Rheological measurements were used to evaluate the particle-particle associations of Na-rich montmorillonite in suspensions, under various electrolyte concentrations. A 2% free electrolyte clay suspension showed pseudoplastic flow behavior and had a high apparent viscosity, attributed at low shear rates to the high volume fraction of the suspended clay platelets, the flexibility of the platelets, and the presence of edge-to-edge association. The breaking of edge-to-edge associations and the progressive orientation of the individual platelets in the direction of flow contribute to the reduction in viscosity with increasing shear rate.

The compression of the diffuse double layer at a NaCl concentration of 10 mEq L-1 contributes to the free movement of the individual platelets, even at low shear rates. The flow behavior changed from pseudoplastic to plastic at an NaCl concentration of 100 mEq L-1. At this electrolyte concentration, face-to-face associations of specific junction points at certain areas of the planar surface are probably occurring.

The apparent viscosity of the clay suspension for the two particle-size ranges (<2 and <0.02 μm) at all shear rates converged to a minimum value of 4.5 mPa s at NaCl concentrations of 10–20 mEq L-1. On both sides of the minimum, the lower the shear rate, the greater the slope. The apparent viscosity of a 2% suspension of Na-rich montmorillonite <0.02 μm particles, however, was significantly greater than that observed for a suspension of <2 u,m particles. This high apparent viscosity is attributed to the increase in edge surface area and the number of clay particles in a unit volume.

We suggest that edge-to-edge association between Na-rich montmorillonite platelets prevails when the NaCl concentration is below the electrolyte critical concentration, for which the apparent viscosity of the suspension is at its minimum value, whereas face-to-face association prevails at NaCl concentrations above this critical value.

The cause of pH and ionic strength-dependent proton and hydroxyl adsorption onto kaolinite is specific binding at edge Al and Si sites, and it can be modeled as a function of temperature with a triple layer model (TLM) of the mineral-solution interface. Exchange of Al for protons and hydroxyls is observed at low pH, with a stoichiometry approaching 1:3 (Al:H+). Adsorption of organic acids from dilute solutions depends on: 1) solution pH; 2) the functionality of the acid; and, to a lesser extent, 3) temperature. Such adsorption may occur primarily at Al sites exposed on kaolinite edges, as indicated by sorption experiments on the constituent oxides, where negligible sorption was observed on SiO2 (quartz), but was significant on Al2O3 (corundum) surfaces. Under similar conditions, oxalate adsorbs more strongly than acetate or formate to aluminol sites.

Montmorillonite dissolution under highly alkaline conditions (pH = 13.3; I = 0.3 M) was investigated by bulk dissolution methods and in situ atomic force microscopy (AFM). In bulk dissolution experiments, initial SiO2 concentrations were high, and a steady state was reached after 136 h. The dissolution rates derived from the edge surface area (ESA) at the steady-state condition at 30, 50 and 70°C were 3.39 x 10−12, 1.75 × 10−11 and 5.81 × 10−11 mol/m2 s, respectively. The AFM observations were conducted under three conditions: (Run I) short-term in situ batch dissolution at RT; (Run II) long-term in situ flow-through dissolution at RT; and (Run III) long-term batch dissolution at 50°C. The observed reductions in montmorillonite particle volume for Runs I and II were due primarily to edge-surface dissolution. The ESA-based dissolution rate for Run I (10−9 mol/m2 s) was three orders of magnitude faster than that for Run II (10−12 mol/m2 s). The rate obtained for Run II corresponded to the rate at the steady-state conditions in the bulk dissolution experiments. A small number of etch pits developed in Run III slightly increased the ESA of montmorillonite since most of the montmorillonite particles were separated into monolayers lacking three-dimensional periodicity. The ESA-based dissolution rate for Run III was 2.26 × 10−11 mol/m2 s. Dissolution rates based on long-term AFM observations could be directly compared with steady-state rates obtained from bulk dissolution experiments. The AFM observations indicated that dissolution occurred at edge surfaces; therefore, the ESA should be used to calculate the dissolution rate for montmorillonite under alkaline conditions. Dissolution rates of individual particles with different morphologies estimated by AFM were similar to rates estimated from bulk dissolution experiments.

Ferrihydrite is a poorly crystalline Fe oxide of which 2-(XRD)line and 6-line varieties are commonly used in experiments, although species with intermediate numbers of peaks have been found in nature. To simulate nature, we synthesized two continuous series of ferrihydrites with between 2 and 6–7 peaks at room temperature in two different ways: (1) by varying the rate of hydrolysis of an Fe(NO3)3 solution (HR series); and (2) by oxidizing an FeCl2 solution containing up to 73 mmol Si/L (Si series), both at pH 7. Mössbauer spectra at 4.2 K showed that the ferrihydrites of the HR series had a constant magnetic hyperfine field (Bhf) at 4.2 K of 48.8 T whereas Bhf in the Si series dropped from 49.4 to 46.7 T as the Si content of the ferrihydrites increased from 0 to 74.7 g/kg of Si. Temperature scans between 4 and 170 K illustrate that the magnetic order breaks down at a temperature which is lower the higher the hydrolysis rate and the Si concentration in the ferrihydrite.

The crystal structure of kaolinite(Pl, a = 5.153(1), b = 8.941 (1), c = 7.403 (1)Å, α = 91.692(3)°, β = 104.860(3)°, γ = 89.822(3)°, specimens from Keokuk geodes) has been refined in detail and that of dickite (Cc, a = 5.1460(3), b = 8.9376(5), c = 14.4244(6) Å, β = 96.761(5)°) has been re-refined, both from powder diffraction data with the Rietveld method. Except for the hydrogen atoms, the layer structures in both clays are very similar and are much as inferred or determined previously by others. The rotation in the tetrahedral sheet is 7(1)°. The two inner hydroxyl O-H bonds in kaolinite are differently oriented; one points into an octahedral vacancy and the other somewhat away from the octahedral sheet and toward the unoccupied center of an oxygen triangle formed by the two apical oxygens and shared basal oxygen of two adjacent SiO4 tetrahedra. All six of the inner surface hydrogen atoms appear to be nearly equally involved in the hydrogen bonding between kaolinite layers in kaolinite.

The purpose of this report is to describe the appropriate use of indices relating to crystallinity, such as the ‘crystallinity index’, the ‘Hinckley index’, the ‘Kübler index’, and the ‘Árkai index’. A ‘crystalline’ solid is defined as a solid consisting of atoms, ions or molecules packed together in a periodic arrangement. A ‘crystallinity index’ is purported to be a measure of crystallinity, although there is uncertainty about what this means (see below). This report discusses briefly the nature of order, disorder and crystallinity in phyllosilicates and discusses why the use of a ‘crystallinity index’ should be avoided. If possible, it is suggested that indices be referred to using the name of the author who originally described the parameter, e.g. ‘Hinckley index’ or ‘Kübler index’, or in honor of a researcher who investigated the importance of the parameter extensively, e.g. ‘Árkai index’.

A structure refinement of kaolinite made using the Rietveld neutron profile refinement technique has given non-hydrogen atom positions which were not significantly different from those given by B. B. Zvyagin in 1960. All of the hydrogen atoms have been located; the three inner-surface hydrogen atoms are involved in interlayer hydrogen bonds with lengths of 2.95(4), 2.95(4), and 3.06(4) Å with O-H ... O angles of 168(4)°, 144(4)°, and 146(4)° respectively. The inner hydrogen atom is located in a position consistent with that found earlier in dickite and muscovite which are the only dioctahedral layer silicates studied by neutron diffraction to date. The O-H vector makes an angle of 34° with the (001) plane, away from the octahedral sheet, and the projection of the vector on to (001) is at ~30° to the b axis.

An overlap of bands produced by the O−H stretching vibrations of H2O (O–Hw) and structural OH (O−Hs) in smectite hampers the study by infrared spectroscopy (IR) of both their layer and interlayer structure. The present study re-evaluated the D2O saturation of smectite as a tool to enable separation of the overlapping bands at ambient conditions. Real-time monitoring by Attenuated Total Reflectance infrared spectroscopy (ATR-IR) was employed during in situ sample drying and H2O or D2O saturation at ambient temperature. Six dioctahedral and one trioctahedral pure smectites in Ca2+-, Na+-, and Cs+-cationic forms were studied to explore variability in total layer charge, charge location, and interlayer cation. The IR data showed the interlayer O−Dw signature at 2700–2200 cm−1 as a proxy for the O−Hw signature in the 3700–3000 cm−1 region. In addition to the expected liquid-like bands of D2O in the interlayer, these O−Dw spectra exhibited an additional sharp stretching feature in the 2695–2680 cm−1 range. No significant cation dependence of the sharp band position was observed between pairs of Ca- and Na-smectites for relative humidity (RH) between 60 and 80%, despite the large difference in the ionic potential between these interlayer cations. The intensity of the sharp band was found to be almost insensitive to changes in water content within the range 60–80% RH. The sharp band frequency decreased linearly with increasing total charge of the 2:1 layer (and can be used as a proxy for it), but no effect of charge location could be discerned. In agreement with early studies, this band was attributed to D2O located on the surface of the interlayer, pointing one O−D group toward the siloxane surface. Based on its high frequency, this band was indicative of free O−D oscillators, with very little or no involvement in hydrogen bonding (“dangling OD”). By analogy to the spectra of D2O-smectites, the spectrum of H2O-smectites also involves a sharp O−Hw analog at ~3630 cm−1 overlapping with typical OHs bands (e.g. Al2OH). As a result of this overlap, the sharp 3630 cm−1 O−Hw contribution was often missed or attributed solely to O−Hs.

A IIb orthochlorite (brunsvigite), from the chloritic metabasalt in the Middletown Valley of Maryland, was altered to a regularly interstratified chlorite—vermiculite after four months reaction in saturated bromine water on the steambath. The artificial weathering product resembles the regularly interstratified chlorite—vermiculite in a soil in Adams County, Pennsylvania, that has developed in greenstone similar to the chloritic metabasalt in the Middletown Valley of Maryland. The results of this and previous investigations on artificial and natural weathering of chlorites indicate that oxidation of structural Fe2+ is an important reaction in the alteration of chlorites into vermiculitic products. The results also show that IIb orthochlorites, apparently structurally similar but having different Fe2+ contents may react differently under the same weathering conditions. For example the acid oxidation treatment altered a IIb chlorite (diabantite) directly into a vermiculite, whereas the same treatment altered another IIb chlorite (brunsvigite) into a regularly interstratified chlorite—vermiculite. Detailed investigations of the chlorite structures are probably necessary to determine whether or not these alteration processes are structurally controlled.

This article emphasizes the influence of crystal-chemical composition on the ease of transformation of micas to vermiculite or more smectite-like minerals. The swelling test, with glycerol, of Mg-saturated mineral is used to characterize the degree of transformation.

The main structural factors of this evolution are tetrahedral substitution of Si by Al, and total charge. There is a relation between these two factors, i.e. the lower the tetrahedral substitution, the greater can be the charge without affecting the smectite swelling behavior.

In this respect there is a contrast between tri and dioctahedral micas. In the first, tetrahedral Al is so high (> 1·20 for Si4O10) that transformation into smectite must imply modification of the tetrahedral layer. For dioctahedral micaceous phyllites (illites, glauconites) where tetrahedral charge is lower, transformation can be easier. Only a lowering of total charge is needed and reduction-oxidation seems to play a very important role in this process.

Because of the large surface area and common occurrence of smectites in the clay fractions, they are important in sorption/desorption reactions of organic pollutants entering soils and sediments. Results of sorption and desorption of atrazine by Al- and Ca-saturated smectite reported here showed that Al-saturated smectite sorbed much higher amounts of atrazine than Ca-saturated smectite. Al-saturated smectite sorbed 3820 mg kg−1 as compared to 1902 mg kg−1 by the Ca-saturated smectite during 5 consecutive sorptions from 10 ppm atrazine solution. Sorption isotherms over 2–10 ppm atrazine concentration range were nonlinear in all cases. Freundich coefficients, Kf, obtained from the sorption isotherms were much higher for Al- than Ca-saturated smectite; Kf varied from 405 to 3035 for Al-saturated and 100 to 306 for Ca-saturated. The pH values of Al- and Ca-saturated smectites were 3–4 units higher than the pKa (1.68) of atrazine, which suggests that atrazine was sorbed as neutral molecules. Stronger H-bonding between the more polarized H2O associated with the trivalent Al ion than the divalent Ca ion is likely responsible for the greater sorption by Al-smectite.

The full structure of dickite from St. Claire, Pennsylvania, including hydrogen atoms, was refined in space group Cc using time-of-flight neutron powder diffraction data obtained at 12 K and Rietveld refinement/difference-Fourier methods (Rwp = 2.62%, reduced χ2 = 1.915, 113 variables, a = 5.1474(6)Å, b = 8.9386(10)Å, c = 14.390(2)Å, and ß = 96.483(1)°). The non-hydrogen structure is essentially identical to published structures for dickite, but the hydrogen positions are distinct. The inner hydroxyl group is approximately parallel to the (001) plane, inclined by 1.3° towards the tetrahedral sheet. Contrary to published low-temperature infrared (IR) spectra, there is no evidence that dickite possess lower symmetry at low temperatures although there is tentative evidence for statistical occupancy of H3 on more than one site. Low-temperature IR spectra of St. Claire and Wisconsin dickites do not show evidence for more than four hydroxyl groups and are consistent with the reported structure. Upon cooling from 300 to 15 K, the position of the OH3 stretching band increased from 3710 to 3731 cm−1. This large, positive shift in frequency was attributed to the increase in the internuclear O-H3 ⋯ O distance upon cooling. The frequency of the 3655 cm−1 band initially decreased by 2 cm−1 to 3653 cm−1 upon cooling from 300–125 K; however, the band increased in frequency by 1 cm−1 upon further cooling to 15 K. This unusual change in frequency upon cooling is consistent with the assignment of this band to OH2 and OH4. The position of the OH1 stretching band decreased from 3622 to 3620 cm−1 upon cooling, which was attributed, in part, to the observed increase in the Al-O(H1)-Al angle at low temperature.

Crystal chemical analysis of various dioctahedral 2:1 phyllosilicates consisting of trans-vacant (tv) and cis-vacant (cv) layers and interstratified cv and tv layers shows that there is compositional control over the distribution of octahedral cations over trans and cis sites. Fe3+ and Mg-rich dioctahedral micas (celadonite, glauconite, leucophyllite and most phengite) occur only as tv varieties. Similarly, the occurrence of tv illites and tv illite fundamental particles in illite-smectite (I-S) does not depend significantly on the cation composition of the 2:1 layers. In contrast, compositional restrictions exist to control the occurrence of pure cv1M illite, which can form only as Fe- and Mg-poor varieties. Similarly, proportions of cv and tv layers in illite fundamental particles depend on the amount of Al in octahedral and tetrahedral sheets of the 2:1 layers.

Simulations of atomic coordinates and interatomic distances for periodic tv1M and cv1M illite structures allow us to reveal the main structural factors that favor the formation of cv layers in illite and I-S. It is shown that in contrast to the tv1M structure, interlayer K in cv1M illite has an environment which is similar to that in 2M1 muscovite. This similarity along with a high octahedral and tetrahedral Al content probably provides stability for cv1M illite in low-temperature natural environments. Because of structural control, the occurrence of monomineral cv1M illite, its association with tv 1M illite, and interstratified cv-tv illite fundamental particles is confined by certain physical and chemical conditions. These varieties are most often formed by hydrothermal activity of different origin. The initial material for their formation should be Al-rich and the hydrothermal fluids should be Mg- and Fe-poor. They occur mostly around ore deposits, in bentonites and in sandstone sedimentary rocks.

The factors governing the formation of tv and cv layers in dioctahedral smectite are probably related to the layer composition and local order-disorder in the distribution of isomorphous octahedral cations, because there is no influence from fixed interlayer cations. In particular, the occurrence of Mg-OH-Mg cation arrangements is more favorable for the formation of cv montmorillonite layers.

The sorption of Co and Zn by layer silicates was studied in dilute mineral suspensions containing, in most cases, about 10-6 M Co or Zn. Electrostatic adsorption was eliminated by the presence of 0.1 N CaCl2. Co and Zn were determined by a radioisotopie technique.

Detailed studies on montmorillonite and to a lesser extent vermiculite, muscovite and biotite revealed the presence of at least two forms of specifically sorbed Co, one of which was exchangeable by certain ions such as Cu, Ni, Zn, Fe, Mn or more Co and the other, occurring in much smaller amounts, was not exchangeable by these cations. The latter form is considered to result from lattice penetration; the former is associated with surface groups. These forms may be separated quantitatively by several desorption procedures including successive extractions with dilute acetic acid.

The total amount and relative proportion of these forms of sorbed Co and Zn depend on the pH of the system, time of reaction, mineral species used, and amount of Co or Zn added. Equilibrium is not readily attained but tends to approach a slow steady state after several days. Isotherms indicate a variation in bonding energy with surface coverage. The sorption of Co and Zn from dilute solutions by any mineral is related to its stability.

These studies, considered in conjunction with published data, suggest that a common mechanism may be involved in the specific sorption of many heavy metal cations by many minerals.

Crystallization processes of the illite-smectite (I-S) mixed-layer mineral series during alteration of felsic vitric materials in volcaniclastic sediments through two drill holes (IT-2 and IT-8) near the Kakkonda active geothermal system, Japan, were examined by optical microscopy, scanning and transmission electron microscopy (SEM and TEM), electron microprobe analysis, X-ray diffraction (XRD), and oxygen isotope analysis. Temperatures measured through the drill holes increased nearly linearly with depth up to 317°C at the bottom (1700 m) of IT-8. Homogenization temperature measurements of fluid inclusions indicated that the alteration occurred at temperatures similar to the present temperatures. In selected volcaniclastic rocks, excluding andesitic rocks and black shales, clay minerals occurred as glass replacements and pore fillings as seen under SEM and optical microscopy, and exhibited predominantly euhedral hexagonal and elongated forms under TEM, implying that they precipitated in situ through hydrolytic reactions of glass and fluid. Based on XRD examination, I-S minerals showed a sigmoidal variation in illite layer percentage (%I) in the range of ∼150 to 220°C and R0 I-S minerals with intermediate %I between 20 and 40% rarely occurred (where R is the Reichweite parameter). The chemical composition also showed a specific variation with depth. Intermediate clays including smectite and I-S minerals are enriched in Al compared to those reported previously from hydrothermal alteration of almost equivalent parent rocks. The oxygen isotope data indicated that the reacting solution was percolating groundwater in the shallow levels and with fossil seawater in the deeper levels. Furthermore, calculating the fluid/rock (W/R) ratio from the isotope variations revealed that the alteration occurred at a nearly constant W/R ratio condition irrespective of %I. Consequently, the observed specific variations in structure and chemical composition of I-S minerals reflect the compositional variations of fluid participating in the crystallization at given temperatures under the conditions of a given original rock and constant W/R ratio. High pH and Na-rich solutions generated by progressive hydrolytic reactions between felsic glass and groundwater favored the precipitation of Al-rich smectite up to ∼150°C and was followed by precipitation of an aluminous R1 I-S mineral with few intermediate R0 I-S minerals at higher temperatures. The crystallization obeys Ostwald’s step rule behavior of smectite illitization processes under a high geothermal gradient.

The effects of exchangeable Na and Ca and the ionic strength of NaCl and CaCl2 solutions on the boron adsorption by the Na and Ca forms of montmorillonite and illite at pH 9 were studied. The boron adsorption by montmorillonite was greater for the Ca form than for the Na form at any boron activity in solution and at any ionic strength. For example, the amount of boron adsorbed by Na- and Ca-montmo-rillonite was 2.9 and 4.3 μmole/g, respectively, at an ionic strength of 0.36 and a boron activity of 0.875 mmole/liter in equilibrium solution. The boron adsorption by illite was much greater than by montmorillonite. For example, the amount of boron adsorbed by Na-montmorillonite and Na-illite was 1.5 and 8.7 μmole/g, respectively, at an ionic strength of 0.02 and a boron activity of 0.5 mmole/liter. However, the nature of the exchangeable cation had little effect on the boron adsorption by illite. The effect of the ionic strength on the boron adsorption by Na-montmorillonite was greater than that observed for either Camontmorillonite or illite. The data suggest that the negative electric field around the clay particles is one of the main factors controlling boron adsorption at alkaline pHs.

The sorption of the uranyl oxo-cation (UO22+)at different types of binding sites on layer silicate mineral surfaces was investigated. Well-characterized samples of vermiculite and hydrobiotite were exposed to aqueous uranyl under conditions designed to promote surface sorption either at fixed charge ionexchange sites or at amphoteric surface hydroxyl sites. The local structure of uranium in the sorption samples was directly measured using uranium L3-edge extended X-ray absorption fine structure (EXAFS). Polarized L1- and L3-edge X-ray absorption near-edge structure (XANES) measurements were used to characterize the orientation of uranyl groups in layered samples. X-ray diffraction (XRD) measurements of interlayer spacings were used to assess the effects of ion-exchange and dehydration upon the mineral structure. The most significant findings are: (1) Under conditions which greatly favor ion-exchange sorption mechanisms, uranyl retains a symmetric local structure suggestive of an outer-sphere complex, with a preferred orientation of the uranyl axis parallel to the mineral layers; (2) Upon dehydration, the ionexchange complexes adopt a less symmetric structure, consistent with an inner-sphere complex, with less pronounced orientation of the uranyl axis; and (3) For conditions which favor sorption at surface hydroxyl sites, uranyl has a highly distorted equatorial shell, indicative of stronger equatorial ligation, and the detection of a neighboring U atom suggests the formation of surface precipitates and/or oligomeric complexes.

Tubular halloysite from Wagon Wheel Gap, Colorado and spheroidal halloysite from Redwood County, Minnesota were examined by transmission electron microscopy. Clay samples were prepared by the following techniques: drop-mounted suspension on carbon support films; thin sections of clay in Araldite epoxy resin; and carbon-platinum-palladium single-stage replicas.

Both types of dehydrated halloysite have interlayer separations between packets of layers. Halloysite tubes are composed of packets as thin as five layers which sometimes reveal a rolled interior configuration in cross-sectional view. Thicker tubes are composed of many layers per packet. Some large tubes appear in cross section as folded packets of layers. The interior morphology of spheroidal halloysite particles is more irregular and the layer structure is more discontinuous than in most tubes. The spheroidal halloysite of this study is characterized by external tangential plates with hexagonal shape suggestive of kaolinite.

Aqueous gels of unaltered (oxidized) and chemically reduced ferruginous smectite (SWa-1 from the Source Clays Repository of The Clay Minerals Society) were characterized by transmission electron microscopy, electron diffraction, and energy-dispersive X-ray fluorescence to establish details regarding their texture, inter-layer and inter-particle arrangements, and chemical composition. Micrographs revealed that the reduction of structural Fe(III) to Fe(II) caused a consolidation of smectite particles from an extensive network of small crystals (1–6 layers thick) to distinct particles of limited size in the a-b direction and about 20–40 layers thick. The interlayer distances in the reduced sample appeared to be more uniform than in the oxidized sample, but both exhibited spacings of about 12.6 A. Chemical analysis showed no qualitative differences as a result of oxidation state. Electron diffraction patterns displayed marked differences. The pattern of the oxidized sample consisted of homogeneous rings, indicating that the stacking order in the a-b plane was turbostratic or disordered, whereas the reduced pattern exhibited much more order as evidenced by distinct spots amid low-intensity rings, suggesting that inter-layer attractive forces were stronger if Fe(II) was present in the clay crystal.

Scan electron micrographs (SEMs), are shown of three representative types of Georgia kaolin: ‘soft’, Cretaceous-age clay; ‘hard’, fine-grained clay; and ‘flint kaolin’.

Sparta Granite and a thick deposit of its overlying saprolite are taken to serve as examples of probable source rock for Georgia kaolin. SEMs are presented to show sequential argillation of this fresh granite through a transition zone to the saprolite. The saprolite contains abundant books of kaolinite similar to those occurring in Cretaceous kaolin, accompanied by some elongates, whereas in the transition zone elongates are predominant. A morphologic type of fibrous, elongate 1:1 clay mineral, hitherto not illustrated, is shown from kaolinizing Sparta Granite, and from Pre-Cambrian saprolitic gneiss 25 km south of Athens, GA.

Whereas kaolin elongates originated in contact with feldspar in the Sparta transition zone, well-formed books of kaolinite are shown to have been formed in contact with feldspar in the saprolite from a Missouri granite. The arkose above the granite in the Missouri occurrence has altered to loosely packed books of kaolinite. The interpretation is made that book-type kaolinite, where it occurs in large deposits of saprolite or kaolin, was formed under geochemical conditions of relative chemical stability between kaolinite and feldspar, as might prevail in a geological ‘weathering crust’. It is interpreted that elongate kaolin morphology may be formed under conditions of limited or lesser geochemical stability.

SEMs show that the established book-morphology of Sparta saprolite is disarranged during secondary Holocene surface weathering and pedogenic processing.

SEMs show that books of kaolinite, although individually fragile, may be protected from destruction or delamination during artificial blunging and transport when moved within naturally coalesced pellets. Possibly original books of weathered kaolinite might similarly survive natural stream transport during sedimentational processes.

Some ‘flint kaolin’ contains molds of sponge spicules coated with low-cristobalite.