TECHNICAL CHEMISTRY. 9 5 T e c h n i c a1 C h e m i s t r y . Separation of Carbon, Silicon, Sulphur, and Phosphorus in the Refining and Puddling Furnace, and in the Bessemer Con- verter. By I. L. BELL (Journal of the Iron aid Steel Institufe, 1877, 390 ; Dinyl. poZyt. J., ccxxv, 264-268, 351-357).-Expolsed to the intensely deoxidising agency of the blast-f urnace, portions of the silica, and probably the greater part of the sulphur compounds, lose theiy oxygen and are taken up by the reduced iron. Practically the whole of the phosphorus is found in the metal. I n the primit'ive low furnace of the Catalan type, the deoxidising agency is insufficient to rcduee the ore completely, and silicon, sulphiir, and phosphorus are almost absent from the resulting iron. Experi- ments on the direct process conducted by Siemens show that the o1.e is imperfectly reduced, and that most of the phosphorus goes into the slag, for the slag was found t o contain 36.51 per cent.of metallic iron, and 2.24 per cent. of phosphorus, whilst the resulting iron gave Fe 99.71, C 0.12, Si 0.065, S 0.027, P 0.074. Cleveland pig puddled in a Danks' furnace gave a close approximation in the percentage of phosphorus. Experiments on the refining of pig iron made a t the Rowling Works with Bowling cold-blast pig containing C 3593, Si 1.25.5 S 0.033, P 0.565, showed that the loss per cent. was as follows :- C 9.33, Si 88.05, S 24.24, P 13.27. A second serics of experiments gave an average loss per cent. of C 8.89, Si 90.12, S 29.77, 1' 48.12. A charge of Clarence No.3 hot-blast pig, containing C 3.12, X i 2.80,96 ABSTRACTS OF CEfEMICAL PAPERS. S 0.11, P 1-87, showed on refining a loss per cent. for the C 19.87 Si 90.57, S 100, P 42.85. The difference between refining and the Bessemer process is one of appearance only ; in principle there is a close resemblance. In both cases the silicon is most speedily driven off, but in the ordinary re- finery it is the phosphorus which is next or perhaps simultaneously attacked, whilst in the Bessemer process the phosphorus remains un- changed. Experiments made at the Weardale Iron Company’s Works with Cleveland pig iron in the Bessemer converter, showed that a pig iron containing C 3.60, Si 1.76, S 0.175 and P 1.64 respectively, lost, after a five minutes’ blow, C. 8.61, per cent., Si 77.72, S 16.63, and gained, P 10.89 per cent.A pig iron containing C 3.452, Si 1.626, S 0,120, and P 1.423 per cent., after a nine miriutes’ blow, lost C 86.03 per cent., Si 98.8, and gained, S 0.005 and P 3.23 per cent. After it sixteen minutes’ blow a pig iron containing C 3.48, Si 2.07, S 0.05, and P 1.46, lost C 98.56, Si 96.61, whilst the S remained unchanged, and the P gained 15.75 per cent. After a blow of twenty minutes, an iron containing C 3.87, Si 1.910 S 0.046, P 1.92, lost C 82.83, Si 99.63, and gained S 34.78 and P 15.78. The general conclusion is, that in refining, the loss per cent. on the original quantity present of Si is 90, C 10, S 30, P 50, whilst in the converter, according to the periods of blowing, the loss of Si is 77 to 99, C 8 to 99, whilst in S there is no change, and in P there is 10 to 16 per cent.of gain. 20 to 30 per cent. of fused ore and cinder was run into a converter before its charge of iron ; after blowing, no phosphorus was found to have been removed. In another case the blast was continued until 2.5 per cent. of the iron was oxidised in the converter ; the iron was quite liquid, but no phosphorus was removed. When molten pig iron was poured slowly through a column of fiised oxide of iron, 44-68 per cent. of the P was removed. There seems reason for believing that the high temperature of the metal in the Bessemer converter inverts this action between oxide of iron and phos- phorus. Carbon is that element which makes pig iron fusible at compara- tively low temperatures, and, except in the converter, carbon is more tenaciously retained than Si, S, or P.If the carbon is removed too rapidly, the phosphorus may not be removed. A portion of Clarence pig, from which nearly all the carbon had been burnt off, was trans- ferred from the converter to a “ fettled ’’ puddling furnace, and was ready for balling in five minutes, giving little time for the oxidation of the phosphorus. Another portion was transferred to a furnace charged with melted oxide-of iron. In the first case the phosphorus was reduced from 1.46 per cent. in the original iron to 0.74 per cent., in the second case from 1.46 to 0.54 per cent. Experience shows that the impurities are more easily removed when the pig is melted before charging the puddling furnace. Bar iron puddled cold contained 0.597 per cent. of P ; puddled hot, 0.299.In another case 0.592 and 0.209 per cent. of P respectively were obtained. Analyses by Proctor show that in the Danks’ furnace, neglecting the small quantity of sulphur, the silicon is mainly removed duringTECHNICAL CHEMISTRY. 9 7 the melting of the charge ; phosphorus coming next in amount, and lastly carbon. Before balling,. the phosphorus and carbon have lost equal amounts (per cent. of originals), and from this point the phos- phorus remains unchanged. I n puddling, the heat required for finish- ing the process may approach that at which the phosphorus is no longer oxidised. In portions the temperature may be so high that the iron takes up phosphorus from the slag, for puddled bar sometimes contains more phosphorus than the same iron a t an earlier sta<ge.Iron free from pliosphorus in a bath of slag coritainirig phosphorus, in a Siemens-Martin furnace, was found t o take up phosphorus towards the end of the heat. Machine puddling removes phospho- rus more completely than hand puddling, and the iron produced is more uniform in quality. Method of imparting Sonorousness to Soft Metallic Alloys. By B. SILLIMAN (Dingl. pohjt. J., ccxxv, 268--270).-B. Silliman, of Newhaven (Conn.), has invented a method of making soft metal alloys, such as pewter, Britannia metal, &c., sonorous. A capacious oil or paraffin bath is heated to a temperature from 5" to 5.5" degrees below the melting point of the alloy t o be treated. The exact point is determined in each case by heating the bath until a sample alloy, freely suspended, just begins to be scratched by a wire which has a slightly higher melting point.Small and thin articles are then immersed f o r from 15 to 30 seconds ; larger articles, such as urns, remain in a minute or more. Careful handling is necessary to prevent distortion. It is immaterial whether the articles be cooled quickly or slowly, and the process may be repeated if necessary ; but it is essential that each article be heated uniformly, or the parts not sufficiently heated will have a prejudicial effect on the tone. Although the alloy may lose something of the compactness imparted to it by rolling, pressing, &c., yet it does not become porous, whilst its hardness and st'iffness are increased. Alloys thus treated can be soldered, plated, &c., but cannot be hammered, pressed, &c., without injuring their souiiding qualities. J.T. 9. T. Crystallisation of Metallic Oxides from Glass. By PAUL EBELL (Dingl. polyt. J . , ccxxv, $0-78, and 168--175).-1n former communications (1874, ccxiii, 53, and 1876, ccxx, G4), the author has shown by a number of examples, that glass in the molten state is a powerful solvent for many bodies, which in consequence play an im- portant part in glass manufacture. The action of the metals silver, gold, and copper, with formation of aventuriiie and hematinone, was noticcd, and the results of fusion of glass with alumina, oxide of iron, manganese, chromium, zinc, were also given. In the present communication he deals with the action of silica, lime, baryta, su1- pbate of sodium, phosphate of calcium, crj-olite and sulphides.1. Uehauiow of Glass fused u*it?b emess of SiliccL.--ln all the experi- ments the glass was formed by fusing a mixture called Hautefenilk's mixture, of the following composition. Sand, 150 parts ; chalk, 35.5 ; calcined soda, 80 ; potash, 14.0 ; nitre, 20. This Hautcfeuille mixture VOL. L X X l I l h58 ABSTRA4CTS OF CHEMICAL PAPERS. was fused with 42.86 per cent., 66.67 per cent., and 118.67 per c:snt. of silica respectively, and the results were as follows :-The Erst two could be fused at high temperatures, but gave no signs of Jrystallisation in the interior of the glass, even when very slowly cooled. The third, though it could be fused, was less fluid than the others.Quickly cooled specimens were transparent, but exhibited a shade of green. Specimens slowly cooled showed long six-sided crystalline forms, of milky whiteness, and resembling the tridymite of blast furnaces. 2. Behaviour of Glass with excess of Znze.-The following mixtures were fused:-(a) 100 parts of Hautefeuille mixture and 100 parts lime. ( b ) 100 parts of Hautefeuille and 200 of lime. Both a and b gave at high temperatares a perfectly clear glass; a, when rapidly cooled, gave a transparent glass. When slowly cooled, the result had a stony appearance. Microscopic examination showed white long needle-like crystals ; b was very similar in its behaviour, but the crystals were more numerous. 3. Behaviour oj* Calcium Phosphnte with Glass.-Phosphate has long been used for giving a milky appearance to glass.The addition of 10 to 20 per cent. of burnt bones causes the glass to assume the form of milk-glass or opal glass. If quickly cooled the glass is transparent, but on gentle warming the whole becomes suddenly white. An ex- periment made by melting 100 parts of glass with 30 parts of burnt bones showed that the bone-ash is dissolved with difficulty, on account of its lightness. In 1863 (Di.i~gZ., clxvii, 27) Schur proposed to sub- stitute guano for bone-ash, and with good results. The author thinks, however, that a compound of definite composition would be better, and prepares a phosphate by precipitating an ammoniacal solution of calcium chloride with sodium phosphate. A mixture of 300 parts of broken white-glass with 30 of this calcium phosphate, melts easily.Quickly cooled specimens are colourless and transparent. but become milk-white on warming, Slowly cooled specimens are white, but the particles in the glass are larger than in the other case. Another mixture of 300 parts of broken pieces of white glass with 60 of calcium phosphate was melted and part poured out. This of course cooled suddenly, and when gently warmed became milk-white as in the former case. The crystals in the glass were very small. The remainder in the crucible cooled very slowly, and had quite a different appearance. Though it was milk-white, the cloudiness was less dense. The crystals were large and lay in a clear glass, and were visible to the naked eye. An experiment was made to discover the nature of these crystals by treating a thin section with hydrochloric acid.The crystals disappeared, and the author therefore considers the crystals to be calcium phosphate. Another fact noticed during these experiments was that calcium phosphate, t'hough generally re- ducible with difficulty, can, in t,he presence of the molten glass, be easily reduced to calcium phosphide by coal, the result being to colour the glass very black. 4. Fusion of Cryolite with Glaas.-In the manufacture of white or milk-glass cryolite ha's often been substituted for bone-ash, and with good results. Benrath in 1869 (Dim$. polyt. J., cxcii, 2S9)TECHXICAL CHEMISTRY. 99 aiialysed the glass and found SiO,, 70.01 ; A1,0,, 10.78, and Na,O, 19-21. As he found no fluorine, he thought the small crystals which made the glass opaque must be A1203.Richters (1869, cxci, 301) found much fluorine, and considered it essential in the production of the milky appearance. It has been already shown that a large quantity of alumina can be dissolved in the glass without affecting its transparency. Cryolite and sand in proportion of 1 to 2 were fused. Fluoride of silicon escaped during the fusion, and the product when cooled rapidly was a transparent colourless glass. The mass left in crucible to cool slowly was less opaque, the crystals, as in case of calcium phosphate of lime, being larger and more scattered in a clear base. On analysis the glass gave 1.74 per cent. of fluorine. On re-melting this opaque glass with silica to drive off all the fluorine, the resulting glass gave no appearance of milkiness.This proves that fluorine is neces- sary to the formation of these crystals. It cannot be in the form of fluor spar, however, as calcium fluoride has long been employed in glass manufacture, and yet this milkiness has not been observed. It cannot be in the form of sodium fluoride, as the cry~t~als are insoluble. The author therefore thinks that it must be some fluoride of alnminium, and he supports this belief by the fol- lowing experiment: 100 parts of broken lime glass and 10 parts of silico-fluoride of sodium were melted together. The product had a greenish appearance, but the glass would not run. This proves that neither sodium nor lime forms part of the crystals. 5. F.usion of Xubphates zrlith Glass.-Felouze (1865, Dingl.polyt. J., clxxviii, 134, and 1867, clxxxiv, 310) has shown that glass can contain 3 per cent. of sodiuni sulphate. The author found 3.6 per cent. of this substance ih some experiments he made. 6. Fusion, with Sulphides.-Sulphides of sodium and calcium give rtn intelzse Ped colour, which shades off into brown. The sulphides are usually obtained in the glass manufacture by reduction with coal, during the fusion of glass of sulphates. When fused with sulphides, the red glass on quickly cooling is transparent, but by slowly cooling, crystals large and small are found in the glass. Instead of adding a sulphide, sulphur alone can be added if the glass be a sodium or potassium glass. The resulting glass is however browner than in the former case. Many experiments showed that when free silica is present, the glass remains colourless, although sulphur or sulphides be added.This proves that to colour a glass, more base must be present than the silica, can unite with. In this case the sulphur can form a sulphide, and the glass is coloured. As a result of several experiments, a neutral glass which colours with sulphides has the following composition : 1 part of base t,o 2.5 parts of silica, or 2RO + 5Si0,. When this proportion of silica is exceeded, the glass does not colour; when it is below this, On warming it became opaque. the colour becomes very intense. S. Toughened Glass. By B O W R ~ E (DiwgZ polyt. J., ccxxv, 360- 365).-The author passed some time at the glass works at Choisy- le-Roi, and had therefore had the opportunity of carefully examining100 ABSTRACTS OF CHEMICAL PAPERS.the process of toughening glass (1875, ccxv, 186-381 ; ccxvi, '75; ccxviii, 181) in its details. The glass, after receiving its form, is cooled in a bath of certain temperature, defined by experiment, and varying 1st with the heat to which the glass vessel or object is heated before immersion ; 2nd, with the size and thickness of the vessel or object, and with the chemical composition of the glass. The chemical composition aEects the amount of softening, and therefore the degree to which the glass can be heated, and so indirectly the temperature of the bath. The author experimented with a glass of 300 parts of sand, 100 of potash and soda, and 50 of red lead. This could be easily toughened, and so can all crystal-glass, in a bath of fat a t temperatures varying from 60" to 138" ; common glass in a mixture of oil and fat betwecn 150" and 315".The amount of soda and potash has a great influence on the power of toughening. The thicker objects require more heating, and a hotter bath. The mixture of the bath is also a point of great importance. All Water always The best results are obtained with pure fats With crystal-glass and low temperatures, fat is the best; The objects must be heated uniforndy, or the results will be un- The objects must be made of homogeneous glass, and heated uw,lformZy. The oil or fat baths must be easily movable, and a t the same time large enough to contain several objects a t the same time. A small tramway on which the bath may run is suggested.There are many points of the utmost importance in the manipulation of the glass while being immersed that only long practice can teach ; and yet the success of the operation depends on them. After the objects are placed in the fat-bath, the bath is placed in a room for four or five hours a t the melting-point$ of the fat. After this, the glass is cleansed by caustic soda. When oil is used, the glass is cooled more slowly, as the bath is kept hot f9r a longer period. substances have not the same value in this respect. makes the glass brittle. and oils. with higher temperatures oil and fat are employed. f avourable. The cost of toughening is- For drinking glasses of all kinds, -21 to -33d. ,, lamp-cylinders, *30tZ. ,, lamp-globes, .50d.S. Analyses of Glass. By I?. PRIMKE (Dingl. polyt. ?J., CXXV, 174). I. Glass bell-jar of a Thomson's electrometer which showed great power of insulation. 11. Glass of very similar composition used for optical purposes, and analysed byBerthier :- SiO,. P O . Na20. PbO. CaO. MgO. F'e2O3MnO. 1. 58.450 9.236 3.745 28.019 0.064 0'0.54 0.4'74 - = 100.042 11. 59.2 9.0 - 28.2 - - 0.4 1.0 = 97.8 Impurities neglected, the composition is as follows :- Si02. KZO. Na20. PbO. 58-77 9.28 3.77 28.18 = 100 S.TECHNICAL CHEMISTRY. 9 5T e c h n i c a1 C h e m i s t r y .Separation of Carbon, Silicon, Sulphur, and Phosphorus inthe Refining and Puddling Furnace, and in the Bessemer Con-verter. By I. L. BELL (Journal of the Iron aid Steel Institufe, 1877,390 ; Dinyl. poZyt.J., ccxxv, 264-268, 351-357).-Expolsed to theintensely deoxidising agency of the blast-f urnace, portions of the silica,and probably the greater part of the sulphur compounds, lose theiyoxygen and are taken up by the reduced iron. Practically the wholeof the phosphorus is found in the metal.I n the primit'ive low furnace of the Catalan type, the deoxidisingagency is insufficient to rcduee the ore completely, and silicon, sulphiir,and phosphorus are almost absent from the resulting iron. Experi-ments on the direct process conducted by Siemens show that the o1.eis imperfectly reduced, and that most of the phosphorus goes into theslag, for the slag was found t o contain 36.51 per cent. of metallic iron,and 2.24 per cent. of phosphorus, whilst the resulting iron gaveFe 99.71, C 0.12, Si 0.065, S 0.027, P 0.074.Cleveland pig puddledin a Danks' furnace gave a close approximation in the percentage ofphosphorus.Experiments on the refining of pig iron made a t the RowlingWorks with Bowling cold-blast pig containing C 3593, Si 1.25.5S 0.033, P 0.565, showed that the loss per cent. was as follows :-C 9.33, Si 88.05, S 24.24, P 13.27. A second serics of experimentsgave an average loss per cent. of C 8.89, Si 90.12, S 29.77, 1' 48.12.A charge of Clarence No. 3 hot-blast pig, containing C 3.12, X i 2.8096 ABSTRACTS OF CEfEMICAL PAPERS.S 0.11, P 1-87, showed on refining a loss per cent. for the C 19.87Si 90.57, S 100, P 42.85.The difference between refining and the Bessemer process is one ofappearance only ; in principle there is a close resemblance.In bothcases the silicon is most speedily driven off, but in the ordinary re-finery it is the phosphorus which is next or perhaps simultaneouslyattacked, whilst in the Bessemer process the phosphorus remains un-changed. Experiments made at the Weardale Iron Company’s Workswith Cleveland pig iron in the Bessemer converter, showed that a pigiron containing C 3.60, Si 1.76, S 0.175 and P 1.64 respectively,lost, after a five minutes’ blow, C. 8.61, per cent., Si 77.72, S 16.63,and gained, P 10.89 per cent. A pig iron containing C 3.452, Si1.626, S 0,120, and P 1.423 per cent., after a nine miriutes’ blow,lost C 86.03 per cent., Si 98.8, and gained, S 0.005 and P 3.23 percent.After it sixteen minutes’ blow a pig iron containing C 3.48,Si 2.07, S 0.05, and P 1.46, lost C 98.56, Si 96.61, whilst the Sremained unchanged, and the P gained 15.75 per cent. After ablow of twenty minutes, an iron containing C 3.87, Si 1.910 S 0.046,P 1.92, lost C 82.83, Si 99.63, and gained S 34.78 and P 15.78.The general conclusion is, that in refining, the loss per cent. on theoriginal quantity present of Si is 90, C 10, S 30, P 50, whilst in theconverter, according to the periods of blowing, the loss of Si is 77 to99, C 8 to 99, whilst in S there is no change, and in P there is 10 to16 per cent. of gain.20 to 30 per cent. of fused ore and cinder was run into a converterbefore its charge of iron ; after blowing, no phosphorus was found tohave been removed. In another case the blast was continued until 2.5per cent.of the iron was oxidised in the converter ; the iron was quiteliquid, but no phosphorus was removed.When molten pig iron was poured slowly through a column of fiisedoxide of iron, 44-68 per cent. of the P was removed. There seemsreason for believing that the high temperature of the metal in theBessemer converter inverts this action between oxide of iron and phos-phorus.Carbon is that element which makes pig iron fusible at compara-tively low temperatures, and, except in the converter, carbon is moretenaciously retained than Si, S, or P. If the carbon is removed toorapidly, the phosphorus may not be removed. A portion of Clarencepig, from which nearly all the carbon had been burnt off, was trans-ferred from the converter to a “ fettled ’’ puddling furnace, and wasready for balling in five minutes, giving little time for the oxidationof the phosphorus.Another portion was transferred to a furnacecharged with melted oxide-of iron. In the first case the phosphoruswas reduced from 1.46 per cent. in the original iron to 0.74 per cent.,in the second case from 1.46 to 0.54 per cent. Experience shows thatthe impurities are more easily removed when the pig is melted beforecharging the puddling furnace. Bar iron puddled cold contained0.597 per cent. of P ; puddled hot, 0.299. In another case 0.592 and0.209 per cent. of P respectively were obtained.Analyses by Proctor show that in the Danks’ furnace, neglectingthe small quantity of sulphur, the silicon is mainly removed durinTECHNICAL CHEMISTRY.9 7the melting of the charge ; phosphorus coming next in amount, andlastly carbon. Before balling,. the phosphorus and carbon have lostequal amounts (per cent. of originals), and from this point the phos-phorus remains unchanged. I n puddling, the heat required for finish-ing the process may approach that at which the phosphorus is nolonger oxidised. In portions the temperature may be so high that theiron takes up phosphorus from the slag, for puddled bar sometimescontains more phosphorus than the same iron a t an earlier sta<ge.Iron free from pliosphorus in a bath of slag coritainirig phosphorus,in a Siemens-Martin furnace, was found t o take up phosphorustowards the end of the heat. Machine puddling removes phospho-rus more completely than hand puddling, and the iron produced ismore uniform in quality.Method of imparting Sonorousness to Soft Metallic Alloys.By B.SILLIMAN (Dingl. pohjt. J., ccxxv, 268--270).-B. Silliman, ofNewhaven (Conn.), has invented a method of making soft metalalloys, such as pewter, Britannia metal, &c., sonorous.A capacious oil or paraffin bath is heated to a temperature from 5"to 5.5" degrees below the melting point of the alloy t o be treated. Theexact point is determined in each case by heating the bath until asample alloy, freely suspended, just begins to be scratched by a wirewhich has a slightly higher melting point. Small and thin articlesare then immersed f o r from 15 to 30 seconds ; larger articles, such asurns, remain in a minute or more.Careful handling is necessary toprevent distortion.It is immaterial whether the articles be cooled quickly or slowly,and the process may be repeated if necessary ; but it is essential thateach article be heated uniformly, or the parts not sufficiently heatedwill have a prejudicial effect on the tone. Although the alloy maylose something of the compactness imparted to it by rolling, pressing,&c., yet it does not become porous, whilst its hardness and st'iffnessare increased. Alloys thus treated can be soldered, plated, &c., butcannot be hammered, pressed, &c., without injuring their souiidingqualities. J. T.9. T.Crystallisation of Metallic Oxides from Glass.By PAULEBELL (Dingl. polyt. J . , ccxxv, $0-78, and 168--175).-1n formercommunications (1874, ccxiii, 53, and 1876, ccxx, G4), the authorhas shown by a number of examples, that glass in the molten state isa powerful solvent for many bodies, which in consequence play an im-portant part in glass manufacture. The action of the metals silver,gold, and copper, with formation of aventuriiie and hematinone, wasnoticcd, and the results of fusion of glass with alumina, oxide ofiron, manganese, chromium, zinc, were also given. In the presentcommunication he deals with the action of silica, lime, baryta, su1-pbate of sodium, phosphate of calcium, crj-olite and sulphides.1. Uehauiow of Glass fused u*it?b emess of SiliccL.--ln all the experi-ments the glass was formed by fusing a mixture called Hautefenilk'smixture, of the following composition.Sand, 150 parts ; chalk, 35.5 ;calcined soda, 80 ; potash, 14.0 ; nitre, 20. This Hautcfeuille mixtureVOL. L X X l I l 58 ABSTRA4CTS OF CHEMICAL PAPERS.was fused with 42.86 per cent., 66.67 per cent., and 118.67 perc:snt. of silica respectively, and the results were as follows :-TheErst two could be fused at high temperatures, but gave no signs ofJrystallisation in the interior of the glass, even when very slowlycooled. The third, though it could be fused, was less fluid than theothers. Quickly cooled specimens were transparent, but exhibited ashade of green. Specimens slowly cooled showed long six-sidedcrystalline forms, of milky whiteness, and resembling the tridymiteof blast furnaces.2.Behaviour of Glass with excess of Znze.-The following mixtureswere fused:-(a) 100 parts of Hautefeuille mixture and 100 partslime. ( b ) 100 parts of Hautefeuille and 200 of lime. Both aand b gave at high temperatares a perfectly clear glass; a, whenrapidly cooled, gave a transparent glass. When slowly cooled, theresult had a stony appearance. Microscopic examination showedwhite long needle-like crystals ; b was very similar in its behaviour,but the crystals were more numerous.3. Behaviour oj* Calcium Phosphnte with Glass.-Phosphate has longbeen used for giving a milky appearance to glass. The addition of 10to 20 per cent. of burnt bones causes the glass to assume the form ofmilk-glass or opal glass.If quickly cooled the glass is transparent,but on gentle warming the whole becomes suddenly white. An ex-periment made by melting 100 parts of glass with 30 parts of burntbones showed that the bone-ash is dissolved with difficulty, on accountof its lightness. In 1863 (Di.i~gZ., clxvii, 27) Schur proposed to sub-stitute guano for bone-ash, and with good results. The author thinks,however, that a compound of definite composition would be better, andprepares a phosphate by precipitating an ammoniacal solution ofcalcium chloride with sodium phosphate. A mixture of 300 parts ofbroken white-glass with 30 of this calcium phosphate, melts easily.Quickly cooled specimens are colourless and transparent. but becomemilk-white on warming, Slowly cooled specimens are white, but theparticles in the glass are larger than in the other case.Another mixture of 300 parts of broken pieces of white glass with60 of calcium phosphate was melted and part poured out.This ofcourse cooled suddenly, and when gently warmed became milk-whiteas in the former case. The crystals in the glass were very small.The remainder in the crucible cooled very slowly, and had quite adifferent appearance. Though it was milk-white, the cloudiness wasless dense. The crystals were large and lay in a clear glass, and werevisible to the naked eye. An experiment was made to discover thenature of these crystals by treating a thin section with hydrochloricacid. The crystals disappeared, and the author therefore considersthe crystals to be calcium phosphate.Another fact noticed duringthese experiments was that calcium phosphate, t'hough generally re-ducible with difficulty, can, in t,he presence of the molten glass, beeasily reduced to calcium phosphide by coal, the result being to colourthe glass very black.4. Fusion of Cryolite with Glaas.-In the manufacture of whiteor milk-glass cryolite ha's often been substituted for bone-ash, andwith good results. Benrath in 1869 (Dim$. polyt. J., cxcii, 2S9TECHXICAL CHEMISTRY. 99aiialysed the glass and found SiO,, 70.01 ; A1,0,, 10.78, and Na,O,19-21. As he found no fluorine, he thought the small crystals whichmade the glass opaque must be A1203. Richters (1869, cxci, 301)found much fluorine, and considered it essential in the production ofthe milky appearance.It has been already shown that a large quantity of alumina can bedissolved in the glass without affecting its transparency.Cryolite andsand in proportion of 1 to 2 were fused. Fluoride of silicon escapedduring the fusion, and the product when cooled rapidly was a transparentcolourless glass. The mass left incrucible to cool slowly was less opaque, the crystals, as in case of calciumphosphate of lime, being larger and more scattered in a clear base. Onanalysis the glass gave 1.74 per cent. of fluorine. On re-melting thisopaque glass with silica to drive off all the fluorine, the resulting glassgave no appearance of milkiness. This proves that fluorine is neces-sary to the formation of these crystals.It cannot be in the form of fluor spar, however, as calcium fluoridehas long been employed in glass manufacture, and yet this milkinesshas not been observed.It cannot be in the form of sodium fluoride,as the cry~t~als are insoluble. The author therefore thinks that it mustbe some fluoride of alnminium, and he supports this belief by the fol-lowing experiment: 100 parts of broken lime glass and 10 parts ofsilico-fluoride of sodium were melted together. The product had agreenish appearance, but the glass would not run. This proves thatneither sodium nor lime forms part of the crystals.5. F.usion of Xubphates zrlith Glass.-Felouze (1865, Dingl. polyt. J.,clxxviii, 134, and 1867, clxxxiv, 310) has shown that glass can contain3 per cent. of sodiuni sulphate.The author found 3.6 per cent. of thissubstance ih some experiments he made.6. Fusion, with Sulphides.-Sulphides of sodium and calcium givertn intelzse Ped colour, which shades off into brown. The sulphidesare usually obtained in the glass manufacture by reduction with coal,during the fusion of glass of sulphates. When fused with sulphides,the red glass on quickly cooling is transparent, but by slowly cooling,crystals large and small are found in the glass. Instead of adding asulphide, sulphur alone can be added if the glass be a sodium orpotassium glass. The resulting glass is however browner than in theformer case.Many experiments showed that when free silica is present, the glassremains colourless, although sulphur or sulphides be added.Thisproves that to colour a glass, more base must be present than the silica,can unite with. In this case the sulphur can form a sulphide, andthe glass is coloured. As a result of several experiments, a neutral glasswhich colours with sulphides has the following composition : 1 part ofbase t,o 2.5 parts of silica, or 2RO + 5Si0,. When this proportion ofsilica is exceeded, the glass does not colour; when it is below this,On warming it became opaque.the colour becomes very intense. S.Toughened Glass. By B O W R ~ E (DiwgZ polyt. J., ccxxv, 360-365).-The author passed some time at the glass works at Choisy-le-Roi, and had therefore had the opportunity of carefully examinin100 ABSTRACTS OF CHEMICAL PAPERS.the process of toughening glass (1875, ccxv, 186-381 ; ccxvi, '75;ccxviii, 181) in its details.The glass, after receiving its form, is cooled in a bath of certaintemperature, defined by experiment, and varying 1st with the heat towhich the glass vessel or object is heated before immersion ; 2nd, withthe size and thickness of the vessel or object, and with the chemicalcomposition of the glass.The chemical composition aEects theamount of softening, and therefore the degree to which the glass canbe heated, and so indirectly the temperature of the bath. The authorexperimented with a glass of 300 parts of sand, 100 of potash andsoda, and 50 of red lead. This could be easily toughened, and socan all crystal-glass, in a bath of fat a t temperatures varying from 60"to 138" ; common glass in a mixture of oil and fat betwecn 150" and315". The amount of soda and potash has a great influence on thepower of toughening.The thicker objects require more heating, and a hotter bath.The mixture of the bath is also a point of great importance. AllWater alwaysThe best results are obtained with pure fatsWith crystal-glass and low temperatures, fat is the best;The objects must be heated uniforndy, or the results will be un-The objects must be made of homogeneous glass, and heated uw,lformZy.The oil or fat baths must be easily movable, and a t the same timelarge enough to contain several objects a t the same time. A smalltramway on which the bath may run is suggested. There are manypoints of the utmost importance in the manipulation of the glass whilebeing immersed that only long practice can teach ; and yet the successof the operation depends on them.After the objects are placed in the fat-bath, the bath is placed in aroom for four or five hours a t the melting-point$ of the fat. After this,the glass is cleansed by caustic soda. When oil is used, the glass iscooled more slowly, as the bath is kept hot f9r a longer period.substances have not the same value in this respect.makes the glass brittle.and oils.with higher temperatures oil and fat are employed.f avourable.The cost of toughening is-For drinking glasses of all kinds, -21 to -33d.,, lamp-cylinders, *30tZ.,, lamp-globes, .50d. S.Analyses of Glass. By I?. PRIMKE (Dingl. polyt. ?J., CXXV, 174).I. Glass bell-jar of a Thomson's electrometer which showed greatpower of insulation. 11. Glass of very similar composition used foroptical purposes, and analysed byBerthier :-SiO,. P O . Na20. PbO. CaO. MgO. F'e2O3MnO.1. 58.450 9.236 3.745 28.019 0.064 0'0.54 0.4'74 - = 100.04211. 59.2 9.0 - 28.2 - - 0.4 1.0 = 97.8Impurities neglected, the composition is as follows :-Si02. KZO. Na20. PbO.58-77 9.28 3.77 28.18 = 100 S