Summary. Pure dry chlorine at 250 to 350 ~ C is shown to have no action on the principal silicate and phosphate phases of meteorites, but reacts quantitatively with the sulphide and phosphide phases and with kamacite ; taenite does not react if it contains more than about 20 to 25 % Ni; some rare meteoritic minerals have not yet been tested. Volatile chlorides (including SIC14) distil, and non-volatile chlorides can be leached out with water. Based on these observations, a method has been devised and applied to the analysis of a number of ehondrites and to the purification of olivine and pyroxene separated by physical methods. The amount of Si present in the metal of enstatite chondrites, and the distribution of P, Ga, Ge, Ti, Cr, Mn, Ca, and Mg between oxidized and non-oxidized phases can readily be determined, and some results are presented. From the limited data so far available, Co appears to be concentrated in kamacite; gallium in the metallic phase is mainly in the taenite ; germanium is strongly siderophile in the common chondrites but some is present in the silicate of the enstatite chondrites ; titanium is markedly chalcophile in the enstatite chondrites, less so in the common chondrites, and the same applies to chromium; phosphorus is present as phosphide in the enstatite chondrites, but as phosphate in the hypersthene chondrites.

S.W. 7 SUM MARY. Previous investigations have been confined to individual members or to small compositional ranges of the amblygonite-montebrasite series. The present study is a survey based on twentytwo specimens in the range Amb~Monso to Amb~o.Mon8 (i.8-11.8 % fluorine), and indicates that the fluorine content may be related both qualitatively and quantitatively to differences in the X-ray powder patterns. Fully indexed powder data to d ~ 1.68 ,~ and derived cell parameters are given for four members of the series. There is no evidence of a break in the continuity of the series. The first British occurrence of both amblygonite and montebrasite is reported from a pegmatite vein in aplite at Meldon, Okehampton, Devonshire.I Y the course of an examination of a number of specimens of presumed amblygonite and montebrasite collected by the late Mr. A. W. G. Kingsbury at the old aplite quarry, Meldon, Devonshire, we noted marked differences between the X-ray powder patterns. It was by no means clear from published data that all the specimens belonged to the series, and variations in powder patterns of different specimens of montebrasite (Quensel, 196z ) were less than we observed. Other workers (Fisher, I958; Kittrick and Hope, I967) distinguish between amblygonite and montebrasite on the basis of spacing and intensity differences of the three strongest lines of the patterns, and on comparisons of single crystal diffraction patterns with powder patterns, but their reference samples are inadequately characterized and their data too scanty. The powder patterns figured by Pough and Henderson (1945) are indistinct, and are not supported by chemical analyses. Twenty-two specimens of amblygonite and montebrasite were selected for apparent homogeneity, and powder prepared from each was used for partial chemical analysis (table l) and both photographic and diffractometer records of the powder patterns. Careful examination of the photographic powder patterns showed that it is possible to distinguish four general types of pattern, here designated A, AB, B, and C ( fig. l). Diffractometer records, mentioned below, permit division of the AB pattern-type into two ( fig. z).For chemical analysis, between 15 and 7o mg of the powder were fused with five times their weight of sodium carbonate, extracted with 4o ml of I : I sulphuric acid, transferred to a distillation flask, and the fluorine steam-distilled. Fluorine was determined on the distillate by the zirconium-eriochrome-cyanine method of Megregian (t954). Alkalis were determined by flame photometry on a sulphuric acid solution.Comparison of the chemical and X-ray results shows that there is a correlation between pattern-type and fluorine content ( fig. 4)-The replacement of lithia by soda

Summary. The estimation of molybdenum and tungsten when present together has proved difficult, as is indicated by the large number of methods put forward for their determination. 1 Many of the methods given are for one or other of these elements only, a note often being added that the presence of the other invalidates the method. The estimation of molybdate and tungstate in the mineral powellite has been briefly described by one of the authors (A. A. M.) :2 molybdate and tungstate were precipitated as oxinates (Mo,W)O~(CgHsON)2 and the molybdate (MoO~") determined by measurement of the optical density at 230 m~ of a solution containing molybdate and tungstate. The amount of tungstate present was then estimated by difference. The property of the molybdate ion, MoO~", to absorb strongly in the ultra-violet has been further investigated and it has been found that, if a wave-length of 240 m/~ is chosen and the proper precautions taken, the MoO~" ion can be estimated without significant interference by the W04 '~ ion. Methods are given for the analysis of the common molybdate and tungstate minerals, together with results obtained during the course of this investigation.T HE analysis of molybdates and tungstates resolves itself into three parts: the separation of the anions from the cations, the estimation of the cations, and the estimation of the anions.The methods used to separate the anions from the cations and to estimate the cations depend on the nature of the nfineral but all methods yield a solution of sodium molybdate and sodium tungstate sensibly free from the mineral cations, and the method of analysis of the anions is the same for all minerals, a Experience has shown that, provided the conditions set out below, under scheelite, are followed, molybdate and tungstate can be precipitated together quantitatively, by means of oxine.A solution of sodium molybdate containing about 10 ppm MoO a shows no appreciable absorption above 275 m~. At wave-lengths less than this the absorption increases very rapidly but does not attain a maximum within the limits imposed by the instrument used (Unieam SP 500) : with similar concentrations of tungstate the absorption begins at much lower

SummarySimple chemical procedures (colorimetric and gravimetric) for the determination of Co, Cr, Cu, Ga, Ni, P, and S in an iron meteorite or in the separated metallic phase of a pallasite are described; the minimum size of sample is discussed. A procedure for the quantitative removal of the metallic, sulphide, and phosphide phases from a stone or siderolite is described, and is applicable to the quantitative isolation of silicate and oxide inclusions from siderites (other than some very high in nickel).

No abstract