because while both Baugh and myself witnessed the crossbeds being formed (see video here), when we were finished, the sediments were so uniform as to appear to be one thick layer with no crossbedding!  Thus, it appears that perhaps some layers within the geological record may very well have been formed by a cross-bedding process, but leaving no distinct crossbedding.  For myself personally, I will be looking at layers and rocks differently in my investigations in the future, though hindsight of all that I've seen has not brought to remembrance any layer anywhere that looked like a solid layer that broke apart into angled layers like crossbedding.

Addendum, April 25:  Only weeks after we completed these experiments, I was out on a field trip with Mike Oard and Andrew Snelling in the Rattlesnake Mountains water gap in Montana.  I stumbled upon this layer which usually appears as a simple layer of sedimentary rock.  However, differential erosion had revealed that it was indeed crossbedded, but the crossbeds are not visible except by differential erosion.

Again remembering our model of tidal formation of layers, we would have a main tidal wave every twelve hours.  Riding on top of this wave would be countless smaller waves; perhaps as big as ocean waves today - which easily achieve 5 to 10 feet high.  In this particular experiment, waves were superimposed on the flow of sediment and water being introduced.  The waves were not deliberate, but rather simply the result of the equipment being used.  As a wave would charge into the standing water, it would displace the standing water with a standing wave.  This wave would then collapse into the "vacuum" left behind at the face of the crossbed, slamming the sediments into the crossbed and producing incredibly steep crossbeds.  Here's a video of it.

In an attempt to make two sets of  crossbeds on top of each other (much like is seen at Zion National Park), we performed two runs.  We produced crossbeds in the first run with the flume merely tilted uphill at 1 degree.  We then blocked the drain end of the flume, creating a 4" high dam, and filled the flume with standing water.

While Andrew and I objected to Berthault's dam at first, we realized that the dam was not the point:  The standing water was the point.  There is a variety of ways that standing water can be produced inland during a global flood:  The rains being trapped, lakes, small seas, etc...  I had proposed that because the east coast had essentially no crossbeds, yet the west (Arizona through Utah) had extensive crossbeds, that perhaps this is the where the two water flows of Noah's flood met (the Rocky mountains having not yet formed)- one flow from the east coast, and one from the west coast.  Andrew shot this idea down in flames by pointing out the dinosaur tracks among and above the crossbeds.  However, later on I also proposed that one big wave will build up a heap of sediments along a shoreline.  When we are dealing with a global flood, I have no qualms envisioning a very large sedimentary build-up forming a dam on the shores of the coasts; thus the dam is not in front of the flow, but rather behind it.  This dam would trap water inland from the last tidal wave.

At any rate, standing water was the key, so we produced some by merely blocking the end of the tank and filling it up, on top of our previously formed crossbedded layer.  We then ran an agressive run, same as before.

Global flood skeptics have argued that wet sand will not produce crossbeds as steep as dry sand.  Such a suggestion is ridiculous:  If one merely takes a moment to ask oneself, "Which can produce a steeper bank?  Dry sand?  Or wet, sticky sand?", the answer becomes quite obvious.  We also had Dr. Floyd with us on the last day of the runs, and he surprised me by saying that the geology textbooks specifically say that water will not produce crossbeds steeper than 30 degrees.  This amazes me because we produced 37 degree crossbeds with little effort, using fairly crude techniques!  I am fairly confident that if we worked at it, we could achieve crossbeds meeting or exceeding 40 degrees.  This photo is from the run we performed for the TV crew:

The grain size had no effect on the angle.  However, in our experiments, because of the equipment we were using, grain size tended to coarsen throughout the run.

Further crossbeds, and the reactions of newts:

We also ran one experiment which produced crossbeds with newts in the water.  This was done to examine their behaviour in flood conditions which produce crossbeds, in hopes that our observations would shed light on the prolific fossil tracks found in the coconino sandstone crossbeds - which I think it did.

To finish off the experiment and produce crossbeds to be left for the next day when a TV crew that was there, we cleaned up the flume and loaded the mixer with a double load of sediments.  We left the 4-inch high "dam" at the end of the flume and put in some standing water, though it was not filled completely.  The newts being as newts are, were quite content in the water and very docile.  It probably would have been better to have creatures (such as lizards which are not amphibian) which are not inclined to "hang out" underwater, but the newts still provided quite an education.

The crossbeds were produced, same as before.  While one newt swam around, the second was quite content to stay at the bottom of the crossbeds being formed.  The answer became obvious:  he was sitting the eddy currents; the place where the water was the slowest.  Thus, the newt really didn't have to move or fight any current.  He was quite content to just sit there.
The encroaching crossbeds would eventually begin to cover him up, so the newt would simple "step up" onto the new crossbed.

Several lessons were learned:

Conclusions of crossbed research:

I'll interject my own, personal opinion here which is not necessarily shared by M.E. or Andrew: I am now quite convinced that the crossbeds of the coconino and navajo formations (as well as gravel crossbeds in various locations) are produced by water; convinced to the point that I will be dogmatic about it.  The evidence overwhelmingly points to a watery origin.

Crossbeds as a paleocurrent indicator:

Water-formed crossbeds are, in my opinion, easy to recognize compared to wind-blown sand dunes.  Wind-blown sand dunes have remnants of the windward and lee sides preserved somewhat in the crossbeds.  For example, this is a photo of a sand dune in eastern New Mexico that had been cut by a bulldozer:

While the dune did have bedding planes (layers), and if one were to look strictly at one side, one might interpret that one side's layers as "crossbeds."  However, looking at the breadth of the dune, one can see the layers curve right over to the lee side (on the left), within only a few feet.  The crossbeds we see throughout the west go on for many, many miles with no windward side evident.  This is exactly what we would expect with a continentally-deposited crossbed layer, and completely contrary to what we see with modern sand dunes.  While a lot of the crossbed layers we see in the stratigraphic record are considerably thinner than the height of the sand dune above, we can see layers on both sides of the sand dune (roughly 12 feet high)- but never see the windward side of the crossbeds.

There is one wildcard here:  Andrew would suggest that there are many, giant sand dunes in deserts today which were laid there during the flood; and I would tend to agree.

Addendum, April 25:  David Lines pointed out that there are clear crossbeds within the White Sands of New Mexico which match our crossbeds identically.  This of course has been used to argue that wind-blown sand dunes produce crossbeds.  However, I would contend that this evidence precisely demonstrates that the white sands were originally laid down by water and are now being reworked by the wind! The above photograph is of a sand dune which has clearly been formed only by wind.  The dune has moved enough that if it had been originally laid down by water, any remnants of the layering left behind by the working of that water has been destroyed by the reworking of the wind.  Thus, what we see are only the effects of wind and not large quantities of water.  Crossbeds are only formed by water.

Thus, water-produced crossbeds which are positively identified within the stratigraphic record (be they sands, gravels or boulders), can be used as a paleocurrent (ancient water flow direction) indicator.  I have personally examined crossbedded layers by the hundreds throughout North America, and I cannot think of a single one that even has the potential to be a wind-blown sand dune.  They are all missing the tell-tale windward side of the dune.  Thus, we can incorporate crossbeds into the mapping of megatrends in paleocurrents:  A valuable study reflecting what went on during the global flood of Noah.

Liquefaction Experiments:

Liquefaction is a state in which sediments are temporarily suspended in water, usually from water percolating up through them.  This effect can be seen by working wet concrete, vibrating mud, or even during earthquakes.

Archimedes: Click on the image to view an introductory video

was built by our late friend, Don Yeager, from Oklahoma.  Sadly, Don passed away literally the day we returned home after performing our research.  Archimedes consists of a sealed acrylic box designed to withstand some pressure.  Spaced off of the bottom is a membrane which allows water to pass through but not sediments.  Beneath this is the inlet from the water pump, and water from this pump goes through a series of baffles to spread out the flow so that it is as uniform as possible throughout the base of the entire unit.
Sediments are loaded into Archimedes on top of the membrane, and the pump intake sticks down from the top of the unit.
In the center of the top is a large, rolling-gasket piston which cycles up and down to induce pressure upon the water and sediments inside the unit.  This is to simulate the pressure of waves during the global flood, and the pump's water flow is to induce liquefaction of the sediments.  The two mechanisms can be used separately, or in conjunction with each other.

I came to the table with a long-standing desire to perform research in this area of liquefaction, as it relates to the global flood of Noah.  I had high expectations that not only would the process produce layers, I had more than one model I had developed in which I used liqeufaction to explain anomolies in the geological and fossil record.  I suspected this research would verify some suspicions I had.

Much to my surprise, it became evident very quickly that liquefaction does not produce layers, it destroys them.

I do need to qualify this statement however:  liquefaction did indeed sort (more or less) the sediments by density.  However, the resulting "layers" were hardly layers at all; they blended together and if the system was to become lithified (cemented, or hardened into rock), it would be one, thick block.  If I saw these layers in the geologic record, they would be interesting and noteworthy, but I wouldn't call them layers; I would call it a layer fining upward.
Futhermore, in an attempt to homogonize (uniformly mix-up) the sediments that were loaded into Archimedes, I stuck a high pressure garden hose into the pump return hole and blasted the sediments with high-speed water.  To my surprise, this made layers!  In fact, try as hard as I could, the only thing that best homogenized the sediments was liquefaction!

Here's a video to show what I mean.

Some have suggested (and I personally believed, until now) that cycles of liquefaction during the flood were what produced layers.  To affirm/refute this, long period cycles were run in Archimedes.  All effects were finished with about 30 seconds, whether liquefying or settling the sediments.  We ran 20 cycles of 1 minute duration, pump-induced liquefaction, followed by 1 minute of settling (no moving water).  The results were virtually identical on each and every cycle - to the point that it was boring, it was so predictable.  It did not produce anything I would call layers, but did definitely (and very, very rapidly) destroy the very definite layers I had inadvertantly produced! 

I did run a few long-period cycles with the piston being operated simultaneously, both during liquefaction and settling cycles.  The pumping action had no visible effect, except to flex the 1/2" acrylic walls in and out.  To be honest, I did not expect the pressure differential to accomplish much.  The flexing was enough that it was producing more of an effect than the pressure difference; so the piston action was abandoned.

In the end, the results were the same, no matter what.  If we ran the pump any longer than 30 seconds, no change was noted, and no layers were recovered.

There has been some question of flow rates, and this is part of future research.  Flow rates will be controlled very accurately, but I strongly suspect this will make no difference on the final outcome except the time required to produce the same results.

Introducing a heresy:

Andrew and I both share a simliar skepticism for the metamorphic interpretation of gneises and schists, and after examining the Llano granite uplift, we both came to the same conclusion:  It's a giant, sedimentary rock dome.  I know for myself, I believe granite simply has a supernatural origin - there is no natural way to produce it.  Contrary to common belief, it is impossible to form it from a melt.  This has been borne out both in the lab in and in nature.  While Andrew and I both agree that the granite batholith was a sedimentary rock, it's formation still requires previously existing granites!  It is granite that has simply been crushed up, transported, and relithified elsewhere.  This is a continuing research which I will not discuss here.

One thing I personally noted with the liquefaction sediments was a stark resemblance to schist, gneiss and granitic outcrops I've examined in so many different places:  They have stratification, but it's a disordered mess, in the midst of giant plumes.  This is precisely what we saw, on a small scale, in the liquefaction tank.

The liquefaction went through several distinguishable phases:  Plumes, which brought lower layers through to the top, which caused a tilting of the upper layers downward.  This led to "boiling" where all of the layers would eventually go to vertical or near vertical, followed by collapse of all of the structures, including the plumes.  I have seen all of these stages within the rock record, namely in the "basement" rocks.
While I cannot be dogmatic on this, it appears as though the granite plume now known as the Llano uplift, was precisely that:  A plume.  However, it was not formed by a melt (as is conventionally believed), as that is impossible - so it must have been formed "cold" or at lower temperature.  I am suggesting that water, supersaturated with silica, produced a liquefaction plume of granitic gravels.  The silica precipitated out of the water, cementing the granite gravels into sedimentary granites.  The cementing silica (quartz) appears to simply be a part of the granite, as quartz is one of the three main constituents of granite.

Andrew and I were supposed to perform considerable research into silica supersaturation and sedimentary cementation while at Glen Rose, however we simply ran out of time.  Andrew has pointed out some processes which are now known which greatly simplify silica super-solubility, even in room termperature water.  This may play a major role in explaining the massive beds of silica-cemented sediments around the world.

Conclusions to liquefaction:

Liquefaction doesn't produce layers, it destroys them.  However, it may very well be the father of plumes (such as those seen at Kodachrome basin) and the presumed metamorphic rocks referred to as the gneiss and schists so common throughout the Canadian Shield and in the bottom of the Grand Canyon.  Some granites and granite "dykes" within these rocks may also very well be simply the cemented sediments from a liquefaction event - layers that were tipped up during the liquefaction process and solidifed before liquefaction destroyed all of the structures.

Addendum, February 2011:
Dr. Walter T. Brown has expressed disagreement with my conclusions on this page, specifically regarding liquefaction.  He claims that what we acheived with "Archimedes" was not liquefaction, but rather disruption. Dr. Brown has a rather large chapter of his book devoted to liquefaction here:  
One of our researchers, Professor M.E. Clark, had built a reproduction of Dr. Brown's liquefaction apparatus (shown here, figure 94), but was unable to replicate Brown's results. It was actually this failure to reproduce results that led to the design and construction of Archimedes, but Dr. Brown has contended that there was an error in Clark's methodology and/or construction of the apparatus.
I do have to agree with at least some of Brown's criticisms, though during our research there were considerably more applicable observations which I did not report here. These other observations, as well as Dr. Brown's critique have raised questions which I'm convinced will lead to more exciting discoveries, but to date, we have not been able to get back to the liquefaction research.  
Sadly, after a year of battling cancer, Professor Clark passed away in December of 2010.

References and footnotes:

1)  M. E. Clark and H. D. Voss, Resonance and Sedimentary Layering in the Context of a Global Flood, Proceedings of the Second International Conference on Creationism, R. E. Walsh and C. L. Brooks, Editors, 1991, Creation Science Fellowship, Inc., Pittsburgh, PA, Vol. 2, pp. 53-63.