MINERALOGICAL CHEMISTRY,GROTH is dead. Paul Heinrich Ritter von Groth (1843-1927)died on December 2nd at the ripe age of 849 years. He was Pro-fessor of Mineralogy in Strassburg during 1872-1883, and since1883 in Munich, and his laboratory was long the centre of trainingfor students from all parts of the world. In 1877 he founded theZeitschrift f u r Krystallographie und Mineralogie (carried on since1921 by Prof. Paul Niggli of Zurich under the title Zeitschrift furKristallographie), in which he obtained the willing co-operation ofworkers in all countries. As marking the jubilee of that periodical,he gave in vol. 66, only shortly before his death, an interestingreview of the work that had been done. His monumental work,‘‘ Chemische Krystallographie ” (5 vols., 1906-1919), collectstogether the crystallographic data for some 7350 substances.Theextent to which this is now being used by X-ray workers couldscarcely have been foreseen even by Groth himself. His well-known text-books “ Physikalische Krystallographie ’’ and “ Tabel-larische Ubersicht der Mineralien ” each passed through foureditions, and only last year he gave an interesting essay on thehistorical development of the mineralogical sciences.1 He held aunique position as a leader, and he did more than anyone else inco-ordinating and stabilising crystallographic nomenclature andmethods. His name will for ever stand out as a landmark in thehis tory of crystallography .The deaths have also to be recorded of two distinguished mineralchemists, who both did valuable work, but on entirely differentlines.William Francis Hillebrand (1853-1925), as chemist onthe United States Geological Survey, had made many analyses ofminerals and rocks, including several rare and new minerals collectedby members of the survey in new country. The careful and detailedanalytical work for which he was noted led to the recognition ofseveral chemical elements not previously suspected to be presentin rocks. In 1888-1892 he made a number of very detailedanalyses of uraninite (pitchblende), in which he found a gas. This“ Entwicklungsgeschichte der mineralogischen Wissenschaften,” Berlin,L. J. Spencer, “ Biographical Notices of Mineralogists Recently Deceased ”1926.(third series), Min. Mag., 1927, 21, 229MJXERALOClICAL CHEMISTRY.293gas evidently puzzled him, and he concluded that it was mainlynitrogen. Sir William Ramsay afterwards, in 1895, identified it ashelium, and uraninite was the first known terrestrial source of thiselement. Hillebrand’s well-known book “ Analysis of Silicate andCarbonate Rocks” was first issued in 1907 as a Bulletin of theUnited States Geological Survey, and has passed through severaleditions. Working as a student in Heidelberg under Bunsen heprepared metallic cerium in 1875 and discovered its pyrophoricproperty, a property which now has an extensive practicalapplication.Gustav von Tschermak (1836-1927), of Vienna, died on May 4th,at the advanced age of ninety-one years. As distinct from the workof Hillebrand, his work consisted largely in correlating the largeamount of data accumulated by analytical chemists and in deducinggeneral principles.The current views of the text-books on theconstitution of many of the main groups of silicates are due to him,and were developed in a series of classical papers extending overmany years : plagioclase felsprs (1865), pyroxenes and amphiboles(1868), micas (1877), scapolites (1884), chlorites (l890), vermiculites(1891), tourmaline (1899), zeolites (1917-1918). In later years,with the help of his pupils (including his daughter Silvia Hillebrand),he endeavoured to determine the composition of silicic acids isolatedfrom natural silicates. The valuable periodical TschermahMineralogische und Petrographische Mitteilungen was commenced byhim in 1872, and his well-illustrated “ Lehrbuch der Mineralogie ’’passed through nine editions.Mineralogical jubilees are now falling due.That of the Miner-alogical Society of Great Britain and Ireland (instituted February3rd, 1876) was celebrated last year: and was attended by severaldistinguished foreign mineralogists. The Socihth frangaise deMinhralogie was founded on March 21st, 1878, and the fifty volumesof its Bulletin show a record of brilliant work. As noted above,Groth’s Zeitschrift and Tschermaks Mitteilungen were commencedin 1877 and 1872 respectively. The Russian Mineralogical Societydates from a much earlier period (1817), whilst those of Vienna(1901), Germany (1908), America (1919), and Switzerland (1924) arelater.Account of the jubilee celebration, Min.Mag., 1926, 21, 99-148. Refer-ence is there made to the earlier British Mineralogical Society (1799-1806),which consisted of a small group of chemists, including Arthur Aikin, whooffered to “ examine, free of expense, all specimens of earths or soils, witha view to determining the nature and proportions of their different contents.’That society led to the foundation of the Geological Society of London in1807294 ANNUAJi REPORTS ON THE PROGRESS OF CHEMISTRY.Geochemical Distribution, of the Elements.The series of elaborate papers under this title has been continuedby V. M. Goldschmidt,4 but in the later numbers the scope of theinvestigation has been gradually changing-dealing with X-raydeterminations of the crystal structure of the rare-earth oxides,the laws of crystal chemistry, etc.It is suggested that elementsshowing some homceomorphous relation between the crystal struc-ture of their compounds should be found in association in nature.However, work on the original lines has been continued by someof his collaborators. G. Lunde 5 has examined a variety of mineralsand basic igneous rocks from Norwegian localities for traces ofplatinum metals, finding 04000074% Pt in an olivine-rock and inhornblende-gabbro, and 0~000006% in tantalite. The wide dis-tribution of traces of iodine has been further investigated by T. vonFellenberg and G. Lunde.g In meteoric irons and stones, amountsof iodine up to 0~000018% have been detected, together with up toOW056% of bromine, the latter especially in the stones.Systematicsearch for such elements, which are widely distributed althoughnever found in concentrated quantities, is, of course, of someeconomic importance.Constitution of Silicates.A comparative study of mineral silicates with the more tractableorganic silicon compounds would no doubt throw some light on theconstitution of the former. Some comparisons of this kind weremade by the late J. Emerson Reynolds, and by the oxidation ofthe compound CaSi,Al, (analogous to CaC,N,) he synthesisedanorthite by an interesting method. A long series of valuablepapers by F. S. Kipping and his co-workers on organic derivativesof silicon has appeared in the Jourml of this Society since 1901.G.N. Ridley,’ in a brief outline of some of the present views on theconstitution of the silicates, has compared ethyl orthosilicate,Si( OC,H,),, with olivine, Mg,SiO,, and ethyl metasilicate,SiO(OC,H,),, with enstatite, MgSiO,. On similar lines, W. Wahlhas compared aluminosilicates with alumino-oxalates , tracing, itwould appear, a close analogy. He had recently proved8 thatcertain alkali aluminium trioxalates of the typeAI,(C,O4)3 3R1&&04 nH20 Y4 “ Geochemische Verteilungsgesetze der Elemente,” Nos. I-VIII, Viden-skapssel. Skrifter, Kristiania (later Skrifter Norske Videnskaps-Akad. Oslo),1923-7 ; compare Ann. Report, 1923, 20, 262.6 2. anorg. Chem., 1927,161, 1; A,, 439.6 Biochem. Z., 1926,175,162; A., 1926,1022; 1927,187,l; Beitr.Cfeophgsik,Chem. News, 1925,131, 305; A., 1925, ii, 1130.8 Ber., 1927, 60, [B], 399; A., 339.1927, 16, 413MINERALOGICAL CHEMISTRY. 295usually regarded as “double salts,” can be split up into opticaIZyactive enantiomorphous isomerides. It therefore becomes necessaryto write a co-ordination formula, witha central sexavalent (“ co-ordimtionnumber ” 6) aluminium atom, the six C20,an octahedron to form the tervalent anion.One of the nine possible types of alumino-oxalates is here shown with two aluminium atoms. In some othercases a central quadrivalent aluminium atom forms with two C,O,groups a univalent anion, giving the complex [C,O4:Alxv:C2O4]R1.The aluminosilicates are suggested9 to be analogous to these alumino-oxalates, and formule on the same lines are given for numerousminerals, a “ silicyl ” group, SiO,, or a “ bisilicyl” group, Si,O,,taking the place of the oxdato-group.Some of the formula?,e.g., for the micas, are written to show the polymerisation of themolecule, and these are so large and complex that they are given astext-figures. The following examples may be quoted :groups being arranged as at the corners ofA1IV :Si20Leucite, Orthoclase, GsI.net,[Al,(Si0,)41KZ* [A1,(SiO*)z(SizOs)*IK,. [AIzOa(SiOa )a ICaa.These formulze suggest an explanation for the breaking down oforthoclase into leucite and silica at a high temperature, for thealteration of felspar to kaolin, of garnet to chlorite, etc. Thealuminosilicates being all high-temperature compounds, suchformulze cannot be tested by the methods of stereochemistry, &E)in the case of organic compounds.They are essentially differentfrom the co-ordination formule proposed by J. Jakob.9aIt is further suggested10 that silicon is not always quadrivalentin the silicates. In the fluosilicates the silicon atom is surroundedby six fluorine atoms and the co-ordination formula is [SiVrE’6]R*2.If the fluorine is replaced by oxygen, the compound [Siv10,]R12,with the same empirical composition as a metasilicate, is obtained.This is assigned to clinoenstatite, which at a high temperaturebreaks down into forsterite and silica. A compound similar tothis, together with “ syntagmatite ” and an addition product ofW. Wahl, Finska Kemistsamfundets Meddelanden, 1927, Nos.1 & 2,40 pp.; 2. Krist., 1927, 66, 33, 173. * Helv. Chim. Acta, 1920, 3, 669; A., 1920, ii, 764; 2. Krist., 1921 56,296; see Ann. Report, 1923, 20, 266.lo W. Wahl, Ann. A d . Sci. Fennicae, Ser. A, 1927, 29, No. 22296 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.jadeite, would give molecules with sufficient structural similarityto form isomorphous mixtures, and so explain the presence ofaluminium in the amphiboles.CI inoenstctti te,[Si20,lMg2. of Tschermak.Addition productof jadeite.Here there is a replacement of silicon atoms with a co-ordinationnumber of 6 by aluminium atoms also with a co-ordination numberof 6. Such a replacement had indeed been suggested for thealuminous amphiboles by P.A. von Bonsdorff in 1821, but thiswas acceptable only before the current ideas of valency haddeveloped, and these, it seems, must now be modified.An interesting attempt to elucidate the constitution of the silicateshas been made by W. L. Bragg,11 who has attacked the problemfrom an entirely new point of view. From the results arrived a t bythe determination of crystal structures by X-ray methods, it issuggested that the structure of silicates is governed by the arrange-ment of the oxygen atoms. These are assumed to be the largestof all the atoms that enter into the composition of the silicates,and the diameter assigned to them is 2.7 8. A symmetrical arrange-ment of such atoms, according to either the cubic or the hexagonalplan of closest packing, is supposed to form the foundation of mostsilicate structures.The smaller atoms of silicon and of the con-stituent metals merely fall into the interspaces between the oxygenatoms. In most cases it appears that the silicon atoms lie at thecentre of a tetrahedral group of four oxygen atoms, as in the buildingunit assigned to g-quartz 12 and tridymite.13 In determining thestructure by X-ray methods of any particular mineral, an attemptis made to locate the precise position of each atom. The relativepositions of the several atoms will then give some idea of the chemicalconstitution.In 1920 W. L. Bragg,f* in a series of calculated atomic diameters,gave that of oxygen (1-30 8.) as the smallest of all the elements.To be told now that it is the largest of all is somewhat disconcerting.If oxygen is to play the governing part in the structure of the silicates,25, 302.406; A,, 1926, 13.11 PTOC.ROY. ~ o c . , 1927, [A], 114, 460; A,, 601; PTOC. Roy. Inat., 1927,12 (Sir) W. H. Bragg and R. E. Gibbs, Proc. Roy. SOC., 1925, [A], 109,18 R. E. Gibbs, ibid., 1926, [A], 113, 351; A,, 1927, 10.14 Ann. Report, 1920, 17, 201MINERALOGICAL CHEMISTRY. 297it should presumably also do so in the many other oxygen salts;and we might consequently argue that there would be only minordifferences in structure between CaSiO,, CaCO,, CaSO,, NaNO,, andNaClO,, or between ZrSiO,, CaSO,, AlPO,, KClO,, etc. As a.matter of fact, these are all very different, the only striking similaritybeing between CaCO, and NaNO,.Further, it is difficult toaccustom ourselves to the idea that silicon is not an essential andreally important constituent of the silicates. In many of theircharacters silicates are markedly different from other oxygensalts. They are, for example, usually distinguished by a highdegree of hardness,l5 as compared with the relatively soft car-bonates, sulphates, etc. The considerable difference in hardness(5 and 7 on Mohs's scale) in two directions on the same face of acrystal of kyanite presents a peculiar problem; and it must notbe forgotten that talc is the softest of minerals.V. M. Goldschmidt,lG on the other hand, points out similaritiesin crystal structure between Z%SiO, (willemite) and Li,BeF,, andbetween CaMgSi,O, (diopside) and NaLiBe,F,.These are evidentlyhomocomorphous relations depending on nearness of molecularvolume, as pointed out for many similar,cases by J. I). Dana in1850. In the same place V. M. Goldschmidt suggests that diopsideis the calcium salt of " diopside " acid, H,MgSi,O,, and jadeite thesodium salt of " jadeite " acid, HAlSi,O,. In another directionhe argues,l' from certain similarities in structure between Mg,SiO,(olivine) and K,SO, on the one hand and A1,BeO4 (chrysoberyl) onthe other, that the first is an orthosilicate. For the metasilicatesMgSiO, (enstatite) and CaSiO, (wollastonite) he finds no othercompounds ABX, of the same crystal type, and it is suggested thatMgSiO, is not a metasilicate, but may be Mg,SiO, + SiO,, and thatorthoclase, KAlSi,O,, may be KAlSiO, + 2Si0,.Crystallographic similarities, based on " comparative externalmorphology" as shown by the development of zones and crystalfaces, are used as a basis of classification by P.Niggli in the secondedition of his " Lehrbuch der Mineralogie " (vol. 2,1926). Here wefind silicates ranged with a variety of other minerals : e.g., phenacitewith cuprite, wollastonite with borax, kaolin with graphite, tour-maline with aragonite, etc. When the imagination is given fullplay, crystallography appears to open out endless possibilities.By the skilful manipulation of axial ratios and neglecting dis-crepancies corresponding with angles up to 74' (i.e., half the1 6 The relation between hardness and crystal structure has been diacuaaedby V.M. Goldschmidt, Skrifter Noreke Videmkaps-Akad., 1927 (for 1926),No. 8, p. 102.16 Ibid., p. 131. l7 Ibid., 1926, No. 1, p. 110.K 298 ANNU& REPORTS ON THE PROGRESS OF CHEMISTRY.difference between 30" and 45"), it is of course possible to demon-strate crystallographically any similarity that may be desired.18Helvine, as dark brown tetrahedra associated with fluorite andgarnet in pegmatite from a new occurrence in Argentina, has beenEtnalysed by W. Fischer,lg who also gives a full discussion of thevarious formula that have been proposed for this and the closelyallied minera1 danalite. This analysis [SiO,, 32.65 ; BeO, 12.20 ;MnO, 30-79; FeO, 14.75; ZnO, 4-89; MgO, 2.24; S, 6.01; total(less 0 for S), 100*54], in showing the presence of some zinc, indicatesa transition from helvine to danalite.When ignited in air themineral gains 3.18% in weight, and water then extracts iron andmanganese sulphates. The formula of helvine was formerlyexpressed as a double compound of the orthosilicate and mono-sulphide of the bivalent metals 3(Mn,Be,Fe),SiO4,(Mn,Be,Pe)S.Since, however, Be : (Mn,Fe) is in the constant ratio 3 : 4, the formula,is better written as3 (Mn,Fe)BeSiO,, (Mn,Fe)S or (Mn,Fe),Be,( SiO,),S,representing an isomorphous mixture of 3hlnBeSi04,MnS and3FeBeSi04,FeS. This formula was written by Brogger and Back-strom (1890) in the form (Mn,Fe)2(Mn,S)UBe,(Si04), in order toshow a relation to the garnet group R1',Rm2(SiO4),, where Be, takesthe place of AI,, etc.J. Jakob (1920) gave a co-ordination formula,which was modified by W. Fischer, the two being, respectively,(Moe,Zn), 0 0p e (Ofii0) , ] 2 and pe(SiO,),] (Fe,Mn)S *But, as pointed out by V. M. Goldschmidt,20 this formula correspondswith the ratios 1Be : 6(Mn,Fe) : 3Si : 120 : lS, whilst the analysesgive Be,(Mn,Fe),Si,012S.are very similar to those given by sodalite, and the unit cubes ofedges 8-19 and 8-85 8. contain two molecules, Be,Mn,Si,O,,S, andA16Na,Si6024C1, (sodalite), respectively. C. Gottfried?, however,finds only one such molecule of helvino in a unit cell of edge 8.52 8.Another mineral belonging to the same group and also crystallisingas regular tetrahedra is the zunyite from Zufiy mine, Colorado. Anew analysis 2, agrees very closely with the earlier analyses ofW.F. Hillebrand (1883) and S. L. Penfield (1893), from which thel* An example that has always acted personally as a warning in this direc-tion is that given by the comparison of the axial ratios of andorite(PbAgSb3S,), aeschynite, coluxnbite (FeNb,OJ, and chalcostibite (CuSbS,)(see Min. Mag., 1897, 11, 286; 1907, 14, 320).lS Bol. Ac&. Nac. Ciencias, Cbrdoba, 1926, 28, 133; Centr. Min., [A],1926, 33.I1 T. Barth, Norsk Gml. Tidsskr., 1926, 9, 40.z2 2. Krist., 1927, 65, 425.es B. Gossner and F. Mussgnug, Centr. Min., [A], 1926, 149.X-Ray powder photographs of helvine20 Centr. Min., [A], 1926, 148MINERALOCIICAL CHEMIS!CRY. 299empirical formula H18A1,,Si,( 0,F,Cl),5 was deduced by Hillebrand.This was modified by Brogger and Backstr8m asto correspond with the garnet formula.B. Gossner and F. Mussgnugdevised a formula, SiO,,AlF( OH),,2A102H or 2SiO2,2Al0F,3A10,H,in which SiO, is partly replaceable by A10,H or A1OF (as suggestedby the hommomorphism of TiO, and MgF,). Later,24 however, onthe ground of X-ray examination of the material analysed, theformula was adjusted so that six molecules3Si0,,3AZO(F,C1),4AlO2H,2Al( OH),shall be contained in the unit cube of edge 13.92 A. Surely thisis only leading to greater complexity, if not confusion.B. Gossner has also employed X-ray methods for the purpose ofdeducing the probable chemical formulae of some other complexsilicates, the formula being readjusted so as to give a whole numberof molecules in the unit cell.For example, for the mineral leifite(of 0. B. Barggild, 1915), as calculated from the original formulaNa2A12Sig0,,,2NaF (M = 788*5), the unit hexagonal cell of dimen-sions a = 14-34, c = 4.93 8. would contain 1-78 molecules. Adop-tion of a formula Na,Si,O,,AlOF (H = 359.5) gives 3433 moleculesin the unit cell, and this is considered a sufficiently close approxim-ation to 4 to justify the new formula.25An exhaustive discussion on the chemical composition of thenumerous minerals of the complex group of chlorites has beengiven by J. Orce1.26 After a detailed historical review of the varioustheories of their constitution, he comes to the conclusion that beforethese theories can be thoroughly tested more data must be accumu-lated. He therefore gives a series of new analyses, with density andoptical determinations, for nineteen chlorites belonging to varioustypes, including two new types-an aluminous sheridanite and amagnesian thuringite.These and earlier analyses (290 in number)are calculated to the ratios SiO,/R,O,, FeO/MgO, Fe,O,/Al,O,, andCr,O,/Al,O,, and only empirical formula are given. On thesoratios is based a new classification of the chlorites : I, Amesite, withSiO,/R,O, = s = 1. 11, Corundophyllite group, s = 1.33 to 1.66.111, Prochlorite group, s = 1-66 to 2.33. IV, Prochloriteclino-chlore group, s = 2.33 to 2-66. V, Clinochlore group, 8 = 2.66 to3.33. VI, Clinochlore-pennine group, s = 3.33 to 3-5. VII,Pennine group, s = 3.5 to 4.5. VIII, Chlorites poor in alumina,s>4*5.Each of these groups is sub-divided according to theratios FeO/MgO and Fe,O3/A&O3; but for the ratio Cr,O,/Al,O,[Al(oH,F,c1)2]~Al2(sio,),24 B. Goswer, Jahrb. Min. Bei1.-Bd., [A], 1926, 55, 319.86 B. Gossner and F. Mussgnug, Centr. Min., [A], 1927, 221.86 Bull. Soc. fraw. Min., 1927, 50, 75; ThAse, Paris, 1927, 380 pp.; A.,1923, ii, 647; 1924, ii, 621; 1926, ii, 821; A., 1926, 42, 933, 1119300 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.the chromiferous chlorites are treated in the text as a special group,IX. To fit this classification certain chlorites (aphrosiderite,thuringite, bavalite, daphnite, delessite, diabantite, leuchtenbergite,kotschubeite) have been re-defined, whilst other terms (rumpfitechloropite, protochlorite, dumasite, pseudophite) are regarded assuperfluous.Optical data are tabulated with the chemical ratiosgiven above, and it is shown that chlorites of different chemicalcomposition may differ in some cases only slightly optically (e.g.,grochanites, clinochlores, and prochlorites). Also dehydrationexperiments (time-heating curves and pressure of aqueous vapourwhen the mineral was heated in a vacuum) led to no definite results.Chlorites are usually thought to be secondary minerals, but theymay also occur as primary constituknts of igneous rocks.The work of H. S. Washington and H. E. Merwin on the chemicalcomposition and optical and other data of the pyroxenes and amphi-boles has been extended to the acmitic pyroxenes, 27 and valuabledata, all determined on the same sample of material, are recorded.They find that the molecules acmite (Ac = Na20,Fe203,4Si02),jadeite (Jd = Na2O,AI20,,4SiO2), vanadous acmite (Vac =Na,0,V20,,4Si02), diopside (Di = Ca0,Mg0,2Si02), and heden-bergite (Hd = Ca0,Fe0,2Si02) may mix in all proportions.Forexample, an “ acmitic diopsidic hedenbergite ” is represented bythe formula AclO,Jd6,Di17,Hd60,A7, where A represents the sumof such molecules as FeO,SiO, ; Fe203,3Si02 ; A120,,3Si02, etc.Excess of sesquioxides is assumed to be present as Rj20,,3Si0,,rather than in the molecules E”e0,Pe,03,4Si02 and Fe0,A1203,4Si02,or in solid solution uncombined with silica. Arfvedsonite fromGreenland has been the, subject of detailed determinations byS.G. Gordon.28 The analyses are interpreted as mixtures of thearfvedsonite molecule (R20,3R0,4Si02), riebeckite molecule(R,0,R20,,4Si0,), and usually an excess of R203, where R20 islargely Na20, and RO and R203 are chiefly FeO and Fe203respectively.Examination of Minerals by X-Ray Methods.The crystal structures of a large number of minerals have nowbeen determined by X-ray methods, and a considerable mass ofdata has been accumulated and in part tabulated.29 Unfortunately,t 7 Amer. Min., 1927, 12, 233.2s L. J. Spencer in “ Tables annuelles de constantes et donndes numdriques,”vol. 6 (for 1917-22), p. 1391, Pans, 1926; vol. 6 (for 1923-4), p. 1226,Paris, 1928. R. W. G. Wyckoff in “ International Critical Tables,” vol. 1,p. 338, Washington, 1926.And in greater detail by P. P. Ewald and C.HermaM, “ Strukturbericht, 1913-26,” issued as separately-paged supple-ments in 2. K~ist., 1927, 65 et seg.PTOC. Acad. Nat. Sci. Philadelphia, 1927, 79, 193MINERALOGICAL CHEMISTRY. 301however, these data have only rarely been correlated with constantsdetermined by other methods on the same sample of material.The majority of original papers give very little idea as to the kindof material that has been used in the experiments, and sometimesit is not clear whether artificially prepared material or a naturalmineral has been employed. Due attention does not seem to havebeen always paid to the careful selection of material, and in a fewcases even the identity of the mineral appears to be open to doubt.Most X-ray workers are content to quote the crystallographicdata and density, and even the chemical composition of the materialin hand, from P.Groth's " Chemische Krystallographie '' (5 vols.,1906-1919). This is a most useful standard work of referencefor the crystallographic constants of artificially prepared chemicalcompounds ; but for minerals it is admittedly incomplete, withonly a few of the more important references to the original literature,the idea being that the full information was already available inthe text-books on mineralogy. For example, under the heading" Berylliumaluminiummetasilicat = SisOleAlzBe3," Groth gives avery inadequate account of the mineral species beryl. There areseveral varieties of this mineral.Some contain up to 5% ofalkalis (usually czsium) and most contain up to 2% of water.The text-book formula of beryl can be regarded as only approxi-mate, and it may be doubted if the mineral is really a meta-silicate. The density of beryl is given by Groth as 24----2.7, butrecent determinations show a range from 2.545 to 2.910. Thehexagonal crystals, although well developed, sometimes show opticalanomalies and a complex intergrowth in sectors. Now, based onthis information taken from Groth, a most elaborate structure hasbeen built up for beryl as deduced from the examination of a crystal(or crystals 1 ) by X-ray methods.30 Not the slightest indication isgiven in the original paper of the kind of material used for thisinvestigation; there is no mention of colour (a useful guide to thevarieties of beryl), density, or other characters." The X-raymeasurements lead to a value 2.661 for the density "-but a directdetermination would have been more useful.The same authors31 have also, on data quoted from Groth,deduced structures to explain the morphotropic relations of thehumite group of m'inerals. This group has frequently been quotedas a classical example of a morphotropic series. These morpho-tropic relations are based on the chemical formulz of Penfield andHowe (1894), which have never been confirmed; and recently they30 W. L. Brrtgg and J. West, Proc. Roy. SOC., 1926, [A], 111, 691; A,,1926, 889.31 Idem, ibid., 1927, [A], 114, 450; A., 501302 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.have been called into question by G.Ce~iiro,~~ who has deducedother formulae from the original analyses. Further, H. Sjogren’s“ prolectite ” (1895), supposed to be an end member of this series,has been proved to have no existence.% The three mineralschondrodite, humite, and clinohumite are not easy of determinationand distinction, and it would not appear that the X-ray methodaffords a practical test.It is evident that in the examination of minerals exclusively bya single method, such as the X-ray method, a too narrow view isbeing taken. The same undue importance is often also attached tooptical methods in mineralogy. The only papers on minerals inwhich the materials have been completely examined by all availablemethods, including X-ray methods, are those by G.Aminoff. Forthe minerals br~mellite,~~ magnetoplumbite,35 swedenborgite,36 andtrimerite,37 the crystallographic constants, crystal structure, opticaldata, density, chemical composition, etc., have all been determinedon the same sample of material. The history, origin, and para-genesis of the minerals are concisely stated, and the main factsand data are clearly set out, for the benefit of the reader or recorder,without being confused with a mass of more or less irrelevantspeculation. Further, the papers are presented once for all in acompleted and finished form. There are no preliminary notices,reprints with minor modifications and corrections, and translationsin other journals. This simplifies the literature and the work ofthe bibliographer.These excellent papers may well be taken as apattern.Several minerals of the pyroxene group have been examined andcompared by the X-ray powder method; 38 and here in severalcases the work was done on material that had been previouslycompletely determined chemically and optically by the sameauthors. These minerals give patterns of four main types :(1) Diopside (including hedenbergite, acmite, jadeite, augite) ;(2) Enstatite (and hypersthene) ; (3) Wollastonite (and bustamite,pectolite, schizolite) ; (4) Rhodonite (and calcium-rich rhodonite).Some others (clinoenstatite, spodumene, alamosite, pyroxmangite,sobralite, babingtonite) give other types of pattern. It was foundthat the replacement of MgO by FeO produces practically no changeBull.Acad. TOY. Belg., 1926, 12, 350; A,, 1927, 336.33 P. Geijer, CTeol. P6r. P6rh., 1926, 48, 86.s4 2. Krist., 1925, 62, 113; Ann. Report, 1925, 22, 277.35 Qeol. F6r. FGrh., 1925, 47, 283.8 8 Ibid., 1926, 48, 19.37 2. Krkt., 1924, 60, 262; Ann. Report, 1925, 22, 279.38 R. W. G. Wyckoff, H. E. Merwin, and H. S. Washington, Amer. J . Sci.,1926 [v], 10, 383; A., 1926, ii, 1126MINERALOGICAL CHEMISTRY. 30 3in the structure; MnO has a slightly greater volume, and CaO anappreciably greater volume. In augite the excess of Al,O, andFe,O, has no appreciable effect.The unit-cell dimensions of the isomorphous members of thegarnet group have been determined by G. Menzer39 and also byC. H. Stock~ell.*~ The latter has determined for 40 garnets therefractive index, specific gravity, and the edge of the unit cube(by the X-ray powder method).In only two cases, however, werethe determinations made on analysed material; but with the aidof the previous work of W. E. Ford (1915) on the refractive indexand specific gravity of the garnets, a correlation was obtainedwith the chemical composition. Plotting refractive index againstspecific gravity, Ford found that the garnets fall into two series(almandine-pyrope-spessartine and grossular-andradite), and thisis emphasised by plots of the cell dimensions against either therefractive index or the specific gravity. From these data, (with,in some few cases, a supplementary qualitative test for manganese),Stockwell was able to determine the nature of a given garnet, andthe percentage molecular composition can be calculated.Thecalculated values for the pure molecules are :Pyrope, Mg,Al,(SiO,), ............... 1.705 3-610 11.430 A.Almandine Fe,Al,(SiO,),. ........... 1-830 4.260 11.493Spessartine, Mn,Al,(SiO,), ......... 1.800 4.180 11.668Grossular, Ca,Al,(SiO,), ............ 1.736 3.630 11.840Andradite, Ce,Fe,(SiO,), ............ 1.895 3.760 12.040A good example of how X-ray methods may come to the aid ofmineralogical description when ordinary crystallographic methodshave failed is given in the case of the new mineral aramay~ite,~~Ag(Sb,Bi)S,. Distinctly developed crystals are not available andfrom the cleavages it could only be concluded &hat the mineralwas perhaps tetragonal. Laue photographs, taken by (Miss) K.Yardley 42 through the perfect basal cleavage, showed an absenceof symmetry, and spectrometer and powder measurements provedthe mineral to be triclinic (pseudo-tetragonal).By these means thedimensions of the unit cell and the crystallographic constants werecompletely determined.So many workers are now engaged on X-ray research that itoften happens that the same mineral has been examined independ-ently several times. For example, the following determinations ofthe unit orthorhombic cell of baryte (BaSO,) show a very satisfactoryagreement.9%. d. a.3s Centr. Min., [A], 1926, 344; 1926, 343.40 Arner. Min., 1927, 12, 327.41 L. J. Spencer, Min. Mag., 1926, 21, 166; A., 1927, 226.42 Min.Mag., 1926, 21, 163; A,, 1927, 190304 ANNUA4L REPORTS ON THE PROGRESS OF CHEMISTRY.a. b. C.L. Pauling and P. H. Emmett 4 5 .................. 8.846 5.430 7.10R. W. G. Wyckoff and H. E. Merwin 46 ......... 8.89 5.45 7-17F. Rinne, H. Hentschel, and E. Schiebold 4 7 ... 8.88 5-45 7.15W. Basche and H. Mark 48 ........................ 8.85 5-45 7-14S. K. Allison 43 ....................................... 4.449 5.448 7.170 A.R. W. James a d W. A. Wood 44 ............... 8.852 5.430 7,132With one exception, all these authors agree in taking a doublevalue for the a-axis, and the axial ratios a : b : c, usually acceptedas 0.8152 : 1 : 1.3136, become 1.6304 : 1 : 1.3136. This means thatthe rarer prism n(120) becomes the unit prism instead of the prismparallel to the perfect cleavage, which by crystallographers isnaturally taken as a primitive form. The reason for the markeddifference in this case is not clear.In this connexion we arereminded of the well-known similarity between the crystal formsof baryte and sulphates, selenates, chromates, perchlorates, andpermanganates with the same type of formula, vix., PbSO,, BaSeO,,BaCr04, KClO,, -no4, etc. F. Rinne 49 has referred to these asisotypes of the baryte type, and he points out a remarkable relationbetween the interfacial angles of the crystals. In baryte the meanof the three angles (110) : (110) = 78" 22', (011) : (01T) = 74" 34',(102) : (T02) = 77" 43' is 76" 53'; and in all the other salts themean of the corresponding angles is, perhaps by a mere coincidence,also 76" 53'.This angle is near to that (77" 19') of the cubicpentagonal-dodecahedron (540), which in combination with thecube closely resembles the baryte habit. The fact that some ofthese salts change into a cubic modification a t a higher tem-perature is perhaps related to this approximation to cubic angles.Based on the above similarity, A. E. H. Tutton 50 has calculatedfrom the topic axes the dimensions of the unit cells of variousperchlorates.In contrast with the above example of close agreement obtainedindependently by X-ray workers in different countries, an examplemay be quoted of lack of agreement. For the indirect determin-ation of density from the structure of mercury telluride (butwhether on artificially prepared HgTe or on the mineral coloradoite43 Amer.J . Sci., 1924, [v], 8, 261 ; A., 1925, ii, 18.44 Mem. Mamhster Phil. SOC., 1925, 69, No. 5; Proc. Roy. SOC., 1925, [A],45 J . Amer. Chem. SOC., 1925, 47, 1026; A., 1925, ii, 485.46 Amer. J . Sci., 1925, [v], 9, 286; 2. Krist., 1925, 61, 452; A., 1925, ii,4 7 2. Krist., 1925, 61, 164.48 Ibid., 1926, 64, 1.49 Centr. Min., 1924, 161.60 Proc. Roy. SOC., 1926, [A], 111, 462; A,, 1926, 888.109, 598; A,, 1926, 13.485MINERALOGICAL CHEMISTRY. 305is not in all cases quite clear) the following values have beenobtained :d.W. Hartwig 61 ....................................... 8.026W. F. de Jong 62 .................................... 8-20W. Zachariasen 53 ....................................8-42Microscopic Examination of Opaque Minerals.The examination under the microscope of polished sections ofore-minerals by reflected light, following the methods of metal-lography, has been much developed during recent years. Thissubject, or rather method of investigation, has been called " mineral-ography " or " mineragraphy '' in America and " chalcography ''(Chalkographie) 54 in Germany. Text-books have been written byJ. Murdoch (New York, 1916), W. M. Davy and C. M. Farnham(New York, 1920), H. Schneiderhohn (Berlin, 1922), and R. W.van der Veen (The Hague, 1925), and a useful outline with detailedbibliography has been given by J. O r ~ e l . ~ ~ Various chemicalreagents are applied to the polished surfaces for the purpose ofdistinguishing one mineral from another.More recently, themethod has been extended by the use of polarised light, it beingpossible to distinguish between isotropic and anisotropic crystalsand to determine the directions of the principal axes of refringenceand of absorption. The method has been extensively applied inAmerica and Germany to the study of ores and ore-deposits. Inaddition to identifying the various minerals present in the ore,much can be learnt from their mutual relations and order ofdeposition.Various obscure and doubtful metallic minerals examined bythis method have been proved to be really mixtures, and the com-plex chemical formulae that have been applied to them are thusreadily explained. The formula of bornite has been usually givenas Cu3FeS3; but the mineral is frequently intergrown with chalco-pyrite, chalcosine, etc., and even well-developed crystals oftencontain a nucleus of chalcopyrite.Recent analyses made onmaterial proved microscopically to be homogeneous have giventhe formula C U ~ F ~ S , . ~ ~ In argentiferous galena it has been shown61 Sitzungsber. Preuss. Aka&. Wiss. Berlin, 1926, 79; A,, 1926, 664.62 2. Krist., 1926, 63, 466; A,, 1926, 996.63 Norsk Geol. Tidsskrift, 1926, 8, 302; also 2. physikal. Chem., 1926, 124,s4 The term chalcography has been in use in English since s t least the66 Bull. SOC. frarq. Min., 1926, 48 (for 1925), 272-361 ; Rev. Mdt., 1926,m J. Orcel, Bull. $oc.frmw. Min., 1926, 48 (for 1926), 340.277 ; A,, 1927, 96, where the value 8.123 is given.year 1661 for the art of engraving on copper.23, 637, 618306 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.that the silver is present as specks of tetrahedrite or argentite, orsometimes, in richer samples, as veinlets of ruby-silver.Theexamination of polished sections of the selenium ores from theHarz Mountains has led to the identification of umangite (Cu,Se,)in addition to the selenium minerals previously known from thislocality. In polarised light the umangite shows a strong pleo-chroism, from cherry-red to grey.57 The complex intergrowths ofarsenides and sulpharsenides of cobalt, nickel, and iron occurringa t Cobalt in Ontario have also been studied by this method.5*The two modifications of silver sulphide, the cubic argentite andthe orthorhombic acanthite, were found by H.Schneiderhohn 59 in1922 to be optically anisotropic and both presumably orthorhombic.Crystals of argentite show a complex lamellar structure and areevidently paramorphs after the high-temperature cubic modification,as is the case with leucite. The inversion temperature for silversulphide is 180". Similarly, cuprous sulphide is dimorphous asrepresented by orthorhombic crystals of the mineral chalcosineand by artificially prepared cubic crystals, the latter being stableabove 91". In the copper ores of Tsumeb in South-West Africa,Schneiderhohn found in 1920 that the chalcosine is of two kinds :(1) a more abundant form with a granular structure and presumablyof secondary formation ; (2) one showing a complex lamellar struc-ture with an octahedral arrangement, very similar to the structureshown by meteoric irons.The latter he concluded was the primaryore which had crystallised as the cubic modification a t a temperatureabove 91", and which on cooling passed over into the orthorhombicmodification. The presence of " lamellar argentite " or of " lamellarchalcosine " fixes two points (180" and 91') on the " geologicalthermometer " in any discussion on the origin of ore-deposits.These conclusions, arrived at by the metallographic method, onthe dirnorphous relations of the silver and cuprous sulphides, havesince been amply confirmed by the X-ray method when the materialswere examined a t different temperatures.60Forms of Calcium Carbonate.While " conchite " and " ktypeite " are evidently compact formsof aragonite, the artificially prepared vaterite is doubtless distinctfrom both calcite and aragonite, and it appears to be identical withthe p-CaCO, of Johnston, Merwin, and Williamson (Ann.Report,6' G. Frebold, Centr. Min., [A], 1927, 16, 196; J. Olsacher, ibid., p. 170.s8 E. Thomson, Univ. Toronto Studies, Geol. Ser., 1926, No. 20, 54.69 Arner. Min., 1927, 12, 210.3'. Rinne, 2. Krist., 1924, 60, 299; L. S. Ramsdell, Amer. Min., 1926,10, 281; R. C. Emmons, C. H. Stockwell, and R. H. B. Jones, ibid., 1926,11, 326; T. Barth, Centr. Nh., [A], 1926, 284MINERALOGICAL CHEMISTRY. 3071917, 14, 241). More recent.ly, Gibson, Wyckoff, and Merwin6lfind by the X-ray powder method that the spherulitic material, towhich the name " vaterite " has been applied, includes two forms :" vaterite A " prepared by the method of Johnston, Merwin, andWilliamson is spherulitic calcite ; whilst " vaterite B " (i.e., thetrue vaterite), obtained from colloidal calcium carbonate at 5" inthe presence of an excess of potassium carbonate, is p-CaCO,.The latter was obtained by Johnston, Merwin, and Williamson ashexagonal plates.F. Heide 62 finds that the gelatinous precipitategiven by solutions of N/2-potassium carbonate and 2N-calciumchloride a t 5" changes after l+-2 hours to minute (3-10 p)radially-fibrous spherulites of vaterite. Heated a t 100" in water,this is transformed into calcite, but the dry material is stlabJe upto 430-440". X-Ray powder photographs show a structuredifferent from that of both calcite and aragonite.The hexagonalcell of vaterite has dimensions a = 4.120, c = 8.556 8., and containstwo molecules of CaCO,; the calculated density is 2.645. Afterheating a t 430--440", the material shows the lines of calcite in theX-ray ph0tographs.6~At higher temperatures the rhombohedra1 calcite is the onlystable modification, but transformation into ct-CaC03 a t 970" underpressure in an atmosphere of carbon dioxide was recorded byH. E. Boeke in 1912. Such a transition point could, however, notbe found by F. H. Smyth and L. H. ad am^,^^ and they placed themelting point of calcite at 1339" under 779,000 mm. pressure.This has an important bearing on the question of the occurrenceof primary calcite in igneous rocks, the so-called magmatic calcite,about which there has recently been much discussion.Magmaticcarbonate rocks (" carbonatites ") were described by W. C. Bragger(1921) from the Fen district in Norway; and E. Schuster (1919)and R. Brauns (1919) have described calcite-pegmatite and calcite-syenite from the Laacher See district in Rheinland. This hasbeen disputed by N. L. B ~ w e n , ~ ~ who considers that such occur-rences represent secondary replacement of silicates, particularlyfelspars, by calcite. When limestone rocks are invaded by igneousmagmas, there appears no reason why calcium carbonate shouldnot be incorporated in the mass and melted.There is, for example,abundant evidence of this in the pegmatite veins and nephelinc-61 Amer. J . Sci., 1925, [v], 10, 325; A., 1925, ii, 1183.62 Centr. Min., 1924, 641.63 F. Heide, ibid., [A], 1925, 198.64 J . Amer. Chem. SOC., 1923, 45, 1167; A , , 1923, ii, 490.66 Amer. J . Sci., 1924, [v], 8, 1; Centr. Min., [A], 1926, 241; Amer. J.Sci., 1926, [v], 12, 499; and replies by R, Brauns, Cen,tr. Min., [A], 1926,1 and 245308 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.syenites of Canada. Some authors66 have further attempted toprove that such calcite shows peculiarities of structure markingits inversion from a-CaCO, as the temperature fell below 970°,this being taken as a point on the "geological thermometer," inthe same way that the change from @-quartz to ct-quartz67 fixesthe point 575".The distinction appears, however, to be basedmainly on lamellar twinning and optical anomalies, such as maybe produced artificially in calcite by pressure. The name elatolite 68has been applied to what is believed to have 'been ct-calciteleached out of tree-like cavities in the nepheline-syenites of theKola peninsula in Russian Lapland.The minerals hydroconite, hydrocalcite (trihydrocalcite, penta-hydrocalcite), and lublinite periodically come to be regarded asdoubtful minerals, because when re-examined on museum materialthey are found to be merely calcite. Hydrated calcium carbonate(CaCO,,GH,O) is stable only below 5", and in nature it is but rarelyobserved in the winter as a mould-like efflorescence on limestoneand chalk.To preserve such a mineral a refrigerator would berequired. In addition t o the he~ahydrate,~~ crystals of the penta-hydrate have also been prepared ; whilst a trihydrate, perhapsstable between 17" and 25", could not be isolated.70Determinative Tables.E. S. Fedorov's method of crystallo-chemical analysis (Ann.Report, 1923, 20, 289), on which great hopes were laid, has notcome into general use. The symbols he uses are not generallyintelligible, and the introduction to his volume " Das Krystall-reich " is so written that from the volume itself it is not possibleto understand his method. A. K. Boldyrev, in a pamphlet pub-lished by the Russian Academy of Sciences,71 explains doubtfulpassages in this introduction and gives supplementary explanationsfor the instruction of the reader. Boldyrev 72 has also published66 T.L. Walker and A. L. Parsons, Univ. Toronto Studies, Ceol. Ser., 1925,No. 20, 14; J. L. Gillson, Amer. Min., 1927, 12, 357.67 Unfortunately there is here some confusion in nomenclature : a-CaCO,or a-calcite is the high-temperature modification, whilst in quartz the high-temperature modification is denoted as &quartz and the low-temperaturemodification &s a-quartz.O 8 A. E. Fersman, 1922-3 ; Ann. Report, 1925, 22, 278.6 ) Ann. Report, 1923, 20, 277.70 J. Hume, J., 1925,127, 1036; A , , 1925, ii, 697; J. Hume and B. Topley,J., 1926, 2932; A., 1927, 12.7 1 " Kommentarien zum Werk von E. S. Fedorow : ' Das Krystallreich ' "[Russian with German r6sum61, Leningrad, 1926, 72 pp.[Min. Mag. (Abstr.),1927, 3, 3261.72. M6m. SOC. Russe Min., 1924, 53, 261 [Min. Mag. (Abstr.), 1926, 3, 1SS-JMINERALOUICAL CHEMISTRY. 309a criticism of the method in which he points out that there is somedifficulty in arriving at the correct complex-symbol. He hastherefore suggested and worked out in some detail another methodof “ crystallo-chemical analysis.” This method is based on inter-facial angles taken in conjunction with other characters. Some9000 substances have been entered on catalogue cards and sortedaccording to the angles in each crystal system. The cards givefor each substance : (1) a list of the common crystal forms in theorder of their importance, cleavage, twinning, etc.; (2) physicaland optical data ; (3) the important crystal angles. In the opticallyuniaxial systems the angles given are those to the basal plane, andin the remaining systems the angles t o the axial planes (fromwhich the latitude and longitude, p and 9, angles of two-circlegoniometry can be deduced if wanted). The method is explainedin the French resume, and sample cards are printed in German inthe Russian text. The data have been largely compiled fromP. Groth’s “ Chemische Krystallographie.” One of the examplesgiven for the orthorhombic system is “ orthorhombic tin ” (“ /%tin ”).If the cards had sorted out properly it would have been noticedthat this is identical with stannous ~ulphide.’~Determinative tables or keys of various kinds have long beenused in mineralogy.They have been based on obvious externalcharacters, such as lustre, colour, and streak, supplemented byhardness, specific gravity, system of crystallisation, optical charac-ters, etc. In one ingenious device a number of perforated sheetseach corresponding with a certain character are laid over a sheeton which are printed the names of minerals until at last the nameof the mineral wanted appears in the only opening left. Rule-of-thumb methods of this kind only lead to error, and any table orkey must be used with a certain amount of understanding andintelligence. The main use of a table is in suggesting what agiven mineral may or may not be, and then some special test mustbe applied. E. S. Larsen’s tables of optical data (Ann.Report,1923, 20, 290) are perhaps the most practical and useful that haveyet been devised for determinative purposes. A table of specificgravities may also be a useful aid for the identification of minerals.A recent table 74 gives a numerical list of 2277 determined valuescollected from the literature for the period 1910-1927; and analphabetical list of mineral names gives the minimum and maximumvalues recorded for each mineral.A useful table might be compiled from the data given by theunit -cell dimensions (in Angstrom units) of crystals as determined73 Compare Min. Mug., 1921, 19, 113.74 L. J. Spencer, Min. Mag., 1927, 21, 337310 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.by X-ray methods. In such a table, with a numerical arrangement,there would be a considerable differentiation of the various kindsof materials.The X-ray patterns given by the powder method ofvarious known minerals have already been used by several workersas standards for comparison for the purpose of identifying theminerals present in intimate mixtures (Ann. Report, 1923, 20, 282).A reference collection of the X-ray powder patterns of chemicallyanalysed samples of minerals has been commenced in the Depart-ment of Geology of the University of Wisconsin.75New Minerals.A considerable number of " new " minerals have been describedmore or less completely since the last Report. The following listis limited to those that appear to be distinctive and well established.The ten lists of new mineral names that have been published a tthe end of each volume of the Mineralogical Magazine since 1897include all names not recorded in Dana's " System of Mineralogy "(6th ed., 1892) and show a total of 1506 names.These are nowall collected together in a general index to the set of volumes.76Undoubtedly many more minerals remain to be discovered,although it is of course not to be expected that all known inorganiccompounds will be found in the native state. P. N. Chirvinsky,77in a note on the prediction of minerals by mineral synthesis, givesa list of 163 chemical elements and compounds that had beenprepared artificially before they were known as minerals. He alsogives statistical data of the new minerals described during the past20 years, and finds that silicates and phosphates predominate.The suggestive work of the late Baron A.de Schulten might beextended in this connexion. He prepared artificially a number ofminerals in a crystallised form and determined the crystallographicand physical constants for the pure compounds. He then proceededto prepare the analogous compounds in the same isomorphousseries. For example, having obtained crystals of artificial monetite,CaHPO,, he then prepared the corresponding compounds in whichstrontium, barium, or lead takes the place of calcium, and arsenicthe place of phosphorus. I n this series he determined the crystal-lographic constants and optical data for CaHPO,, SrHPO,, BaHPO,,PbHPO,, SrHAsO,, and PbHAsO,. Crystals of CaHAsOp werealso obtained but they were too small for measurement. Any ofthese compounds might be expected to occur in nature, but hitherto76 A.N. Winchell, " ' Finger Prints ' of Minerals," Amer. Min., 1927, 12,261.76 L. J. Spencer, Min. Mag., 1926, General Index to vols. 11-20 (1896-1925).?7 P. Tschirwinsky, 2. KT~s~., 1926, 64, 644MINERALOGICAL CHEMISTRY. 311only monetite has been known. Recently his " arsenical leadmonetite " has been found as a mineral and named schultenite(p. 313 below).Two recently found minerals, not yet completely described,appear to represent new compounds of palladium. One of them,of sparing occurrence in the diamond washings of British Guiana,was determined by the late Sir John Harrison 78 to be a palladiummercuride with the probable formula PdHg.The small silver-white nuggets and grains have a density up to 15.82, i.e., consider-ably higher than that of either palladium or mercury. This mineralhas since been named potarite, from the Potaro River in BritishGuiana. A palladium antimonide,79 Pd,Sb, has been found, inassociation with fine large crystals of sperrylite (PtAs,), in thePotgietersrust platinum fields, Transvaal, where minute silver-white grains were detected in pannings of the platinum ore.Ammoniojarosite 81 is one of the few minerals containing ammon-ium and is interesting in illustrating the wide range of isomorphousreplacement in the jarosite group.Jarosite ....................................... K20,3Fe20s,4S0,,6H20Natrojarosite .................................Na20,3Fe,0,,4S0,,6H20Plumbojarosite ........................... Pb0,3Fe20,,4S0,,6H,0Argentojarosite .............................. Ag20,3Fe,0s,4S0,,6H20Ammoniojarosite ........................ ( NH4)20,3Fe203,4S0,,6H20Ammoniojarosite was found as small ochre-yellow nodules withtschermigite (ammonium alum) in lignitic shale; and, like allthe other members of the group, it comes from Utah. Stillanother member of the same group is probably carphosiderite,H20,3Pe2O3,4S0,,6H2O. Replacing ferric oxide by alumina,another section of this isomorphous group is represented by aluniteand natroalunite. Further, the sulphate may be partly or whollyreplaced by phosphate, giving several sulphato-phosphates (beud-antite, svanbcrgite, etc.) and phosphates (hamlinite, florencite,etc.), all of which are closely related crystallographically to jarositeand alunite.Aramyoite, 82 sulphantimonite and sulphobismuthite of silver,Ag(Sb,Bi)S,, is found in a silver-tin vein in Bolivia as iron-blackplaty aggregates with brilliant metallic lustre on the perfect basalcleavage.It breaks up into square (or nearly square) platesJ. B. Harrison and C. L. C. Bourne, OJ'. Gazette Brit. Uuiana, Feb. 27,1925, No. 71 ; A . , 1926, ii, 693.?@ H. R. Adam, J . Chem. Met. SOC. S. Africa, 1927, 27, 249; A., 861.*O L. J. Spencer, Min. Mag., 1926, 21, 94.89 L. J. Spencer, Min. Mag., 1926, 21, 156; A,, 1927, 226; E. Kittl,E. V. Shannon, Amer. Min., 1927, 12, 424.Revista Minera de Bolivia, 1927, 2, 53312 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.bounded a t the sides by fibrous pyramidal cleavages.In theabsence of definite crystal faces for goniometric measurement, itappeared to be tetragonal, but as determined by X-ray methods 83it is shown to be triclinic (pseudo-tetragonal), and the crystallo-graphic constants have been completely determined.Avogadrite, potassium fluoborate, KBF4, occurs as minute ortho-rhombic plates (homeomorphous with baryte, BaSO,, KMn04,and KCIO,) in saline sublimations from the floor of the crater ofVesuvius. The detection 84 spectroscopically of notable amountsof cmium together with potassium in the aqueous extract of thesesublimations led to the discovery of the new mineral, and itscharacters were more completely determined on material recrystal-lised from this solution.85 A comparison of the data so obtainedwith the physical constants of pure potassium fluoborate and ofpure casium fluoborate suggested that the crystals of avogadritecontained KBF,, 90.5yo, and CsBF4, 96y0.A later crop of re-crystallised material gave data (d 2.498, nsa 1.325) agreeing moreclosely with those for pure potassium fluoborate.86 The refractiveindex is less than that of water.B~ltgenbachite,~~ a complex copper salt,2CuC12,Cu(N03),, ~~CU(OH)~,~H,O,differs from connellite [2CuC1,,CuS04,15Cu(OH),,4H,0] in contain-ing nitrate in place of sulphate. It is found as a fine felt of sky-blue needles with native silver in cavities in cuprite a t Likasi,Belgian Congo.The crystals are hexagonal, and buttgenbachiteis optically negative, whilst connellite is positive. The two mineralsare the end members of an isomorphous series in which mixedcrystals occur; previously, W. E. Ford and W. M. Bradley, in1915, had found o.72y0 of N,O, in connellite from Arizona.FZuoborite,88 a fluoborate of magnesium, 3Mg0,B203+ 3Mg(F,OH),,has been found as colourless hexagonal prisms in an iron mine a tNorberg, Sweden.IanthiniteYs9 hydrated uranous oxide, 2U0,,7H20 ( ?), represent-ing an intermediate alteration product of pitchblende at the Kasolo83 (Miss) K. Yardley, Min. Mag., 1926, 21, 163 ; A,, 1927, 190.84 F. Zrwnbonini and L. Coniglio, Atti (Rend.) R. Accud. Lincei, 1926, [vi],86 C. Carobbi, ibid., 4, 382; A,, 1927, 129.87 A.Schoep, Compt. rend., 1925, 181, 421; Bull. SOC. chim. Belg., 1926,34, 313; A . , 1925, ii, 1196; H. Buttgenbach, Ann. SOC. gkol. Belg., 1926, 50,Bull. 35; A. Schoep, ibid., 1927, 49 (for 1926), Bull. 308; 1927, 50 (for88 P. Geijer, Geol. F6r. F'cirh., 1926, 48, 84 ; Arsbok Sveriges Geol. Unders.,8s A. Schoep, Natuurwetensch. Tijds., 1926, 7 (for 1925), 97; ibid., 1927,3, 521; A., 1926, 816. 86 F. Zambonini, ibid., p. 644; A., 1926, 934.1926-7), Bull. 216.1927, 20 (for 1926), No. 4.9, 1; Ann. SOC. gdol. Belg., 1927, 49 (for l926), Bull. 188, 310MINERALOGICAL CHEMISTRY. 313mine, Katanga, Belgian Congo, forms minute orthorhombic crystalswith a micaceous cleavage in one direction. The crystals areblack with a violet tinge and semi-metallic lustre ; the pleochroismis intense-dark violet to colourless. Some of the crystals arebordered by a yellow zone and others are completely changed to ayellow material. A crystal of ianthinite heated a t 50" in a drop ofwater changes from violet to brown and finally to yellow. Theseyellow alteration products are perhaps becquerelite and schoepite(U0,,2H20), although they differ from these in their optical char-acters. The refractive indices of ianthinite are wa = 1.674, n;e =1.90, n, = 1.92. A specimen acquired in 1922 for the mineralcollection of the British Museum has since been identified asianthinite. It shows a velvet-like pile of minute needles on pitch-blende. When acquired, the colour was purple, but now (1927) itis greenish-yellow; crystals that had been mounted in Canadabalsam still retain their original colour and intense pleochroism.The change is evidently due to oxidation in the air of a hydrateduranous oxide to a hydrated uranic oxide.Kernite 91 is hydrated sodium borate, Na2B40,,4H20, containingless water of crystallisation than borax. As large orthorhombiccrystals and as clear cleavage masses, it has been found in somequantity in bore-holes in Kern Co., California.MaZk~Zrite,~~ sodium fluosilicate, Na2SiF,, occurs as minutehexagonal prisms, together with hieratite (K,SiF,) in materialcollected from fumaroles on Vesuvius. The recrystallised materialshows regular growths of the cubic potassium salt on the hexagonalsodium salt. Another fluosilicate, cryptohalite [ (NH,),SB',], hadbeen previously observed on Vesuvius.Sch~Ztenite,~~ lead hydrogen arsenate, PbHAsO,, colourless mono-clinic crystals of platy habit, is found on pseudomorphous crustsafter mimetite and anglesite a t Tsumeb, South-West Africa. Thetable of angles for the several crystal forms is set out to give theangle from each face to the three axial planes, this method com-bining the advantages of both the arrangement of interfacial anglesin zones and the latitude and longitude angles of two-circleg~niornetry.~~?L. J. SPENCER.90 V. Billiet, Natuurwetemch. Pijds., 1926, 7 (for 1925), 112; Bull. Soc.91 W. T. Schaller, Amer. Min., 1927,12, 24; H. S. Gale, Engin. Mining J.,92 F. Zambonini and G. Carobbi, Atti (Rend.) R. Accad. Lincei, 1926, [vi],g3 L. J. Spencer, Nature, 1926,118, 411 ; A., 1926, 1022; Min. Mag., 1926,fraw. Min., 1926, 49, 136.1927, 123, 10.4, 171; A., 1926, 1119.21, 149; A., 1927, 226. s p Compare A. K. Boldyrev, p. 309 above