I NORG A NIC CH EM ISTRY. Inorganic Chemistry. ii. 107 Reactions in the Presence of Nickel. ( a ) Inability of Nitrogen and Hydrogen to Combine in Presence of Nickel. (6) Reduction of Oxides of Nitrogen Sulphur and Phosphorus in Presence of Nickel. PA~CH~NAN NEOGI and BIRENDRA BHUSAN ADHIC~RY (Zeitsch. anorg. Chem. 1910 69 209-214).-A mixture of nitrogen and hydrogen (1 3 or 1 6 by volume) does not give rise to the formation of ammonia when passed over heated reduced iron the temperature varying from dull to bright red heat. Similar negative results are obtained when the reduced iron is replaced by ‘pure iron wire or by ferric oxide no matter whether the gases are dried or not (compare Ramsay and Young Trans. 1884 45 93). No ammonia could be detected when the iron was replaced by reduced nickel nickel wire or electrolytic nickel.When a mixture of nitric oxide and hydrogen (1 3 or better 1 4) is passed over reduced nickel the nitric oxide is almost quantitatively converted into ammonia. The reaction begins at 300° but once it has set in the temperature may be lowered to 120’. Hydrogen sulphide is formed when a mixture of sulphur dioxide and hydrogen is passed over nickel heated to a dull red heat. The hydrogen sulphide is probably formed partly by direct reduction of the sulphur dioxide and partly by reduction of nickel sulphide formed by the interaction of nickel and sulphur dioxide. Hydrogen phosphide is produced when hydrogen is passed over a mixture of phosphoric oxide and reduced nickel at a dull red heat. T. S. P. Catalysis of Hydrogen Peroxide.ERNST H. RIESENFELD (Ber. 1911 44 147-150. Compare Abstr. 1905 ii 951).-The reaction between chromic acid and excess of hydrogen peroxide can be represented by the equation 4H2Cr,0 -i- 7H,02 = Cr2(Cr207)3 + 1 1H,O + 50,. This agrees with Spitalsky’s statement that only about 28% of the chromic acid is reduced b u t this author did not notice that all the free acid is used up. The reaction is not a catalytic process (compare Spitalsky this vol. ii 36 37). J. J. S. Raschig’s Nitrososulphonio Acid ‘‘ Blue Acid.” WILHELM MANCHOT (Zeitsch. angew. Chern. 19 11 24 13-1 4. Compare Abstr. 1910 ii 956 1055).-Mainly a reply to Xaschig (Zeitsch. angew. Chem. 1910 23 2248). Raschig’s ferric nitrososulphonate should have the ratio NO Fe = 1.5 1 whereas the actual proportion is 2 1.It is pointed out that there is no trace of evidence for the existence of a blue compound formed by the union of nitric oxide with sulphuric acid. J. J. S. Products Formed when Phosphoric Oxide Dissolves in Water. D. BALABEFF (Zeitsch. nnory. Chem. 191 0 69 2 15-2 16). -Phosphoric oxide was allowed to deliquesce over (1) 75% sulphuricii. 108 ABSTRACTS OF CHEMICAL PAPERS. acid (2) 9% sulphuric acid ; some was also thrown into water. I n the first two cases the reaction was considered to be complete when the snow-like appearance of the phosphoric oxide has disappeared. I n all cases metaphosphoric acid was the only product of reaction. After eighteen hours the metaphosphoric acid had completely changed into the ortho-acid in the last two cases. The velocity of hydration of metaphosphoric acid produced by the deliquescence of phosphoric oxide is much greater than that of the metaphosphoric acid obtained by heating orthophosphoric acid.T. S. P. Rate of Hydration of Pyrophosphorio Acid. A Correction. G. A. ABBOTT (J. Amer. Chem. /~’oc. 1910 32 1576-1577).-1n a paper on this subject (Abstr. 1909 ii 661) it was stated that the specific conductivity of a mixture of pyro- and ortho-phosphoric acids is a linear function of its composition. It has been pointed out to the author that the experimental data quoted were not consistent with this statement and it has been found that the wrong data were inadvertently given. The correct data are now recorded. Correc- tions are also given for certain typographical errors which occurred in Abbott and Bray’s paper (Abstr.1909 ii 660) on the ionisation relations of ortho- and pyro-phosphoric acids and their sodium salts. E. G. New Determinations of Some Constants of the Inert Gases. CLIVE CUTHBERTSON (Phil. Mag. 1911 [vi] 21 69-’77).- Recent determinations of (I) the viscosity at difierent temperatures and (2) the refraction and dispersion of the argon gases afford independent sets of measurements of the fraction of the volume containing the gas which its molecules actually occupy. For argon krypton and xenon the values in absolute measure so found by the two methods are of the same order and there is a constant ratio between them; the volumes calculated from the viscosity being about 1.75 times that calculated from the refractivity.For helium and neon the ratios are 3.0 times and 2.53 respectively. For oxygen and nitrogen the ratios are similar to that found for argon but hydrogen departs widely from it. For all the gases the numbers of ‘ 0 dispersion elec- trons ” in the atoms calculated from the refractivities and dispersions by means of Sellmeier’s formula bear a linear relation to the reciprocals of the radii of the spheres of action of the atoms as calculated from the viscosity. The squares of the numbers of “ dispersion electrons ” are for all the gases proportional to the radii of the spheres of action diminished by a constant equal to 0.95 of the radius of the sphere of action of helium a t 0’. The squares of the relative numbers of ‘‘ dispersion electrons ” plotted against the critical temperatures of the gases fall on a straight line passing near the origin.From this relation the unknown critical temperature of neon is calculated t o be about 4 6 O A. The radii of the spheres of action also bear a linear relation to the critical temperatures. These results suggest that the electric charges which influence dispersion control also the critical temperature and the temperature-coefficient of viscosity and show thatINORGANIC CHEMISTRY. ii. 109 Sellmeier's formula which is in the manner of its derivation tentative can be safely employed to give results comparable with thoce obtained from the kinetic theory. F. S. A General Method for the Preparation of Anhydrous Chlorides. EDOUARD CHAUVENET (Compt. rend 19 11 152 87-89.Compare Abstr. 1909 ii 53).-The method consists in heating the oxide of the metal in a slow current of carbonyl chloride at a tempera- ture varying from about 350' in the case of vanadium oxide to 650' in the case of thoria. Chlorides of the following elements have been prepared in this way vanadium tungsten tantalum titanium zirconium thorium tin barium magnesium zinc glucinum aluminium iron chromium manganese nickel uranium cerium yttrium lanthanum. An excellent yield of the anhydrous higher chloride was obtained in each instance except with tungstic oxide when the oxychloiide was formed arid in the case of titanium when a mixture of chloride and oxychloride was produced. The method is specially suitable for preparing the chlorides of the rare-earth metals Silica is not attacked by carbonyl chloride. w.0. w. The Causes of the Differences in the Action of Sodium and Potassium on Water. MANINDRANATH BANERJEE (Ckem. News 1910 102 319-380).-When the metals are throwo on water the hydrogen which is evolved is charged with the vapours OF the metals and so the motion is affected by their densities and molecular volumes and those of their oxides and hydroxides. Thus in the cme of potassium the high density and large molecular volume of its vapour counteract the motion of the hydrogen its movement is retarded the heat is not dissipated and so the gas bursts into flame. I n the case of soditim the density and molecular volume being very low the hydrogen keeps in a state of motion and does not ignite the heat being dissipated.N. c. The Miscibility of Glaserite with Sodium Sulphate and its Dependence on the Temperature. RICHARD NACKEN (Xilzungsbe?*. X. Akad. Wiss. Berlin 1910 1016-1026. Compare van't HOE and Barschall Abstr. 1903 ii 434).-From a study of the solidification of fused mixtures of sodium and potassium sulphates it has been found that the hexagonal modifications of the enantiotropic dimorphous components which separate out first form a complete series of mixed crystals. As the temperature falls these primary mixed crystals undergo transformation and the diagrammatic representation of the resulting relationships shows that the formation of the hexagonal mixed crystals which crystallise from aqueous solutions at low temperatures is confined to certain concentrations of the components.The limiting concentration on the one side is represented by glaserite Na2S0,,3K2S04 (76%K,SO,). The other limiting concentration corre- sponds with about 49%K2S04 at 180° but with fall of temperature this proportion of potassium sulphate diminishes. Crystalline forms corresponding with these limiting concentrations are obtained together with potassium or sodium sulphate when fusedii. 110 ABSTRACTS OF CHEMICAL PAPERS. mixtures containing excess of potassium or sodium sulphate respectively are allowed to cool. The relationships indicated by the concentration-temperature diagram have been supplemented by crystallisation experiments a t 60' and 34'. When the aqueous solutions contain excess of potassium sulphate glaserite and potassium sulphate are obtained.If excess of sodium sulphate is present the crystals which separate consist of sodium sulphate together with mixed crystals containing glaserite and sodium sulphate and the composition of these mixed crystals approximates towards glaserite as the temperature of crystallisation is lowered. H. M. D. Fusions of Alkali Metaboratee and Metaphosphates. H. 8. VAN KLOOSTER (Zeitsch. amorg. Chern. 1910 69 122-1 34).-The freezing-point curve of mixtures of potassium metaborate and meta- phosphate possesses two eutectics at 681' and 770' respectively the former corresponding with 90% and the latter with 30% of potassium metaphosphate. The portions of the curve between the eutectics rises to a very flat maximum extending between 50 and 60% of potassium metaphosphate. The eutectic arrest is only noticeable in the neigh- bourhood of tbe eutectics so that it cannot be used to determine the position of the maximum. Investigation of the fusions showed that free borate could be detected by Tammann's reaction (characteristic red precipitate with mercuric chloride solution) up to 55% KPO so that the composition of the compound lies between 55 and 60% KPO and is probably 59% KPO corresponding with KPO,,KBO,.Fusions of this composition give neutral solutions whereas with higher and lower percentages of potassium metaphosphnte they are re$pectively acid and alkaline. The presence of the compound KPO,,KBO is also indicated by the microscopic examination of thin sections. Between 65 and 85% potassium metaphosphate the fusions mould not crystallise solidifying to a vitreous mass.Potassium metaphosphate bas m. p. 810° and potassium metaborate melts at 947O. The freezing-point curve of mixtures of sodium metaphosphate (m. p. 610') and sodium metaborate (m. p. 966') could only be followed between 0 and 30% and between 50 and 80% of sodium metaphosphate the other mixtures solidifying to vitreous masses. The curve between 50 and 80% of sodium metaphosphnte shows a flat maximum which is probably due to the existence of the compound NaPO,,NaBO further evidence in support of the existence of which is given by chemical and optical investigations similar to those described for the potassium compounds. Conductivity measurements showed that this compound also exists in solution to some extent.Sodium metaborate and potassium metaborate give a continuous series of mixed crystals the freezing-point curve showing a flat minimum at 50% sodium metaborate. Indications of a decomposition of these mixed crystals at 522-553' were obtained in mixtures containing 40-60% of sodium metaborate. Lithium metaborate (m. p. 843') and sodium metaborate do not form a compound with each other the freezing-point curve showing a eutecticINORGANIC CHEMISTRY. ii. 111 at 650' with 52% of lithium metaborate. The miscibility i i n the crystalline condition is very limited the lithium metaborate dissolving 2% of the sodium metaborate and the latter dissolving 3% of the lithium metaborale. T. S. P. Efflorescence of Washing Soda Crystals. ALEXANDER C. GUMMING (Chern. News 1910 102 311).-The author examined a specimen of large crystals of washing soda which had been for at least twenty years in a glass case with a wooden floor.The case fitted closely but was not air-tight. It was a t first thought that the crystals now consisted of the trihydrate the existence of which had not been previously known. Further analysis however showed that the crystals consisted almost entirely of pure sesquicarbonate Na,CO,!NaHCO 2H,O. The loss of weight on the ignition of the sesquicarbonate if calculated as due entirely to water would lead to the formula for the trihydrate. I T r( Il. w. The Binary Syatems Li,O-SiO Li,SiO,-ZnSiO ZnSi0,-CdSiO Li2Si0,-LiB02. Na,SiO,-NaBO and Na,SiO,-Na,WO,. H. S. VAN KLOOSTER (Zeztsch. anorg. Clhem. 1910 69 135-157).-The binary system Li,O-SiO forms two compounds namely lithium ortho- and meta-silicate with m.p.'s 1243' and 1188' respectively ; they are only slightly miscible in the crystalline state. Lithium metasilicate forms mixed crystals with silica up to 24.3% SiO,. The existence of a n acid silicate Li,Si,O, has not been confirmed. No compound is formed in the binary system Li,SiO,-ZnSiO,. Mixed crystais exist from 0 to 7('c)% and 71(?) to 100% of lithium meta- silicate. The eutectic temperature is approximately 990° and the eutectic composition 52(?)% of lithium metasilicate. That portion of the freezing-point curve lying between 10 and 70% Li,SiO could not be determined thermally and reliance had to be placed on the optical investigation. The system ZnSi0,-CdSiO gives an isomorphous series of mixed crystals the minimum point of the curve being at 25% ZnSiO,.The optical investigation did not completely verify the results obtained thermally. I n the system Li,SiO,LiBO mixed crystals exist from 0 to 24% and 91 t o 100% of lithium metasiliaate; no compound is formed. The eutectic temperature and composition are respectively 803' and 22% Li,SiO,. No compound is formed in the system Na,SiO,-NaBO but mixed crystals exist from 0 to 5% and 96 t o 100% of sodium metasilicate. The eutectic temperature and composition are respectively 81 5" and 55% Na,SiO,. Below 1100' sodium tungstate is practically immiscible with sodium metasilicate. The two components are quite immiscible in the crystalline condition. Sodium tungstate has m. p. 700' and transition temperatures at 589' and 572". Throughout this investigation optical methods were more trust- worthy than the thermal ones for determining the compositions of the saturated mixed crystals. Zinc metasilicate has m.p. 1419'. Cadmium metasilicate has m. p. 1155'. Sodium metasilicate has m. p. 1056'. T. 5. P.ii. I12 ABSTRACTS OF CHEMICAL PAPERS. Revision of the Atomic Weights of Silver and Iodine. 11. Ratio of Silver to Iodine. GREGORY P. BAXTER (J. Amer. Chem. Soc. 1910 33 1591-16C)2).-From determinations of the value of the ratio 2Ag I,O Baxter and Tilley (Abstr. 1909 ii 225) calcu- lated the atomic weights of iodine and silver by the aid of the value 0.849943 for the ratio A g I (Baxter Abstr. 1905 ii 81 579) and found them to be 126.891 and 107.850 respectively (0 = 16).Richards and Willard (Abstr. 1910 ii 292) however have obtained a value of 107.871 for the atomic weight of silver. As it was thought possible that this discrepancy might have been due to an error in the ratio Ag I this ratio has now been re-determined. Weighed quantit,ieu of iodine were reduced to hydriodic acid by means of a solution of hydrazine. The product was diluted and treated with a slight excess of a very dilute solution of silver nitrate. The clear supernatant liquid was carefully filtered and concentrated by evaporation and the excess of silver was estimated gravimetrically as silver iodide. Three samples of iodine and several specimens of silver were employed each of which had been carefully purified The results of thirteen experiments gave an average value for the ratio A g I 0,849906 and i t is therefore considered probable that the silver iodide obtained in the earlier determinations was contaminated with occluded impurities. On combining this ratio with that of ZAg:T,O the atomic weights of silver and iodine are found to be 107.864 and 126.913 respectively (0= 16).E. G. Revision of the Atomic Weight of Calcium. I. Analysis of Calcium Bromide. THEODORE W. RICHARDS and OTTO HONIG- SCHMID (J. Amer. Chern. Soc. 1910 32 1577-1590; Monatsh. 1910 31 1203-1226).-Determinations of the atomic weight of calcium by the analysis of the pure chloride (Richards Abstr. 1902 ii 394) gave a value of 40.126 (0 = 16 ; C1= 35*455) which agrees fairly well with that obtained by Hinrichsen (Abstr.1902 ii 137). In the present paper an account is given of a further study of this constant by the analysis of calcium bromide. The calcium bromide was prepared in the following manner. Calcium nitrate was carefully purified by repeated crystallisation and was converted into the carbonate by precipitation with ammonium carbonate. The carbonate was dissolved in hydrobromic acid prepared by the action of hot platinum on a mixture of bromine vapour and hydrogen and the solution was slightly acidified and afterwards concentrated. The bromide was repeatedly crystallised in quartz vessels and was dried with special precautions and fused in a platinum boat first in a current of hydrogen mixed with hydrogen bromide and afterwards in an atmosphere of nitrogen. The salt was then dissolved in water and when necessary the solution was carefully neutralised the deviations from exact neutrality being estimated by comparison with the pure crystallised salt with the aid of methyl-red.The analysis of the bromide was effected either by determining the amount of silver equivalent to the calcium present or by weighing the precipitated silver bromide. From the results of six experiments in each way values for the two ratios CaBr 2Ag and CaBr 2AgBrIN 0 RCI A N I C CH E 1cI I ST R Y. ii. 113 were obtained which gave essentially the same value for the atomic weight 40.070 (Ag= 107.88) or 40.066 (Ag= 107.87). Two different specimens of the salt gave almost identical results. The density of fused calcium bromide was found to be 3.353 at 25”.E. G. Electro-deposition of Lead from Perchlorate Solutions. FRANK C . MATHERS (Chem. Zeit. 1910 34 1316-1318 1350-1351 ; Trans. Amer. Elektrochem. Xoc. 1910 17 261-272).-Experiments with the lead perchlorate plating and refining bath are described. The properties of lead perchlorate which are of special value in plating or refining solutions are (I) Great solubility. (2) Cathode deposits which are smooth dense and free from “treee.” (3) Ap- proximately theoretical corrosion of the anode and deposition upon the cathode. (4) Absolute stability under all conditions to which it is subjected in a plating or refining batb. (5). No polarisstion from the formation of lead peroxide on the anode. (6) Very high electrical conductivity. The bath should contain about 5% of lead 2-5% of free perchloric acid and 0.25% of peptone.A current density of from 2-3 amperes per sq. dcm. (18-2‘7 amperes per sq. ft.) may bo used. The peptone is gradually used up and after about four dLiys a quantity equal to the original amount should be added. The free acid which is very slowly neutralised by the chemical solution of the lead must be restored by treatment of a suitable portion of the solution with the right amount of sulphuric acid thus precipitating lead sulphate and leaving perchloric acid in solution. The bath gives excellent purification the cathode being about 99*9S% pure and shows no deterioration with use giving as good deposits after two months as at the beginning if the concentration acidity and the required amount of peptone are maintained. Chlorides and barium salts must be absent.A bath that has been giving good deposits will form very bad “trees” if a quantity of hydrochloric acid or some barium perchlorate is added to it. The filtrate is returned t o the bath. T. S. P. Red Lead. IV. JAROSLAV MILBBUER (Chem. Zeit. 1910 34 1341-1342. Compare Abstr. 1910 ii 294).-The oxidation of litharge to red lead in air a t 460’ follows a course similar t o t h a t already observed in the case of lead. Litharge obtained in the manufacture of nitrites is much better for this purpose than ordinary litharge; at 500” the former gave t h e same percentage of red lead after one hour as the latter after fifteen hours at 460’. Thus although 460’ is the optimum temperature for the formation of red lead the increased velocity attained at 500° gives better results.The rate at which red lead is formed is conditioned more by the origin or by the kind of lead oxide used than by the size of the particles. The percentage of red lead obtained depends on the partial pressure of the oxygen in the gas used but even with pure oxygen a t 450° it has hitherto been impossible to obtain 100% red lead. T. S. P VOL. c. ii. 8ii. 114 ABSTRACTS OF CHEMlCAL PAPERS. Studies in Tapour Pressure. VI. Quantitative Study of the Constitution of Ualomel Vapour. ALEXANDER SMITH and ALAN w. c MENZrEs (proc. Boy. 8oc. Edin. 1910 31 183-185; Amer Chem. Soc. 1910 32 1541-1555).-A review is given of previous work on the constitution of mercurous chloride vapour and it is shown that there are no experimental data in existeuce from which the proportion of dissociated (Hg i- HgC1,) t o non-dissociated molecules (HgCI) in the vapour can be deduced.An investigation has now been made based on the principle of partial vapour pressures. Determinations have been made of the vaporir pressures of mercury mercurous chloride and a mixtiire of theie substances between 360' and 400' by means of the static isoteniecope (Abstr. 1910 ii 1036 1037). The results show t h a t mercurous chloride vapour even when saturated is completely dissociated into Hg and HgCl and that molecules of the formula HgCl or Hg,CI are not present. The b. p. of mercurous chloride is 382-5' and its molecular weight when dissolved in mercury corresponds with the formula HgCl. E. G. Double Nitrates of the Rare Earths.I. Double Nitrates of the Rare Earths with the Alkali Metals. GUSTAV JANTSCH and 8. WIGDOROW (Zeitsch. unorg. Chem. 1911 69 221-231).-To prepare the double nitrates of the rare earths with the alkali metals t h s procedure generally adopted was to dissolve the oxide of the rare earth together with the necessary quantity of the nibrate of the alkali metal in concentrated nitric acid and evaporate until crystals formed. Wyrouboff's statements (Abstr. 19OS ii 385) as to the hydration of the crystals of the double nitrates of sodium potassium and cesium with lanthanum and cerium are not confirmed. The formulz of the various compounds are written so as to indicate that the rare earths are tervalent in their stable forms of combination.Lanthanum sodium nitrate [La(N O,),]Na,,H,O forms blender white needles ; D = 2.63 and molecular volume = 195.08 ; it is not completely dehydrated at 150". has D:= 2.54 and molecular volume = 281.76 ; hygrcscopic white shining crystals which lose 2H,O at 60'. Acid Eunthanum rubidium mnitrute [La(N03),]Rb,HN0,,6H20 obtained when lanthanum and rubidium nitrates are taken in the molecular proportion of 1 2 ; forms colourless plates which lose 5H20 and 1HN03 a t 120'; the resulting [La(N0,)4]Rb,H,0 is not dehydrated at 200" ; D = 2.377 and molecular volume = 270.6. When lanthanum and rubidium nitrates are taken in the molecular proportion of 1 4 lanthanum rubidium nitrate results j monoclinic crystals m. p. 8 6 O DZ = 2.497 and molecular volume = 277.1 ; it loses 4H,O on prolonged heating at 60".Lanthanum csesium nitrate [La(N0,),]Cs2,2H,0 forms small tabular crystals D = 2.827 and molecular volume = 265.5. Lunthanum thalloua nitrate [La(NO,),]TI2,4H,O forms hygroscopic crystals m. p. 72" ; Dt = 3.318 2nd molecular volume = 280.0 ; it loses 4H,O a t 100'. Cerous sodium Lanthanum potassium nitrate [La(NO,),I K,J H,O [La(N@,),]Rb 4H,O,INORGANIC CHEMlSTRY. ii. 115 nitrate [Ce(N0,)5]Na2,H,0 consists of hygroscopic slender needles which are not corupletely dehydrated at 150" ; D = 2.65 and molecular volume = 194.0. Cerous rubidium nitrate [Ce(N0,),]Rb2,4H20 has D = 2.497 and molecular volume = 277.6 ; hygroscopic monoclinic needles m. p. 70° which lose 4H,O at 60'. Cerous thaltous nitrate [Ce(NO3),]T1,,4H,O forms hygroscopic crystals m.p. 64s53; Di = 3.326 and molecular volume = 279-7 ; it loses 4H20 a t 60°. Praseodymium rubidium nitrate [Pr(N0,)5]Rb2,4H,0 green hygroscopic monoclinic crystals m. p. 69.5' ; D = 2 5 U and molecular volume = 277.4 ; it loses 4H20 a t 60°. Neodymium rubidium nitrate [Nd(N0,),]Rb,,4H20 consists of hygroscopic bright reddish-violet plates m. p. 47' ; D! = 2.56 and molecular volume = 272.3 ; it loses 4H,O a t 60". The temperature a t which the above compounds melt in their water of crystallisation falls with increasing atomic weight of the rare-earth metal. T. 8. P. A New Element Accompanying Lutecium and Scandium in Gadolinite Earths Celtium. GEORGES URBAIN (Compt. rend. 1911 152 141-143. Compare Abstr. 1907 ii 956; 1908 ii 283; 1909 ii 735).-During repeated fractionation of the nitrates in the isolation of lutecium from gadolinite earths a few drops of a mother liquor were obtained which did not crystallise.This contained a new oxide belonging to the rare earths and characterised by a magnetic susceptibility three or four times less than that of lutecia. The name celtium is given to the corresponding element and the symbol Ct assigned to it. Spectroscopic examination of the oxide showed the presence of luteciurn scandium a trace of neoytterbium and negligible traces of calcium and magnesium. The new element shows a large number of lines in the arc ; the following are very intense A = 2685.2 2765.8 3080.7 3118.6 3197.9. The chloride is somewhat more volatile than t h a t of lutecium but less volatile than scandium chloride.The hydroxide is less basic than lutecium oxide and more basic than scandium oxide Celtium either appears t o be entirely absent from xenotime or else it occurs in very faint traces. w. 0. w. Electrical Properties of Aluminium-magnesium Alloys. WITOLD BRONIEWSKI (Compt. rend. 191 1 152 85-S7. Compare Abstr. 1910 ii 128; Grube Abstr. 1905 ii 523).-From an examination of aluminium-magnesiam alloys by the electrical method already described the author comes to the conclusion that two definite compounds AlMg and Al,Mg probably exist. These form a con- tinuous series of solid solutions with one another preventing their recognition by the thermal method. The existence of the compounds A1,Mg and AlMg could not be confirmed and alloys of the metals in these proportions showed a heterogeneous structure under the microscope.w. 0. w 8-2ii. 116 ABSTRACTS OF CHEMICAL PAPERS. Formulm of Aluminium Salts. GERRIT H. COOPS (Chem. Weekblud 1910 '7 1071-10'76. Compare Coops Abstr. 1910 ii 506 ; and Olivier ibid. 507).-Polemical. A reply to Olivier. A. J. W. Colloidal Solubility of Metals in Distilled Water in Presence of Air and in a Vacuum. MARGHERITA TBAUBE-MENGARINI and ALBERTO SCALA ( A t t i R. Accad. Lincei 1910 [v] 19 ii 505-508. Compare Abstr. 1909 ii 809).-Distilled water acts on aluminium in the warm and in the presence of air and zinc and iron are attacked in the cold colloidal solutions being formed in each case. Lead and iron which alone were experimented with! yield colloidal solutions when treated with distilled.water in a vacuum. The clear solutions become turbid in air that of 'iron turning a greenish and finally a reddish colour whilst the red solution becomes milky. The colloidal iron solution when kept i n a vacuum forms black green and red deposits. Of these the red and black ones are permanent in air but the green deposit becomes red even in a vacuum. The bacteriform colloidal corpuscles of all these metals change (without passing into true solution) into the leaf-like crystals characteristic of colloidal solutions. It. v. 8. Solid Solutions of Iron and Manganese Borides. JOSEF HOFFMANN (Chem. Zeit. 1910 34 1349-1350).-The heterogeneous nature of the borides prepared by the thermite process has already been shown by optical methods (Abstr. 1910 ii 5081 and is confirmed by the chemical behnviour of the various products The composition of the saturated solutions obtained is 7 atoms of iron to 9 of boron for the iron boride and 10 atoms of manganese to 28 atoms of boron for the manganese boride. Mineral acids extract compounds such as Fe,B FeSB4 and FeB from the iron boride leaving a residue consisting mainly of Fe with a little FeB,.The soluble portions of manganese boride consist chiefly of MnB toget,her with some MnB the undissolved residue being elementary boron mixed with some higher borides. T. S. P. Iso- and Hetero-poly-acids. I. Metatungstic Acid. ARTHUR ROSENEIEIM and FRANZ KOHN (Zeitsch. anorg. Chenz. 1911 69 247-260. Compare this vol. i log).-The authors distinguish between '' isopoly-acids " and '' hetero-poly-acids." The former are compounds containing the acid anhydride and the acid hydrate of one and the same element for example the polychromates polytungstates etc. whilst the latter are compounds in which acid anhydrides of one or more elements are combined with a hydrate or salt of the acid of another element for example the phosphomolybdates etc.which have hitherto been called complex acids. According to Copaux's views of the constitution of the meta- tungstates boro- and silico-tungstates (Abstr. 1909 ii 318) the metatungstates must contain water of constitution so that they should belong to the hetero- and not to the iso-poly-acids. It order to find out how much water of constitution is contained in the meta-INORGANIC CHEMISTRY. ii. I17 tungstates the authors have prepared by double decomposition various insoluble salts since these generally do not contain water of crystal- lisation.XiZuey rnetatungstate Ag2W,0,,,3H20 forms small white crystals which lose l*lH,O a t 16U0 and 1.3H20 at 200’; 2 mole- cules of water and probably 3 are therefore firmly combined. I n t?haZZium rnetatungstate T1,W,01,,3H20 only 1 molecule of water is firmly combined. Guanidine metatungstate (CHSN,)2H,W40,3,3H,0 was obtained from guanidine carbonate and metatungstic acid as a white microcrystalline powder. Between 90’ and 150° i t loses 2H20 so that probably only 1 molecule of wat,er is firmly combined. The normal lead and mercury metatungstates although frequently mentioned in the literature could not be obtained. A solution of an alkali metatungstate gives with lead nitrate a precipitate of the double salt PbW40,3,Pb(N0,),,10H20 which loses SH20 at l l O o so that 3 molecules of water are firmly combined. From the above results combined with Friedheim’s statement (In- aug.Dissert.) that the metatungstntes of sodium barium manganese and cadmium still cont,ain 3H20 a t looo and especially since insol- uble salts are generally anhydrous the authors draw the conclusion that the metatungstates contain 3 molecules of water of constitution that is they are aquo-salts and must probably be formulated as n,[wo[~,3jq although they may be R2[WO(wo4)3] When heated at such a temperature that water of constitution is lost they are decomposed. Metatungstic acid was prepared by treating a concentrated aqueous solution of ammonium metatungstate with ether and concentrated hydrochloric (or sulphuric) acid.Of the three layers formed the lower yellow one contains the free acid together with ether and the mineral acid from which the free metatungstic acid is obtained by evaporation in a current of air. It readily euoresces ; crystals were obtained corresponding with H,W4O,,,8K2O and H2W40,,,6H20. It is quite insoluble in ether (compare Abstr. 1896 ii 477) although in the presence of a mineral acid it is possible that a molecular compound of metatungstic acid and ether is formed. I n aqueous solution it behaves as a normal electrolyte and conductivity measure- ments after the addition of varying quantities of sodium hydroxide show it to be a dibasic acid.I n absolute alcoholic Eolution it acts as a colloid. P2O) - Esters of metatungstic acid could not be obtained. The above results cannot be brought into accordance with Copaus’ formulation of metatungstic acid as H,,W,,07,,3H20 + aq. T. S. P. Atomic Weight of Vanadium. D. J. M ~ A D A M . ~ ~ ~ . (J. Amel.. Chem. Xoc. 19 10 32 1603-1 615).-The values previously obtained for the atomic weight of vanadium show considerable discrepancy and a re-determination has therefore been made by a new method. This method is based on the observation of Smith and Hibbs (Abstr. 1894 ii 455) that vanadium can be completely removed from sodium metavanadate by volatilisation in a current of dry hydrogen chloride,ii. 118 ABSTRACTS OF CHEMICAL PAPERS Five samples of sodium metavanadate were used in the experiments The apparatus employed is described with the aid of a diagram.A weighed quantity of the anhydrous salt was placed in a weighed quartz flask and was heated in 8 current of hydrogen chloride containing a little chlorine. When as much as possible of the vanadium had been removed a little water was introduced into the flask and the mixture was again heated in the current of hydrogen chloride. The whole of the vanadium was thus expelled and w residue of sodium chloride obtained. The flask and residue were weighed and the weight of the sodium chloride obtained by subtracting that of the flask. The results of five experiments gave an average value for the atomic weight of vanadium 501967 +_ 0.006 (Na = 23-00 ; C1= 35.46). This value agrees with that obtained by Prandtl and Bleyer (Abstr.19 10 ii 135) by the analysis of vanadium oxychloride. Anhydrous sodium metavanadate has D 2.79. E. G. Bismuth. LUDWIU VANINO and EMILIE ZUMBUSCH (Arch. Pharm. 1910 248 665-669).-Trials with the various methods described for the preparation of bismuth hydroxide showed that it was difficult to prepare a product free from nitrate. Good results were obtained with Thibault’s process (Abstr. 1901 ii 106) but only when a very large excess of potassium hydroxide was employed. A satisfactory preparation was obtained eventually by dissolving bismuth nitrate (20 grams) in water (100 c.c.) containing mannitol (7.5 grams) adding 50 C.C. of ice-cold potassium hydroxide solution (22 grams in 100 C.C. of water) and finally dilute sulphuric acid until the mixture was only slightly alkaline (compare Abstr.1902 i 8). Previous work by Vanino and Treubert (Abstr. 1898 ii 435 598 ; 1899 ii 428; compare Herz and Guttmann Abstr. 1907 ii 274) has shown that bismuth suboxide probably does not exist but the authors have made experiments with the process described by Jaworososki for the preparation of this substance (Pharm. Zeit. BUSS. 1896). This method consists in warming a mixture of ferrous sul phate sodium potassium tartrate and sodium hydroxide in water with basic bismuth nitrate. The brownish-black precipitate so obtained in the author’s experience was never free from iron even when the reacting ingredients were used in calculated proportions for the production OF the suboxide so that they do not regard Jaworososki’s preparation as a definite substance. T. A. H. Brown Gold. MAURICE HANRIOT (Compt. rend. 1910 151 1355-1357).-This name is given to the residue obtained when nitric acid is allowed t o act on an alloy of gold and silver containing about 20% of gold It always contains a small quantity of silver and a considerable amount of nitric acid. The latter is lost a t 175-200°; on further heating it changes colour and undergoes contraction ; at 900’ the substance evolves gas and at 1040O it melts changing into red gold. The author has measured the contraction undergone by strips of alloy containing 1-3.5% of silver on treatment with nitric acid andMINERALOGICAL CHEMISTRY. ii. 119 also the further contraction t h a t ensues on heating. Results are also quoted showing the further contraction t h a t occurs on a second and third heating. ViT. 0. w.