Search

SummaryKerolites from Goles Mountain (Yugoslavia), Wiry (Poland), Madison Co. (North Carolina), and Kremze (Czechoslovakia) are compared with talc and stevensite. Chemical analyses give a composition for kerolite near R3Si4O10(OH)2.nH2O with R mainly Mg and n about 0·8–1·2. Infra-red data and dehydration-rehydration experiments suggest that the additional water is partly surface-held hydrogen-bonded molecular water, lost up to about 300 °C and easily recoverable, and ‘water’ held as surface hydroxyls, lost at temperatures up to about 650 °C and less easily recoverable. The surface area of kerolite from North Carolina by nitrogen absorption measurements is 196 m2/g.X-ray data show broad basal reflections, a basal spacing (after Lorentz-polarization correction) of about 9·6 Å and a crystallite size of about five structural layers, which is consistent with the large surface area. Two-dimensional hk bands indicate a randomly stacked layer structure with b ≈ 9·14 Å and a crystallite ‘diameter’ of about 150 Å. The basal spacing shows no expansion in water and no thermal contraction up to 500°C. A partial and irregular swelling with ethylene glycol is obtained only after prolonged cxposure (several weeks). In both respects, kerolites differ from stevensite. Kerolites come close to talc in structure and composition but differ in having a highly random layer arrangement, a slightly enlarged basal spacing, which could be due to misfit of layers caused by random stacking, and weaker interlayer bonding.Kerolite is considered to be a useful varietal name for this talc-like mineral in agreement with the views of D'yakonov and of Maksimovic. It cannot be defined as serpentine + stevensite.

SummaryAn X-ray and chemical examination of deweylites reveals that they are intimate mixtures of very poorly ordered trioctahedral 2:1 and 1:1 layer silicates. The 2:1 mineral exhibits no swelling in the presence of ethylene glycol or water and is best described as an extremely fine-grained and highly disordered form of talc, for which the term kerolite is often used. Stevensite is not a component of deweylites. The 1:1 component most closely resembles a disordered chrysotile.Evaluation of the chemical data, including values for H2O +, shows that both end-members are hydrated to various degrees. This hydration is most probably attributable to the presence of water molecules and hydroxyls associated with the large surface areas and unbalanced surface bonds. Formulae for the end-members are approximately R3Si4O10(OH)2·0·3 0·7 H2O and R3Si2O3(OH)4·0·3–0·7 H2O where R is principally Mg, but samples with more or less H2O are entirely possible due to variations in crystallinity. No specific formula can be given to deweylites as they are mixtures with variable proportions of the components. The name is useful, however, as a field or ‘box’ term, similar to the use of garnierite and limonite.

SummaryMotukoreaite resembles hydrotalcite and has positively charged brucite-like layers separated by carbonate and sulphate anions and water molecules. Four phases are recognized. Phase I, the normal mineral with an 11.1 Å basal spacing, readily changes to a mixture of phase I and a partially dehydrated phase II. Phase III with a basal spacing 8.7 Å persists from 50–170°C. Phase II is a pseudo-regular interstratification of phases I and III, with a spacing 11.1 + 8.7 = 19.8 Å; second- and third-order reflections are observed. Phase IV with a 7.5 Å spacing is obtained by treating motukoreaite with sodium carbonate solution to replace SO4 by CO3 anions. Motukoreaite itself shows only slight tendency to transform to a 7.5 Å phase by heat-treatment alone. Comparison is made with the mineral carrboydite, a NiAl sulphate carbonate hydrate. The possibility is considered that sulphate anions may substitute partially in the brucite-like layers to compensate for a deficiency of hydroxyl ions.

SummaryThe composition of scarbroite, written in the form Al5(OH)13CO3·5H2O, suggests close crystal-chemical relations to gibbsite and to hydrotalcite and kindred minerals. Parametral relations a(S) ≃ 2b(G) and 3b(S) ≃ 5a(G), S = scarbroite, G = gibbsite, are obtained and are consistent with a hydroxy Al layer structure and interlayer carbonate anions and water molecules. Thermal transformations show two discrete hydrated phases with layer spacings near 8.5 and 7.65 Å and a third phase with initial spacing 6.65 Å which decreases progressively to 5.5 Å over the temperature range 100°–250 °C as decomposition proceeds. No clear break is observed between dehydroxylation and decarbonation.

In a previous paper (Brindley, 1951) chamosites from a number of localities were examined, and their structure was shown to be of kaolin-type. Furthermore, it was shown that many of the variations in X-ray powder diagrams of chamosites can be accounted for by the existence of two kaolin-type structures, one hexagonal and one monoclinic, with simply related ceil dimensions.

Previous studies of the kinetics of dehydroxylation of kaolinite and halloysite point to first-order reactions, in approximate conformity with the Arrhenius relation. Isothermal weight-loss measurements have shown that the rate constants are markedly dependent on factors such as specimen size, shape, and compaction. A technique has been developed for determining the reaction kinetics of infinitely thin disk-type specimens. The reactions are then strictly first order and the Arrhenius relation is obeyed. Activation energies of 65 and 55 kcal. per mole are obtained for kaolinite and halloysite, respectively. Comparison is made between the behavior of kaolinite and halloysite on the one hand and of macrocrystalline anauxite on the other. For anauxite, nucleation and growth of nuclei produce a sigmoid-type reaction curve, but for the fine-grained minerals it is believed that nucleation alone is the rate-controlling process. The dependence of the reaction rate on geometrical factors is attributed to the retention of water vapor within the powder specimen. The influence of water vapor on these reactions is discussed generally.

A possible structure for kaolinite was first put forward by Pauling (1930) in a paper dealing generally with the structures of layer-type minerals, chlorites, micas, &c. He suggested that the structure consisted of a sheet of Si–O tetrahedra arranged in a hexagonal network with a superposed sheet of Al–(O,OH) octahedra, the two together forming a layer of composition Al2Si2O5(OH)4 (fig. 2). Ross and Kerr (1931), from a detailed microscopic study of the kaolin minerals, concluded that the symmetry of kaolinite is ‘probably monoclinic’. The first attempt at a detailed structure analysis was made by Gruner (1932a) from X-ray powder photographs. The reflections, some 40 altogether, were indexed on the basis of a monoclinic cell with a 5·14, b 8·90, c 14·51Å., β 100° 12′.

Many solid state reactions can be represented by equations of the type F(α) =kt, where α is the fraction of material reacted in time, t. These equations can be expressed in the form F(α) =A(t/t0.5) where t0.5 is the time for 50% reaction and A is a calculable constant depending on the form of F(α). Numerical tables are given of F(α) in relation to α, and to (t/t0.5), for nine equations corresponding to reactions which are diffusion controlled, or are reaction-rate controlled, or obey first order kinetics, or follow the equations of Avrami and Erofe'ev. The application of the tables to the analysis of experimental data is described.

Clays containing kaolinite, diaspore, and boehmite are analyzed chemically and mineralogically and their high-temperature reactions in the range 1000° to 1500°C are followed quantitatively by X-ray diffraction measurements. It is shown that the development of mullite progresses toward the equilibrium value as the temperature approaches 1500°C. When considered in detail, the reactions reflect the source of the alumina. Diaspore and boehmite first form corundum at widely different temperatures which results in the reactivity of the corundum from boehmite being greater than that from diaspore, and, in consequence, mullite develops more readily from the boehmite-containing clays. Relative expansions of clays rich in diaspore or boehmite when fired in the range 1200° to 1400°C are correlated with the manner in which mullite forms in this temperature range.

This paper concludes the account of our experiments on halloysite and is to be read in conjunction with the two previous parts, to which frequent reference will be made.The effect of heat-treatment on metahalloysite was examined with the primary object of ascertaining if a recrystallization or ‘kaolinization’ of metahalloysite could be achieved.

SummaryForsterite formed from serpentine powders at various temperatures up to 800° C has been measured quantitatively by X-ray diffraction. The results are considered in relation to a reaction mechanism proposed by Ball and Taylor and are shown to be inconsistent with the details of this mechanism. An alternative mechanism is proposed by which all or nearly all the Mg ions together with some of the Si ions liberated in the reaction zones where water is formed migrate into regions where forsterite is formed. This mechanism is in closer accord with the experimental results. The retarded development of enstatite until temperatures above 1000° C are used is attributed to the nature of the reaction by which it is formed, namely a reaction between forsterite and silica.

Single crystals of magnesiowustite (Mg,Fe)O containing up to 79 at.% Fe were prepared, and the proportions of ferrous and ferric ions were determined by chemical analysis. The rate of oxidation of individual crystals was studied at temperatures from 1000° to 1100°C and oxygen partial pressures of 2 × 10−1 to 3 × 10−6 atm. The reaction follows an equation based on diffusion through a product layer. The rate constant varies with temperature according to the Arrhenius relation, and activation energies range from 52 kcal/mole with 13 at.% Fe to about 17 to 20 kcal/mole when Fe atoms occupy one third or more of the cation positions. The rate constant is independent of oxygen partial pressure within the pressure range studied. Spectrographic measurements show that Fe in magnesiowustite occupies mainly octahedral sites, but some tetra-hedral positions appear to be occupied. The mechanism of the oxidation reaction is considered.

X-ray diffraction data for the high-temperature phases show that a spinel-type structure develops with marked orientation at about 925°C. This phase is considered to be an aluminum silicon spinel with vacant cation sites. Mullite is thought to be formed by the decomposition of the spinel. Silica is eliminated progressively as metakaolin transforms to the spinel phase and thence to mullite. The X-ray data show variation in the mullite parameters between 1200° and 1400°C.; at 1400°C., the composition probably approximates 3A12O3·2SiO2. The nature of the intermediate spinel-type phase is discussed in relation to the crystal chemistry of spinels.

Magnesium oxide prepared by calcining the carbonate at 1000°C was sintered at 1800°C and ground under methanol. Well-sized 2 to 5 and 5 to 10μ fractions were separated. After outgassing at 550°C in vacuum, the powders were hydrated at various temperatures to 98°C in a wide range of water vapor pressures, and the reaction was followed gravimetrically with a silica helix balance. At low temperatures and vapor pressures, where nucleation is exceedingly slow, the hydration of prenucleated samples was studied. The results are interpreted on the basis of a model in which an interface reaction controls the progress of the reaction. The reaction rate varies with the water vapor pressure p according to a function (p/p* - 1), where p 0.3 psat. and appears to correspond to the equilibrium adsorption pressure. This suggests that the reaction depends on an adsorbed water vapor film which has free access to the reacting surface. The reaction rate constant varies with temperature according to an Arrhenius-type equation, with an activation energy 16,100 cal per mole.

The experiments described below on the dehydration of halloysite were largely carried out in the laboratoire Central des Services Chimiques de l'État, Paris, where we enjoyed not only the excellent facilities of the laboratory but also the advice of Monsieur J. Méring and Mile R. Glaser who had made similar studies of montmorillonite. The use of a Guinier-type focusing camera with strictly monochromatic radiation was especially useful. These experiments were undertaken because little was known about the dehydration process beyond the recognition that it occurs very readily in dry atmospheres and at low temperatures. The main experimental difficulty lies in differentiating between water adsorbed oil external surfaces of the clay particles (adsorbed water) and water internally absorbed between the kaolin layers (interlayer water).

In X-ray studies of fireclays we have frequently observed the presence of anatase to the extent of 1 or 2% with little or no rutile, and no brookite. Chemical analyses of these clays usually record 1–2% of TiO2 and occasionally more (Ennos and Scott, 1924). The mineralogical form of the TiO2 is usually not stated, but a number of references to the observance of rutile needles give the impression that rutile is the form in which titania occurs in these deposits. Our own observations based on X-ray powder photographs lead us to conclude that anatase is commonly present and that rutile is either absent or present in smaller quantities. We have not detected brookite.

In recent papers on the intensity of reflexion of X-rays from rocksalt crystals it has been shown that, from the temperature of liquid air up to about 500° abs., the dependence of the intensity of reflexion upon temperature is in accord quantitatively with a formula of the type originally deduced by Debye, if the modification suggested later by Waller is introduced, although the decrease of intensity for higher temperatures is much greater than that indicated by the law. In dealing with quantitative results of experiments on reflexion from crystals, it is convenient to consider the quantity usually denoted by F, which is a measure of the scattering power, in a given direction, of an atom for X-rays. In the course of experiments with rocksalt it has been possible to determine F both for Na and Cl, and the values so obtained, when corrected for temperature, agree very closely with the F factors calculated from Hartree’s Schrodinger density-distribution for the ions Cl - and Na + . The calculation is based upon the theoretical result, due to Wentzel and Waller that, to obtain the coherent scattering from an electron in an atom, the electron must be represented by its corresponding Schrodinger charge-density, each element of which must be supposed to scatter classically. In order to get agreement between the calculated and observed F curves, it is necessary to assume the existence of zero-point energy of amount half a quantum per degree of freedom, which is of course required by the new quantum mechanics. The agreement between theory and experiment in the case of rocksalt is extremely interesting, but, in order to place the quantitative treatment of X-ray reflexion on an entirely satisfactory basis, it appears to be of some impor­tance to see whether a similar agreement can be obtained with other crystals. This is not quite so easy as might be supposed.

Nacrite is the rarest member of the kaolin group of minerals. The name was first applied by Brongniart in 1807 to a mineral (talc granuleux of Haüy, 1801) the properties of which were not then adequately described. In 1832 Breithaupt recorded a mineral from an ore vein near Freiberg in Saxony, which has since been called nacrite. In the past, however, 'nacrite' has been applied quite frequently to different nacreous and micaceous minerals, not necessarily of kaolin composition, and the authenticity of the species was not definitely established until 1931 when Ross and Kerr, on the basis of an X-ray study of nacrite from the locality Brand near Freiberg in Saxony, proved the mineral to be a distinct species of the kaolin group.

It is now well established that from the point of view of the theory of X-ray reflection, the majority of crystals can be divided into those which are relatively perfect and those which are relatively imperfect or mosaic. The intensity of reflection of X-rays by the former has been much less extensively studied than by the latter and hitherto no really satisfactory agreement appears to have been found between the observed intensities of reflection from highly perfect crystals such as diamond and the results predicted by the theoretical treatment of the subject. It will be shown in what follows that this lack of agreement is very largely removed when the atomic scattering factor, f , which plays such an important part in the theory of reflection by mosaic crystals, is taken into account for perfect crystals.

1. It has already been shown that the observed variations with temperature of the intensity of reflexion of X-rays from crystals of rock-salt and sylvine agree closely with those predicted by the theory of Debye as modified by Waller, from the lowest temperature at which experiments have been made, that of liquid air, up to about 500° abs. Moreover, the absolute intensities of reflexion agree closely with those calculated theoretically, if the atomic scattering factors, F, are calculated from the Schrödinger charge-distributions for the atoms, obtained by the method due to Hartree. To obtain agreement, it is necessary to assume the existence of zero-point energy of an amount half a quantum for each degree of freedom, and the experiments may perhaps do considered as furnishing direct confirmation of such energy, since the differences between the intensities calculated with and without it are considerable. The work to do described in this paper was undertaken with a view to extending the investigations to a crystal of a metallic element. Aluminium at ones suggested itself as suitable for this purpose, since it can do obtained in largo single crystals, and since its coefficient of adsorption for the X-rays employed in the experiments, Mo K a , is small.

employed. The orientation of micaceous minerals is most easily investigated by considering the intensities of X-rays reflected from their basal (001) planes in different directions with respect to the bedding or cleavage plane of the material. With X-ray powder cameras of conventional type in which the reflected X-rays are recorded through an aperture on a strip of film, a series of photographs for different angular settings of the material will be required. If, however, a complete Debye-Scherrer diagram is recorded, then a single photograph taken with one angular setting of the material suffices to give the required information. Custers (i948) has analysed in detail this method for studying orientation with particular reference to rolled metal sheets. The practical problems are somewhat different with micaceous materials, since the fragility of many specimens makes the use of small blocks of material almost essential and these can be studied only by surface reflection methods. With thin sheets both surface reflection and transmission arrangements may be employed. With metals more penetrating radiations of shorter wave-length can be usefully used than are practicable

OR many years, a number of problems have persisted concerning the reactions which occur and the products which form when kaolinite and kaolin clays are heated to progressively higher temperatures. Between approximately 550 ~ and 900 ~ C., X-ray powder diagrams of kaolinite show mainly a broad band indicating either an amorphous, or a highly disordered, or a very finely dispersed state. The product is generally called metakaolin. At temperatures above 950 ~ C. y-alumina and mullite are formed. In some eases they appear together ; in others, the appearance of mullite is delayed. The relations of these products to each other and to the original kaolinite have received various interpretations and the course of the reaction is still disputed. Following an examination of nacrite from Hirvivaara, Finland (yon Knorring, Brindley, and ttunter, 1952), we considered that a single crystal study of its thermal reactions might throw further light on those of kaolinite. The two minerals have a similar lamellar morphology and are composed of silicate layers having the same structure and the same chemical composition; they differ only in the stacking arrangements of the layers and in crystal size. The dehydration of nacrite follows a similar course to that of kaolinite and the ultimate products of decomposition are the same for the two minerals. It seems reasonable to suppose that their thermal reactions will be closely similar, though tile particular temperatures at which reactions proceed may depend on crystal size.

The mineralogy and ceramic properties (linear shrinkage and porosity) of fired compacts of quartz, kaolinite, and mica containing 25% of quartz and variable proportions of kaolinite and mica were studied systematically in relation to composition and firing temperature. A procedure for the quantitative determination of components by X-ray diffraction measurements is outlined and applied to the determination of quartz and mullite in the fired samples. For a mixture containing 25%, quartz, 75% kaolinite, the amount of mullite developed at 1300°C. is 41% and this contains 85% of the total Alsoa available. In micarich mixtures, mullite develops at lower temperatures and in smaller proportions. The shrinkage and porosity vary systematically with the percentage of mica in the samples and with firing temperature. The formation of cristobalite depends on the kaolinite content and is not related to the quartz content. Mullite and cristobalite develop at about 1100°C. from the transitory Si-Al spinel-type phase derived from kaolinite.

About forty specimens of talc have been examined by X-ray powder methods. Variations in the degree of crystallinity of the mineral have been observed, but no evidence has been obtained for polymorphic structural varieties. Single-crystal data have confirmed the two-layer mono-clinic cell. Little evidence has been obtained of talcs containing mixed-layer sequences. Electron micrographs have revealed a variety of platy and fibrous forms, and in the latter a close structural relation between talc and tremolite has been established. Talcs prepared synthetically in the presence of aluminum showed small lattice parameter differences from talcs prepared in the absence of aluminum, and the results are consistent with the substitution of small percentages of A1 ions in the talc structure. New X-ray powder data and unit-cell parameters for talc and tremolite are listed.

This reaction series has been a subject of extensive investigation since the work of Le Chatelier in 1887. Nevertheless, major problems have remained concerned with the nature of metakaolin, the manner in which it transforms into a spinel-type phase and mullite, and the relation of this spinel-type phase to mullite. The present survey brings these problems into sharp focus and provides the necessary historical background fora new approach to their solution.

Reactions between corundum and cristobalite in mixed powders and at interfaces between sintered pellets at 1500°C and in various forms of reaction couples at 1550°C were studied by quantitative X-ray diffraction, electron microprobe analysis, and optical microscopy. The results indicate that a noncrystalline product is an important transitional phase in reactions under subsolidus conditions. When corundum is used as a coarse powder (particle size 20 to 30 μm), the amount of noncrystalline material produced can be as much as 30 to 40% of the total reactants, and the amount of mullite is correspondingly small. When corundum is used as a single crystal, the development of a noncrystalline phase is notably increased. Microprobe analyses of coarse powders reacted at 1500°C show a noncrystalline phase of the approximate composition Al2O3·12SiO2. Calculations based on the X-ray analyses indicate a higher proportion of alumina in the noncrystalline material, and it is suggested that alumina-rich zones may form adjacent the corundum surfaces where they are not resolved in the microprobe analysis.

The data so far published on the diamagnetic susceptibilities of the alkaline halides, measured for the salts in the crystalline state, are very discordant and incomplete, as reference to Table I will show. The aim in carrying out these experiments has been twofold: firstly, to obtain a complete set of values for these salts and secondly to examine more closely than has hitherto been possible how rigorously the susceptibilities of simple crystalline salts are additive. It has already been established that the susceptibilities are approximately additive, but it has not been possible to test this with exactitude because of ( a ) the large discrepancies between the results obtained by previous observers, and ( b ) the lack of data for many crystals. The discrepancies may have arisen to some extent from the different experimental methods, some of which are more accurate than others and some of which may introduce errors peculiar to themselves. We have therefore made a complete re-determination of the susceptibilities of all the alkaline halides, using the same apparatus and method under the same conditions. Since any systematic experi-mental errors will affect all our results to approximately the same extent, we shall be in a stronger position for testing the additivity of the susceptibilities than if we rely partly on our own and partly on other observers’ results. Previous investigators have measured the susceptibilities of some compounds in the crystalline state and others in solution; the latter are of no help in connexion with our problem, for an examination of the available data suggests that solutions have susceptibilities higher by several per cent, than the corresponding crystals. We cannot, therefore, arrive at any certain conclusion by using results obtained partly for crystals and partly for solutions. 2—Method The method previously described has been used to obtain a complete set of values for the susceptibilities of the alkaline halides. Although slight modifications have been made from time to time, the apparatus has remained, in essentials, the same as when used for the measurement of the susceptibilities of the sodium and potassium halides.

Metahalloysite is one of the kaolin group of minerals with a composition approximating to the formuia Al2O3.2SiO2.2H2O, and its structure is generally considered to be built of composite layers of atoms of the kind found in other minerals of this group. It differs from them most strikingly in that it can be produced by the dehydration of the hydrated mineral halloysite, which has the approximate composition Al2O3.2SiO2.4H2O. Both minerals give comparatively poor X-ray diagranm showing relatively few lines attd bands. The lines correspond to reflections from a basal spacing of about 7·2 Å. in metahalloysite and 10·1 Å. in halloysite.

Chamosite is a commonly occurring hydrous ferrous silicate frequently found in close association with chalybite in marine mudstone deposits, the composition of different beds ranging from almost pure chamosite to pure chalybite. In oolitic ironstones chamosite is found both in the ooliths themselves and also in the groundmass. It occurs as a constituent in shales, as a cementing mineral in sandstones, and is occasionally found in lateritic and other deposits. Mineralogically it is usually regarded as a chlorite, or akin to the chlorites, but evidence has been accumulating which tends to show that some varieties may not be chlorites at all, but are probably closely related to the kaolin group of minerals (Brindley, 1949).

In the preceding papers of this series, Part I (Hoare 1934) gave an account of the experimental method and results for sodium and potassium halides, Part II (Brindley and Hoare 1935) gave results for lithium, rubidium and caesium halides and a general discussion of this group of salts. The present paper deals with the halides of the alkaline earth metals. Although a few values have previously been obtained for these salts, this is the first time that a comprehensive systematic series of measurements has been made. When comparing the results for the halides of Mg, Ca, Sr and Ba with those for the alkaline halides, the effect of the positive ions being doubly charged must be considered. The positive and negative ions will be drawn closer together than if both ions were singly charged, but the change of interatomic distance on this account is not large owing to the intrinsic repulsion between adjacent ions varying as a high inverse power of their separation; the change will in fact be about 6-10% for the ions under consideration. The effects of the ionic charges on the electron distributions of the adjacent atoms have also to be considered; the positive ions will be subjected to approximately the same effects as the positive ions in the alkaline halides, but the negative ions will be subjected to larger effects owing to the positive ions being both doubly charged and at relatively smaller distances than in the alkaline halides. Whatever the effects of positive and negative ions on each other in crystal lattices may be, it seems unquestionable that these effects will be greater for the negative ions in the alkaline earth halides than in the alkaline halides, but will be of the same order for the positive ions in the two series of crystals.

Little exact information exists on the relation of mineralogical composition of clay materials to their properties after firing. Property-composition diagrams are constructed for linear shrinkage and apparent porosity of quartz-kaolinitemica mixtures fired at 1100° and 1200°C.; contours of equal shrinkage and equal porosity are derived. Sixty natural clays and shales are analyzed semiquantitatively for mineral content, and their shrinkage and porosity values are obtained after firing to 1100° and 1200°C. The ceramic properties of the natural materials are compared with those of the prepared mixtures. Conclusions are drawn regarding the influence of mineralogical composition on the ceramic properties. The mica content is shown to be particularly important because of its role in the for mation of a glassy component.

The two methods by which the intensities of X-ray reflections from fine powders can be measured absolutely by comparison with reflections from a standard substance are critically examined. The applicability of the mixture method is limited by the necessity for having powders of extremely small particle size, of the order of 10-5 cm. or less. In the substitution method, the powders are kept separate and the use of such fine powders is not essential; the absolute accuracy of this method is limited by the fact that the absorption coefficients of the two powders have to be known. Experimental results obtained by the two methods for powders of different particle sizes are compared with theoretical results by Schafer for the effect of particle size on intensities measured by the mixture method. The agreement between the experimental and the theoretical results is sufficiently close to show the usefulness of Schafer's analysis in determining the magnitude of the error which may occur when the mixture method is used with powders which are not sufficiently fine. A general account is given of a camera designed for use with the substitution method. It allows an alternating interchange of the two powders with a corresponding movement of the photographic film in order to obtain a separation of the photographs corresponding to the two powders.

Summary. The layer structure of the kaolin mineral dickite has been shown to differ significantly from the idealized arrangement of regular tetrahedra and octahedra. Since the structure of kaolinite can be studied only by X-ray powder methods, it cannot be refined beyond the idealized arrangement. By applying the diekite type of layer structure to the ease of kaolinite, an explanation for the triclinic symmetry of kaolinite has been obtained, and also an improved agreement with the observed X-ray intensities. A DETAILED structure analysis of kaolinite was first attempted by Gruner (1932), based on X-ray powder diagrams of the mineral. He concluded that the unit cell was monoclinic, with probable space group C~-Cc, and contained two structural layers related by a c-glide plane. However, further examination of powder diagrams of various kaolinites by Brindley and Robinson (1946) revealed that there were reflections that could not be explained in terms of this monoclinic cell, and they put forward a triclinic cell containing only one structural layer. So far, it has been proved that this triclinie cell fits very well with the observed features of the X-ray powder patterns of the mineral as regards both spacings and intensities of reflections. The structure of the composite layer of Si-O and AI-OH was assumed to have the geometrically idealized arrangement first proposed by Pauling (1930). At that time, there were no means to establish a refined structure for the layer itself in the absence of suitable single crystals. In terms of this idealized layer, it is impossible to account in the actual structure of kaolinite for the s-angle and for a small departure of the fl-angle from an idealized displacement of --a/3. Recently, Newnham and Brindley (1956, 1957) have made a detailed analysis of the mineral dickite, a polymorphic form of kaolinite, and have shown that the actual layer structure is distorted to some extent from the idealized arrangement. This has opened the way to an interpretation of the triclinic nature of kaolinite, since there are good reasons for thinking that the structures of the layers in dickite and kaolinite

SummaryElectron micrographs of scarbroite show thin platy crystals of about 1 μ size, having rhombic outlines with angles 66° ± 1° and 113° ± 1°. Single-crystal electron-diffraction patterns show rectangular net patterns, with d100 = 9·90 Å., d010 = 14·67 Å., γ* = 90° ± 0·05°. Strong hk0 reflections show a pseudohexagonal arrangement, but true symmetry is probably orthorhombic or monoclinic. Faces outlining rhombic forms are of type {11l}, .

MANY solid-state reactions involving oxide systems take place by re-organization of cations within a framework of oxygen ions and the rate of the reaction is controlled by ion diffusion rather than by nucleation and grain growth processes. This general concept has been well illustrated by recent work of Bernal et al. 1 on oriented transformations of iron oxides and hydroxides, and of Steadman and Youell2 on the thermal transformation of an iron silicate of layer-lattice type.

The AlKα emission wave length is determined by the state of coordination of Al atoms and a clear distinction can be made between Al atoms in fourfold and in sixfold coordination to oxygen and also between Al atoms in fourfold coordination to oxygen and to nitrogen. Measurements of Δ(2θ) (the difference in diffraction angle for AlKα radiations from a sample and from metallic aluminum) are used to distinguish Al in fourfold and in sixfold coordination in normal and in heat-treated kaolin minerals. The Al atoms in meta-kaolin are shown to have fourfold coordination to oxygen in agreement with the structure proposed by Brindley and Nakahira. After heat-treatment at 1024°C., the coordination of Al reverts toward sixfold, and this is more marked for disordered kaolinite and halloysite than it is for well-crystallized kaolinite and dickite. Infrared absorption measurements by Stubican and Roy provide further confirmation of the coordination changes of the Al ions in these reactions.

THROUGH the kindness of the Editor, I have been allowed to see Prof. Herrmann's letter prior to its publication. I entirely agree that the X-ray results do not exclude the possibility that the outermost electrons of the atoms in metallic zinc and cadmium may depart from having spherical symmetry ; one would not expect the completed inner electron shells to have appreciable asymmetry. The possibility of the outermost electrons being asymmetrical, however, is not to be regarded as an alternative explanation of the experimental results, but as something additional to the explanation already given in terms of asymmetrical lattice vibrations. The contributions of the outermost electrons to the scattering factors of zinc and cadmium at all angles for which measurements can be made are so small that one could scarcely hope to measure an asymmetry in their distribution by means of X-rays. On the other hand, any asymmetry in the lattice vibrations will affect the entire atom, and at large scattering angles, measurable effects may be expected if the asymmetry is sufficiently great, and this appears to be so for zinc and cadmium.

A summary is given of a structure analysis of chloritoid. The structure contains continuously linked octahedral sheets cross-linked by discrete SiO4 units. Comparisons are made with mica structures and with corundum.

The dehydroxylation reaction of kaolinite in a vacuum at 425 degrees C is halted by introducing a water vapor pressure of 47 mm-Hg, and is resumed when the vacuum is reestablished. The sample gains weight corresponding to an approximately monomotecular layer of water on the kaolinite surface. At the temperatures and vapor pressures involved, the sorption is considered to be a chemisorption process.

DURING the past few years, it has been shown that the diamagnetic susceptibilities of ions in crystals are approximately additive, and their values are in fairly close agreement with values calculated theoretically from the wave-mechanical structures of free ions1. Attempts to interpret the experimental results in more detail have been handicapped by a lack of essential data, for although results are now available for many substances, some have been obtained for solutions and others for crystals, and the two are not directly comparable. With the view of obtaining more precise information for a particular group of crystals, we have studied the susceptibilities of the alkaline halides, using a method of measurement previously described2.

THE axial ratio, c/a, of the majority of metals having a hexagonal close-packed type of structure differs appreciably from the value 2 ” = 1·633 which corresponds to a hexagonal close-packed system of spheres. The two most notable deviations are for zinc and cadmium, for which c/a = 1·86 and 1·89 respectively, while for beryllium, titanium, zirconium, hafnium, yttrium, ruthenium and osmium, c/a lies between 1·58 and 1·59. These facts have sometimes been expressed by describing the atomic domains as having the forms of prolate and oblate spheroids when c/a is respectively greater than and less than 1·633 ; but it is only recently that experimental evidence has been obtained which throws light on the nature of the asymmetry in these metals.

IN a recent investigation of the ionisation processes in methane, Hogness and Kvalnes (Phys. Rev., 32, December 1928), using a mass-spectrograph method, find that at 14.5 volts only CH4+ ions are formed, but at 15.5 volts two processes occur; either stable CH4+ ions are formed or unstable CH4+ ions which dissociate spontaneously into CH3+ ions and neutral hydrogen atoms, the probabilities of the processes occurring being approximately equal over a wide range of pressure.

CHAMOSITE is a hydrous ferrous silicate, of common occurrence especially in bedded ironstone deposits. While it has generally been regarded as belonging to the chlorite family of minerals, evidence has accumulated tending to show that it is not a normal chlorite. To obtain more certain information on this question, a detailed analysis has been made of the X-ray powder diagrams of a number of chamosite specimens ; the results obtained lead me to conclude that chamosite is not a chlorite but has a structure of the kaolin type. Two other iron-bearing minerals, namely, amesite and cronstedtite, originally classified as chlorites, have also in recent years been shown by Gruner1 and by Hendricks2 respectively to be kaolin-type minerals. While these are comparatively rare minerals, chamosite is a very common mineral. Its recognition as a kaolin-type structure is of general interest in that it brings the kaolin group of structures more nearly into line with other groups of clay minerals in which iron-bearing members are quite common.

X-RAY studies of these minerals in relation to (a) their reactions with organic liquids (MacEwan), (b) the effect of heat treatment (G. W. B. and K. R.) have led us to certain conclusions which can best be expressed in a joint communication. Meta-halloysite has the same formula as kaolinite, Al2Si2O5(OH)4, and from the work of Hendricks and others1 is generally considered to have a closely similar structure, consisting of layers of Si-tetrahedra and Al-O,(OH) octahedra which in kaolinite2 are arranged regularly above each other, but which in meta-halloysite are stacked with “considerable disorder”1 or even, as we believe, with almost complete disorder as regards the orientation of the layers. In what follows we set out briefly the evidence for this view of the meta-halloysite structure.