INORGANIC CHEJIISTRY. 13 I n o r g a n i c Chemistry. List of Elementary Substances announced from 1877 to 1887. By H. C. BOLTON (Chem. News, 58, 188).-The names of 58 substances, announced as being elementary, together with their sources and the names of their discoverers, are given in tabular form. By E. ALLARY (Bull. XOC. Cliim., 49, 865-867) .-The author cites the isomorphism of potassium chloride and potassium cyanide as fresh evidence in support of the views of Brodie, Dumag, Lockyer, and others that chlorine is composed of two elements. If the atomic weight of chlorine is dirided into two parts proportioiial to the atomic weights of carbon and nitrogen, the two constituents of chlorine would have the atomic weights 19.1 and 16.4 respectively. From the study of certain series of organic com- pounds, Dumas came to the conclusion that chlorine consists of two elements, the atomic weights of which are 19 and 16.5 respectively. If this view were correct, chlorine would probably be composed of fluorine and oxygen, and the oxygen is perhaps intimately united H with --.Meyer’s observation that oxygen is produced when chlorine 2 is strongly heated, may be adduced in support of this theory. Apparatus for a Constant Supply of Chlorine. Chlorine and Cyanogen. F. S. K. By .A. VOSMAER ( Z a i t . anal. Chem., 27, 638-640).-The manganese dioxide is used in fragments of the size of peas, andis placed in a two-necked bottle, a t the bottom of which there is a layer of broken glass o r pumice. This stands in a water-bath. Hydrochloric acid is supplied irom a reservoir a t a higher level by a tube reaching to the bottom of the layer of glass, a T-piece and stopcocks allowing the same tube to Serve for the removal of the manganese solution.The corks should be soaked in paraffin. Suitable drying apparatus can be attached, and will not require replenishing for a long time. The chlorine begins to come oft’ when the temperature of the bath reaches 50°, and by means of a stopcock on the outlet its rate is completely under control. The evolution can speedily be arrested by closing t,he stop- cock a t the outlet of the d q i n g apparatus and emptying the water- bath. The apparatus is then left full of chlorine, and is ready a t any moment to give a supply of the gas completely free from oxygen. M. J .S. Dissemination of Sulphur and Phosphorus in Masses of Metal. By H. N. WARREN (Chenz. Ncws, 58, 177--178).--When sulphur is well mixed with excess of molten iron, and a rod cast from14 ABSTRACTS OF BHEMICAL PAPERS. the mass is made the positive electrode of an electriccircuit, in a bath of ferrous chloride, with a platinum 01- copper negative electrode, tile metallic iron is dissolred and is precipitated on the negative elec- irode, whilst ferrous sulphide, FeS, remains attached to the positive electrode. I n the same way, iron phosphide, Fe4P, may be separated from a mam of the metal. Other metals of the iron group behave in a similar manner. The author r e p r d s the above compounds, FeS, Fe4P, and the compound Fe,Si. obtained by him, as the lowest forms of the respective iron sulphides, phosphides, or silicides, although ordinary analytical results may indicate still lower forms, owing to the intimate dissemination of these compound< throughout the mass of the metal.The sulphides of the metals of the second group are not disseminated in this way ihrough the mass of the metal (compare Abstr., 1888, 555-556). Preparation of Hydrogen Iodide. By A. ~ A R D (Bdl. XOC. Cltim., 49, 742-743) .-In preparing- hydrogen iodide from iodine and amorphous phosphorus, the authw places the iodine in a flitsk provided with a bent, neck and connected with the vessel contailling the phosphorus and water by means of a bent tube. By turning the flask round the bent tube, fresh quantities of iodine can be added when requisite without admitting air into the apparatus.Hydrogen Sulphide Apparatus. By J. H. J. DAGGER (Chem. News, 58, 127).-The apparatus figured and described consists of two glass globes connected by india-rubber tubing. The acid is put in one of the globes, the ferroiis sulphide (zinc or marble) in the other ; they are fi.tted with tubes and corks and then fixed in a suit- able position. D, A. L. Apparatus for the Preparation of Hydrogen Sulphide. By P. CHANTEMILLE (Bzdl. Roc. Chim., 50,170-171).-The iron sulphide is contained i n an kpouvrette, and the acid in an ordiiiary flask fitted with a doubly-bored cork, through the one hole of which is passed a tnbe reaching to the bottom of the flask, and closed a t its upper end by an india-rubber tube and pinch-cock ; the second hole is fitted with x short# tube bent a t right angles and fitted to the lower end of the bpouvrette.The force of gas is increased or diminished by raising or lowering the flask. By E. J. MAUMENB (Bull. SOC. Chin?., 49, t350-85~).-Ch!/dr(iza~ne is evolved when a fiolntion of potassium permanganate (158 grams) and sulphuric acid (40 prams SO,) is added to dried, crystallised ammonium oxalate (141.2 grams), the whole well mixed and gently heated until it begins to boil. The gafeous product is absorbed in hydrochloric acid and a neutral solution of the salt can thus be obtained. The hyrl~o- chlwide is crystalline and very readily soluble in water, but only sparingly in alcohol. The sublimed salt has the composition N2H,O,2HC1, but the crystals dried by means of the anhydrous salt contain one-fifteenth of their weight of water.When a solution of D. A. L. F. S. K. The gas is purified in the usual manner. N. H. &I. Chydrazaine or Protoxide of Ammonia.INdRGANIC CHEJJISTRP. 35 the hydrochloride is mixed with platinic chloride, a pZatinochloride is obtained, the composition of which varies with the conditions of the experiment ; with excess of the hydrochloride, a yellow salt is formed, the composition of which is approximately NzH,0,H2PtC16, but if excess of platinic chloride is added, the proportion of platinum is sensibly increased. The suZpphate is crystalliiie and soluble in water, but only very sparingly so in absolute alcohol ; it forms a double salt with aluminium sulphate. The nitrate is crystalline. When a solu- tion of the nitrate is evaporated, nitric acid, nitric peroxide, nitrogen, and a compound having the composition NzHz are evolved.Action of the Electric Spark on Mlxtures of Nitric Oxide with Hydrogen, with Methane, &c. By S. COOKE ( C h e w News, 58, 130--131).-Under the influence of sparks (from a coil capable of giving a 4-inch spark), with the eudiometer wires half an inch apart, a mixture of hydrogen and nitric oxide always explodes, pro- vided the proportion of hydrogen to nitric oxide does not exceed 6 : 10 ; but with the wires closer together, or with a feeble coil, or if the pressure is diminished to 300 mm. of mercury, no explosion occurs. The carefully dried gases explode quite as readily as when they are not dried. The nature of the gaseous mixture after explosion varies very considerably, but there is little doulnt that much of the nitric oxide is converted into oxygen aiid nitrogen.Explosions have also been obtained with nitric oxide and hydrogen sulphide ; with nitric oxide and methane with the production of carbonic anhydride and oxygen : other hydrocarbons i n proper proportions and suitable spa1.k make an explosive mixture with nitric oxide. Carbonic and nitric oxides mixed do not explode under the influence of the spark, but combination goes on gradually with the production of carbonic anhy- dride and nitrogen. Decomposition of Nitric Oxide in Contact with Water and with Potash. By S. COOKE (Chem. News, 58, 155--156).--Nitric oxide exposed in tubes over water in the dark undergoes gradual decomposition with the production of nitrous acid, nitrogen, and a little nitrous oxide.The change is always slow, but is more active at the commencement than a t the end of an experiment; it is also accelerated by the presence of platinum and by heat, whilst admix- ture with hydrogen retards it. The actionof potassium hydroxide on nitric oxide is also aided by platinum and heat (oompare this Journal, 18i7, ii, 37). BJ- L. W. I?. S. K. D. A. L. D. A. L. Action of Hydrogen Sulphide on Arsenic Acid. MCCAY (Zeit. an/iZ. Chern., 27, 632-634; compare Brauner and Toinihek, Trans., 1888, 145).-When a slow stream of hydrogen sulphide is passed through an acidified solution of an arsenate a t 7u0, besides arsenic pentasulphide there is also formed Home free thioxy- arsenic acid, H,As03S.This, under the influence of mineral acids and heat, decomposes into free sulphur and arsenious acid, the latter of which then yields arsenic trisulphide with the hydrogen sulphide. A solution of thioxyarsenic acid may be obtained by passing hydro-16 ABSTRACTS OF CHEMICAL PAPERS. gen sulphide not in excess into a cold, dilute, acidified solution of potassium arsenate. If R larger quantity of hydrogen sulphide is employed, the excess may be removed either by immediate addition of copper sulphate or by a vigorous stream of air bubbles. An opalescence caused by free sulphur may be removed by shaking with asbestos. The clear, strongly acid liquid obtained, exhibits the fol- lowing pPoperties. It remains clear for a long time after addition of sulphuric or hydrochloric acid ; it gives no immediate precipitate with hydrogen sulphide, but ultimately yields one.When boiled, i t gives a precipitate of pure sulphur, without evolution of hydrogen sulphide or sulphurous anhydricle. With hydrogen sulphide, the boiled and cooled liquid gives an immediate precipitate of arsenic trisulphide ; it gives no precipitate with copper sulphate ; with mercuric chloride it gives immediately a heavy yellowish-white precipitate ; with silver sulphate it gives a heavy black precipitate, the filtrate from which contains no arsenious acid. The potassium thioxyarsenate of Bouquet and CloGz agrees with this solution in all the above particulars. M. J. S. Barium Sulphite. By E. R. HODGES (Chem. News, 58, 128); G. S. JOHNSON (ibid., 155).-Hodges' experiments led him to infer that barium sulphite is insoluble in hydrochloric acid ; but 3ohnsoii proves it to be readily and completely soluble in that solvent.He, moreover, shows that pure aqueous barium chloride is not precipitated by sulphrnrous acid, but that i n the presence of dissolved oxygen a precipitate of barinm sulphate forms. n. A. L. Solubility of Gypsum. By G. A. RAUPFNSTRAUCH ( C h ~ m . Cent?,, 1888, 891-822, from Pharin. Centrulhal., 29,229-233).-The author finds that a saturated solution of gypsum is readily obtained, but supersaturation of the solution can only be obviated by shaking the solution for some time. The solubility of gypsum increases up to the temperature of 32", between 32" and 38" it, remains almost constant, and at, higher temperatures than 38" decreases.Natural gypsum comports itself like artificial, provided it be pure. After heating, gypsum takes up water of crystallisation more or less rapidly, and shows the normal solubility again. J. W. L. Ancient Mortar from a Roman Wall in London. By J. SPILLER (Chem. News, 58, 189).-While examining some mortar from a Roman wall, exposed when sinking the foundations of the new Post Office buildings in St. Martin's-le-Grand, the author found, after extracting as much silica as possible by means of dilute hydro- chloric acid, that the residue yielded nearly 11 per cent. of silica to cold dilute sodium hydroxide. Under similar treatment with cold soda, builder's sand and pulverised flints yield a mere trace of silica in solution, whilst mortars 20, 160, and many hundreds of years old yielded quantities of silica increasingwith the age of the mortar.The author suggests that perhaps the Romans used a puzzuolana in com- pounding their mortar, or perhaps this soluble silica or silicate is the direct result of long contact of plain sand and lime. The mortarIN0 RGANIC CHEMISTRY. 17 from St. Martin's-le-Grand had the following composition per cent: :-Sand and brick, 46.48 ; acid-soluble, SiO, 0.52 ; alkali-solublt~, SiO, 10.44, A1,0, 3.00, Fez03 0.48, CaO 20.02, MgO 0.76, CO, 13-03, SO3 0.37,NaCl trace, H,O and loss 4.90. D. A. L. Analysis of Money. By 5. C. WELCH (Chem. News, 58, 164- 165) .-The author has analysed some specimens of " matlilla money'' made in Birmingham. It is yello wish-red, reddish-yellow, brass or bronze-like in colour, and in shape like a G.They proved to be variable alloys of lead and copper with small quantities of iron, tin, zinc, antimony, and arsenic. Some contained pieces of originally unmolten metal, and some had a semi-fused appearance. Another coin resembling gold in colour contained per cent. : Cu, 62-58 ; Zn, 37.26 ; Fe, 0.11 ; Pb, 0.013, and had been silvered to pass for a 6d.-piece. D. -4. L. New Hydrated Cupric Chloride. By E. CHUARD (Chem. Centr., 1888, 887, from Arch. sci. Phys. Nat. Gendve [3], 19, 477).-A hydrated cupric chloride of the formula CuCI, + 3H,O crystallises from the solution of the green hydrate when cooled down to 0". The existence of this compound explains the change in the colour of the solution which takes place when it suffers dilntion, for the chloride of the green hydrate contains only 2 mols. HzO, and by diluting or by cooling below O", the chloride with 3 mols.H,O is formed, and gives a blue colour to the solution. The latter chloride again loses 1 mol. H,O when the solution is boiled. J. W. L. Purification of Mercury. By J. M. CRAFTS (BuZ7. Xoc. Chim., 49, 856--860).-Mercury can be completely freed from lead, zinc, tin, and other impurities by placing it in a slightly inclined glass tube provided with a funnel at the lower extremity, and aspirating a gentle stream of air through the apparatus for about 48 honrs. The oxides of the metals collect at the upper end of the tube, and after about 24 hours, as a rule, the surface of the mercury is quite clean and the operation is finished. Large quantities can be treated in this way, but mercury which has been used for amalgamating zinc contains such a large amount of impurity that this method cannot be suitably employed.A stream of piire air was passed through pure mercury contained in the apparatus described above, but even after 10 days' time no appreciable quantity of oxide was formed. Platinum, in the form of thin foil, is not attacked by cold mercur-, but when the latter is boiled, air being excluded, the platinum is gradually acted on. Only a small quantit? is dissolved, as a large proportion separates in the form of a black powder, and almost the whole is simply held in suspension. The surface of the mercury remains bright, but when a stream of air is passed, the platinum collects at the surface aft,er some time as a black powder, and on dis- tilling the separated mercury only a very small quantity of platinum remains.F. S. K. Silver is not removed by this process. VOL. LVI. C18 ABSTRACTS OF’ CHEMICAL PAPERS. Yttrium-Potassium and Y ttrium-Sodium Phosphates. Bg A. DUBOIN (Compt. rend., 107, 622-624) .-YYtt&ma-pofassiurn pyro- phosphate, K2O,YZO3,2PzO5, is obtained by saturating potassium meta- phosphate with yttrium oxide a t a bright red heat, and then keeping the mixture a t a somewhat lower temperature for a considerable time. The cooled product is extracted with water, when the phosphate remains undissolved. It forms small, colourless, highly birefractive prisms. Yttrium-potassium orthophosphate, 3K20,YZO3,2P2O5, is obtained by adding excess of yttria to a fnsed mixture of potassium pyrophos- phate with nine times its weight of potassium chloride, and heating over a Bunsen burner f o r 20 minutes. It crystallises in brilliant, hexagonal lamellae; sp.gr. at 20” = 3.3. The same product is obtained on heating amorphous yttrium phosphate to redness, o r even to a much higher temperature, with excess of potassium sulphate. Another orthophosphate, 3Kz0,5Y20,,6P205, is obtained by heating potassium sulphate to a high temperature with a much larger propor- tion of yttrium phosphate. It forms brilliant, trausparent, colour- less, hexagonal prisms. If the mixture of yttrium phosphate and potassium sulphate con- tains 10 per cent. of the former, and is heated t o a very high tempe- rature for about 10 hours, yttrium phosphate is obtained i u a form identical with xenotime.C. H. B. So-called (( Crackle” China. By C. LAUTH and G. DUTATLLY (Bull. Xoc.. Chim., 49, 948--956).-The name crackle china is given to china, the glazing of which is cracked in a regnlai- manner so as to form a sort of network. If the appearance is similar to that of fish scales, the china is said to be “ troutod.” The Chinese colour the interstices by means of smoke or Indian ink, and the articles thus produced are valued very highly. The crackle effect is due to the unequal contraction of the glaze and of the paste. When the firing is continued for a certain time, the coefficients of expansion of the glaze and of the paste gradually become identical, and i f the firing is stopped at this point the crackle effect is not produced ; if, however, the burning is continued, the co- efficients of expansion again become different, and fracture of the glaze takes place on cooling.If the temperature which is required for a certain paste and a certain glaze to give perfect porcelain is known, and if the baking is always arrested a t this point, the crackle effect can be obtained by altering the composition of the paste or that of the glaze. The latter alternative is the more practical, but i t was found necessary to alter the degree of fusibility of the glaze at the same time. As, for various reasons, i t is disadvantageous to employ a readily fusible glaze, the temperature of fusion was raised by adding a larger pro- portion of silica.The paste experimented on is that employed at Sdvres ; its composition is- Silica, 66 ; alumina, 27 ; alkalis, 7 ; and the ordinary glaze employed a t Sbvres has the composition-INORQANIC CHEMISTRY. 19 Silica ................ 66-18 Alumina. ............. 14-55 Alkalis, .............. 3.55 Chalk.. .............. 15-90 The glaze which was found to give the best crackle effecb has the following composition :- Silica ................ 79.42 Alumina, ............. 11 -80 Alkalis. .............. 5-51 Chalk.. .............. 2-88 Crackle china is also produced if the pEoportion of alumina in the glaze is increased, for example with a glaze composed of- Silica, 69-92; alumina, 18.13 ; alkalis, 11.95, but the effects produced are not so good, and it seems that fracture more frequently occurs.Whichever method is adopted, it is always nnrticularly advantageous to substitute the chalk in the glaze for kknlis. If the propartion of alumina is increased, the necessary degree of fusibility is obtained by addimg other bases ; if the propor- tion of alumina is decreased, the quantity of other bases is diminished, and an additional quantity of silica is added. It was found that with a given glaze the crackle china can be pro- duced by making the paste more readily fusible. This can be done by increasing the quantity of felspar, and the results are better the smaller the proportion of quartz contained i n the paste. The extreme composition of the unbaked paste employed for these experiments was- Silica, 58.5 ; alumina, 28.0 ; alkalis, 5.5 ; water, 8.0.The glaze should be of medium thickness, and to obtain a close mesh it should have the composition given in one of the above two examples. If a larger mesh is required the " crackle " glaze is mixed with the ordinary glaze, and the mesh is larger the greater the pro- portion of the latter. It is advisable to make crackle china of sufficient thickness t o avoid all chance of fracture. F. S. K. Atomic Weight of Tin. By T. BONGARTZ and A. CLASSEN (Bey., 21, 2'300-2909) .-The authors made determinations of the atomic weight of tin (1) by oxidising the pure metal with nitric acid; (2) by the electrolysis of ammonium stannic chloride ; ( 3 ) by the electroljsls of potassium stannic chloride ; and (4) by the electrolysis of stannic bromide. The results are given in &he following table :- Number of Difference between experiments.Atomic weight. maximum and minimum. (1.) ...... 11 11 8,7606 u-459 (3.) ...... 10 118-7975 0.1 63 (4.) ...... 10 11 8.7309 0.144 ...... ( 2 . ) 16 118.8093 0.228 - Total ... 47 Average 118.7745 0 *;I 4s 5 c 220 ABSTRACTS OF CHEMICAL PAPERS. The average of the 26 results obtained by the electrolysis of the ammonium and potassium double salts is higher than the average of all the results. This is probably due to the fact that in oxidisinq the tin, the platinum vessel is always attacked to a slight extent, and that in dissolving the stannic bromide, a small portion may be lost by volatilisation. If, therefore, the average of the 26 experiments is taken as the most trustworthy result, the atomic weight of tin is 118.8034 (0 = 15.96) or 119.1 (0 = 16). Experiments in which it was sought to determine the atomic weight by converting pure tin into stannic sulphide, and estimating the sulphur in the latter, did not give satisfactory results (118.676 as the average of eight experiments). It was also found that when the oxide is reduced with pure hydrogen, it small quantity of tin is always volatilised.F. S. K. Action of Incandescent Platinum Wire on Gases and Vapours. By W. R. HODGKINSON and F. I(. S LOWNDES (Chern. News, 58, 187).-When a spiral of platinum wire is exposed to the action of chlorine in a glass globe and rendered incandescent by 2n electric current, a white glow is observed round the heated wire, the sides of the globe become covered with platinous chloride, arid very fine crystals of platinum are formed on the wire.With bro- mine or iodine vapour, the flame round the wire is greatly increased, but only a very small quantity of platinous bromide or a trace of the iodide is formed, and no crystalline metal is produced. With dry silicon fluoride, crystals of silicon are deposited, and the top of the globe is deeply corroded, presumedly by the liberated fluorine. D. A. I;. Normal Platinum Chloride. By R. ENGEL (BUZZ. HOG. Chi~z., 50, 100-102) .--Norm:tl platinum chloride, PtCla + 4H20, can be obtained by dissolving the calculated amount of platinum oxide in a solution of platinum chloride hydrochloride; the liquid is filtered and evaporated. The crystals are deliquescent like those of the hydro- chloride.When dry h j drogen chloride is passed over the crystals heated a t 50°, the hydrochloride is formed ; a t a lower temperature no combination takes place. N. H. M.INORGANIC CHEJIISTRY. 13I n o r g a n i c Chemistry.List of Elementary Substances announced from 1877 to1887. By H. C. BOLTON (Chem. News, 58, 188).-The names of 58substances, announced as being elementary, together with theirsources and the names of their discoverers, are given in tabular form.By E. ALLARY (Bull. XOC. Cliim., 49,865-867) .-The author cites the isomorphism of potassium chlorideand potassium cyanide as fresh evidence in support of the views ofBrodie, Dumag, Lockyer, and others that chlorine is composed oftwo elements.If the atomic weight of chlorine is dirided into twoparts proportioiial to the atomic weights of carbon and nitrogen, thetwo constituents of chlorine would have the atomic weights 19.1 and16.4 respectively. From the study of certain series of organic com-pounds, Dumas came to the conclusion that chlorine consists of twoelements, the atomic weights of which are 19 and 16.5 respectively.If this view were correct, chlorine would probably be composed offluorine and oxygen, and the oxygen is perhaps intimately unitedH with --. Meyer’s observation that oxygen is produced when chlorine 2is strongly heated, may be adduced in support of this theory.Apparatus for a Constant Supply of Chlorine.Chlorine and Cyanogen.F. S. K.By .A.VOSMAER ( Z a i t .anal. Chem., 27, 638-640).-The manganese dioxideis used in fragments of the size of peas, andis placed in a two-neckedbottle, a t the bottom of which there is a layer of broken glass o rpumice. This stands in a water-bath. Hydrochloric acid is suppliedirom a reservoir a t a higher level by a tube reaching to the bottom ofthe layer of glass, a T-piece and stopcocks allowing the same tube toServe for the removal of the manganese solution. The corks shouldbe soaked in paraffin. Suitable drying apparatus can be attached,and will not require replenishing for a long time. The chlorinebegins to come oft’ when the temperature of the bath reaches 50°, andby means of a stopcock on the outlet its rate is completely undercontrol. The evolution can speedily be arrested by closing t,he stop-cock a t the outlet of the d q i n g apparatus and emptying the water-bath.The apparatus is then left full of chlorine, and is ready a t anymoment to give a supply of the gas completely free from oxygen.M. J . S.Dissemination of Sulphur and Phosphorus in Masses ofMetal. By H. N. WARREN (Chenz. Ncws, 58, 177--178).--Whensulphur is well mixed with excess of molten iron, and a rod cast fro14 ABSTRACTS OF BHEMICAL PAPERS.the mass is made the positive electrode of an electriccircuit, in a bathof ferrous chloride, with a platinum 01- copper negative electrode, tilemetallic iron is dissolred and is precipitated on the negative elec-irode, whilst ferrous sulphide, FeS, remains attached to the positiveelectrode.I n the same way, iron phosphide, Fe4P, may be separatedfrom a mam of the metal. Other metals of the iron group behavein a similar manner. The author r e p r d s the above compounds,FeS, Fe4P, and the compound Fe,Si. obtained by him, as the lowestforms of the respective iron sulphides, phosphides, or silicides,although ordinary analytical results may indicate still lower forms,owing to the intimate dissemination of these compound< throughoutthe mass of the metal. The sulphides of the metals of the secondgroup are not disseminated in this way ihrough the mass of the metal(compare Abstr., 1888, 555-556).Preparation of Hydrogen Iodide. By A. ~ A R D (Bdl. XOC.Cltim., 49, 742-743) .-In preparing- hydrogen iodide from iodineand amorphous phosphorus, the authw places the iodine in a flitskprovided with a bent, neck and connected with the vessel contaillingthe phosphorus and water by means of a bent tube.By turning theflask round the bent tube, fresh quantities of iodine can be added whenrequisite without admitting air into the apparatus.Hydrogen Sulphide Apparatus. By J. H. J. DAGGER (Chem.News, 58, 127).-The apparatus figured and described consists oftwo glass globes connected by india-rubber tubing. The acid is putin one of the globes, the ferroiis sulphide (zinc or marble) in theother ; they are fi.tted with tubes and corks and then fixed in a suit-able position. D, A. L.Apparatus for the Preparation of Hydrogen Sulphide. ByP. CHANTEMILLE (Bzdl. Roc. Chim., 50,170-171).-The iron sulphideis contained i n an kpouvrette, and the acid in an ordiiiary flask fittedwith a doubly-bored cork, through the one hole of which is passed atnbe reaching to the bottom of the flask, and closed a t its upper endby an india-rubber tube and pinch-cock ; the second hole is fitted withx short# tube bent a t right angles and fitted to the lower end of thebpouvrette. The force of gas is increased or diminished by raising orlowering the flask.By E.J. MAUMENB(Bull. SOC. Chin?., 49, t350-85~).-Ch!/dr(iza~ne is evolved when afiolntion of potassium permanganate (158 grams) and sulphuric acid(40 prams SO,) is added to dried, crystallised ammonium oxalate(141.2 grams), the whole well mixed and gently heated until itbegins to boil.The gafeous product is absorbed in hydrochloric acidand a neutral solution of the salt can thus be obtained. The hyrl~o-chlwide is crystalline and very readily soluble in water, but onlysparingly in alcohol. The sublimed salt has the compositionN2H,O,2HC1, but the crystals dried by means of the anhydrous saltcontain one-fifteenth of their weight of water. When a solution ofD. A. L.F. S. K.The gas is purified in the usual manner.N. H. &I.Chydrazaine or Protoxide of AmmoniaINdRGANIC CHEJJISTRP. 35the hydrochloride is mixed with platinic chloride, a pZatinochloride isobtained, the composition of which varies with the conditions of theexperiment ; with excess of the hydrochloride, a yellow salt is formed,the composition of which is approximately NzH,0,H2PtC16, but ifexcess of platinic chloride is added, the proportion of platinum issensibly increased.The suZpphate is crystalliiie and soluble in water,but only very sparingly so in absolute alcohol ; it forms a double saltwith aluminium sulphate. The nitrate is crystalline. When a solu-tion of the nitrate is evaporated, nitric acid, nitric peroxide, nitrogen,and a compound having the composition NzHz are evolved.Action of the Electric Spark on Mlxtures of Nitric Oxidewith Hydrogen, with Methane, &c. By S. COOKE ( C h e w News,58, 130--131).-Under the influence of sparks (from a coil capableof giving a 4-inch spark), with the eudiometer wires half an inchapart, a mixture of hydrogen and nitric oxide always explodes, pro-vided the proportion of hydrogen to nitric oxide does not exceed6 : 10 ; but with the wires closer together, or with a feeble coil, or if thepressure is diminished to 300 mm.of mercury, no explosion occurs.The carefully dried gases explode quite as readily as when they arenot dried. The nature of the gaseous mixture after explosion variesvery considerably, but there is little doulnt that much of the nitricoxide is converted into oxygen aiid nitrogen. Explosions have alsobeen obtained with nitric oxide and hydrogen sulphide ; with nitricoxide and methane with the production of carbonic anhydride andoxygen : other hydrocarbons i n proper proportions and suitable spa1.kmake an explosive mixture with nitric oxide. Carbonic and nitricoxides mixed do not explode under the influence of the spark, butcombination goes on gradually with the production of carbonic anhy-dride and nitrogen.Decomposition of Nitric Oxide in Contact with Water andwith Potash.By S. COOKE (Chem. News, 58, 155--156).--Nitricoxide exposed in tubes over water in the dark undergoes gradualdecomposition with the production of nitrous acid, nitrogen, and alittle nitrous oxide. The change is always slow, but is more active atthe commencement than a t the end of an experiment; it is alsoaccelerated by the presence of platinum and by heat, whilst admix-ture with hydrogen retards it. The actionof potassium hydroxide onnitric oxide is also aided by platinum and heat (oompare this Journal,18i7, ii, 37).BJ- L.W.I?. S. K.D. A. L.D. A. L.Action of Hydrogen Sulphide on Arsenic Acid.MCCAY (Zeit. an/iZ. Chern., 27, 632-634; compare Brauner andToinihek, Trans., 1888, 145).-When a slow stream of hydrogensulphide is passed through an acidified solution of an arsenate a t 7u0,besides arsenic pentasulphide there is also formed Home free thioxy-arsenic acid, H,As03S. This, under the influence of mineral acidsand heat, decomposes into free sulphur and arsenious acid, the latterof which then yields arsenic trisulphide with the hydrogen sulphide.A solution of thioxyarsenic acid may be obtained by passing hydro16 ABSTRACTS OF CHEMICAL PAPERS.gen sulphide not in excess into a cold, dilute, acidified solution ofpotassium arsenate. If R larger quantity of hydrogen sulphide isemployed, the excess may be removed either by immediate additionof copper sulphate or by a vigorous stream of air bubbles.Anopalescence caused by free sulphur may be removed by shaking withasbestos. The clear, strongly acid liquid obtained, exhibits the fol-lowing pPoperties. It remains clear for a long time after addition ofsulphuric or hydrochloric acid ; it gives no immediate precipitatewith hydrogen sulphide, but ultimately yields one. When boiled, i tgives a precipitate of pure sulphur, without evolution of hydrogensulphide or sulphurous anhydricle. With hydrogen sulphide, the boiledand cooled liquid gives an immediate precipitate of arsenic trisulphide ;it gives no precipitate with copper sulphate ; with mercuric chlorideit gives immediately a heavy yellowish-white precipitate ; with silversulphate it gives a heavy black precipitate, the filtrate from whichcontains no arsenious acid.The potassium thioxyarsenate of Bouquetand CloGz agrees with this solution in all the above particulars.M. J. S.Barium Sulphite. By E. R. HODGES (Chem. News, 58, 128);G. S. JOHNSON (ibid., 155).-Hodges' experiments led him to inferthat barium sulphite is insoluble in hydrochloric acid ; but 3ohnsoiiproves it to be readily and completely soluble in that solvent. He,moreover, shows that pure aqueous barium chloride is not precipitatedby sulphrnrous acid, but that i n the presence of dissolved oxygen aprecipitate of barinm sulphate forms. n. A. L.Solubility of Gypsum. By G.A. RAUPFNSTRAUCH ( C h ~ m . Cent?,,1888, 891-822, from Pharin. Centrulhal., 29,229-233).-The authorfinds that a saturated solution of gypsum is readily obtained, butsupersaturation of the solution can only be obviated by shaking thesolution for some time. The solubility of gypsum increases up to thetemperature of 32", between 32" and 38" it, remains almost constant,and at, higher temperatures than 38" decreases. Natural gypsumcomports itself like artificial, provided it be pure. After heating,gypsum takes up water of crystallisation more or less rapidly, andshows the normal solubility again. J. W. L.Ancient Mortar from a Roman Wall in London. ByJ. SPILLER (Chem. News, 58, 189).-While examining some mortarfrom a Roman wall, exposed when sinking the foundations of thenew Post Office buildings in St.Martin's-le-Grand, the author found,after extracting as much silica as possible by means of dilute hydro-chloric acid, that the residue yielded nearly 11 per cent. of silica tocold dilute sodium hydroxide. Under similar treatment with coldsoda, builder's sand and pulverised flints yield a mere trace of silicain solution, whilst mortars 20, 160, and many hundreds of years oldyielded quantities of silica increasingwith the age of the mortar. Theauthor suggests that perhaps the Romans used a puzzuolana in com-pounding their mortar, or perhaps this soluble silica or silicate is thedirect result of long contact of plain sand and lime. The mortaIN0 RGANIC CHEMISTRY. 17from St. Martin's-le-Grand had the following composition percent: :-Sand and brick, 46.48 ; acid-soluble, SiO, 0.52 ; alkali-solublt~,SiO, 10.44, A1,0, 3.00, Fez03 0.48, CaO 20.02, MgO 0.76, CO, 13-03,SO3 0.37,NaCl trace, H,O and loss 4.90.D. A. L.Analysis of Money. By 5. C. WELCH (Chem. News, 58, 164-165) .-The author has analysed some specimens of " matlilla money''made in Birmingham. It is yello wish-red, reddish-yellow, brass orbronze-like in colour, and in shape like a G. They proved to bevariable alloys of lead and copper with small quantities of iron, tin,zinc, antimony, and arsenic. Some contained pieces of originallyunmolten metal, and some had a semi-fused appearance. Anothercoin resembling gold in colour contained per cent. : Cu, 62-58 ;Zn, 37.26 ; Fe, 0.11 ; Pb, 0.013, and had been silvered to pass for a6d.-piece. D.-4. L.New Hydrated Cupric Chloride. By E. CHUARD (Chem. Centr.,1888, 887, from Arch. sci. Phys. Nat. Gendve [3], 19, 477).-A hydratedcupric chloride of the formula CuCI, + 3H,O crystallises from thesolution of the green hydrate when cooled down to 0". The existenceof this compound explains the change in the colour of the solutionwhich takes place when it suffers dilntion, for the chloride of thegreen hydrate contains only 2 mols. HzO, and by diluting or bycooling below O", the chloride with 3 mols. H,O is formed, and givesa blue colour to the solution. The latter chloride again loses 1 mol.H,O when the solution is boiled. J. W. L.Purification of Mercury. By J. M. CRAFTS (BuZ7.Xoc. Chim.,49, 856--860).-Mercury can be completely freed from lead, zinc, tin,and other impurities by placing it in a slightly inclined glass tubeprovided with a funnel at the lower extremity, and aspirating agentle stream of air through the apparatus for about 48 honrs. Theoxides of the metals collect at the upper end of the tube, and afterabout 24 hours, as a rule, the surface of the mercury is quite cleanand the operation is finished. Large quantities can be treated inthis way, but mercury which has been used for amalgamating zinccontains such a large amount of impurity that this method cannot besuitably employed.A stream of piire air was passed through pure mercury containedin the apparatus described above, but even after 10 days' time noappreciable quantity of oxide was formed.Platinum, in the form of thin foil, is not attacked by cold mercur-,but when the latter is boiled, air being excluded, the platinum isgradually acted on.Only a small quantit? is dissolved, as a largeproportion separates in the form of a black powder, and almost thewhole is simply held in suspension. The surface of the mercuryremains bright, but when a stream of air is passed, the platinumcollects at the surface aft,er some time as a black powder, and on dis-tilling the separated mercury only a very small quantity of platinumremains. F. S. K.Silver is not removed by this process.VOL. LVI. 18 ABSTRACTS OF’ CHEMICAL PAPERS.Yttrium-Potassium and Y ttrium-Sodium Phosphates. BgA.DUBOIN (Compt. rend., 107, 622-624) .-YYtt&ma-pofassiurn pyro-phosphate, K2O,YZO3,2PzO5, is obtained by saturating potassium meta-phosphate with yttrium oxide a t a bright red heat, and then keepingthe mixture a t a somewhat lower temperature for a considerable time.The cooled product is extracted with water, when the phosphateremains undissolved. It forms small, colourless, highly birefractiveprisms.Yttrium-potassium orthophosphate, 3K20,YZO3,2P2O5, is obtained byadding excess of yttria to a fnsed mixture of potassium pyrophos-phate with nine times its weight of potassium chloride, and heatingover a Bunsen burner f o r 20 minutes. It crystallises in brilliant,hexagonal lamellae; sp. gr. at 20” = 3.3. The same product isobtained on heating amorphous yttrium phosphate to redness, o r evento a much higher temperature, with excess of potassium sulphate.Another orthophosphate, 3Kz0,5Y20,,6P205, is obtained by heatingpotassium sulphate to a high temperature with a much larger propor-tion of yttrium phosphate.It forms brilliant, trausparent, colour-less, hexagonal prisms.If the mixture of yttrium phosphate and potassium sulphate con-tains 10 per cent. of the former, and is heated t o a very high tempe-rature for about 10 hours, yttrium phosphate is obtained i u a formidentical with xenotime. C. H. B.So-called (( Crackle” China. By C. LAUTH and G. DUTATLLY(Bull. Xoc.. Chim., 49, 948--956).-The name crackle china is givento china, the glazing of which is cracked in a regnlai- manner so as toform a sort of network.If the appearance is similar to that of fishscales, the china is said to be “ troutod.” The Chinese colour theinterstices by means of smoke or Indian ink, and the articles thusproduced are valued very highly.The crackle effect is due to the unequal contraction of the glazeand of the paste. When the firing is continued for a certain time,the coefficients of expansion of the glaze and of the paste graduallybecome identical, and i f the firing is stopped at this point the crackleeffect is not produced ; if, however, the burning is continued, the co-efficients of expansion again become different, and fracture of the glazetakes place on cooling.If the temperature which is required for a certain paste and acertain glaze to give perfect porcelain is known, and if the baking isalways arrested a t this point, the crackle effect can be obtained byaltering the composition of the paste or that of the glaze.The latteralternative is the more practical, but i t was found necessary to alterthe degree of fusibility of the glaze at the same time. As, forvarious reasons, i t is disadvantageous to employ a readily fusibleglaze, the temperature of fusion was raised by adding a larger pro-portion of silica. The paste experimented on is that employed atSdvres ; its composition is-Silica, 66 ; alumina, 27 ; alkalis, 7 ;and the ordinary glaze employed a t Sbvres has the compositionINORQANIC CHEMISTRY. 19Silica ................ 66-18Alumina. .............14-55Alkalis, .............. 3.55Chalk.. .............. 15-90The glaze which was found to give the best crackle effecb has thefollowing composition :-Silica ................ 79.42Alumina, ............. 11 -80Alkalis. .............. 5-51Chalk.. .............. 2-88Crackle china is also produced if the pEoportion of alumina in theglaze is increased, for example with a glaze composed of-Silica, 69-92; alumina, 18.13 ; alkalis, 11.95,but the effects produced are not so good, and it seems that fracturemore frequently occurs. Whichever method is adopted, it is alwaysnnrticularly advantageous to substitute the chalk in the glaze forkknlis. If the propartion of alumina is increased, the necessarydegree of fusibility is obtained by addimg other bases ; if the propor-tion of alumina is decreased, the quantity of other bases is diminished,and an additional quantity of silica is added.It was found that with a given glaze the crackle china can be pro-duced by making the paste more readily fusible.This can be doneby increasing the quantity of felspar, and the results are better thesmaller the proportion of quartz contained i n the paste. The extremecomposition of the unbaked paste employed for these experimentswas-Silica, 58.5 ; alumina, 28.0 ; alkalis, 5.5 ; water, 8.0.The glaze should be of medium thickness, and to obtain a closemesh it should have the composition given in one of the above twoexamples. If a larger mesh is required the " crackle " glaze is mixedwith the ordinary glaze, and the mesh is larger the greater the pro-portion of the latter.It is advisable to make crackle china of sufficient thickness t oavoid all chance of fracture.F. S. K.Atomic Weight of Tin. By T. BONGARTZ and A. CLASSEN (Bey.,21, 2'300-2909) .-The authors made determinations of the atomicweight of tin (1) by oxidising the pure metal with nitric acid; (2) bythe electrolysis of ammonium stannic chloride ; ( 3 ) by the electroljslsof potassium stannic chloride ; and (4) by the electrolysis of stannicbromide. The results are given in &he following table :-Number of Difference betweenexperiments. Atomic weight. maximum and minimum.(1.) ...... 11 11 8,7606 u-459(3.) ...... 10 118-7975 0.1 63(4.) ...... 10 11 8.7309 0.144...... ( 2 . ) 16 118.8093 0.228-Total ... 47 Average 118.7745 0 *;I 4s 5c 20 ABSTRACTS OF CHEMICAL PAPERS.The average of the 26 results obtained by the electrolysis of theammonium and potassium double salts is higher than the averageof all the results. This is probably due to the fact that in oxidisinqthe tin, the platinum vessel is always attacked to a slight extent, andthat in dissolving the stannic bromide, a small portion may be lost byvolatilisation. If, therefore, the average of the 26 experiments istaken as the most trustworthy result, the atomic weight of tin is118.8034 (0 = 15.96) or 119.1 (0 = 16).Experiments in which it was sought to determine the atomic weightby converting pure tin into stannic sulphide, and estimating thesulphur in the latter, did not give satisfactory results (118.676 as theaverage of eight experiments). It was also found that when the oxideis reduced with pure hydrogen, it small quantity of tin is alwaysvolatilised. F. S. K.Action of Incandescent Platinum Wire on Gases andVapours. By W. R. HODGKINSON and F. I(. S LOWNDES (Chern.News, 58, 187).-When a spiral of platinum wire is exposed to theaction of chlorine in a glass globe and rendered incandescent by 2nelectric current, a white glow is observed round the heated wire,the sides of the globe become covered with platinous chloride, aridvery fine crystals of platinum are formed on the wire. With bro-mine or iodine vapour, the flame round the wire is greatly increased,but only a very small quantity of platinous bromide or a trace of theiodide is formed, and no crystalline metal is produced. With drysilicon fluoride, crystals of silicon are deposited, and the top of theglobe is deeply corroded, presumedly by the liberated fluorine.D. A. I;.Normal Platinum Chloride. By R. ENGEL (BUZZ. HOG. Chi~z., 50,100-102) .--Norm:tl platinum chloride, PtCla + 4H20, can be obtainedby dissolving the calculated amount of platinum oxide in a solutionof platinum chloride hydrochloride; the liquid is filtered andevaporated. The crystals are deliquescent like those of the hydro-chloride. When dry h j drogen chloride is passed over the crystalsheated a t 50°, the hydrochloride is formed ; a t a lower temperatureno combination takes place. N. H. M