Many attempts have been made to discover the causes of the various colours displayed by minerals. If we except fluorescence and allied phenomena, the diiteren~ theories advanced may in general be grouped under one or other of the following classes:

1. (1) Colour may be due to an intrinsic property of atoms, ions, or molecules. Examples are copper atoms, chromate ions, and ferric oxide molecules.

2. (2) Colour is sometimes attributed to structural causes. The blue tints of certain halites have been explained in this way; but a fresh suggestion is made in this communication.

3. (3) A pigmentary impurity may cause the observed colour. The colour of the ruby is believed to be caused by the presence of a red chromic oxide. The blue tint of a Yorkshire aragonite is shown later to be due to a copper salt.

4. (4) Some minerals show plays of colours due to interference and/or grating effects. Labradorite and opal are cases in point. Into this group also fall minerals like bomite which become iridescent from tarnish.

The cell parameters of fifteen arfvedsonites and four ferrohastingsites were determined using powder and diffractometer techniques and a least squares method of computation. Methods are proposed of distinguishing between calcium- and sodium-rich amphiboles, and between iron-rich and iron-poor varieties of these two groups, by their powder data. Within the arfvedsonite group the b parameter was found to increase with increasing Ca, Fe″, and (OH) and to decrease with increasing Na, F, and Li. The c parameter was found to decrease with increasing F and increase with increasing (Fe‴+Al+Ti). The parameters a, β, and a sin β all tended to increase with increasing Ca, Fe″, and F content. The cell volumes were found to vary with the b parameter. The ferrohastingsites were found to have much larger cell parameters than the arfvedsonites. An attempt was made to explain these variations from structural considerations.

Jurupaite was discovered at Crestmore, California, by A. S. Eakle in 1921. The mineral was found in a quarry which was rapidly being enlarged, and Eakle stated that it was probably represented only by the one specimen which he had collected. He showed that it was a hydrated calcium silicate containing magnesia, with the composition 2(Ca,Mg)O. 2SiO2. H2O, the ratio of lime to magnesia being approximately 7 : 1.

This specimen passed into the keeping of Professor A. Pabst, who kindly made a portion available to the writer. He confirmed that it was unlikely that any other specimen existed. The jurupaite consisted of rosettes of white needles or fibres, about a centimetre in diameter. A brown discoloration was observed on the exposed outer surfaces of the specimen, but not on freshly cut surfaces. Calcite was present in contact with the jurupaite.

In the course of some researches on the relation between the chemical constitution and the crystalline form of substances, I found it necessary to attempt an accurate determination of the densities of crystals.

Messrs. Playfair and Joule, in 1843, in the course of similar researches, determined the specific gravities of numerous salts by the use of an instrument which showed the actual volume of the solid investigated. This method was found to be subject to many errors, and they abandoned it for one which is the prototype of that ordinarily used in accurate work at present.

A new universal stage extinction technique is presented. It is based on the special case of coplanar and conical extinction curves where an optic axis lies on the wave-normal locus. It may be used in favourable circumstances to determine 2V when neither optic axis is directly observable.

The behaviour of a mass of molten but cooling igneous rock under moderate pressure may be compared to that of liquid cupelled silver, which at a certain temperature can no longer retain the oxygen, many times its volume, that it has occluded. Thus, during the submarine eruption off the island of Pantelleria in 1891, blocks of lava that swam for a while on the surface of the sea exploded with violence. Chilling by contact with sea-water had occasioned exactly what Lowthian Bell observed to happen on the sudden cooling of masses of slag by contact with water. There was a discharge within the set outer envelope, at something near a red heat, of previously dissolved water and gases, converting the floating ejectamenta into a species of bombs.

Pyrolusite pseudomorphous after manganite occurs at Serro do Navio, Amapa, Brazil. It consists of wedges, flattened on (100) and elongated along [010], twinned on (011) and showing forms d{210}, , b{010}, and f{021} (all indices referred to manganite, in the setting of Dana, 7th edn). The difficulty of distinguishing between pyrolusite pseudomorphous after manganite and after groutite is discussed and also the possible existence of the latter. The distinction was made here by measuring the twinning angle between the pyrolusite lattices using (0kl) precession photographs. Specimens of manganite intergrown with pyrolusite are described and their relative orientations determined. On heating, manganite partly decomposes to a monoclinic phase, a 9·9 Å, b 5·73 Å, c 4·89 Å, β 98°, space group I2, Im, or I2/m.

Relianceite-(K), K4Mg(V4+O)2(C2O4)(PO3OH)4(H2O)10, is a new mineral species from the Rowley mine, Maricopa County, Arizona, USA. It occurs in an unusual bat-guano-related, post-mining assemblage of phases. Other secondary minerals associated with relianceite-(K) are antipinite, dendoraite-(NH4), fluorite, mimetite, mottramite, rowleyite, salammoniac, struvite, vanadinite, willemite, wulfenite and at least one other new mineral. Crystals of relianceite-(K) are sky blue prisms up to ~0.1 mm in length. The streak is very pale blue and lustre is vitreous, Mohs hardness is 2½, tenacity is brittle and fracture is splintery. The calculated density is 2.111 g⋅cm–3. Relianceite-(K) is optically biaxial (+) with α = 1.528(2), β = 1.529(2), γ = 1.562(2) (white light); 2Vmeas = 22(1)°; orientation Z = b; pleochroism: X = colourless, Y = pale blue, Z = pale blue; X < Y ≈ Z. Electron microprobe analysis gave the empirical formula [K2.21(NH4)1.79]Σ4.00Mg0.96(V4+0.95O)2(C2O4)[P1.03O3.03(OH)0.97]4(H2O)10, with the C, N and H contents constrained by the crystal structure. Raman spectroscopy confirmed the presence of NH4 and C2O4. Relianceite-(K) is monoclinic, Pc, with a = 12.404 (7) Å, b = 9.014 (6), c = 13.260 (8) Å, β = 100.803(10)°, V = 1456 (2) Å3 and Z = 2. The structural unit in the crystal structure of relianceite-(K) (R1 = 0.0540 for 3751 Io > 2σI reflections) is a (V4+O)2(C2O4)(PO3OH)4 chain in which VO6 octahedra are bridged by an oxalate group to form [V2C2O12] dimers, PO3OH tetrahedra form a double bridge between the VO6 octahedra of the dimers, and additional PO3OH tetrahedra decorate the chain. Topologically, this is the same chain found in the structure of davidbrownite-(NH4). The MgO(H2O)5 octahedron can be considered a distant decoration on the chain. The chains are linked to each other through an extensive system of K/NH4–O bonds and hydrogen bonds.

Cross-hatched twinning is characteristic of two alkali-felspar species, microclines and anorthoclases. Because of the difference in the crystallographic angles of these two species, the cross-hatched twinning is most clearly seen in (001) sections of microcline and in sections perpendicular to the a crystallographic axis in anorthoclases. If pericline twinning is present it is generally seen in (010) sections of microclines, but if the microcline is microperthitic the twin lamellae may not be seen. Strings or lamellae of perthitic albite can be distinguished from pericline twin lamellae by refractive index differences and also by the angle that their trace makes with the trace of (001) in (010) sections.

Arsenoflorencite-(La), ideally LaAl3(AsO4)2(OH)6, was studied at high pressure by single-crystal X-ray diffractometry. The unit cell was determined at nine pressures up to 7.471(8) GPa; no evidence of a phase transformation was found in this range. The pressure volume data (refined simultaneously) were fitted to a third-order Birch Murnaghan equation of state which gave V0 = 710.71(8) Å3, KT0 = 106(2) GPa and K' = 9.2(9). These values were confirmed independently from an FE–fE plot. The crystal structure was refined at 1.596, 3.622, 5.749 and 7.471 GPa, the first time this has been done for a member the alunite supergroup. The compressibility of arsenoflorencite-(La) is strongly anisotropic, with βc > βa. The main compression mechanism was found to be governed by the internal angle O3 La O2 of the La polyhedron, where the O2 and O3 atoms move toward the c axis during compression.

The Environment Agency Guidance on Requirements for Authorisation (GRA) of a geological disposal facility (GDF) requires a demonstration that "the possibility of a local accumulation of fissile material such as to produce a neutron chain reaction is not a significant concern." A neutron chain reaction that is just self-sustaining is also known as critical.

Waste packages can be designed to ensure that criticality is impossible during the transport and operational phases of a GDF, and for a significant period post-closure. Over longer times, however, packages may degrade, and groundwater flows could lead to a localized accumulation of fissile material. Hence, even though the initial distribution of materials would need to change substantially, criticality cannot be ruled out completely.

This paper describes how an accumulation of fissile material could, hypothetically, lead to a critical configuration; how such a system could evolve; what the local consequences could be; and how the engineered and geological barriers could be affected. The conclusion from studies to date is that, even for large (and very unlikely) fissile accumulations, the consequences of a post-closure criticality event are not a significant concern.

The clinozoisite to be described is found in veins traversing a metamorphosed basic intrusion exposed at the top of Camaderry Mountain, Glendalough, Co. Wicklow. This intrusion is a small boss of Lower Palaeozoic age which has been metamorphosed both by earth movement and by the intrusion of the Leinster granite. The rock exposed is now an amphibolite (fig. 1) within which are found the veins carrying clinozoisite with albite and a little prehnite, sphene, and tremolite. These veins, which are up to an inch in width, were probably introduced during the late stages of the basic intrusion as veins of lime-rich plagioclase.

Preliminary examination under the microscope shows that the clinozoisite is well crystallized, fresh, and free from inclusions. The colour in the hand-specimen is grey-white; in thin section it is colourless. Individual crystals are found up to 5 mm. in length. Twinning and intricate zoning are common.

Two new isostructural minerals radekškodaite-(La) (CaLa5)(Al4Fe2+)[Si2O7][SiO4]5O(OH)3 and radekškodaite-(Ce) (CaCe5)(Al4Fe2+)[Si2O7][SiO4]5O(OH)3 were discovered in polymineralic nodules from the Mochalin Log REE deposit, South Urals, Russia. Radekškodaite-(La) is associated with allanite-(Ce), allanite-(La), bastnäsite-(Ce), bastnäsite-(La), ferriallanite-(Ce), ferriallanite-(La), ferriperbøeite-(La), fluorbritholite-(Ce), törnebohmite-(Ce) and törnebohmite-(La). Radekškodaite-(Ce) is associated with ancylite-(Ce), bastnäsite-(Ce), bastnäsite-(La), lanthanite-(La), perbøeite-(Ce) and törnebohmite-(Ce). The new minerals form isolated anhedral grains up to 0.35 × 0.75 mm [radekškodaite-(La)] and 1 mm × 2 mm [radekškodaite-(Ce)]. Both minerals are greenish-brown with vitreous lustre. Dcalc = 4.644 [radekškodaite-(La)] and 4.651 [radekškodaite-(Ce)] g cm–3. Both minerals are optically biaxial (+); radekškodaite-(La): α = 1.790(7), β = 1.798(5), γ = 1.825(8) and 2Vmeas = 60(10)°; radekškodaite-(Ce): α = 1.798(6), β = 1.806(6), γ = 1.833(8) and 2Vmeas = 65(10)°. Chemical data [wt.%, electron-microprobe; FeO:Fe2O3 by charge balance; H2O by stochiometry; radekškodaite-(La)/radekškodaite-(Ce)] are: CaO 3.40/2.74, La2O3 27.68/22.23, Ce2O3 20.39/24.30, Pr2O3 0.94/1.48, Nd2O3 1.71/3.18, ThO2 0.23/0.24, MgO 0.85/1.04, Al2O3 10.35/10.84, MnO 0.64/0.69, FeO 2.55/2.76, Fe2O3 3.12/2.57, TiO2 0.13/0.04, SiO2 26.03/26.10, F 0.10/0.09, H2O 1.62/1.63, –O=F –0.04/–0.04, total 99.70/99.89. The empirical formulae based on O28(OH,F)3 are: radekškodaite-(La): (Ca0.98Th0.01La2.75Ce2.01Nd0.16Pr0.09)Σ6.00(Al3.28Fe3+0.63Fe2+0.57Mg0.34Mn0.15Ti0.03)Σ5.00Si7.00O28[(OH)2.91F0.09]; radekškodaite-(Ce): (Ca0.79Mn0.16Th0.01Ce2.39La2.20Nd0.30Pr0.14)Σ5.99(Al3.43Fe2+0.62Fe3+0.52Mg0.42Ti0.01)Σ5.00Si7.00O28[(OH)2.92F0.08]. Both minerals are monoclinic, P21/m; the unit-cell parameters [radekškodaite-(La)/radekškodaite-(Ce)] are: a = 8.9604(3)/8.9702(4), b = 5.7268(2)/5.7044(2), c = 25.1128(10)/25.1642(13) Å, β = 116.627(5)/116.766(6)°, V = 1151.98(7)/1149.68(11) Å3 and Z = 2/2. The crystal structures are solved based on single-crystal X-ray diffraction data; R = 0.0554 [radekškodaite-(La)] and 0.0769 [radekškodaite-(Ce)]. Both minerals belong to the epidote–törnebohmite polysomatic series and represent first members of ET2-type: their structure consists of regular alternating modules, one slab of the epidote (E) structure and two slabs of törnebohmite (T). The rootname radekškodaite is given in honor of the Czech mineralogist Radek Škoda (born 1979), Associate Professor at Masaryk University, Brno, Czech Republic. The suffix-modifier -(La) or -(Ce) indicates the predominance of La or Ce among REE in the mineral.

Uranium contents ranging from 0.03–1.18 ppm have been determined by delayed neutron activation analysis in a suite of 19 ultrabasic xenoliths (mostly of the common garnet lherzolite type) from the kimberlites of southern Africa. Consideration of these and previous data lead us to conclude that the U content of uncontaminated garnet lherzolite in the uppermost mantle is on average roughly 0.04–0.05 ppm, levels which are sufficient to account for the observed U contents of fresh basalts. Higher U contents in mantle-derived ultrabasic xenoliths are the result of secondary alteration processes, most probably related to infiltration metasomatism stemming from contact with kimberlitic or carbonatitic magma fractions. This introduced U is concentrated along the boundaries of the primary mineral grains and is mostly readily removed by acid leaching. A partial analysis is given for a secondary uraniferous calcium phosphate phase observed in kelyphitic rims around garnets.

The fall of a shower of meteorites numbering several hundred fragments at Wiluna, Western Australia, on the night of 2 September 1967, has been investigated. Although the dispersion ellipse had been largely obscured by removal of fragments before a party of scientists were able to make a field investigation, it has, nevertheless, been possible to make a reasonable estimate of the shower distribution pattern. In spite of this removal of fragments, a number of pieces of meteorite were still found in situ. The bulk of the total recovery is in the collections of the Western Australian Museum, and the physical characteristics of these masses and their petrography is described. In all, some 490 individual fusion-crust coated stones and a large number of broken stony fragments are known to have been recovered. The meteorite is an olivine bronzite chondrite remarkably rich in discrete nodules of nickel iron, up to an inch across, commonly aggregated with troilite. A full chemical analysis of this fresh meteoritic material has been supplied by the British Museum (Natural History).

The new mineral flaggite (IMA2021-044), Pb4Cu2+4Te6+2(SO4)2O11(OH)2(H2O), occurs at the Grand Central mine in the Tombstone district, Cochise County, Arizona, USA, in cavities in quartz matrix in association with alunite, backite, cerussite, jarosite and rodalquilarite. Flaggite crystals are lime-green to yellow-green tablets, up to 0.5 mm across. The mineral has a very pale green streak and adamantine lustre. It is brittle with irregular fracture and a Mohs hardness of ~3. It has one excellent cleavage on {010}. The calculated density is 6.137 g cm–3. Optically, the mineral is biaxial (+) with α = 1.95(1), β = 1.96(1), γ = 2.00(1) (white light); 2V = 54(2)°; pleochroism: X green, Y light yellow green, Z nearly colourless; X > Y > Z. The Raman spectrum exhibits bands consistent with TeO6 and SO4. Electron microprobe analysis provided the empirical formula Pb3.88Cu2+3.89Te6+2.08(SO4)2O11(OH)2(H2O) (–0.03 H). Flaggite is triclinic, P1, a = 9.5610(2), b = 9.9755(2), c = 10.4449(3) Å, α = 74.884(1), β = 89.994(1), γ = 78.219(1)°, V = 939.97(4) Å3 and Z = 2. The structure of flaggite (R1 = 0.0342 for 5936 I > 2σI) contains hexagonal-close-packed, stair-step-like layers comprising TeO6 octahedra and Jahn-Teller distorted CuO6 octahedra. The layers in the structure of flaggite are very similar to those in bairdite, timroseite and paratimroseite.