ii. 120 ABSTRACTS OF CHEMICAL PAPERS. Inorganic Chemistry. Preservation of Hydrogen Peroxide by means of Acet- aoilide. A. M. CLOVEH (Bmev. J. Phavrn,. 19 13,85,538-545) -The investigation has been undertaken with the object of determining the justification for using acetanilide as a preservative for com-INORGANIC CHEMISTRY. ii. 121 mercial hydrogen peroxide solutions since it has been claimed that the usual instability of the latter is due to the presence of certain impurities introduced in the course of manufacture and that if these impurities could be eliminated a stable product would result. The gradual decomposition of hydrogen peroxide solution in the presence of hydrochloric sulphuric phosphoric or boric acid a t different concentrations has been studied whilst the effect of various salts (sodium chloride potassium chloride calcium chloride sodium silicate ferric chloride and aluminium sulphate) alone and with acetanilide on solutions of hydrogen peroxide which have been made N / 100 in hydrochloric sulphuric and phosphoric acids respectively has been investigated. The author is led to the conclusion that pure hydrogen peroxide is a very unstable substance and that its stability is greatly in- creased by the addition of small amounts of acid.Addition of salts of the alkali and alkaline-earth metals does not appear to have any marked effect. Of the acids used phosphoric acid gave the best results a t all concentrations but the best preserved solution had lost nearly 30% in strength after seven months. I n the presence of both acids and salts the decomposition in those solutions containing acetanilide is only a small fraction of that in the corresponding solutions which do not contain the preservative.As to the mineral impurities the salts of the alkali and alkaline-earth metals and all other salts used except those of copper and iron appear to have no influence whatever on the stability of the solutions when acetanilide is used. Traces of copper and iron have a very deteriorating effect but this is prevented to a great extent by acetanilide. H. w. Ignition of a Glowing Splint of Wood in Oxygen Mix- tures. P ANEMA (Chem. Weekblud 1913 10 1056).-A mixtnre of air and oxygen containing 30% of the latter does not ignite a glowing splint 3f wood. Ignition of Glowing Match-Sticks and the Extinction of Their Flame in Mixtures of Oxygen and Nitrogen.W. P. JORISSEN (Cham. Weekbhd 1913 10 1057. Compare preceding abstract).-The ignition limit is 28-29% of oxygen by volume and the extinction limit about 1676 of oxygen by volume. A. J. W. A. J. W. A Third Form of Sulphur. 111. A. H. W. ATEN (Zeitsch. physihl. Chum. 1913 86 1-35. Compare A .. 191 3 ii 580).-The third form of sulphur S has been further investigated. A method is devised by means of which the quantity of S may be estimated in a mixture of SA S and S,. The transformation of S into S is investigated and it is shown that the change S -+ S is at first very rapid but the velocity quickly decreases and becomes extremely small when there is only a small quantity of S present. The relative quantities of Sh S and S present in sulphur which has been heated to various temperatures are determined. The amount of .S is a t a maximum in sulphur which has been heabd to 180° and a t this point the amount present is 6.5% The quantityii.1%2 ABSTRACTS OF CHEMICAL PAPERS. of S increases as the temperature increases up to 445O the greatest rate of increase being between 170° and 180O. SA decreases as the temperature increases; a t 180° the composition is given by Sr = 6*5% S = 20*4% and S h = 73.1%. The influence of the cata- lysts sulphur dioxide ammonia and iodine on the equilibria S t S Z Sh is studied and i t is shown that ammonia increases the transformation of S more than it does that of S,. The change S,-+S takes place a t the same rate in the presence of iodine as i t does in the absence of iodine.The hard and soft varieties of S are considered together with S and evidence advanced t o show that S is in all probability contained in the soft variety of S which is really a mixture of S Sh and S containing very little SA. A number of properties of sulphur for example coefficient of ex- pansion and viscosity are considered and it is shown that the changes in these properties with temperature are more easily ex- plained when three modifications of sulphur are admitted than when two only are considered. A table of the composition of various forms of solid sulphur is given; in most cases the quantity of S is extremely small but S is found in sulphur which has been melted but not in that which has been crystallised.J. F. S. The Change of Sulphites into Sulphates. EMILE SAILLARD (Z+?itsch. Vev. deut. Zuckerind. 191 3 1035-1 043. Compare Titnff A . 1904 ii 113; Lange A. 1912 ii 550).-The change of sulphites into sulphates is retarded in sucrose solutions the amount of retard- ation being the greater the greater the concentration of the sucrose. Rise in temperature increases the velocity of change. Nitr- ogenous substances for example asparagine aspartic acid glutamic acid and potassium lactate also exert a retarding action whilst sodium chloride has no effect. T. S. P. Invert sugar has approximately the same effect as sucrose. Active Nitrogen. ERICH TIEDR and EMIL DOMCHE (Ber. 1913 46 4095-4103).-1n a previous communication (A. 1913 ii 210) Tiede has stated that when pure nitrogen is used the active modi- fication of Strutt cannot be obtained; this has been disputed by Koenig and Elod (A.1913 ii 316) and by Strutt (A. 1913 ii 316 696). The authors now give an account of experiments in support of the statement of Tiede. I n the preliminary experiments the apparatus was made com- pletely of glass no rubber joints being used; the various connexions were either fused together or else cemented with marine glue. Taps were made tight. with mercury in order to prevent the use of any fat and mercury vapour was prevented from diffusing into the apparatus by the interposition of vessels filled with gold leaf and cooled in liquid air. The nitrogen used was freed from oxygen by a combination of the copper method with that of Gehlhoff (A.1911 ii 487) and moisture and carbon dioxide were removed by appropriate absorbents. All parts of the apparatus were freed from adsorbed gases by heating during exhaustion. Under theseI PIT ( ) llGANIC C H EM ISTRY. ii. 123 conditioiis when it was certain that all oxygen had been removed the phenomena described by Strutt could not be observed. The final experiments were made with a smaller apparatus com- pletely of glass which contained fused on to it two glass tubes containing respectively barium azoimide and silver oxide from the former of which pure nitrogen is readily obtained by heating at 170°. After all parts of the apparatus had been freed from gas by repeated heating and exhausting the barium azoimide being main- tained a t 120° it was filled with pure nitrogen from the azoimide.There were then no signs of Strutt’s phenomenon but if the silver oxide were heated in order to add oxygen to the contents of the apparatus the characteristic glow was observed. The apparatus could then be cleaned from oxygen filled with pure nitrogen and again there would be no sign of Strutt’s phenomenon. The authors therefore conclude that the phenomena observed with so-called “active nitrogeu” are due to traces of oxygen in the nitrogen. T. S. P. Oxidation of Nitrogen during Electric Discharge. FRANZ FISCHER (Bey. 1913 46 4103).-The arithor maintains his opinion that the oxidation of nitrogen during electric discharges is pre- ceded by an activation of the oxygen (compare Fischer and Hene A. 1913 ii 132 317). The views of Koenig and Elod (A.1913 ii 1049) are invalidated by the recent communication of Tiede and Domcke (compare preceding abstract). T. S. Y. Combustion of Air in the Voltaic Arc. A. GOHBOV and V. The authors criticise the manner in which the results of Saposhni- kov Gudima and Kutovski (A. 1913 ii 950) are expressed the hourly volumes of air passing through the furnace being referred to varying numbers of kilowatts whereas the important point is not the absolute volume of air traversing the furnace per unit of time but a magnitude indicating the amount of energy applied to such volume of air. Calculation from the results given by the above authors shows good agreement with the formulae elaborated by the present authors (A. 1913 ii 950). MITKEVITSCH (tJ. KUS8. P/L:oS. Cht?m..L ~ O C . . 1913 45 16!n-1697).- T. H. P. Reduction of Hydronitric Acid [Azoimide]. 11. Structures of the Trinitride Radicle. J. W. TURRENTINE (J. Amsr. Chenr. Soc. 1914 36 23-35. Compare A. 1912 ii 448).-Further evidence is adduced in favour of the formula. H*N:NiN for azoimide (compare Turrentine and Moore A. 1912 ii 449). It is shown that Fischer’s interpretation of the reaction yielding diazobenzeneimide (A. 1878 305) on which is based the conception of the cyclic structure of the trinitride radicle is erroneous. It is pointed out that this reaction is essentially an oxidation of hydrazine and that Browne and Shetterly (A. 1907 ii 863; 1908 ii 373; 1909 ii 233 658) in their study of the oxidation of hydrazine have failed to observe any reaction analogous to that proposed by Fischer.E. G.ii. 124 ' ABSTRAC'I'S OF CHEMICAL PAPERS. The Intermediate Formation of Nitrogen Trioxide by the Action of Oxygen on Nitric Oxide. Behaviour of Nitrogen Trioxide towards Potassium Hydroxide. GABRIEL KLINGER (Zeitsch. angew. Chem. 1914 2 7 7-8).-Both RA schig and Lunge have shown that a mixture of nitric oxide and nitrogen dioxide dissolves in concentrated sulphuric acid quantitatively as if it were the compound N,O whereas sodium hydroxide dissolves only about 85% of the mixture relatively more of the dioxide dissolving than of nitric oxide. Raschig (A. 1905 ii 700) gives an explanation of this which postulates the existence of the compound N,O in the gaseous mixture whereas Lunge (A. 1906 ii 438) gives a different explanation and denies the existence of nitrogen trioxide.The author points out that both explanations are unsatisfactory and that it is probable that water is the disturbing factor when sodium hydroxide is used. Nitrogen trioxide would give nitrous acid with even traces of water and this would be further oxidised to nitric acid by the excess of nitrogen trioxide with evolution of nitric oxide and thus cause the observed discrepancies. I n support of this theory it is found that dry nitrogen trioxide is quantitatively absorbed by dry potassium hydroxide (compare A. 1913 ii 619). The question as to whether nitrogen trioxide is formed as an intermediate product when a mixture of nitric oxide and nitrogen dioxide is acted on by dry potassium hydroxide can be decided by measuring the ratio of the contraction which takes place to the volume of nitric oxide taken when a known mixture of nitric oxide with excess of oxygen is treated with the dry hydroxide.Experi- ments which were carried out agreed quantitatively with the inter- mediate formation of nitrogen trioxide which was then absorbed by the potassium hydroxide giving potassium nitrite. The Reaction of Metals and Alloys with Nitric Acid. J. H. STANSBIE (J. Soc. Chem. Ind. 1913 32 1135-1136).-Nitric acid dissolves copper silver mercury and bismuth much more rapidly when the metals remain a t rest in the acid than when they are rapidly rotated or the acid is stirred owing to the fact that in the former case nitrous acid accumulates in the neighbourhood of the metal (compare A. 1913 ii 982).Similar results are obtained with alloys of copper and zinc as long as they contain more than 48% of copper the alloy dis- solving as a whole. Below this percentage of copper the zinc either dissolves faster in proportion than the copper or precipitates that metal from solution with the result that more copper dissolves in the stirred solutions than in the solutions a t . rest. A neutral solution of copper nitrate has no corrosive action on alloys con- taining more than 80% of copper. T. s. P. T. S. P. The Action of Garbonyl Chloride on Phosphates and Oxides. J. RIBAN (Compt. . r e d . 1913 157 1432-1433. Compare A. 1883 287).-A claim for priority over Barlot and Chauvenet (compare this vol. ii 49). W. G.INORGANIC CHEMISTRY. ii. 125 Preparation of Solid Alkali Perborates from Boric Acid and Alkali Peroxides without the Use of Water 80 Solvent.CHEMISCHE FABRIK REISHOLZ (D.R.-P. 262144) -Boric acid is mixcltl with the amount of ice necessary for hydration and the alkali peroxide is added. Examples are given of the preparation of sodium perborates of the formulae Na,B,0,,2NaB03,10H20 NaB03,4H20 and Na,B,08,10H20. Experimental Demonstration of the Variability of the Molecule and the Atom. P. DE HEEN (Bull. Acud. ~ o y . ndg.. 1913 680-694).-Observations are described which seem to show that the sensitiveness of silver chloride to light can be varied by the action of reagents although the physical condition of the chloride remains constant. If the silver salt is triturated with a very concentrated solution of potassium hydroxide washed with water and boiled with nitric acid it is found that the residuai silver chloride is very much more sensitive to light than the original silver salt.A similar effect is obtained if the silver chloride is spread in a thin layer over the surface of a platinum cathode and subjected to the action of a current f o r several days. I f on the other hand the silver salt is subjected to the action of the current a t the surface of the anode its photo-sensitiveness is found to diminish. I n this case the action is less rapid and the current must be passed for a t least ten days. I n another series of experiments a quantity of silver chloride was divided into three portions one of which was subjected t o the action of the current a t the cathode the second portion to the anodic action of the current whilst the third was unacted on.The three portions of silver chloride were then reduced to metal and the three samples of metallic silver re-converted into chloride by dissolving in nitric acid and precipitating with hydrochloric acid. The silver chloride obtained from the first sample of silver was found t o be much less sensitive and that from the second sample much more sensitive to light than the chloride prepared from the third sample of silver. The above facts are supposed to show that the silver chloride molecule can be modified by suitable treatment and that the dif- ferent forms of silver salt are to be regarded as derived from silver atoms which are not identical. I n other words the experiments afford evidence of transmutation of normal silver into its meta- elements.H. M. D. New Compounds of Nitrogen and Hydrogen with the Alkaline-earth Metals. F. W. DAFERT and R. MIKLAUZ (Monatdh. 1913 34 1685-1712).-The calcium used by the authors in their experiments was obtained pure by distilling the commercial article in an apparatus similar to that used by Guntz in the preparation of strontium (A. 1910 ii 1064); the strontium and barium were obtained by distilling a mixture of the respective oxides with the equivalent quantity of aluminium powder (compare Guntz A. 1906 ii 669; 1910 ii 1064). J. C. C.ii. 126 AESTRACTS O F CHEMlCAL PAPERS. The pure nitrides and hydrides of calcium strontium and barium are readily obtained by heating the respective metals in the pure gases.When the nitrides are heated in a current of hydrogen compounds having the formula= M//,N,H are formed (compare A. 1909 ii 8821 but only the calcium and strontium compounds could be obtained pure since the barium compound even a t relatively low temperature reacts with hydrogen in accordance with the equation Ba,N,H + H = 3BaH + N,. When hydrogen is passed over heated barium nitride or more correctly over the impure compound Ba,N,H ammonia is formed. The barium hydride which is thereby produced is readily trans- formed back to the nitride by the action of nitrogen so that a process is given for the fixation of atmospheric nitrogen. When a mixture of equal volumes of hydrogen and nitrogen is passed over the heated alkaline-earth metals or over their hydrides or nitrides imides M'INH are produced which similarly t o lithium imide (A.1912 ii 253) darken on exposure to the light. Calciumlimide is most easily prepared but it has not been obtained pure; the formation of barium iinide is incomplete. The following table giving the temperatures a t which reaction occurs with the various gases shows that the tendency of the alkaline-earth metals to combine with nitrogen and hydrogen increases with the atomic weight? whilst the tendency of the nitrides to combine with hydrogen decreases with increase in atomic weight of the metal. Ca ......... 410" 300" Cn,N .......... 230" Ba ........ 260 170 Ra,N,.. . . . . . . . 300 N. H. H. Sr ........ 380 21 5 Sr,N ........... 270 T. S. P. Formation of Magnesium Barium and Strontium Com- pounds Analagous t o Apatite and Wagnerite.HANS WINTER (Diss. Leipzig 1913 1-46),-The phosphates and halogen salts were fused together in an electric oven and the equilibrium diagrams are given for several pairs. Barium chloride (m. p. 958O D 3.789) is optically biaxial and positive whilst barium fluoride (m. p. 1289O) strontium chloride (m. p. 874O D 3*054) and strontium fluoride (m. p. 1400O) are all cubic in crystallisation. These do not form mixed crystals but the double salts BaCl,,BaF (m. p. 1008O D 5.931) and SrCl,,SrF2 (m. p. 962O D 4.616) both of which are tetragonal and optically negative. Magnesium fluoride and phosphate yield wagnerite MgF2,Mg3P,08 and only doubt- fully a compound corresponding with apatite (magnesium-fluor- apatite). The barium and strontium compounds corresponding with wagnerite were not obtained but the four apatites of the composition BaC1,,3Ba,P,08 BaF2,3Ba3P208 SrCI2,3Sr3P,O and SrF2,3Sr3P,0 ; these were obtained in a crystallised condition and their melting points density and refractive indices determined.L. J. S.IN( 1RGANIC CHEMISTRY. ii. 127 Allotropy of Zinc. ERNST COHEN and W. D. HELDERMAN (Proc. K Akad. Wetensch. Amsterdam 191 3 16 565-568).-Experi- ments are described which show that zinc which has been obtained by rapid cooling of the liquid metal undergoes slow changes in respect of its density. The molten metal was poured into a cylinder of asbestos paper cooled by a mixture of solid carbon dioxide and alcohol. The density of the solid was then found to be 7.130 a t 25O.After heating for a fortnight a t looo in a solution of zinc sulphate the density was found tyo be only 7.102. These observations show that ordinary zinc is in a metastable condition and that the modification which is formed a t high temperatures changes only very slowly into that which is stable at the ordinary temperature. The stable form was probably obtained in an almost pure condition by Kahlbaum Roth and Siedler (A. 1902 ii 259). H. M. D. Atomic Weight of Oadmium by the Investigation of Cadmium Chloride and Cadmium Bromide. ELTON L. QUINN aod GEORGE A. HULETT (J. Physical Chew. 1913 17 780-798).- Weighed quantities of the carefully purified chloride or bromide were converted into sulphate by evaporating solutions of the salts to dryness after the addition of the calculated quantity of sulph- uric acid.The cadmium sulphate was then dissolved and subjected to electrolysis in an amalgamated platinum crucible which served as the cathode. As shown by previous experiments (Perdue and Hulett A. 1911 ii 433) this method can be applied very con- veniently i n the case of metals like cadmium which are readily . soluble in mercury and sufficient evidence has already been obtained that the electrolytic method affords very exact results in the estima- tion of cadmium (compare also A. 1911 ii 397). The results obtained with the chloride give for the atomic weight of cadmium 112.32 +0.01 whilst the bromide results lead to the value 112.26 +0*005. The mean of these is 112.29 which agrees very well with the value 112.30 obtained by Perdue and Hulett (Zoc.cit.) from the analysis of cadmium sulphate and also with the results obtained by Laird and Hulett in their work on the cadmium coulometer (Trans. Amer. Electrochem. SOC. 1912 22 385) which lead t o the value 112.31. The authors consider that these observations show that the atomic weight of cadmium is nearer 112.3 than the value of 112.4 which is accepted as the most probable value according to previous measurements. H. M. D. Cadmium. MANUEL VERES (Compt. rend. 1914. 158 39-40)- Using the methods of Lepierre and Lachaud (compare A. 1892 943 1282) and Klobb (compare A. 1892 941 1399) the author has prepared a new double salt of cadmium sulphate and ammonium sulphate having the constitution 2CdS04,(NH4),S04. It is ob- tained in microscopic crystals yellow when hot white when cold.It is v0ry hygroscopic and has D22 3-11. It is very soluble in water and from its solution on evaporation crystals of CdSO4,(N€&),S0,,6H,Oii. 123 ABSTRACTS OF CHEMICAL PAPERS are deposited. The salt 2CdS04,(NH,),S0 is decomposed by warming with concentrated sulphuric acid at looo giving anhydrous cadmium sulphate in rhombic crystals. W. G . Egyptian-Blue. A P. LAURIE W. F. P. MCLINTOCK and F. D. MILES (Proc. Roy. Soc. 1914 A 89 418-429).-Experimerits have been made to determine the nature and conditiolns of formation of Egyptian-blue. Acco'rding to Fouqu6 (BUZZ. SOC. Mines 12 36; Compt. rend. 1889 108 325) this substance is a crystalline double silicate of copper and calcium of the formula Ca0,Cu0,4Si02.That the substance is really crystalline in character has been shown by the examination of a number of samples of real Egyptian- blue between crossed nicols. I n order to ascertain the conditions under which the compound is formed a mixture containing 36 grams of quartz sand 4 of fusion mixture 8.6 of copper carbonate and 7.2 of calcium carbonate was heated for several hours in an electric resistance furnace a t temperatura ranging from 760° t o over 900°. These experiments show that the temperature should be kept between narrow limits (830-900O) if the blue compound is to be obtained. The mixture may however be heated to the temperature of the oxyhydrogen blow-pipe provided that the mass is subsequently maintained for a considerable time at about 850O.If the tempera- ture is too high or too law the product is an olive-green glass. Further experimenta show that the formation of the blue com- pound is not dependent on the presence of sodium or potassium carbonate or other alkali salts although if these are absent the mass is so infusible that reaction takes place with great difficulty. On the other hand if the amount of fusion mixture is increased very much above that correspondipg with the mixture referred to above the calcium copper silicate does not crystallise out of the mass but remains in solution as a green glass. From analyses of Egyptian-blue prepared by the authors i t appears that in presence of an alkali a little of the copper and calcium is replaced by the alkali metals. Apart from this the analyses are found to correspond with the formula Ca0,Cu0,4Si02.H. M. D. Polymorphism of Mercuric Iodide. MEINHARD HASSELBLATT (Zeitsch. physikal. Chem. 1913 86 61-64).-Polemical against Smits (A. 1910 ii 400). Mercuric Oxide. GUY B. TAYLOR and GEORGE A . HULETT (J. Physicatl Chm. 1913 17 755 -761).-Mercuric oxide waR prepared by heating carefully purified mercury in an atmosphere of oxygen at about 420° and a pressure of 2-3 atmospheres. The apparatus was arranged so that fresh supplies of oxygen could be admitted from time t o time. It was found possible to obtain from 10 to 15 grams of the oxide in thel course of an experiment which lasted from five to Beven days. The oxide was finally freed from all J. F. S.Ih’ ORGAXIC CHEMISTRY. ii. 12:) trac;es of unconibined metal by heating to 400° iu a rapid current of oxygen a t atmospheric pressure.The pure mercuric oxide obtained in this way was reduced to metal by heating with pure finely divided iron the temperature being maintained a t 275-300O for two to three hours and then at about 600° for twenty-four to thirty-six hours. During the last half-hour the protruding end of the tube was cooled in ice and the globule of condensed mercury transferred to a porcelain crucible and weighed. Nine analyses of the oxide were made according to this method the results agreeing exceedingly well with one another and giving for the mean value of the percentage of mercury in the oxide 92.6053 k 0*0008. This corresponds with an atomic-weight value for mercury of 200-37 k0-025. This value is considerably smaller than that recently obtained by Easley and Brann (A.1909 ii 1013; 1910 ii 957; 1912 ii 257) from the analysis of mercuric chloride and bromide namely? 200.62. I f the higher value is correct the lower value might be explained by the presence of a small quantity of a higher oxide. The authors consider however that the discrepancy calls for a further investigation of the atomic weight of mercury. H. M. D. The Alloys of Cerium with Silicon and Bismuth. RUDOLF VOGEL (Zeitsch. anorg. Chem. 1913 84 323-339).-The fact that cerium combines with lead and tin with development of heat forming several compounds indicates that it does not belong to the same chemical family of the fourth group. It is now found that it combines with silicon at high temperatures with great violence.It has not been found possible to prepare alloys con- taining more than 70% of cerium as combination does not take place until 1400° and the heat developed brings about the destruc- tion of the containing vessel or if carbon crucibles are used there is a considerable production of cerium carbide in which case the alloys rapidly disintegrate in air. Solid silicon floats on molten cerium and a t a sufficiently high temperature combines explo- sively. Between 0 and 70% Ce the freezing-point curve has two branches intersecting at l24Oo and 53% Ce. The eutectic times indicate that the maximum on the curve must be at 83% Ce corresponding with a compound CeSi melting above 1500O. This compound forms yellow rounded crystallites and the eutectic has a distinct lamellar structure.Free silicon crystallises in needles which are harder than the compound. The alloys are brittle very stable in air and are not pyrophoric. Cerium and bismuth combine with great development of heat. Porcelain tubes are rapidly corroded but carbon tubes may be used and the alloys are not seriously contaminated with carbide. The thermal effects are ofteri small and the micro-structure has been largely employed in determining the form of the diagram. The sections must be polished with wet alumina but they oxidise very rapidly and the polished surfaces cannot be preserved. VOL. CVI. ii. 9ii. 130 ABSTRACTS OF CHEMICAL PAPERS. Four compounds are fornied BiCS Bi,Ce BiCe and BizCe. The compound Bi,Ce melts a t 1630° and appears as a maximum on the freezing-point curve.It crystallises in polygonal grains. BiCe is formed at 1400° and is softer than Bi,Ce,. It forms a eutectic with cerium a t 757O. BiCe is formed a t 1525O and Bi&e a t 882O whilst the second eutectic point practically coincides with the melting point of bismuth. All the alloys are more readily attacked by water than cerium. Between 25 and 75% Bi the action of water may even raise the alloys to incandescence. C. H. D. The Resolution of Ytterbium into its Elements. C. AUER VON WELSBACH (Monatsh. 1913 34 1713-1728).-1n the resolution of ytterbium into its elements (compare A. 1908 ii 591) the fractionation proceeds at first very slowly but after a large number (ZOO) crf fractionations have been carried out there is a very marked increase in the rate of fractionation.This increase could be ascribed to the existence of an unknown element occurring between the elements aldebaranium and cassiopeium but the spectroscopic examination afforded no definite evidence in this direction. The cause of the increased rate of fractionation is therefore a t present inexplicable. The atomic weights of aldebaranium and cassiopeium have been determined by a new method. The respective hydrated sulphates 3X2(SO4),,8H2Oy were dried in a platinum crucible in a vacuum desiccator; heating on a water-bath is not permissible in the case of cassiopeium sulphate. They were then dehydrated by careful heating an excess of oxalic acid added and the crucible twethirds filled with water. The crucible and its contents were then heated on a water-bath t o dissolve the excess of oxalic acid after which the insoluble oxalates were collected washed and then converted into oxide. Any salt.remaining in the mother liquor and washings was precipitated by making use of the fact that the ammonium oxalates of these metals are insoluble in a saturated solution of ammonium hydrogen oxalate and allowed for. The results gave Cp = 175.00 Ad = 173'00. T. S. P. The Perchlorates of Aluminium Chromium and Mag- nesium. R. F. WEINLAND and FR. ENSGRARER (Zeitsclh. anorg. Chem. 1913 84 368-372).-AZumirtium percchlorate [ Al(€€20)61(C104)q is ft derivative of the hexa-aquo-base and thus completely resembles the ferric compound (this vol. ii 132). Its solution yields with sodium perchlorate a crystalline sodium aluminotetraperchlorate [A1(C104),]Na,12H,0 which loses 6H20 over sulphuric acid. Chromium yields two perchiEorates of the hexa-aquo-base [Cr(H,0),]C10,)3 and [Cr(H,O),!(C!10,),,3R,O but no sodium chromiDerchlorate.Both of theae salts are bluish-preen. Mag&iam perchJornte [Mg(H,O),](ClO,) is al& a hexa-aquo- salt. C. H. D.INORGANIC CHEMISTRY. ii. 131 The Critical Ranges of Pure Iron. H. c'. H. CARPENTER (J. Iron Sted Inst. 1913 i 315 -326).-Electrolytic iron sheet contain- ing 99.967% of iron gives cooling curves which are consistent with the view that the critical point Ar is merely the retarded termina- tion of Br3. The point -4c is only faintly marked on the heating curves when dissolved gas has been removed. The results support the view of Benedicks (A. 1913 ii 599) that P-iron is a solid solution of y-iron in a-iron (compare Miiller A.1909 ii 485; Burgess and Crowe 9. 1913 ii 711). C. H. D. The Tenacity Deformation and Fracture of Soft Steel at High Temperatures. WALTER ROSENHAIN and J. C. W. HUMFREY (J. fron Steel Inst. 2913 i 219-271. Compare A. 1910 ii 128).- Tensile tests with strips of soft steel in a very high vacuum indicate that the tenacity falls rapidly with increasing temperature from 600° to 730O; there is then a break in the curve and the tenacity passes through a minimum between 800° and 900° an entirely distinct curve starting from a relatively high tenacity at 900') and then falling gradually representing the y-phase. The first break is considered to represent the a+/3 change.The influence of size of crystal grain and rate of loading on the resulta has also been determined and the results are regarded as supporting the " amorphous cement " theory of the constitution of metals (Rosen- liain and Ewen A 1913 ii 119). C. H. D. Influence of Sulphur on the Stability of Iron Carbide in the Presence of Silicon. W. H. HATFIELD (J. I T O ~ Steel Inat. 1913 i 139-156).-Sulphur increases the stability of iron carbide (cementite) a t high temperatures entering in small quantity into the carbide. The influence of sulphur is not purely mechanical through the formation of sulphide films. Its influence in cast iron is neutralised by the presence of manganese which forms an insoluble sulphide and of silicon possibly owing to the formation of a silicon sulphide although such a compound was not actually isolated.C. H. D. The Rusting of Iron in Water. W. A. BRADBURY (Chem. News 1913 108 307-308).-Two flasks were filled with well-boiled Manchester tap-water some coils of bright iron binding wire added and the flasks securely corked. No rusting took place whereas rusting readily occurred in unboiled tap-water. I f rusting takes place according to the equation Fe + 2H,C03 = Fe(HCO,) + H the ferrous hydrogen carbonate then being oxidised by oxygen present hydrogen should be evolved. It was found however that no hydrogen was liberated when rusting took place in ordinary tap-water. This could be accounted for by the nascent hydrogen liberated being oxidised by the oxygen present in the water or else by the hydrogen remaining dissolved in the water.This was tested by using water into which carbon dioxide had been passed for fifteen minutes. At first no hydrogen was evolved; the iron 9-2ii. 131 ABSTRACTS OF CHE3lIC.A J d PAPERS. reiuained quite bright but ferrous iron was present ill solutiorJ. After some days hydrogen was evolved but the iron remained quite bright. Apparently no hydrogen was evolved until all the oxygen present in the water had been reduced; owing to the removal of this oxygen no rust could form although a large quantity of iron was present in solution as ferrous bicarbonate. These experiments confirm the view that rusting is due t o dissolved oxygen and carbonic acid present in the water. Contrary to what is usually supposed to be the case magnesium chloride was found not to have a deleterious effect on iron.T. S. P. New Method for the Preparation of Colloidal Ferric Hydroxide. 'hEoDoRE C'OHEN (J Amer. Chern soc. 1914 36 19-23).-Experiments are described which show that a colloidal solution of ferric hydroxide can be obtained by the hydrolysis of a ferric nitrate solution which takes place in a nitric acid solution containing copper; for example if 3 grams of iron filings contain- ing copper as an impurity are added to 10 C.C. of concentrated nitric acid and the solution is diluted filtered and dialysed a deep red liquid is obtained from which on treatment with a little sulphuric acid or with the electric current ferric hydroxide separates. A similar solution can be prepared by boiling a solution of ferric nitrate with copper filings or with zinc dust.R. DE FORCRAND (Compt. rend. 1914 158 20-23).-A btudy of the hepta- tetra- and mono- hydrates of ferrous sulphate and the anhydrous salt. The hepta- hydrate can be obtained perfectly pure and dry by powdering an ordinary sample and repeatedly pressing it between folds of filter- paper. It neither oxidises effloresces nor deliquesces. It has heat of solution -4.323 Cal. a t 13.5O. The tetrahydrate has heat of solution t1.599 Cal. a t 13*5O the monohydrate +7.538 Cal. at 13.5O and the anhydrous salt +14.901 Cal. at +13*5O. From these results the author calculates for the monohydrate a b. p. 300° for the tetrahydrate b. I>. 118'5O and for tho lieptahydrate 118.3O. E. G. Ferrous Sulphate and Its Hydrates. The product thus obtained is quite stable in air a t 1 5 O .W. G. Salts of Ferri-phosphoric -sulphuric and -perchloric Acids. R. F. WETNLAND and FIR. ENSGRABER (Zeifsch. anorg. C h m 1913 84 340-367).-An ammonium ferriphosphate has been obtained by Cohen (A. 1907 ii 552) and there is other evidence of a series of complex salts. Either ferric hydroxide or ferric chloride is mixed with an excess of phosphoric acid and the complex alkali salts may then be obt'ained by the addition of alkali hydr- oxide phosphate or chloride. Definite proportions must be used and the solutions must be heated for a t least twenty-four hours otherwise the products although well crystallised are not homo- geneous. Sodium f erridiphosphate [Fe(PO,),1R,Na,H2O is obtained in pale pink crystals by heating a solution of ferric hydroxide inINORGANIC CHEMISTRY. ii.133 phosphoric acid with sodium hydroxide (Fe P Na = 1 24 8) for three days on the water-bath. After six months a t the ordinary temperature i t is obtained with 3H,O. The corresponding am- monium salt [Fe(PO,),]H,-NH consists of pale pink microscopic hexagonal crystals. By using less ammonia an acid ammonium salt [Fe(P0,),H3],NH,,7H,0 may be prepared in light red micro- scopic crystals. The pyridinium salt [Fe(PO,),]H,C,NH is white. Sodium ferritriphosphate [Fe(PO4),]H,Na,H2O is obtained when sodium phosphate o r chloride is added to the solution of ferric phosphate instead of sodium hydroxide. Thus a mixture in the proportions (Fe P NaCl= 1 6 4) yields the salt after heating for twenty-four hours on the water-bath.It is a very pale red crys- talline powder. A mixed ammonium ferri-di- and tri-phosplutte 3[ Fe( P04),]H,,[Fe( PO,),]H 1 *5NH3,1 OH20 crystallises in the course of six months With a large excess of ammonium chloride and a deficiency of phosphoric acid an entirely different salt of unknown constitution is obtained as a greenish-yellow microcrys- talline powder. It has the empirical composition 2FePO4,NH,,4H,O. All of these compounds are very 'sparingly soluble in water. It has not been found possible to prepare corresponding potassium salts. Ferric phosphate prepared by heating molecular proportions of ferric chloride or acetate with phosphoric acid for two days on the water-bath is a pink microcrystalline powder and may be regarded as ferric ferridiphosphate [Fe(P04),1Fe,5H,0.It is very sparingly soluble in water and dilute acids whilst the phosphates of variable composition obtained by precipitating ferric solutions with alkali phosphates dissolve readily in dilute acids. Ammonium ferridisulphate [Fe(SO,),]NH obtained on heating a solution of ferric ammonium asluin with sulphuric acid for twenty- four hours is a white microcrystalline powder sparingly soluble in water and depositing ferric hydroxide rapidly on heating. The potassium salt [Fe(SO,),]K,H,O and the pyridinium salt have similar properties. Trisodium ferritm'sulphate [Fe(SO,),]Na,,SH,O is a white crys- talline salt as is the previously known ferridisulphuric acid [ Fe( S04)2]H,4H,O- Sodium f erritetraperchlorate [Fe(C1O4),]Na,6H,O forms large pink hygroscopic crystals.Ferric perchlorate Fe(C10,),6H20 is to be regarded as a hexa-aquo-salt [Fe(H,O),](ClO,),. [Fe(S0*)2]H,C,NH,,2H20 C. H. D. Rinmann's Green. A RVID HEDVALL (Arkiu. Kern. illin. Geol. 1913 5 No. 6 1-27).-!Che author's further investigations (com- pare A. 1912 ii 846) show that Rinmann's green is not a definite substance but rather a series of solid solutions of the components zinc oxide and cobalt oxide since its composition may vary con- siderably. The series of mixed crystals is isodimorphic since zincii. 134 ABSTRACTS OF CHEMICAL PAPERS. oxide is hexagonal and cobalt oxide generally regular although in one case the latter oxide has probably been obtained in hexagonal crystals. T. S. P. Chromic Oxide Jellies.E. H. BUNCE and L. S. FINCH (J. Physical Chew. 1913 17 769-779) -Experiments have bwn made to determine the conditions of formation of chromic oxide jellies. If a sufficient amount of sodium acetate is added to a solution of chromic sulphate or chloride the subsequent addition of a suitable amount of alkali metal hydroxide or ammonia causes gelatinisa- tion. The chromic oxide jelly is violet if prepared by the addition of ammonia or of a slight excess of alkali metal hydroxide. If this is added in larger quantity the jelly is green in colour. The con- centrations of the chromic salt and sodium acetate may be varied within fairly wide limits without interfering with the formation of the jelly. If after the addition of sodium acetate the solution is heated the time required for gelatinisation is found to diminish appreciably.Apart from this prolonged heating of the solution has no effect on the subsequent gelatinisation provided acetic acid is not driven off during the process. Chromic oxide jellies dissolve in hydro- chloric acid but are reformed when the solution is neutralised if sufficient sodium acetate is present. The addition of sodium or potassium chloride has no appreciable influence on the gelatinisa- tion but this is prevented by the freezing or stirring of the solutions. Chromic oxide jellies can also be obtained by adding sodium or potassium hydroxide to a solution of chrome alum but not by the addition of ammonia. The addition of sodium acetate to the chrome alum solution appears to be without influence on the result.Although jellies are obtained by the addition of potassium hydr- oxide to solutions of chromic acetate and chrome alum this is not the case when the alkali is added to solutions of chromic nitrate chloride or sulphate. The acetate method of preparing jellies gives no result in the case of salts of manganese aluminium copper and cadmium. H. M. I). Chromic Silicofluoride. Its TraneformationR. Fluopenta- nquochromic Silicofluoride. A. RECOURA (Compt. rend. 191 3 157 1525-1528. Compare A. 1913 ii 603).-In explanation of the fact that the normal ferric silicofluoride is decomposed as formed giving the compound Fe2F,,2SiF the author expresses the view that ferric fluoride itself is really a complex t o be represented by (Fe2F?)F4 and thus the double compound prepared is a true silicofluoride (F%F2)(SiF4),. I n support of this he shows that the latter compound reacts with potassium chloride t o give potass- ium silicofluoride and a compoicnd (Fe,F,)Cl in which the fluorine is masked.I n the case of chroniiuu1 the silicofluoride Cr2F,(SiF& is obtained in solution although i t is not possible to isolate it as the violet solution spontaneously and gradually turns bluish-green andINORGAPilC CHEMISTRY. ii. 135 from i t a green solid can be isolated to which the author assigns the constitution (CrF,5H20)SiF6 and compares i t with the two silicofluorides - (CrC1,5NH3)SiF6 and (CrCl,H,0,4NH3)SiF6 pTe- pared by Jorgensen. The violet solution immediately after its preparation reacts normally with potassium chloride giving potassium silicofluoride and chromic chloride but on remain- ing it reacts to give potassium silicofluoride and the com- pound (Cr2F2)C14 from which the fluorine cannot be precipitated by addition of barium chloride.The compound (CrF,5H20) SiF is perfectly stable when kept in a desiccator even in a vacuum but on exposure to air it slowly loses silicon fluoride ultimately yielding a green chromic fluoride Cr,F,,7H20. Aluminium silicofluoride gives analogous results. Metastability of Metals Prepared by an Electrolytic Method. ERNST COHEN and W. D. HELDERMAN (Chem. Weekblad 1914 11 83-84).-At 1 8 O grey tin is converted into the white modification. The white modificatlon is also deposited by electrolysis of a solution of a tin salt a t -2OO. The Action of the Silent Electric Discharge on a Mixture of Hydrogen and Titanium Tetrachloride Vapour.11. A Poly- morphic Titanium Trichloride. F. BOCK and L. MOSER (I?lonntsh. 1913 34 1825-1849).-1n a previous communication (A. 1913 ii 9) a brown substance was described which was considered to be a polymorphic form of ordinary violet titanium trichloride. I n the present communication the authors describe an improved apparatus f o r preparing and handling this substance. The change from brown to violet trichloride is not reversible; the heating curve of the brown modification the temperature being gradually raised above that a t which transformation takes place did not differ appreciably from that of the cooling curve of the violet modification thereby formed. The heats of solution in water of various known mixtures of the brown and violet trichlorides with titanium tetrachloride were determined as also the heat of solution of pure titanium tetra- chloride.From the results the following figures were obtained (TiCl,,aq.) 59030 cal. ; (brown TiCl,,aq.) 48150-48837 cal. ; (violet TiCl,,aq.) 44287-45800 cal. These figures are taken to prove that the brown and violet titanium trichlorides are mono- tropic modifications. T. s. P. H A N ~ BUCHTALA ( J . pr. Chern. 1913 [ii] 88 '771-'785).-hn account of the prepara- tion of a number of thallous borates. ThJaZZous tetraborate Tl,B40 prepared by dissolving thallous carbonate (1 mol.) and boric acid (1-4 mols.) in water crystallises with 2H,O. It is also obtained together with the hexnborate T12B60,,,3H20 when the carbonate is fused with boric acid (6 mols.) and the product crystallised from water; if the fusion is carried out with a greater excess of boric acid (8 mols.) a mixture of tho hexaborate and decaborate T12B,,0,,,8H,0 is produced. W. G. A. J. W. Compounds of Thorium with Boric Acid.ii. 136 ABSTRACTS OF CHEMICAL PAPERS. Thallozis octaborate T12B,0,,,4H,0 and thallous dodecaborate T12B,,01~,7H20 are prepared by dissolving thallous carbonate in an excess of aqueous boric acid (10 mols.); the last-named borate also crystallises with 5H20 in twinned monoclinic crystals (a b c = 1.556 1 1.920. 11 = 945?0°). ThaZEoiLs perborate is obtained as a white powder by the addition of 30:h hydrogen peroxide to an aqneous solution of any of the preceding borates. It shows the usual reactions of a per-salt and decomposes explosively when rapidly heated. The amount of oxygen liberated by hating the perborate with water corresponds with the formula I >O; the aqueous solution on evaporation yields a red thdious metaborate which becomes blackish-red on exposure t o air and crystallises from water in radially-arranged colourless wodge-shaped crystals of the corn- position Tl,B,O,,H,O. F. B. TlO*B-O*O TIWB*O*O